Performance Analysis of Doubly-Fed Induction Generator (DFIG) - Based Wind Turbine With Sensored and Sensorless Vector Control

Performance Analysis of Doubly-Fed Induction Generator
(DFIG)- Based Wind Turbine with Sensored and

Sensorless Vector Control
Amer Obaid Kareem
B.Sc., M.Sc
Student Number: 089112111
A thesis submitted for the degree of
Doctor of Philosophy
November 2016
School of Electrical and Electronic Engineering
Newcastle University
United Kingdom
ABSTRACT
Conventional energy sources are limited and pollute the environment. Therefore
more attention has been paid to utilizing renewable energy resources. Wind energy is
the fastest growing and most promising renewable energy source due to its
economically viability. Wind turbine generator systems (WTGSs) are being widely
manufactured and their number is rising dramatically day by day. There are different
generator technologies adopted in wind turbine generator systems, but the most
promising type of wind turbine for the future market is investigated in the present
study, namely the doubly-fed induction generator wind turbine (DFIG). This has
distinct advantages, such as cost effectiveness, efficiency, less acoustic noise, and
reliability and in addition this machine can operate either in grid-connected or
standalone mode. This investigation considers the analysis, modeling, control, rotor
position estimation and impact of grid disturbances in DFIG systems in order to
optimally extract power from wind and to accurately predict performance. In this
study, the dynamic performance evaluation of the DFIG system is depicted the power
quantities (active and reactive power) are succeed to track its command signals. This
means that the decouple controllers able to regulating the impact of coupling effect in
the tracking of command signals that verify the robust of the PI rotor active power
even in disturbance condition.
One of the main objectives of this study is to investigate the comparative

estimation analysis of DFIG-based wind turbines with two types of PI vector control
using PWM. The first is indirect sensor vector control and the other type includes two
schemes using model reference adaptive system (MRAS) estimators to validate the
ability to detect rotor position when the generator is connected to the grid. The results
for the DFIG-based on reactive power MRAS (QRMRAS) are compared with those of
the rotor current-based MRAS (RCMRAS) and the former scheme proved to be better
and less sensitive to parameter deviations, its required few mathematical computations
and was more accurate. During the set of tests using MATLAB® /SMULINK ® in
adjusting the error between the reference and adaptive models, the estimated rotor
position can be obtained with the objective of achieving accurate rotor position
information, which is usually measured by rotary encoders or resolvers. The use of
these encoders will conventionally lead to increased cost, size, weight, and wiring
i
complexity and reduced the mechanical robustness and reliability of the overall DFIG
drive systems. However the use of rotor position estimation represents a backup
function in sensor vector control systems when sensor failure occurs.
The behavioral response of the DFIG-based wind turbine system to grid

disturbances is analyzed and simulated with the proposed control strategies and
protection scheme in order to maintain the connection to the network during grid
faults. Moreover, the use of the null active and reactive reference set scheme control
strategy, which modifies the vector control in the rotor side converter (RSC)
contributes to limiting the over-current in the rotor windings and over-voltage in the
DC bus during voltage dips, which can improve the Low Voltage Ride-through
(LVRT) ability of the DFIG-based wind turbine system.
ii
ACKNOWLEDGMENTS
In the Name of God, the most Gracious, the most Merciful, say Glory to Allah,
who has given me the strength, patience, ability, and knowledge to finish my PhD
study.
From the depths of my heart, I would like to express my sincere gratitude to my

supervisor, Dr. Shady Gadoue, for his guidance and support, intellectual advice and
patience thoughtout the study process. I give him heartfelt thanks for his time and
assistance throughout my PhD journey. I wish you the best and further success and
achievement in your life. I also wish to thank my co-supervisor, Dr. Mohamed
Elgendy, for his support and helpful suggestions.
Of course, no acknowledgments would be complete without giving thanks to my

parents, my wife Nadeen, my brothers (Mr. Kareem and Dr. Nagham) and my uncle
(Mr. Bassil) to whom I would like to express my sincere gratitude. They provided me
with love, guidance, prayers, support and wisdom. Thank you very much for being the
motivational element in my life, enhancing me in every moment and giving me the
ability to complete this work successfully. My wife’s insights and words of
encouragement have often inspired me and renewed my hopes for completing my
PhD research.
Special appreciation goes to the School of Electrical and Electronic Engineering,

Newcastle University, for providing me with this opportunity to complete my study
and research.
I would also like to express special thanks to my home country of Iraq and its
Ministry of Planning for providing a scholarship for my study.
iii
TABLE OF CONTENTS
Abstract i
Acknowledgements iii
Table of Contents iv
List of Symbols
ix
Acronyms and Abbreviations

xii
Superscripts
xiii
Units of Measure
xiii
List of Figures xiv
List of Tabels xxi
Chapter 1 Introduction 2
1-1 Overview of Wind Energy Conversion System Technology 2
1-2 Wind Power Generators 5
1-3 Research Methodology and Challenges 13
1-4 Objectives and Contributions of the Study 14
1-5 Thesis Layout 16
Chapter 2 Literature Survey 18
2-1 Introduction 18
Modeling and Control System of the DFIG

2-2 18
2-3 Sensorless Control of the DFIG-Based Wind Turbine 22
2-4 Maximize Power Point Tracking (MPPT) in DFIG-Based Wind 25
iv
Turbines
2-5 Fault Ride-Through (FRT) of the DFIG-Based Wind Turbine 29
2-6 Wind Turbine Topologies 34
2-7 Conclusions 36
Chapter 3 Mathematical Model of the Doubly-Fed Induction Generator System 38
3-1 Introduction 38
3-2 Dynamic Model of Doubly-Fed Induction Generator 38
3-3 Vector Control System 42
3-3-1 Vector Control of the Rotor Side Converter (RSC) 43
3-3-2 Vector Control of the Gride Side Converter (GSC) 51
3-3-3 Control of the DC-Bus Voltage 57
3-4 Drive Train Model 60
3-5 Wind Turbine Power Characteristics 61
3-5-1 Annual Wind Distribution 61
3-5-2 Aerodynamic Characteristics of the Wind Turbine 62
3-5-3 Tip Speed Ratio (TSR) of the Wind Turbine 64
3-6 Test System 65
3-7 Phase Locked Loop (PLL) System Performance 67
3-8 DFIG System Model Confirmation 72
3-9 Simulation Results of Modelling the DFIG system 77
3-10 Conclusions 84
v
Chapter 4 Rotor Position Estimation Using The Model Reference Adaptive System 87
4-1 Introduction 87
4-2 Sensorless Control of DFIG System Using MRAS Observer 88
4-2-1 Rotor Current-Based MRAS (RCMRAS) 89
4-2-2 Rotor Reactive Power-Based MRAS (QRMRAS) 90
4-3 Simulation Results 91
4-4 Conclusions 98
Chapter 5 Behavioral Analysis of DFIG System with Grid Disturbances 100
5-1 Introduction 100
5-2 Grid Codes Requirements of DFIG System 100
5-3 Fault Ride-Through (FRT) of DFIG System 102
5-4 Performance Analysis of the DFIG System with Voltage Dips 106
5-5 DFIG Analysis During Voltage Swell 108
5-6 Reactive Current Constraint of GSC During Grid Voltage Dip and 110
Swell
5-7 Improved FRT Control Scheme for DFIG Wind Turbine 112
5-8 Detection of Disturbances in the DFIG System 114
5-9 Performance of Grid Connected DFIG System Grid Without 117

Protection
5-9-1 Case of 80% Voltage Dips 117
5-9-2 Case of Voltage Swell of 1.3 p.u. 125
5-10 Performance of Sensored and Sensorless Vector Control of DFIG 130
vi
System with PQ Null Modification during Disturbances
5-11 Conclusions 136
Chapter 6 Conclusions and Further Works 138
6-1 Research Summary 138
6-2 Contributions 139
6-3 Further Work 139
References 142
Appendix A Model Parameters 159
A-1 DFIG 159
A-2 Grid Side Filter 159
A-3 DC- Link 160
A-4 Wind Turbine 160
A-5 Simulation Controller Parameters 161
Appendix B Transformation Process of the Currents 162
Appendix C Pulse Generation of the Controlled Switched Modulation 164

Technology
Appendix D Simulink Model 166
D-1 Complete DFIG System 166
D-2 Aerodynamic Model 167
D-3 Pitch Angle Controller Model 168
D-4 Generator Model 169
vii
D-5 PWM Voltage Source Converter Model 170
D-6 GSC and RSC Model 171
D-7 Stator Angle Estimator Model 172
D-8 Rotor Reactive Power MRAS (QRMRAS) Model 172
D-9 Rotor Current MRAS (RCMRAS) Model 173
D-10 Crowbar Model 173
Appendix E Matlab Initialisation Code 174
E-1 Appendix Model Parameter 174
E-2 Plotting Tip Speed Ratio Variation with Power Coefficient and 174
Pitch Blade Angle
E-3 Calculation of the Annual WEIBULL Distribution of the Wind 175
Appendix F PI Controller Design in Z domain 176
viii
LIST OF SYMBOLS
� = �⁄�� First derivative with respect to time.

� Capacitance of the DC link.
� Fundamental frequency.
� Carrier frequency.
� Grid frequency.
� Modulation frequency.
� Rotor frequency.
� Stator frequency.
, Three-phase, abc, currents.
, Two-phase DC rotor currents, rotating dq-axis.
, Two-phase DC stator currents, rotating dq-axis.
, , Three-phase, abc, rotor currents in the stationary reference frame.
, , Three-phase, abc, stator currents in the stationary reference frame.
, Two-phase AC rotor currents, alpha/beta, in the stationary frame.
, Two-phase AC stator currents, alpha/beta, in the stationary frame.
, Two-phase DC rotor currents, rotating dq-axis synchronous frame.
, Two-phase DC stator currents, rotating dq-axis synchronous frame.
, Two-phase DC grid currents, rotating dq-axis synchronous frame.
, Polar currents relating to the imaginary axis of rotor and stator q-axis.
� _ d-axis rotor error current in synchronous reference frame.
� q-axis rotor error current in synchronous reference frame.

�
d-axis rotor reference current in synchronous reference frame.
�
q-axis rotor reference current in synchronous reference frame.
, � Proportional and integral gain of PI controller.
Inductance.
, Rotor and stator leakage inductance.
, Rotor and stator self-inductance.
Magnetizing inductance.
DC link inductance.
Frequency modulation index.
ix
Rotor speed of induction generator, RPM.
Rotor speed of induction generator, RPM.
� Pole pairs of induction machine.
Active Power.
Rotor active power.
Stator active power.
Mechanical active power.
� Active power error.
Reference stator active power.
Reactive power.
Rotor reactive power.
Stator reactive power.
� Reactive power error.
Reference stator reactive power.
� ,� Rotor and stator resistance.
s Induction machine slip.
� ,� Electrical and mechanical torque of the DFIG.
�,� ,� Three-phase, abc, voltages.
� DC link bus voltage.
� Modulated voltage.
�,� Two-phase alpha/beta voltages.
� ,� Two-phase DC rotor voltages, rotating dq-axis.
� ,� Two-phase DC stator voltages, rotating dq-axis.
� ,� , � Three-phase, abc, rotor voltages in the stationary reference frame.
� ,� ,� Three-phase, abc, stator voltages in the stationary reference frame.
� ,� ,� Three-phase, abc, sinusoidal modulation waveform, PWM.
� ,� Two-phase AC rotor voltages, alpha/beta, in the stationary frame.
� ,� Two-phase AC stator voltages, alpha/beta, in the stationary frame.
� ,� Two-phase DC rotor voltages, rotating dq-axis synchronous frame.
� ,� Two-phase DC stator voltages, rotating dq-axis synchronous frame.
� , � Polar voltages relating to the imaginary axis of rotor and stator q axis.
� ,� Rotor and stator flux.
x
� ,� Rotor dq-axis flux.
� ,� Stator dq-axis flux.
� ,� Polar flux relating to the imaginary axis of rotor and stator q-axis flux.
�
= �� Flux differential with respect to time.
�� , �� Rotor flux differential with respect to time.

�� , �� Stator flux differential with respect to time.
� Angular rotor speed (rev/s).
� Angular synchronous peed (rev/s).
� Angular slip speed (rev/s).
� Rotor angular displacement position.

� Stator angular displacement position.
� Angular slip displacement position.
�2
� Leakage factor of induction machine. � = 1 − � ��
∫ dt Integral with respect to time.
xi
ACRONYMS AND ABBREVIATIONS
AC Alternate current
DFIG Doubly-fed induction generator
DC Direct current
Dy Delta star connection
FRT Fault ride-through
GSC Grid-side converter
LVRT Low voltage ride-through
MPPT Maximum power-point tracking
MRAS Model reference adaptive system
IG Induction generator
PF Power factor
PMSG Permanent magnet synchronous generator
PWM Pulse with modulation
RSC Rotor-side converter
SCIG Squirrel cage induction generator
SG Synchronous generator
WECS Wind energy conversion system
THD Total harmonic distortion
WRIG Wound-rotor induction generator
WRSG Wound-rotor synchronous generator
WT Wind turbine
xii
SUPERSCRIPTS
Avg Average
max Maximum
min Minimum
s Stator-oriented reference frame
ref Reference
UNITS OF MEASURE
A Amps
V Volts
MW Megawatts
Ω Ohms
rad/s Radian per second
RPM Revolution per minute
�� Mega volt ampere reactive
xiii
LIST OF FIGURES
Figure 1.1 Atmospheric cells and latitudinal wind belts. 3

Figure 1.2 (a) Annual wind installed global capacity 1996-2013; (b) annual 4
wind installed capacity by region 2005-2013; (c) top ten cumulative
installed capacity in 2013.
Figure 1.3 Installed renewable power generation capacity by type, end-2011. 5
Figure 1.4 Typical components of horizontal axis, three-bladed wind turbine. 6
Figure 1.5 Types of tower used in wind conversion system (a) monopole 7
tower; (b) lattice tower; (c) tube tower
Figure 1.6 Induction generator feeding to a utility grid with excitation capacitor. 7
Figure 1.7 Double-output induction generator system. 9
Figure 1.8 Double-output system with direct current link. 9
Figure 1.9 Power flow in slip power control scheme with DC link voltage. 10
Figure 1.10 Self-excited induction generator feeding a load. 11
Figure 2.1 Types of DFIG: (a) standard doubly-fed induction machine; (b) 19
brushless doubly-fed induction machine; (c) cascaded doubly-fed
induction machine.
Figure 2.2 General structure of a phase locked loop 21
Figure 2.3 Block diagram of stator voltage angle calculation 22
Figure 2.4 Typical model of reference adaptive system (MRAS) block diagram 25
Figure 2.5 Ideal power curve for wind turbines 26
Figure 2.6 Maximum power point tracking (MPPT) of the DFIG system 27
Figure 2.7 FRT requirements for various grid codes 30
Figure 2.8 Algorithm to support the FRT in DFIG wind turbine 31
Figure 2.9 DFIG system equipped with a crowbar 31
Figure 2.10 Configuration of wind turbine topology using: (a) DC generator; (b) 35
switch reluctance generator; (c) PMSG; (d) Squirrel cage induction
generator.
Figure 3.1 System configuration of the DFIG-based wind turbine. 39
Figure 3.2 Ideal three-phase windings (stator and rotor) of the DFIG. 40
Figure 3.3 Reference frames and angles. 44
Figure 3.4 Block diagram of RSC controller. 45
xiv
Figure 3.5 Closed loop current control of RSC in s-domain 47
Figure 3.6 Closed loop GSC current-control in z-domain 47
Figure 3.7 RSC controller tuning PI parameters: (a) root locus, open-loop bode 50
plot; (b) step response.
Figure 3.8 The grid side system. 51
Figure 3.9 Eequivalent electric circuit of GSC system in: (a) dq coordinates; 52
(b) αβ coordinates.
Figure 3.10 Orientation with d-axis of the grid voltage space vector in GSC. 53
Figure 3.11 Current control loop of the GSC in S-domain 55
Figure 3.12 Closed-loop GSC current-control in z-domain. 55
Figure 3.13 GSC controller tuning the PI parameters: (a) root locus, open-loop 56
bode plot; (b) step response.
Figure 3.14 Back-to-back power converter. 57
Figure 3.15 DC link loop control of the GSC in S-domain. . 59
Figure 3.16 Closed-loop DC link voltage control in z-domain. 59
Figure 3.17 DC-link controller tuning the PI parameters: (a) root locus, open 59
loop bode plot; (b) step response.
Figure 3.18 Two-mass model for the drive train. 61
Figure 3.19 Block diagram of wind energy conversion system. 61
Figure 3.20 Annual wind distribution ( c =11.38, k =2). 62
Figure 3.21 Power coefficient curve verses tip speed ratio and pitch blade angle. 63
Figure 3.22 Power characteristics of wind turbine versus rotational speed. 64
Figure 3.23 Screen shot of the DFIG system MATLAB/SIMULINK® model. 66
Figure 3.24 Instantaneous measurements of: (a) grid voltage; (b) PLL response 68
after -40Hz frequency step change.
after +40 Hz frequency step change.
Figure 3.26 Instantaneous measurements of: (a) grid voltage; (b) PLL 69
response during 3rd order harmonic (+ve sequence, 20%
amplitude, and -25 degree phase angle step) parasitic with the 2nd
order harmonic ( +ve sequence, 10% amplitude, and 0 degree
phase angle step).
xv
response during 3rd order harmonic (+ve sequence, 20%
amplitude, and -25 degree phase angle step) parasitic with the 7th
order harmonic ( +ve sequence, 2% amplitude, and 0 degree
phase angle step).
during 80% voltage dip, having harmonics ( 3rd (+ve sequence,
20% amplitude, and -25 degree phase angle step) parasitic with
the 7th ( +ve sequence, 2% amplitude, and 0 degree phase angle
step)).
response during 40% voltage swell, having harmonics of order
3rd (+ve sequence, 20% amplitude, and 25 degree phase angle
step) parasitic with 2nd order harmonic ( +ve sequence, 10%
amplitude, and 0 degree phase angle step).
response single phase fault, having harmonics ( 3rd (+ve
sequence, 20% amplitude, and -25 degree phase angle step)
parasitic with the 2nd order harmonic ( +ve sequence, 10%
amplitude, and 0 degree phase angle step)).
Figure 3.31 The DFIG wind turbine model test system in PLECS®. 73
Figure 3.32 Instantaneous comparison measurements of DC-link voltage 74
regulator in MATLAB® and PLECS®.
Figure 3.33 Instantaneous measurements of: (a) d-component rotor current 75
controller MATLAB® and PLECS®; (b) q-component rotor
current controller in MATLAB® and PLECS®
Figure 3.34 Instantaneous measurements of: (a) generator speed in MATLAB® 76
and PLECS®; (b) reactive power control in MATLAB® and
PLECS®.
Figure 3.35 Inside the block of RSC and GSC control of PLECS® DFIG model. 77
Figure 3.36 Instantaneous measurements of the rotor current in synchronous 78
reference frame .
Figure 3.37 Instantaneous measurements of the three-phase rotor current 79
Figure 3.38 Instantaneous measurements of the stator current in synchronous 79
xvi
reference frame .
Figure 3.39 Instantaneous measurements of the three-phase stator current. 80
Figure 3.40 Instantaneous measurements of the grid voltage: (a) three-phase; (b) 80
in synchronous reference frame coordination.
Figure 3.41 Instantaneous measurements of three-phase GSC current (A). 81
Figure 3.42 Instantaneous measurements of quadrature voltage component of 81
the RSC.
Figure 3.43 Instantaneous measurements of direct voltage component of the 81
RSC
Figure 3.44 Instantaneous measurements comparison of reference and measured 82
stator active power.
Figure 3.45 Instantaneous comparison reference and measured stator reactive 82
power.
Figure 3.46 Instantaneous measurements phase-to-phase voltage of rotor 83
voltage.
Figure 3.47 Instantaneous measurements phase-to-phase GSC output voltage. 83
Figure 3.48 Instantaneous measurements DC-link voltage regulator in the GSC. 83
Figure 3.49 Instantaneous comparison reference and measured current regulator 84
in the inner loop GSC control with reference frame oriented along
the grid voltage vector (a) (b) .
Figure 4.1 Simulink model of rotor current-based MRAS (RCMRAS). 90
Figure 4.2 Simulink model of rotor reactive power-based MRAS (QRMRAS). 91
Figure 4.3 Comparison results of the measured and estimated rotor 92
position computed in normal operating conditions.
Figure 4.4 Instantaneous comparison reference and measured value of rotor 92
reactive power in the QRMRAS sub-system computation.
Figure 4.5 Instantaneous error between reference and adaptive model output in 93
rotor RCMRAS and QRMRAS.
Figure 4.6 Variable wind speed signal. 93
Figure 4.7 Instantaneous results of generator speed (rad/sec). 94
Figure 4.8 Comparison results of the measured and estimated rotor position in 94
step change wind speed operating conditions.
xvii
condition of 30% increase of. .
Figure 4.10 Instantaneous measurement of three-phase voltage. 95
Figure 4.11 Instantaneous measurement of three-phase current during voltage dip 96
condition without protection.
Figure 4.12 Instantaneous comparison between outputs of reference model and 96
adaptive model in RCMRAS sub-system computation in fault
condition.
Figure 4.13 Instantaneous comparison reference and measured value of rotor 96
reactive power in the QRMRAS sub-system computation in fault
condition.
Figure 4.14 Instantaneous error between reference and adaptive model output in 96
RCMRAS and QRMRAS in fault condition.
voltage dip condition.
Figure 5.1 Typical voltage dip limit for LVRT. 104
Figure 5.2 Spatial relationship of GSC steady-state voltage vectors. 110
Figure 5.3 Diagram of null the active and reactive power method. 114
Figure 5.4 Inverse time relay characteristics. 116
Figure 5.5 Relay characteristic and over-current. 116
Figure 5.6 MATLAB/SIMULINK® subsystem of: (a) over-current trip model; 117
(b) over/under-voltage trip model [MATLAB/ 1.5 MW DFIG Demo
model].
Figure 5.7 Instantaneous rms grid voltage. 119
Figure 5.8 Instantaneous three-phase stator voltage. 119
Figure 5.9 Instantaneous three-phase stator current. 119
Figure 5.10 Instantaneous dq stator current. 120
Figure 5.11 Instantaneous three-phase rotor current. 120
Figure 5.12 Instantaneous dq rotor current. 120
Figure 5.13 Instantaneous measurements of rms grid current. 121
Figure 5.14 DC-link voltage regulator in the GSC. 121
Figure 5.15 Instantaneous regulation of the stator active power in the RSC. 122
Figure 5.16 Instantaneous regulation of the stator reactive power in the RSC. 123
Figure 5.17 Instantaneous measurement of generator speed. 123
xviii
Figure 5.18 Instantaneous measurement of three-phase current of the GSC. 123
Figure 5.19 Electromagnetic torque and aerodynamic torque during voltage dip. 125
Figure 5.20 Generator speed and rotor during voltage dip. 125
Figure 5.21 Instantaneous rms stator voltage. 127
Figure 5.22 Instantaneous measurement of the three-phase stator voltage. 127
Figure 5.23 Instantaneous three-phase stator current. 127
Figure 5.24 Instantaneous dq stator current. 128
Figure 5.25 Instantaneous three phase rotor current. 128
Figure 5.26 Instantaneous dq rotor current. 128
Figure 5.27 DC-link voltage regulator. 129
Figure 5.30 Instantaneous rms grid current. 130
Figure 5.31 Instantaneous rms grid voltage. 132
Figure 5.32 Instantaneous three-phase stator voltage. 132
Figure 5.33 Instantaneous measurements of phase c stator current. 133
Figure 5.34 Instantaneous value of the d components stator current. 133
Figure 5.35 Instantaneous value of the q components stator current. 133
Figure 5.36 Instantaneous measurements of phase c rotor current. 134
Figure 5.37 Instantaneous d-component rotor current. 134
Figure 5.38 Instantaneous q-component rotor current. 134
Figure 5.41 DC-link voltage regulator. 135
Figure 5.42 Instantaneous three-phase grid current. 136
Figure A.1 Simulink model of grid side filter. 160
Figure A.2 Simulink model of the DC-link 160
Figure B.1 Current transformation process 162
Figure C.1 Output voltages of two level converters with sinusoidal SPWM 165
Figure D.1 Block diagram of the DFIG wind turbine model. 162
Figure D.2 Diagram of the aerodynamic block in the DFIG wind turbine model. 167
Figure D.3 Diagram of the pitch angle control model in the DFIG wind turbine. 168
Figure D.4 Diagram of the mathematical model of DFIG. 169
xix
Figure D.5 Block diagram of the PWM. 170
FigureD.6 Diagram of inner and outer loop control blocks in RSC and GSC 171
Figure D.7 Diagram of PLL used to compute the stator voltage angle. 172
Figure D.8 Diagram of rotor reactive power-based MRAS (QRMRAS). 172
Figure D.9 Diagram of rotor current MRAS (RCMRAS) subsystem. 173
Figure D.10 Diagram of crowbar subsystem. 173
Figure F.1 PI control in z-domain using backward transform. 176
xx
LIST OF TABELS
Table 1.1 Comparison between different wind turbine systems. 12
Table 2.1 Brief description of different order models. 19
Table 2.2 Transitory time response of the vector control, DTC, and DPC 21
Table 2.3 Basic comparison of different MRAS schemes. 25
Table 2.4 Overall comparison of WT generator topologies 35
Table 3.1 Possibilities of vector control in RSC controller. 43
Table 3.2 The Electrical and Mechanical parameters of the doubly-fed 73
induction machine used in the PLECS platform
Table 4.1 Response of performance index of rotor position estimation (at t = 97
1.5 seconds) with ramp variation applied to � � � , � .
Table 5.1 Summary of the conventional modes of DFIG-based WT operating 105
in normal and fault conditions.
Table A.1 PI controller parameters. 161
xxi
Chapter 1: Introduction
CHAPTER ONE
1
CHAPTER ONE
INTRODUCTION
1-1 Overview of Wind Energy Conversion System Technology

Renewable energy conversion systems have become increasingly popular over
recent years; and there has been a strong penetration of renewable energy into power
generation systems due to the prioritizing of secure energy supplies, environmental
concerns and the recognition of climate change, peak oil and the finite supplies of other
fuel. All these factors have contributed to strengthening demand for wind power and
other renewable energy solutions. Wind power is one of the most important renewable
energy sources and it has contributed in historical developments since the ancient times
of Babylon and Egypt where the wind was used to irrigate crops (1700 B.C). Some other
civilizations, like the Persians (500–900 A.D.), used wind power to grind grain, while
others have used the wind to propel ships and other early industrial applications [1]. The
rapid growth of the use of wind energy has not only been stimulated by financial
enhancement mechanisms of different types, but also by the maturing of the technology.
Wind is free energy and does not pollute unlike the traditional fossil energy sources. The
emissions reduction involved depends on the technology that wind replaces. For
example, coal creates about 1 kg of CO2 for each kWh, oil creates 0.75 kg for each kWh
and gas creates 0.5 kg for each kWh [2]. Wind energy depends on weather patterns
caused by the sun’s uneven heating of the earth’s atmosphere. Variations in atmospheric
pressure produced by differential heating propel air from high-pressure to low-pressure
regions, generating winds that are mainly affected by the earth’s rotation and surface
geography as illustrated in figure 1.1 showing the major wind belts encircling the planet
[3].
The kinetic energy of the wind represents a clean source of electrical power when
the force of the wind is employed to drive a turbine which rotates the generator to
produce electricity.
2
Figure 1.1: Atmospheric cells and latitudinal wind belts [3].
The cost of electrical power produced by a wind power conversion system is

decreasing dramatically due to developments in the manufacturing technology of power
electronics the increases in the average level of power produced. Figure 1.2 shows the
growth in wind power systems installed around the world. In the sector statistics of
cumulative wind capacity categorized by country shown in figure 1.2. (c), China and the
USA hold the leading positions in installed wind capacity of (152) GW which has
broken all previous records. As illustrated in figure 1.3, the total installed wind power
capacity at the end of 2011 had grown to around 240 GW, making it the second largest
contributor to renewable generation capacity after hydropower. On the other hand, even
through wind energy represents environment-friendly source of electrical power, it has
limitations due to its discontinuous nature as compared to other renewable power
sources.
The doubly-fed induction generator (DFIG) represents one of the most attractive
technologies to be adopted in wind farms and it has contributed to the wind turbine
market. A DFIG-based wind turbine may suffer from various situations in both normal
condition, and during grid disturbances as with other conventional generators that reduce
system efficiency and/ or reliability. This thesis presents a comprehensive performance
evaluation of a DFIG-based wind turbine model with different scenarios of operation
after the confirmation of a DFIG model system built mathematically using the
3
MATLAB/ SIMULINK® platform. The mathematical models for DFIG sub-systems are
then derived. These mechanical, electrical, and control sub-systems are employed to
determine the system performance. The generator control sub-systems with different
reference frame orientations are presented to design the cascade control loops of the
rotor-side converter (RSC) and grid-side converter (GSC). Two schemes are employed
in this research to estimate the generator’s rotor angle position are investigated the
accuracy and sensitivity of these estimators in different operation conditions. A control
modification for the RSC power loop is introduced to enhance the fault ride-through
(FRT) ability of the DFIG during short voltage dip conditions and illustrated it
capability is demonstrated to reduce stator/rotor over-current oscillation, thus satisfying
the desired system’s grid code specifications.
(a)
(c)
(b)
Figure1.2: (a) Annual wind installed global capacity 1996-2013; (b) annual wind installed
capacity by region 2005-2013; (c) top countries cumulative installed capacity in 2013 [4].
4
Figure 1.3: Installed renewable power generation capacity by type, end 2011[5].
1-2 Wind Power Generators
As shown in figure 1.4, the typical wind energy conversion system (WECS) consists
of a wind turbine (WT) and nacelle [2]. The electrical generator, bearing, gearbox, rotor
shaft, yaw, break system and control system are included inside the nacelle, while the
anemometer is usually located upon the cover of the nacelle to measure the wind speed
[6, 7]. Today, the most common criteria adopted to design wind turbines focus on the
available options for blade number, the orientation of rotor rotation (downwind or
upwind of the tower), fixed or variable rotor speed, gearbox or a direct drive generator,
synchronous or induction generator, hub design (rigid, teetering or hinged), power
control via aerodynamic stall control or variable pitch blade control, turbine tower type
(Lattice tower, monopole tower, or tube tower: as shown in figure 1.5), axis of rotation
as horizontal or vertical, orientation by self-align action (free yaw), or direct control
(active yaw) [6, 8]. The rotor is used to collect the energy from the wind and it is
composed of blades which are attached to a hub and is regulated by pitch control system
to achieve either the maximum use of the wind or to brake the rotation of the rotor. The
rotor shaft transfers the motion from the low speed side via the gearbox which raises the
speed of the rotational shaft connected to the generator, and then it converts the
mechanical, rotational movement into electricity. Then the induced current, power
5
converters are used to create electrical power to match the criteria of the grid in terms of
voltage, current, frequency, power factor, active and reactive power of system [9-11].
Wind turbines are designed to operate either at fixed speed (equipped with induction
generators directly connected to the grid) or variable speed with induction or
synchronous generators, directly or indirectly connected to the grid. As the name
suggests, fixed speed wind turbines rotate at almost a constant speed, which is
determined by the gear ratio, the grid frequency, and the number of poles of the
generator in addition to its maximum conversion efficiency which can be achieved only
at a given wind speed. On the other hand, variable speed wind turbines can achieve
maximum energy conversion over a wide range of wind speeds [6, 12]. The converter in a
variable speed wind turbine gives the possibility to decouple control over the power output of
the wind turbine, which is important for the integration of wind turbines into the grid [11].
Figure 1.4: Typical components of horizontal axis, three-bladed wind turbine [13].
6
(a) (b) (c)
Figure 1.5: Types of tower used in wind conversion system: (a) monopole tower; (b)
lattice tower; (c) tube tower [7].
According to their method of excitation, induction generators are classified into two
basic categories which are constant voltage constant freuncy and variable voltage
variable frequency generators. Constant-voltage, constant-frequency generators is
illustrated in figure 1.6. This generator uses the grid to excite its drive. This type of
system called the grid connected induction generators (GCIG). The power generated is
fed to the supply when the rotor is driven above the synchronous speed. In the cage rotor
type, the power will flow through the stator at low negative slip, and when the rotor is
the wound type the machine can feed power through the stator and rotor windings across
a wide range of operating speeds [14].
Pig
Wind
IG Turbine
Grid
Qgrid Qig
Qc
Capacitor
bank
Figure (1.6): Induction generator feeding to a utility grid

with excitation capacitor [14].
7
Grid-connected induction generators can be divided into two types: single-output

and double-output systems. The single output system generally utilizes the squirrel cage
induction generator (SCIG), and the power feeds to the grid only through the stator
winding which is run in either fixed-speed or semi-variable-speed. In the semi-variable
speed system, a winding type induction generator is used in which the rotor resistance
can change by using power electronics [15- 17]. With changes in rotor resistance, the
torque/speed characteristics of the generator are changed, giving a decrease in rotor
speed of about 10% from the nominal rotor speed. In this generating system, a limited
variable speed capability is achieved at relatively low cost [18]. The optimal value of the
power coefficient must be achieved in harvesting power by choosing the ratio to perform
this condition for the most frequent wind speed, while the wind turbines must limit the
power by using either blade pitch regulation or stall regulation [19]. The generator
always draws reactive power from the network and capacitors are conventionally used to
compensate this lagging VAR.
The basic configuration of the Double-Output System is illustrated in figure 1.7. In

this type, the power can be tapped both from the stator and from the rotor windings over
a wide range of speed by suitable control of the exchanging rotor power. This can be
achieved by the use of a slip-ring induction motor with a bidirectional AC-DC-AC
converter connected between the slip-ring terminals and the grid [20]. Double-output
systems can be categorized into two types according to the technology of the
bidirectional converter: current converters and voltage converters [14, 21-23]. The solid-
state system is used in the double-output system with current converter to control the
slip power at variable speed through current converters as shown in Figure 1.8. The
main function of using the smoothing reactor is to regulate the current’s continuity and
to reduce ripples in the link circuit. Rotor and supply-side converters are operated to
exchange electrical power from the rotor circuit to the supply in the rectification and
inversion modes respectively. However, for power flows in the reverse direction, the
supply side converter acts as a rectifier and the rotor side converter as an inverter. The
step-down transformer between converter and supply extends the control range of the
firing delay angle of the supply side converter [24].
Double-output system with voltage converter is utilize the dual PWM voltage-fed,
current-regulated converters, connected back to back, in the rotor circuit of wound rotor
8
induction generator gives the ability to overcome the problems of naturally commutated,
line-commutated converters and low-frequency forced-commutated converters. As
depicted in figure 1.9 the PWM converters with DC link offer the ability to decoupled
control of the generator’s active and reactive power, with low stator, rotor, and supply
current distortion in, owing to the shift of the harmonic spectra from lower to higher
order, requiring a small-sized filter for attenuation of higher harmonics. This system will
improve the overall system power factor through the control of the displacement factor
between the voltage and current of the grid side converter (GSC) [7].
Utility
system
Bidirectional
Power flow
Converter
system
Wound rotor
To prime Induction
mover motor
Figure 1.7: Double-output induction generator system [20].
grid
Step-down
transformer
Smoothing
reactor
Id
Vd2
Vd1
Slip-ring DC link
Induction rotor side Supply side
generator converted converter
Figure 1.8: Double-output system with direct current link [24].
9
P Q
ref ref
Ps P
ref
s Q
ref
V dc Q g
s
Qs Pr Pl
i2 i1
Vdc
Qr Ql
Rotor side Grid side
converter I converter II
Figure 1.9: Power flow in slip power control scheme with DC-link voltage.
The squirrel cage machine is usually used in variable-voltage, variable-frequency

generators with the capacitor connected across the terminal as shown in figure 1.10
to enhance the voltage built up. This is because when the load current increases, the
current pass through the capacitance also increases, and more magnetizing reactive
power is furnished to the machine. Hence the voltage drop with load will be less
severe and achieved safety voltage regulation. Such induction generators are called
self-excited induction generators (SEIG). The possibility to use the SEIG where the
three-phase capacitor bank are connected to feed the reactive power requirement of
the load and generator was discovered by Basset and Potter in the 1930s [25]. If an
appropriate capacitor bank is connected across the terminals of an externally driven
induction machine, and if the rotor has sufficient residual magnetism, an
electromotive force (EMF) is induced in the machine windings due to the excitation
provided by the capacitor and if the EMF if sufficient to circulate leading currents in
the capacitors [26-27], the flux produced due to these currents would boost the
residual magnetism. This will increase the machine flux and a larger EMF will be
induced. This in turn increases the currents and the flux. The induced voltage and
current will continue to rise until the reactive power supplied by the capacitor is
balanced by the reactive power demanded by the machine and this condition is
essentially decided by the saturation of the magnetic circuit. The induced voltage
keeps on rising until saturation is reached [28].
10
Pig
Load
IG Wind
Turbine
Qig
Qc
Capacitor
bank
Figure 1.10: Self-excited induction generator feeding a load.
Table 1.1 summaries the technical comparison of different wind turbine system
configuration in terms of fixed and variable speed conversion systems and presents their
benefits and drawbacks with the comments on grid control, cost, maintenance, internal
turbine systems and some other important issues for wind turbines. According to the
table and based on industrial reports, the wind power industry is seriously considering
the use of the fully-rated power converter-based generator due to the fault ride-through
(FRT) limitations of the DFIG.
11
Table 1.1: Comparison between different wind turbine systems. ++ Very positive (low
cost); + positive; 0, not competitive in cost [11].
Wind Turbine Type
Fixed Speed Variable Speed
WRIG +
Generator type SCIG DFIG SCIG SG PMSG
variable rotor
resistance
Power converters No Partial Partial Full Full Full
Converter
N/A Small Reduced Full Full Full
capacity
Full, Full, Full,

Speed range ±30% of 100% 100% 100%
<1% of rated <10% of rated
rated of rated of rated of rated
Soft starter Yes Yes No No No No
Flicker
Yes Yes No No No No
(sensitive)
Gearbox Yes Yes Yes Yes Optional Optional

I – Pitch
Aerodynamic II – Stall
power control III – Active Pitch Pitch Pitch Pitch Pitch
Stall
Grid side
reactive power Yes Yes No No No No
compensator
Active power
control and N/A Limited Yes Yes Yes Yes
MPPT
Short circuit
No No No/Yes Yes Yes Yes
(fault active)
Short circuit
Contribute Contribute Contribute Limit Limit Limit
power
Standby function No No Yes + Yes ++ Yes ++ Yes ++
Investment cost ++ ++ + 0 0 0
Maintenance ++ ++ o + + +
Efficiency rating Low Low / reduced Good Good Good Good
12
1-3 Research Methodology and Challenges
This study presents a research methodology to investigate the dynamic comparative

behaviour of the grid connected doubly-fed induction generator (DFIG) with different
scenarios of operation. At the beginning, the study focuses on modelling the mechanical,
electrical and control systems and examines the system results with another software demo
station model to double checked the system built in MATLAB/SIMULINK®. Then after
this step, a comparative estimation analysis of the DFIG-based wind turbine is conducted
with two types of PI vector control using a pulse width modulation (PWM) modulator, the
first is indirect sensore vector control and the other type uses two schemes of estimators
which are utilized to validate the ability to detect the rotor position angle when the generator
is connected the to the grid. The behavioral response of the DFIG system to grid
disturbances is analyzed and simulated using the control strategies and the protection
scheme in order to maintain the connection to the network during grid faults. Moreover the
null active and reactive reference set scheme control strategy modifies the vector control in
the rotor-side converter contributing to limiting the over-current in the rotor windings
during short voltage dips, which can improve the low voltage ride-through ability of the
DFIG system. Furthermore, a reconfiguration scheme of control strategies for the system is
proposed to meet the latest grid code requirements. With this specific methodology, the
system can stay connected to the grid in the case of grid faults. In addition, the fault ride-
through capability of the DFIG system can be greatly supported by providing ancillary
reactive power under voltage dips of different durations, which can increase the voltage at
PCC during the fault and thus provide support to the grid. The system can resume normal
operation immediately after the clearance of the fault. Thus the system can meet the
requirements of the grid code.
SIMULINK has become a fundamental element of most studies and applications in this
field. Utilizing simulink can save money and time. Simulink also offer flexibility, and
enable repeatability, control and the pushing of boundaries. This method proves to be very
efficient for testing and running the proposed model, and for testing and observing its
dynamic behaviour in different scenarios. The software used in this study is
MATLAB/SIMULINK®. These tools are used for modelling, simulating and examining the
proposed DFIG wind turbine system with its controller, sensorless schemes, grid system,
aerodynamic system and protection models. This includes MATLAB/SIMULINK®
13
software with all its modules and library models, and in addition the PLECS®/DFIG demo
model platform is adopted for the confirmation of the system.
As the popularity of the DFIG wind turbine continues to increase, the effect of changes
in wind speed, grid voltage disturbances, and variation in generator parameters represent
severe challenges in assessing the behaviour of the model reference adaptive system
(MRAS) observer. In addition the, grid codes have been revised challenges to demand from
the DFIG based wind turbine to operate as conventional power plants, with focus on fault
ride-through behavior. In this study, a control strategy is designed to extract power from
wind as well as to satisfy reactive power constraints and to decouple control of active and
reactive power. However, in the presence of grid faults, wind turbines should be controlled
so that they can remain connected to the grid and supply adequate reactive power to help the
grid voltage recover. In addition, the estimation of rotor position should remain accurate and
be less sensitive during DFIG operation. Therefore, reconfiguration of the control strategies
for the wind turbine system must be performed in accordance to the operation condition.
1-4 Objectives and Contributions of the Study

In summary, the main objectives of this research are as follows:-
- To model in detail the doubly-fed induction generator (DFIG)-based wind turbine using
MATLAB/SIMULINK® software. The model should be as real as possible, and simulating
the power converter in detail.
- To investigate the dynamic behaviour of the grid connected DFIG-based wind turbine in
addition confirm the system built in MATLAB/SIMULINK® using PLECS® platform.
- To demonstrate the ability of the phase lock loop (PLL) used in the DFIG-based wind
turbine system to track changes in grid frequency in different operating conditions
(frequency responses, input harmonics responses, and grid voltage disturbances).
- To estimate rotor angle position with two model reference adaptive system (MRAS)
schemes and to compare the robustness of these schemes in different scenarios. In this
study, rotor current-based MRAS (RCMRAS) and rotor reactive power-based MRAS
(QRMRAS) are used in the estimating of the rotor position angle because the currents and
voltage are available as measured quantities in coordination, and therefore the
implementation of RCMRAS and QRMRAS observers is simple.
14
- To provide a better understanding of the dynamic comparative performance of the DFIG-

based wind turbine with sensored and sensorless vector control during different operating
conditions in order to contribute to the development of a better controlled, viable DFIG
system.
- To analyze the voltage disturbance response of the DFIG-based wind turbine with and
without the modified vector control scheme to show its influence on the dynamic behaviour
of the wind turbine system.
- To utilize the active and reactive power null technique which enhance the fault ride-
through (FRT) ability of the DFIG with voltage dips. In addition this strategy will keep the
back-to-back converter connected to the DFIG even the grid suffer from the short time
voltage drop also it can be equipped with the conventional crowbar against severe grid
faults to order to reduce the activated time of the crowbar as much as possible.
The main contributions of this thesis are summarized as following:-
- A complete dynamic comprehensive performance assessment is conducted for grid-

connected variable-speed wind turbines based on a DFIG which is built mathematically
in the MATLAB/SIMULINK® platform and the system is verified with PLECS®.
Although the wind turbine model and its control system are designed based on
conventional machine theories, the modelling has prove to be able to support the study
the sensorless vector control and FRT issues of such grid-connected wind turbines and
their interactions with the grid.
- Rotor angle estimation using rotor reactive-based MRAS (QRMRAS) is identified to

be superior in terms of accuracy and robustness, as compared with the rotor current-
based MRAS (RCMRAS) and it is found to be promising for rotor position estimation
for DFIG-based wind turbines.
- This study improves the methodology for utilizing the available electrical
measurements to detect rotor angle position and to enhance the fault ride-through ability
of grid connected DFIG-based wind turbines. Therefore, such a system will be more
reliable in remote areas and stable against disturbances when the control modification is
adopted. This structure more satisfies all of the standards of connection to electrical
networks imposed by the most demanding operators.
15
1-5 Thesis Layout

The objective of the thesis is to investigate the dynamic behaviour integrated control,
rotor angle estimation and protection system of the doubly-fed induction generator-
based wind turbine to fulfill the revised grid requirements. The following parts of the
dissertation are organized as follows:
In chapter 2, the literature review describes doubly-fed induction generator (DFIG)-based
wind turbine systems and more specifically addresses the modeling, control, sensorless
control, and interaction of DFIG systems with grid disturbances.
In chapter 3, a whole model is built of the grid connected doubly-fed induction generator
(DFIG)-based wind turbine and the back-to-back PWM voltage source converters in the
rotor circuit. The model includes the wind speed model, the aerodynamic model of the
wind turbine, the mechanical model of the transmission system and models of the
electrical components, namely the DFIG.
Chapter 4 includes complete analyses of two schemes based on the model reference
adaptive system (MRAS) which have been employed in this study to estimate the rotor
position of doubly-fed induction generator (DFIG) system and utilized in which are the
control of the stator flux orientation of the rotor side converter (RSC).
Chapter 5 analyzes in detail the dynamic responses of the control schemes and the DFIG-
based wind turbine during a voltage dip across the point common coupling of the grid. In
this disturbance condition, a control strategy is employed without the use of the crowbar to
assist in the recover of the DFIG-based wind turbine’s terminal voltage, which is verified by
the simulation results.
Finally, chapter 6 summarizes the most important results obtained in this work and presents
the conclusions of the study and makes recommendations for future work.
16
Chapter 2: Literature Survey
CHAPTER TWO
17
CHAPTER TWO
LITERATURE SURVEY
2-1 Introduction
Recently, due to the continuous and rapid increase in the cumulative installed power
of grid-connected wind turbines, the control strategies used methods for estimation of
the rotor position, and system response with its FRT ability during issues of grid
connected wind turbines disturbances have attracted considerable interest. This has
led to the publication of numerous papers, investigations and reports dealing with
these aspects.
In this chapter, a literature review is presented describing doubly-fed induction
generator (DFIG)-based wind turbine systems.
2-2 Modeling and Control System of the DFIG
Only the standard doubly-fed induction machine and brushless type of systems have
been employed in the manufacture of wind turbine generator systems, although
another type of DFIG system is the cascaded doubly-fed induction machine as shown
in figure 2.1[29-32]. Conventionally, the synchronous reference frame orientation is
adopted in DFIG modelling to simplify the design of the control system, since the
currents and voltages will be expressed as DC vectors. On the other hand some studies
have utilized the stationary stator reference frame and rotor reference frame [7, 33-
37].
Some papers have presented models of the DFIG system of different order so as to
reduced the required for computational time for the evaluation of system behaviour
during different operating conditions. The mean of reduce order model is summarized
by neglect the differential terms and some parameters in the voltage equation of the
machine as illustrated in table 2.1 [38]. Many studies have performed comparisons
between the full order model and reduced order models with various scenarios [39-
43]. A comparative study has been introduced of three different order models of DFIG
(1st, 3rd and 5th) using mathematical modeling with considering and ignoring
saturation effects [39]. The third order model of DFIG is characterized by neglecting
the derivative terms of the stator flux and is widely used in power system transient
18
stability analysis, while the first order model is presented by neglecting both the
derivative terms of the stator flux and rotor flux and where the only differential
equation left is the swing equation which is suitable for long-term studies of power
system dynamics [44].
Table 2.1: Brief description of different order models [38].
5th order 3rd order 1st order

� � �
≠0 =0 =0
� � �
≠0 =0 =0
� � �
Assumption ≠0 ≠0 =0
� � �
≠0 ≠0 =0
��
≠0 ≠0 ≠0
(a) (b)
(c)
Figure 2.1: Types of DFIG: (a) standard doubly-fed induction machine; (b) brushless
doubly-fed induction machine; (c) cascaded doubly-fed induction machine [29-32].
Based on the dynamic mathematical model of the DFIG, many publications have
presented different control strategies, such as vector control, direct power control,
direct torque control, predictive control, sensorless control and non-linear control.
Conventional designs of DFIG control systems are based on rotor current vector
19
control techniques. The control system is usually defined in the synchronous dq frame
oriented to the stator flux [22, 45-47], the stator voltage (or grid flux) [48-51] or the
gap flux [52]. On the other hand a detailed DFIG-based wind turbine model in
MATLAB/SIMULINK® using field-oriented control-based on a rotor flux-oriented
reference frame has been proposed and compared with the conventional stator flux
vector control in [53]. The stator-flux oriented control is basically adopted in DFIG
control designs in which the q-axis current component is used for active power control
and the d-axis component is used for reactive power control. On the other hand, in
stator voltage-oriented control, the opposite is the case and the d-axis component is
used for active power control and the q-axis current component is used for reactive
power control. The steady state and dynamic analysis associated with controllers are
based on the small signal stability criteria in the detailed modeling of the DFIG-based
wind turbine to evaluate system stability. The system has a very extensive stability
region and intrinsic stability of electromagnetic torque has investigated [54, 55]. The
dynamic variations in DFIG flux, wind turbine, power converter and phase-locked
loop (PLL) have been studied [56] and a non-linear mathematical model of DFIG
system connected to a series-compensated transmission line was developed. Direct
power control (DPC) and closed loop stator flux orientation have been proposed by
Yongchang and Kostyantyn [57, 58] respectively to control the brushless doubly-fed
induction generator. Basic direct torque control has also been applied to a doubly-fed
induction generator [59-61]. Direct torque control (DTC) was achieved without a
current regulator and without coordinate transformations. Knowledge of grid voltages,
rotor currents, and rotor position is needed in this type of controller [62]. Rotor flux
linkage magnitude and generator torque can be controlled by specifically selecting the
inverter switching states [63]. Starting from the basic rules of DTC, direct power
control (DPC) has been introduced [64] and applied to DFIG systems [65-68]. Etinne
and Ambrish [69] have introduced a comparative performance assessment using
vector control, DTC and DPC strategies to control DFIG-based wind turbine system.
The authors concluded that vector control (VC) has lower THD while the direct
schemes are faster than the vector control in their transitory response. Because this is
in the direct control strategy shown in table 2.2, the only vectors required are
electromagnetic torque stator active power and stator reactive power which are used
during the control computations without pulse width modulation (PWM) and in
20
cascaded loop control (i.e. inner and outer loops as in vector control). Thus the
converter switching states in the direct schemes are selected by the switching table
based on the instantaneous errors between the commanded and estimated values of
stator flux and active and reactive power, while the rotor control voltage in vector
control scheme is generated after a cascade of PI controllers [69].
Table 2.2: Transitory time response of vector control, DTC, and DPC [69].
Transitory time response (ms)

Reference value
VC DTC DPC
-ve step: � /� / 10 2.5 2.5
+ve step: � /� / 10 1.5 1.5
-ve step: � /� / 10 2.5 2.5
+ve step: � /� / 5 2.5 2.5
Since the phase locked loop (PLL) is a central component in a control structure for
converters in the DFIG, it is discussed here in the context of grid component models.
The PLL is used to estimate the grid angle and therefore to generate the control
signals for the grid-side converter. A generalized PLL structure is shown in figure 2.2.
Figure 2.2: General structure of a phase locked loop [70].
The phase-locked loop (PLL) as implemented on study [7] is typically employed to

measure the stator voltage angle as illustrated in figure 2.3 where this signal is
important in vector control. The PLL consists of a voltage-controlled oscillator (VCO)
which is initially tuned to a frequency close to the desired frequency. A circuit called
a phase comparator causes the VCO to seek and lock onto the desired frequency.
21
sin
Vsα ×
- Theta
PI K 1/s
+
Vsβ
×
cos
Figure 2.3: Block diagram of stator voltage angle calculation [7].
The three-phase voltage inputs are transformed into αβ components referring to the
grid voltage phasor. Using these components, sufficient information about the phase
angle is obtained. A phase detector range of (-π/2; π/2) is obtained using a simple
inverse tangent function, while if a four quadrant inverse tangent function is used in
this range can be expanded to (-π, π). The loop filter contains a low-pass filter to
suppress noise and high frequency terms in the signal from the phase detector; for
example, asymmetry in the three-phase voltages. To avoid stationary error in phase
after a step in the input signal frequency, a PI controller is added. The performance of
the PLL depends on the phased detector and the selected bandwidth of the loop filter.
With slow dynamic of the PLL loop, higher rejection of the disturbances can be
achieved, although with a poor tracking of the grid angle. In power system
applications, the voltage-controlled oscillator is usually implemented as an integrator.
The loop is closed by feeding the estimated angle into the αβ-transformation. As long
as the phase angle is correct, the output of the phase detector is zero, and thus the
frequency input to the integrator is constant and this means that the PLL is in lock. If
the estimated angle is not correct, the frequency is adjusted and the phase angle of the
grid phasor voltage is changed.
2-3 Sensorless Control of the DFIG-Based Wind Turbine

Several sensorless control strategies have been proposed in DFIG-based wind turbine
systems with the main challenges being designing rotor position scheme estimation at
stable and well detection and in any speed of operating range. The vector control
strategy in the back-to-back power converter of DFIG-based wind turbine generators
essentially requires rotor position information which is conventionally provided by
22
mechanical sensor coupled to the rotor shaft [71]. The position of the rotor is
employed to transfer rotor variables from their natural reference frame coordination to
the stator flux-oriented reference frame or the stator voltage-oriented reference frame
and vice versa. In addition, the rotor’s angular speed could be detected using the
position sensor signal by calculating the position difference over a small sampling
period. Recently, researchers have attempted to remove the position sensor from the
DFIG system in an effort to improve the reliability of the system without affecting the
dynamic behaviour of the controller [72]. The elimination of the mechanical sensor
will benefit the DFIG system in several ways: reduced hardware complexity, reduced
size of the drive machine, elimination of the sensor cable, better noise immunity,
increased reliability, and less maintenance requirements. Operation in hostile
environments mostly requires motors without position sensors [73].
Open-loop estimator approaches have been implemented in [74-79], where the rotor
position is directly computed from measured voltages and currents by reference frame
transformation, and the rotor speed is obtained via differentiation. A position
sensorless scheme based only on rotor variables as has been introduced [80], but this
gives poor position detection around synchronous speed because during the rotor slip
frequency, the accuracy of detection is very low during this condition of operation
which effected by the integration of the rotor back electromotive force (back-EMF).
Open loop methods have been used [75-78] which compare the estimated and
measured currents so as to compute the rotor position using open-loop algebraic
calculation. However, these methods have no error correction mechanism, low
accuracy, instable to modify speed convergences in addition its sensitive to
parameters variations. The rotor angular speed is calculated via differentiation the
position signal which may produce noisy speed estimation [76]. Alternatively it has
been based on the comparison of stator current values calculated in different ways to
estimate rotor position with good speed transient characteristics [78] but the results
are influenced by the operational state of the machine. The phase comparison of an
estimated air-gap power vector and the measured rotor current in a common reference
frame has been implemented with a closed loop PI adjuster controller [79]. This has
the advantage of performing the estimation in the rotor or in field reference frame;
however this has sensitive to stator no load power. The rotor current in the stator flux-
23
oriented reference has also been estimated from the stator current in the stationary
reference frame [76].
The model reference adaptive system (MRAS) method is used to estimate speed and
position using the outputs of two different machine models, as illustrated in figure 2.4.
One of these is position dependant, which is the adaptive model (AM), while the other
types dose not involve the estimated position in computation and is named the
reference model (RM) [78, 80-82]. Based on Popov’s hyperstability criterion, the
difference between the RM and the AM outputs is used to derive a suitable adaptive
mechanism to generate the estimated speed used to minimize the error between the
reference and estimated values [83]. This scheme for estimation can be used for
position or speed detection, which are investigated using a proportional-integral (PI)
or hysterias controller in DFIG systems implemented in different stator or rotor
reference frames [79, 84-86].
Different techniques have been adopted to form error vectors employed to detect
rotor position. The proposed sensorless methods [78, 81, 82, 86, 87] are based on
stator flux MRAS, rotor flux MRAS, rotor current MRAS , stator current MRAS,
rotor active power and rotor reactive power, and a comparison of these methods
briefly summarized in table 2.3 with different operating conditions including grid
connected and stand-alone DFIG system. The closed-loop MRAS observers has an
adaptive models based on static flux-current relations, therefore this makes the
estimation process very sensitive to machine inductances according to sensitivity
analysis. In addition the stationary reference frame is adopted in the transformation of
input variables and this may become inaccurate or even unstable in digital
implementations. A non-linear control approach which uses non-linear static and
dynamic state feedback controllers with a wind speed estimator in a wind turbine-
generator system has been proposed [88] while predictive control in DFIG system has
also been discussed [89-91].
24
Table 2.3: Basic comparison between different MRAS schemes.
MRAS based on
Stator Rotor Rotor Stator Active Reactive
flux flux current current power power
Good Best choice in
estimation grid tide and Stable only at
Best choice with have good Grid tide Independent on
Simple to
Advantages in stand different stability and condition and
implemented
rotor resistance
alone operation simple have simple variation
condition reference reference model
with � ≠0 model
PI gain is
complex Sensitive to Complex PI
Instable at Sensitive to rotor
and incorrect gain
grid tide resistance Reasonable
Disadvantages sensitive to estimation of implementation
with � = ��
variation and low accurate
incorrect and instable at
0 ��
accuracy
machine stand alone.
parameters
Actual o/p
I/p RM
Error
AM
Adaptive o/p
Estimated rotor speed
Adaptation
mechanism
Figure 2.4: Typical model of reference adaptive system (MRAS) block

diagram [83].
2-4 Maximize Power Point Tracking (MPPT) in DFIG-Based Wind

Turbines
Variable-speed wind turbines are designed with a strategy to achieve maximum
aerodynamic efficiency over a range of wind speeds. However, this technique requires
a power control scheme to track the existing maximum power and to limit the power
captured when the wind speed exceeds a certain level. For variable-speed wind
turbines, there are two types of power control: aerodynamic and generator control.
Aerodynamic control aims to limit the power in very high winds in order to avoid
25
damage to the wind turbine. The most commonly used type of aerodynamic power
control is to adjust the attack angle of the turbine blades (i.e. pitch blade angle)
according to the wind speed. On the other hand, generator control is realized by
adjusting rotor speed in order to capture the energy from the wind in an optimal
manner. Any change the rotor speed induces a change in the turbine’s power capture
[7, 92, 93]. Figure 2.5 illustrates the typical maximum power curve of a wind turbine
with maximum power point tracking (MPPT) as a function of wind speed (�� ). MPPT
is a method of assessment adopted to recover the maximum of the turbine’s power in
adjusting a new set point torque with slip at the optimum level [94, 95]. The
operation of the wind turbine is considered in four speed regions. In region I, from
zero to the minimum speed of generation (CUT-IN), the wind speed is usually below
3 m/s. Up to this limit, the power generation just supplies the friction losses.
Therefore, the turbine is shut down. In region II, the turbine operates with fixed-pitch
and variable-speed, and the generator speed is controlled in order to obtain the
maximum power available from the wind [92, 96]. In this operating region, the
objective of speed control is to follow the path of maximum power extraction [7]. In
the literature, different methods have been proposed to regulate the wind turbine to
follow the maximum power extraction trajectory, as explained later.
Figure 2.5: Ideal power curve for wind turbines [41].
26
Ideally, aerodynamic control only starts to operate when the power generated achieves
its rated value, which characterizes the beginning of region III. In cases where the
wind speed is above the rated value, turbine speed and power must be limited to its
rated value in order to prevent mechanical damage. Therefore, operation with MPPT
is no longer applied. The pitch blade angle mechanism is activated in order to reduce
power capture with increasing of the wind speed, and this allows the turbine to work
even with wind speeds above the rated level up to certain limit. In the fourth region
IV, the wind speed is considered too high and the turbine is shut down. The
combination of generator and pitch control comprises the MPPT, limited to rating
values of turbine and generator. Figure 2.6 illustrates the combination of generator
and aerodynamic control used to form the MPPT through several wind speeds [6].
The values of the wind speed and turbine power are represented in per unit for
simplicity.
Figure 2.6: Maximum power point tracking (MPPT) of the DFIG system [92].
For each wind speed, there is an optimal turbine rotational speed which takes the
maximum power from the wind. The combination of turbine speed and mechanical
27
power form, therefore, the operating point of the turbine. In other words, MPPT
consists of maintaining the operating point along the maximum power trajectory
presented in figure 2.6 by means of variable-speed control. To achieve variable-speed
operation, a back-to-back converter is placed between the generator and the grid. The
power converter will apply voltages of variable frequency and amplitude to the
generator in order to control the rotor currents, which in turn allows control of the
generator torque and speed [7]. The wind turbine used with the technique of
maximum power tracking is a very appropriate power source for the grid [93].
Basically, there are three traditional methods of achieving MPPT: tip speed ratio
(TSR) control, power signal feedback (PSF) control and hill-climb searching (HSC)
control. In TSR, wind velocities are difficult to obtain accurately, while the PSF
control method requires knowledge of the maximum power curve of the wind turbine.
This method can effectively avoid fluctuations in output power and is easy to execute
[97]. Wind speed measurement and estimation is required in most types of variable
speed vector control to track the desired wind power with changes the rotational speed
of the turbine generator based on the variation in wind velocities [98].
Conventionally, a number of anemometers surrounding the wind turbine were used to
measure wind velocities, but this method is usually costly and unreliable [99].
Ahmed and Kalil [100] have used an algorithm of support vector regression (SVR) to
estimate an unknown system’s wind speed depending on the power of the turbine and
its speed based on training data of the optimal tip speed ratio. However, this study is
not purely sensorless because the turbine speed is measured with an encoder. A power
coefficient polynomial method has been proposed [101] to compute the wind velocity
iteratively by deducing the polynomial roots numerically. This has the advantage of
being applicable to all doubly-fed machines. However, this method would increase the
complexity computation, consume more time, and adversely affect system
performance. The rotor position phase and PLL have been applied track the maximum
power point of a DFIG [98] and this method has the advantage of parameter
independent and insensitive to measurement noise. Based on the two dimensional
look-up table of the power coefficient with power mapping, which is used to track the
MPPT with or without wind speed measurement to estimate the optimal torque from
the estimating on rotating speed. This has the drawback of requiring greater memory
size which influences on the accuracy of estimation [102, 103].
28
Another method of sensorless vector control has employed an artificial neural network
(ANN) with a three-layer radial basis function network (RBFN) used for estimating a
shaft speed [104]. In addition, an extended Luenberger and another neuronal observer
were used to estimate the wind velocity so as to impose MPPT.
2-5 Fault Ride-Through (FRT) of the DFIG Based Wind Turbine

Fault ride-through (FRT) represents the ability of the DFIG wind turbine to remain
connected to the grid even when the voltage level across the point common coupling
(PCC) is drops to 20% [2, 105]. This requirement is essential in the new grid code of
power systems in different countries since the isolation of the wind turbine (WT) from
the grid system in conditions of sever voltage dip will lead to the loss of stability in the
power system [7]. The utilities adopted to enhance FRT capability in wind turbines
deals with issues of over-current in the stator and rotor circuit windings of generators
in addition to over-voltage across the DC-link capacitor in the back-to-back power
converter of the DFIG system [2, 38]. The low power rating of the converter in the
DFIG system configuration means priority is given to protecting the converter from the
risk of damage due to short-circuit fault effects producing a fluctuating high currents in
the stator and rotor windings of the generator. Therefore it is essential to protect the
converter so as to keep the DFIG operating and connected to the grid in order to fulfil
the requirements of the grid code. Figure 2.7 illustrates a summary of the voltage
profile for ride-through capabilities in the grid codes, and its condition that only when
the grid voltage goes below the curves are turbines permitted to be isolated from the
grid. Moreover, when the voltage is in the special area, the turbines should supply
reactive power [36, 105, 106]. Different countries have issued grid codes for wind
turbines with a special focus on their ride-through capability, including fault ride-
through (FRT) at the point of common coupling (PCC) for wind farms and low voltage
ride-through (LVRT) for wind turbine generators.
29
Figure 2.7: FRT requirements for various grid codes [106].
The basic definition of a voltage dip is a sudden reduction in grid rms voltage in the
range of 0.1-0.9 pu for a short period of time between 0.5-30 cycles. The duration of a
voltage dip is generally evaluated based on the clearing time of the protection devices.
Voltage dips represent one of the most harmful power quality problems and the major
concern for the industry. The main causes of this problem as follows [37]:-
 Short-circuit in the grid

 Switching operation associated with a temporary disconnection of the supply
 Flow of peak current caused by starting a large motor
 Peak current caused by arc furans
 Peak current caused by transformer in saturated
Based on the space vector transformations, seven types of voltage dips in the DFIG-
based wind turbine have been listed [107], and this classification enables the
evaluation of the severity and characteristics of each type of voltage dip. Choudhury
[108] has observed the main effects of a sudden change in grid voltage on the
behaviour of DFIG wind turbine systems under steady state conditions. The
configuration of the DFIG has a stator circuit directly connected to the grid while the
rotor winding is connected to the grid via the back-to-back converter, and this is the
reason for all of the efforts to develop the fault ride-through (FRT) capability. Where a
power grid system is with a voltage drop at the PCC; the stator voltage of the DFIG
30
will immediately change, and if it not suddenly varies the rotor voltage will be
compensate the dropping in stator voltage and this will drive to oscillate the stator and
rotor currents with high incremental in its values [36, 106]. This disturbance in the
stator and rotor currents will cause damage to the rotor converter and increase the
mechanical stress on the wind turbine as a result of the transient in the electromagnetic
torque of the generator [2]. An algorithm for support and safety operation requirements
for the DFIG system to ride this condition is presented in figure 2.8, which describes
the main aims of FRT ability enhancement in DFIG system.
Protect the RSC

from over-current
Grid voltage
Protect the DC-link
support by
capacitor from
injecting reactive
over-voltage
current
Ride-through Reduce the

ability for all types mechanical stress
of grid faults on the WT
Figure 2.8: Algorithm to support FRT in DFIG wind turbines.
Figure 2.9: DFIG system equipped with a crowbar [31].
31
Traditionally, the protective systems utilized to enhance the behaviour of DFIG-

based wind turbines in severe voltage dips can be summarized as crowbar circuits,
energy storage systems, stator switches, and auxiliary parallel grid side rectifier (stator
reactive power converter) [109-112]. The auxiliary parallel grid side rectifier is
represent another method to enhance the FRT capability of the DFIG system which is
characterized by connect the GSC in series with the stator voltage rather than in
parallel as in the conventional DFIG. This configuration can be accomplished alone or
with adding extra rectifier. It can be used to control two different variables which are
the stator voltage and stator flux control. In stator flux control it has been used to
remove the oscillations in the stator flux and so regulating the stator current and the
rotor current. The disadvantage of this method is regulating the stator flux at a low
value during the fault consequently; the stator power will be regulated at a lower value
during the fault which differs from the operating value, leading to lose the maximum
power tracking during faults. Another problem with this technique is using a flux
estimator, which adds a time delay, inaccuracy and system complexity. When it used
to control the stator voltage, this technology has the same targets of containing the
stator current, rotor current and the DC voltage within their safe limits has been
achieved. Moreover, the stator power is regulated at its operating value which keeps
maximum power tracking even during faults [112].
. The crowbar system is the most common technique used in protects the RSC during
voltage dip, and it’s concept of operation is based on the rotor current flowing through
an external resistor during faults to dissipate the unbalanced power which results from
the rotor over-current as shown in figure 2.9. [2, 7, 31] it is used to avoid the voltage
bus exceeding the maximum value once the RSC loses current control, providing a
path for the rotor currents [111]. The crowbar short-circuits the rotor and the machine
operates as a squirrel cage machine. There are two types of crowbar which are passive
(use a Thyristor to connecting the crowbar resistor to the rotor winding with no gate
turn-off feature and it will not be able to resume operation as soon as the transient
current and voltage have decayed) and active (a three phase rectifier and power
semiconductor device with gate turn off- capability and external resistor and this
configuration is fast response and is able to resume to the normal operation condition
state) [7]. The crowbar and DC chopper protection schemes have been used to enhance
32
the FRT ability in DFIG-based wind turbine systems [113, 114]. The main demerits of
utilizing the crowbar in a DFIG wind turbine system is that when activated, the DFIG
runs as a squirrel cage machine where the rotor windings are short circuited [115-117]
.The DFIG start to absorb reactive power from the grid and the decoupled control of
active and reactive power is no longer available [118]. This means that the situations
deteriorate during a grid fault and the fault recovery process will slow down or it may
even result in voltage collapse [119]. Thus, it is better to reduce the connection time of
the crowbar and enable the wind turbine to return to normal operation as soon as the
transients have decayed [116].
Some of the most important studies relating to DFIG modeling, simulation and
analysis are reviewed in this section. Lihui [36] proposed a mathematical modification
to the control system for both the rotor side converter (RSC) and the grid side
converter (GSC) to improve the low voltage ride-through (LVRT) capability of the
DFIG wind turbine during voltage dip based on an increase in the generator’s rotor
speed and adding a compensation term temporally when faults occur to absorb the
excessive energy and damp the oscillation of DC-link voltage. The simulation results
were verified by comparison of the modification with crowbar protection. It was
shown that the speed, electromagnetic torque, DC-link voltage, current variation and
pitch control with longer duration of voltage dip in crowbar has better LVRT
behaviour upon the proposed method even the later have better enhancement to ride
through grid faults at the short period conditions. In addition, theoretical analysis has
utilized the stator feedback and power (PQ) null method to enhance the DFIG’s ability
to ride-through the voltage dip [120, 121]. [122] have considered the theoretical
transient responses of DFIG wind turbines subjected to symmetrical and
unsymmetrical voltage sags. An 80% voltage sag was used when employing the full-
order and reduced order models to measure the power quality impact on the DFIG
system, and it was found that the flicker emission is very low, the reactive power is
close to zero and the current total harmonic distortion (THD) is always less than 5%.
Another study [123] simplified dynamic model for DFIG-based wind turbines was
presented, which was tested theoretically with different conditions during both the
steady state and the transient of the grid voltage by means of PDCAD/EMTDC
simulations.
33
2-6 Wind Turbine Topologies

Many choices are available to the academics and industry for a wind turbine with no
assent among them for the best wind turbine generator technology. In conventional,
wind turbine generators (WTGs) categorized in three main types, which are the direct
current (DC), alternating current (AC) synchronous and AC asynchronous generators.
Figure 2.10a shows the DC wind generator system which consist of a wind turbine, a
generator, inverter, system control, and transformer. The power is injected through
brushes connecting the commentator which is utilized to rectify the generated AC
power into DC output [162]. This type of technology is costly as a result of using the
commutators and brushes. Furthermore, this wind turbine generator is preferred to use
where the load is closed to the generator. Secondly, switched reluctance-based WTs
illustrated in figure 2.10b which includes stator and salient rotors and it is
characterized by laminated steel sheets. Stator current induced as a result of magnetic
circuit linking the stator and rotor changes through shaft rotation. The primary
advantages of this WT are the ability to operate even in harsh environment, high
reliability, and simplicity. However, switched reluctance-based WTs suffer from the
high weight of generator. Permanent magnet excitation is greatly adopted in newer
smaller scale wind turbine designs, because of its higher efficiency and reduced the
diameter of blade. Furthermore, this generator does not need external excitation
current. The permanent magnets (PMs) are the source of DC excitation in this
generator where these rugged PMs are installed on the rotor to generate a constant
magnetic field and the generated electrical power will transfer to grid from the stator
[162, 163]. The permanent magnet synchronous generator (PMSG) shown in figure
2.10c which unable to produced electrical power with fixed frequency to the grid.
Therefore, the stator required to connect to the power grid through full rated AC-DC-
AC power converters because the converter is connected to the stator. Squirrel-cage
induction generator (SCIG) depicted in figure 2.10d where the schematics of the
system including the wind turbine, pitch control, and reactive power compensator.
This type of WT generally exchanges energy with the interconnected grid via only the
stator winding. The rotor winding in this category almost is short-circuited. A rotating
magnetic field is established across the air-gap [162]. If the rotor rotates at a speed
different to synchronous speed, a slip is created and the rotor circuit is energized. This
type of WT is simple, reliable, and cost effective. However, induction machines draw
34
reactive power from the grid and thus some form of reactive power compensation is
needed such as the use of capacitors or power converters. Table 2.4 is summarized
different wind turbine topologies.
(a) (b)
(c) (d)
Figure 2.10: Configuration of wind turbine topology using: (a) DC generator; (b)
switch reluctance generator; (c) PMSG; (d) Squirrel
cage induction generator [162].
Table 2.4: Overall comparison of WT generator topologies [162]

DC Asynchronous Synchronous
DC generators SCIG DFIG PM Reluctance
Speed variable fixed variable variable variable
partially
directly to the directly to totally via totally via
Power supply stator-
grid the grid converters converters
converter
Controllability poor poor good good good
Active-reactive power
no dependent separate separate separate
control
Grid-support capability low low high very high medium
Reliability poor medium high high very high
Fault response slow slow high high high
Cost low low medium high medium
Mass saving low low high very high low
low power, medium- direct drive;
small wind
Suitability residential large wind small-medium early stage
turbines
application turbines wind turbines
35
2-7 Conclusions
In this chapter, an overview has been given of different approaches to modelling the
doubly-fed induction generator (DFIG)-based wind turbine system in addition to
different control strategies with their various advantages and disadvantages.
In the second part of a chapter, different methods developed in the literature for the
detection of the rotor position of DFIG systems have been described. Although there
are methods estimate the rotor position of the DFIG with different phases of
configurations and electric quantities measured to provide position information for the
rotor, it appears that the most attractive techniques are categorized based detecting
position with acceptable error in addition to the insensitivity to variation in the
system’s parameters.
A brief overview of maximum power point tracking (MPPT) methods for DFIG
systems in the literature has also been presented.
Based on the literature survey, several types of improvements designed to enhance the
fault ride-through (FRT) capability of the DFIG-based wind turbine have been
introduced in the fourth part of this chapter.
36
Chapter 3: Mathematical Model of DFIG System
CHAPTER THREE
37
CHAPTER THREE
MATHEMATICAL MODEL OF DOUBLY- FED INDUCTION

GENERATOR SYSTEM
3-1 Introduction
In this chapter a dynamic model of different parts of the DFIG based wind turbine
including both the mechanical and electrical components are presented in order to fully
understand the dynamic performance of the system in any specific operating conditions
and to design the control system. Firstly, the main mathematical equations, that
characterize the relationship between voltage and fluxes in the machine. These are the
basic equations used for deriving the dynamic model. Models of the aerodynamic and
mechanical parts are also presented. The decoupling control active and reactive power
in this model is also explained. The vector control strategy for the grid-side converter
is studied based on the dynamic models of the grid-side system formed by the grid, the
grid-side filter, and the grid-side converter itself. This control strategy highlights the
two main objectives of the grid-side converter: control of both the bus voltage of the
DC-link and the active and reactive powers exchanged bi-directionally between the
rotor of the machine and the grid. Based on the final modeling of the machine and
control system equations, a detailed evaluation of the behaviour of the system is
carried out, resulting in performance curves that can reveal the current, voltage, or
different magnitude requirements, depending on the specific operating conditions of
the machine.
In Chapters Four and Five respectively, the simulation results obtained from the
Simulink model developed in this chapter will be employed with the corresponding
rotor position estimation and in the performance comparison results of the system
under unbalance dynamic behavior condition.
3-2 Dynamic Model of Doubly-Fed Induction Generator
The equivalent circuit of a DFIG shown in figure 3.1 can be characterized by
different reference frames such as the stationary frame, rotor frame, or the synchronous
frame oriented to either stator flux [22] or stator voltage [47]. The simplified DFIG
model can be described as three windings in the stator and three windings in the rotor,
38
as shown in figure 3.2. The instantaneous stator and rotor voltages, current, and flux of
the machine are given as follow:
Figure 3.1: System configuration of the DFIG-based wind turbine [7].
� = � + � (3.1)
� = � + � (3.2)
� = � + � (3.3)
� = � + � (3.4)
� = � + � (3.5)
� = � + � (3.6)
where:-
: is the stator resistance.
: is the rotor resistance referred to the stator.
� ,� , and � : are the applied stator voltages.
� ,� , and � : are the stator currents of phases.
� ,� , and � : are the stator referred rotor voltages.
� ,� , and � : are the stator referred rotor currents of phases.
� ,� , and � : are the stator fluxes.
� ,� , and � : are the rotor fluxes.
At steady state condition the following hold:-
- The stator side electric magnitudes have a constant sinusoidal angular
frequency (� ).
39
- The rotor side electric magnitudes have a constant angular frequency (� ).

The relationship between the stator angular frequency and the rotor angular frequency
can be expressed as:
� +� =� (3.7)
where (� ) is the electrical angular frequency of the machine.
.
Figure 3.2: Ideal three-phase windings (stator and rotor) of the DFIG [7].
In generalized electrical machine theory, the inclusion of space vector analysis is based
on the following hypotheses shown in:
1. The distributions of flux and magneto-motive force (MMF) are represented by

their fundamental harmonic component alone.
2. The effect of slotting winding distribution can be neglected.
3. Commutation effects and brush-connections are considered to be ideal.
4. The influences of eddy currents and hysteresis can be neglected.
5. Magnetic saturation is not present.
Magnetic linearity is assumed in order to produce linear superimposition magnetic
fields, which simplifies model development considerably. In practical, if magnetic
saturation has an effect on the machine, it will limit the transient over-shoot current and
power in the step response of the vector control, and restrict the instant excesses of fault
response. The linear magnetic analysis will therefore derive a worst-case response in
terms of transient response to both control and faults. Therefore, the stator and rotor
40
voltage and magnetic flux equations of the DFIG in the stationary reference (αβ) frame
are as follows [7, 31, 34]:
� = � + � (3.8)
� = � + � (3.9)
� = � +� � + � (3.10)
� = � −� � + � (3.11)
In the (dq) frame, the voltages can describe as:
� = � +� � + � (3.12)
� = � −� � + � (3.13)
� = � + � −� � + � (3.14)
� = � − � −� � + � (3.15)
where: = / , and the stator and rotor flux linkage equations in the (αβ) frame are:
�
� = � + � (3.16)
�
� = � + � (3.17)
�
� = � + � (3.18)
�
� = � + � (3.19)
The stator and rotor flux linkage equations in the (dq) frame are:-
� = � + � (3.20)
� = � + � (3.21)
� = � + � (3.22)
� = � + � (3.23)
where, � ,� ,, � ,� ,� ,� are respectively the stator and rotor voltages,

currents, and flux vector in the synchronous reference frame , represent the stator
and rotor resistance; , the stator and rotor self-inductance; the mutual
inductance; and � , � the synchronous and rotor angular speed.
41
The stator and rotor active and reactive power in the (αβ) frame are calculated as
follows:
= {�̅ . �̅∗ } = � � +� � (3.24)
= {�̅ . �̅∗ } = � � +� � (3.25)
= {�̅ . �̅∗ } = � � −� � (3.26)
= {�̅ . �̅∗ } = � � −� � (3.27)
The stator/rotor active and reactive power and electromagnetic torque expressions in
(dq) frame:-
= {�̅ . �̅∗ } = � � +� � (3.28)
= {�̅ . �̅∗ } = � � −� � (3.29)
= {�̅ . �̅∗ } = � � +� � (3.30)
= {�̅ . �̅∗ } = � � −� � (3.31)
where the superscript * represents the complex conjugate of a space vector, as used in
phasors. Finally, the electromagnetic torque can be found from:
= {�̅ . �̅∗ } = � � −� � (3.32)
3-3 Vector Control System

The control strategy utilized to control the RSC and GSC in this study is vector-
oriented control in the synchronous reference frame since it has the several merits of a
good dynamic response in regular and fault operating conditions, it is robust to
parameter variation and measurement noise, and has fixed switching frequency.
Furthermore, a wide range of reference material is available. On the other hand, this
method has drawbacks when compared with DTC, DPC, and vector-oriented control in
the rotor/stationary reference frame, including slower response time compared to
direct methods and the need to estimate the flux angle [97].
42
3-3-1 Vector Control of the Rotor Side Converter (RSC)

The main objective of the use RSC is to decouple control of the active and reactive
power in normal and fault operating conditions, in addition to controlling and
synchronizing the stator voltages fed to the grid in the start-up phase. To give better
differentiation between the RSC and the GSC controllers, the control variables utilized
to conduct the vector orientation in this study are based on the stator active and
reactive power. Table 3.1 demonstrates the different scenarios which may be adopted
in the vector control of the RSC [22, 98, 124].
Table 3.1: Possibilities of vector control in RSC controller.

Synchronous Frame Rotor Frame
Stator/grid flux Stator/grid flux
Stator voltage Stator voltage
Vector of Control
Air-gap flux Air-gap flux
Rotor flux Rotor flux
Appearance of Control
DC AC
Variables
Adaptive Proportional-
Controller Classical PI
Resonant controller
The generic expressions for the DFIG can be simplified by using a reference frame
aligned with the stator flux (i.e.� =� , � = as shown in figure 3.3 to decouple
the control of the rotor quantities. In these reference frames, the control variables
appear as DC values, and classical PI control can be employed.
Under the hypothesis of stator flux orientation and neglecting the stator phase
resistance, the equations of the currents, fluxes and the voltages, may be written as
follows [7]:
� = � − � (3.33)
� =− � (3.34)
� = � +� � (3.35)
43
� =� � (3.36)
After simplifications of equations 3.12 - 3.15 by assuming that stator resistance is small,
and because the direct stator flux component is constant in the case of stator flux
orientation (i.e. � = , therefore:
� = (3.37)
� =� � (3.38)
� = � +� � − � −� �� (3.39)
� = � +� � + � −� �� + � (3.40)
In addition, active and reactive stator power and electromagnetic torque can be
formulated after simplification as:
=− (� � ) (3.41)
= � − � (3.42)
=− � � (3.43)
where � = − /
Figure 3.3: Reference frames and angles [76].

The q-axis current component is used to control the active power using a maximum
power point tracking (MPPT) strategy to calculate the active power reference [125].
44
The measured active power is compared with the reference active power and the error
result is fed to the PI controller to estimate the reference value for the q-axis rotor
current. This signal is then compared to its actual value and the error is passed through
a second PI controller determining the reference voltage for the q-axis component.
The d-axis is used to control the reactive power exchanged with the grid, which in
normal operation is set to zero in order to operate with unity power factor. Similar to
the control strategy for the q-component, when the sensed reactive power is compared
with the reference reactive power, an error is produced on processing through the PI
controller, computes the reference d-axis component of the rotor current. This signal is
compared to the d-axis current value and the error is sent to a second PI controller
which determines the reference voltage for the d-axis component. Finally, the dq-
reference voltages are passed through the PWM module and the modulation indices for
the control of the RSC are determined, as shown in figure 3.4.
The inner and outer control loops shown in figure 3.4 have been judge by four PI
controllers in determining the reference dq rotor currents in the outer loop and the
reference dq voltages in the inner loop. The transfer functions of the stator active and
reactive power (outer loop) control are given as follows:
From equations 3.39 and 3.40, if it is assumed that:
=− � −� �� (3.44)
= � −� �� + � (3.45)
 
    L 
e r
m
ds
  idr Lr 

L  s 
ref  i
ref
 
P s Ʃ PI
qr
Ʃ PI Ʃ v
*
  v qr1
 qr
 
 
ref
Q
ref
Ʃ PI
i dr
Ʃ PI Ʃ
*
s
  v dr1  v dr
 
cal cal
Q s P
cal
s i
cal
dr i qr
Computation
the stator active
Computation
rotor current in
    i L 
e r qr r
and reactive synchronise
power reference frame
Figure 3.4: Block diagram of RSC controller.
45
Substituting equations 3.44 and 3.45 in equations 3.39 and 3.40 respectively gives:
� = � +� � + (3.46)
� = � +� � + (3.47)
From equations 3.46 and 3.47, the plant model includes the voltage drop in the
transitory inductance � , cross-coupling from the orthogonal loop and the emf
induced by the stator flux. These couplings are nulled at the controller output by the
compensating terms. This results in the same closed-loop transfer function for the two
current control loops. Furthermore simplifications of equations 3.46 and 3.47, gives:
−
� = (3.48)
+ �
( − )
� = + �
(3.49)
Figure 3.5 shows the closed loop system using the rotor current control by a PI
controller. The inner control loop has a significant advantage for the protection of the
DFIG. It can naturally protect the system from over-current, since current limiters can
easily be inserted in the control system shown below. Because general PI controllers
are widely used and have proven to be effective, they are also applied in the following
analysis [7, 126]:
Hence, the plant for the current control loops is given by:
� = = =
+ �
(3.50)
The ( � ) and ( � ) current control loop forms are then:
�
� = +� � = (� + ) (� −� )
} (3.51)
�
� = +� � = (� + ) (� −� )
Then the transfer functions between the reference and actual currents are changed to
the following:-
� ( � +� )
�
= } (3.52)
� ( +� ) �
+ +�
�
46
� ( � +� )
�
=
� ( +� ) �
+ � +�
The design of the current controllers follows directly from equation 3.50, which can be
written in the z-domain as:
−�
� � = �−�
(3.53)
− ⁄ �
where: =
( ), is the sample time 0.5ms. The control of the current in RSC in z-domain is shown
in figure 3.6.
e
1


 k i1
1
r L
ref
i Ʃ k Ʃ  
  s
p1
dr s r r
i dr
  1
i qr
k
ref
Ʃ Ʃ
i r r  s L
i1
qr k p1
s  r

e
2
Figure 3.5: Closed loop current control of RSC in S-domain.
Out put
Ts v current
Reference k p  ki 0.0596
current + 1
1 Inverter
z  0.708
- z
PI Plant
Figure 3.6: Closed loop GSC current-control in z-domain.
47
Choosing a compromise control for the parameters is necessary to gain good

performance even through the whole system might be able to run with a wide range of
parameters. Most previous studies have selected the gains of control using experience
or trial and error, but this method is not good, particularly when the control system is
implemented for different system. Furthermore the most important objective is to keep
the system in a stable condition by choosing appropriate parameters for the control to
use to tune up according to the specific behavioural requirements. The transfer function
of the outer control loop of the rotor-side converter control is given by the following
procedure:
From equations 3.20 - 3.23, the following expression can be deduced for the stator and
rotor currents:
� �
� = − (3.54)
� �
� �
� =� − �
(3.55)
−� �
� = �
+� (3.56)
−� �
� = �
+� (3.57)
Substituting equations 3.38 and 3.56 in equation 3.42 gives, the stator reactive power
formula as equation:
�
= �
� − � � (3.58)
Differentiating equation 3.58 with respect to time gives:

− � � �
= (3.59)
�
Solving equation 3.59 in term of � then gives:

− � �
= �
� − � −� � − � (� ) (3.60)
In addition, the PI controller formula will be:
�
= � − ( − ) (3.61)
where:-
�
= (3.62)
�
48
The transfer function of stator active power control is derived as the same procedure in
equations 3.58-3.61, equations 3.38 and 3.57 are substituted in equation 3.41 and
differentiation is applied to both sides of the equation with respect to time, giving:
− � � �
= (3.63)
�
Solving equation 3.63 in terms of ( � ) gives:
− � � �
= �
� − � −� � −� � + (3.64)
�
The PI controller formula for the stator power loop is formulated as follows:
�
� = � − ( − ) (3.65)
And it is know:
�= �
(3.66)
The rotor side controller will give the DFIG model feedback control, with the system
outputs constantly inputted back into the system, giving a more efficient and stable
system. The tuning of the discrete PI synchronization controller defined in equation
3.50 has been implemented by the MATLAB/SISO Tool, by assuming acceptable
bandwidth, phase and gain margins as can be observed in the controller design plots in
figure 3.7. The root locus of the RSC current control loop gives a stable loop as
illustrated in figure 3.7. The open loop test of the current control loop is stable because
the poles on the left hand plane (LHP). In addition, this is confirmed by simulation
results of RSC control loop in section (DFIG system model verification) with very
good tracking of the reference values to the measured values and with minor
oscillation.
The frequency response of a system can be represented by a Bode diagram, where one
graph shows magnitude versus frequency and the other shows phase versus frequency.
The root-locus provides information on the closed-loop system's transient response,
while the frequency response shows the steady-state or low-frequency performance
and stability margins. Several parameters need to be considered in the design using
frequency response analysis. In this investigation, the phase margin (PM) and gain
margin (GM) have been used. The observation of the open loop frequency response
49
results of the system control are illustrated in figure 3.7. There is no frequency
oscillation with phase margin of 60 deg and frequency at 79.6 rad/sec, and in addition
the step response is characterized by an overshoot of 11.6% at 0.038 sec with a peak
amplitude of 1.12, rise time of 0.0184 sec and settling time of 0.0612 sec. The
significant overshoot in this step response analysis is present due to the derivative term
in the numerator. This is not desirable as it may cause damage to the converter.
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
250 0.9
40
0.81 0.7 0.56 0.4 0.2
200
20
Magnitude (dB)
0.955
150
0
100
0.988
-20
50
Imag Axis
G.M.: Inf
500 400 300 200 100 -40 Freq: NaN
0 Stable loop
-90
-50
0.988
Phase (deg)
-100 -120
-150 0.955
-150
-200
P.M.: 60 deg
0.9 0.81 0.7 0.56 0.4 0.2 Freq: 79.6 rad/s
-250 -180
0 1 2 3 4
-500 -400 -300 -200 -100 0 100 10 10 10 10 10
Real Axis Frequency (rad/s)
(a)
1.4
1.2 Peak amplitude: 1.12

Overshoot (%): 11.6
at time (sec): 0.0382 Settling time (sec): 0.0612
1
Amplitude
0.8
Rise time (sec): 0.0184
0.6
0.4
0.2
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Time (seconds)
(b)
Figure 3.7: RSC controller tuning of the PI parameters: (a) root locus, open-loop Bode
50
3-3-2 Vector Control of the Grid Side Converter (GSC)

The grid-side converter, grid side filter, and grid voltage terminal represent the grid-
side system as shown in figure 3.8. The grid side converter is characterized as ideal
bidirectional switches that convert voltage and currents from DC to AC given that the
exchange of power can be conducted in both directions from AC to DC (rectifier
mode) and from DC to AC (inverter mode) [7]. The main objective of the grid side
filter is to produce a fast-acting response of the current controller of the GSC with the
changes in current during a time of milliseconds and to maintain the stability of the
controllers output.
GSC
v ag
AC
L r
f f v af i dc i
RSC
dc
i 
i dc
v bg
AC ag
v bf
RSC
GSC v dc
v cg
AC i bg
v cf
i cg
Controller
Figure 3.8: The grid-side system.
Modelling of the inductive filter and the grid voltage model in the GSC system is
performed using a simple and reliable assumption which adopts an inductive filter and
locating an inductance in each phase, as shown in figure 3.9. The grid voltage is
modelled as an ideal three-phase balanced voltage. The three-phase system can be
modelled as three independent, but equivalent, single-phase systems. Therefore, the
balance across the inductor is:
� � � �
� �
[ ]−[ ]= [� ] + [� ]
� � (3.67)
� �
Consequently, for modelling purposes, it is necessary to isolate the first derivative of

the currents as follows:
51
� � � �
[� ] = [ � ] − [� ] − [� ] (3.68)
� � � �
Equation 3.67 can be written in the stationary reference (αβ) frame by transforming
the coordination from (abc) to (αβ), which gives:
� = � + +�
} (3.69)
�
� = � + +�
The cconversion of the coordination of equation 3.69 is performed by multiplying it

−�
with the factor , hence:
� = � + +� −� � (3.70)
� = � + +� +� � (3.71)
The active and reactive power flow between the grid and the grid side converter are
expressed as:
= ̅̅̅. �̅∗ } =
{� � � +� � (3.72)
= ̅̅̅. �̅∗ } =
{� � � −� � (3.73)
For voltage and current coordinate transformations, the angle of the grid voltage is
needed. The angular position of the supply voltage is calculated as [22, 127]:
−
� = ∫� = � (3.74)
i - i e qg
+ i
dg
r f
L f
g
r f
L f
+ + + +
v df v dg v f v g
- - - -
i i e dg
i
qg
r f L f + - g
r f
L f
+ + + +
v v qg v
qf
v f
g
- - - -
(a) (b)
Figure 3.9: Equivalent electric circuit of the GSC system in:
(a) dq coordinates; (b) αβ coordinates.
52
In this study, the vector control approach is used in the grid-side converter (GSC) with
a reference frame oriented along the grid voltage vector position as shown in figure
3.10 this enables independent control of the active and reactive power flowing between
the grid and the GSC. The pulse width modulation (PWM) voltage source converter is
a current regulator with d-axis current used to regulate the DC-link voltage and the q-
axis current used to control the reactive power.
q
d

v

g
 e
i g
Figure 3.10: Orientation with d-axis of the grid voltage space vector in GSC [30].
In the GSC system, the vector control approach is designed with a reference frame
oriented along the grid voltage vector � = ). Therefore, according the alignment
mention above, the system equations will be as follows:
� = � + +� −� � (3.75)
� = � + +� � (3.76)
( )
= (3.77)
( )
=− (3.78)
By assuming the following terms in equations 3.75 and 3.76:
� = + � (3.79)
� = + � (3.80)
Equations 3.79 and 3.80 are then converted to the S-plane by taking the Laplace
transformation, giving:
53
� =( + )� (3.81)
� =( + )� (3.82)
Assume that:
� = � + ∆� (3.83)
� =� + ∆� (3.84)
where � and � are the DC components of � and � respectively (i.e.
� =� ,� =� ). In addition ∆� and ∆� are the ripple current components.

Then by substituting equations 3.81 and 3.82 in equations 3.83 and 3.84 respectively,
and rearranging the equations knowing that ∆� ≪� and ∆� ≪� , the transfer
function is given from � to � and � to � using the first order transfer
function without cross-coupling. Hence the plant model for the current control loops is
given by:
= = (3.85)
+
Therefore, it is possible to design a feedback loop and PI controller to generate the

reference value of voltage in the current control regulation formula using a PI
controller as follows [22, 126]:
�
� = � + (� − � ) (3.86)
�
� = � + (� − � ) (3.87)
Substituting equations 3.86 and 3.87 in equations 3.75 and 3.76 respectively gives:
�
� = � + (� − � )+� −� � (3.88)
�
� = � + (� − � )+� � (3.89)
The transfer function of GSC closed-loop control with the PI controller is given by:
( . + � )
= = +
(3.90)
+ + �
54
Therefore, the scheme of the current loop control is shown in figure 3.11. The design
of the current controller plant follows directly from equation 3.85, which can be
written in the z-domain as:
−�
� = (3.91)
�−�
−( ⁄ )
where: � =
where ( ), is the sample time which is equal to 0.5ms. Then equation 3.91 expressed
as follows:
.
� = (3.92)
�− .
The control of the current in the GSC in the z-domain is shown in figure 3.12 have
parameters used in the standard design techniques of the DFIG system may be applied.
v dg
 Li e f qg
ref 
i dg
k v df 1

1
Ʃ k p2  i2
Ʃ
s
 s  r f L f
i dg

ref
i qg v qf 1  1 i qg
Ʃ k Ʃ
k p2  i2
s  r f
s L f

Li e f dg
Figure 3.11: Current control loop of the GSC in S-domain [126].
Out put
Ts v current
Reference k p  ki 0.1938
current + 1
1 Inverter z  0.737
- z
PI Plant
Figure 3.12: The closed loop GSC current-control in z-domain.
The results of the root locus analysis of the inner loop control are shown in figure 3.13
for the dq-axis current of the GSC. The poles of the model lies on the-left hand plane
(LHP) in the root locus graph thus confirming that the loop control is stable without
any oscillation in the frequency response. Hence, there are two closed-loop poles in the
55
system: the first one has value of pole = -0.283, damping =1 and frequency = 0.823Hz.
The second pole has values of = -44.9, damping=1, and frequency = 44.9Hz. In
addition, the step response has an overshoot to 1dB at 0.084 sec with, rise time of
0.047sec and settling time of 0.0812sec and with stability margin values of phase
margin is 89.6 deg and frequency response of 45.4 rad/sec.
0.4 120
1 1 1 0.999 0.999 0.994
100
0.3 1
Magnitude (dB)
80
0.2
60
1
40
0.1
Imag Axis
20 G.M.: Inf
Freq: NaN
45 40 35 30 25 20 15 10 5
0 Stable loop
0
-90
P.M.: 89.6 deg
Freq: 45.2 rad/sec
-0.1
1
Phase (deg)
-0.2
-135
1
-0.3
1 1 1 0.999 0.999 0.994

-0.4 -180
-45 -40 -35 -30 -25 -20 -15 -10 -5 0 -2 -1 0 1
10 10 10 10
Real Axis Frequency (rad/sec)
(a)
1.4
1.2
Amplitude
0.8
0.6
0.4
0.2
0
0 0.02 0.04 0.06 0.08 0.1
time (seconds)
0.12 0.14 0.16 0.18 0.2
(b)
Figure 3.13: GSC controller tuning the PI parameters: (a) root locus, open loop Bode
56
3-3-3 Control of the DC-Bus Voltage

The primary objective of the GSC shown in figure 3.14 is to control the value of the
DC bus voltage, which the direct current component of the GSC acts on DC-bus
voltage. The DC voltage control is adjusted by control of the power exchanged by the
converter. An increase or decrease in the DC voltage level is obtained by taking more
or less power from the grid with respect to what is required by the DC load, thus
varying the value of the reference for the AC inner current control loops. The DC
voltage loop is the outer loop and the current control loops are the inner loops in the
GSC. These internal loops are implemented so as to achieve short settling times. On
the other hand, the main goals of the outer loop are optimum regulation and stability
and thus the voltage loop could be designed to be somewhat slower (by 5-20 times).
Therefore, the internal and the external loops can be considered to be decoupled [128].
dc dc
irsc i
gsc
i
dc
P r
P g
Q Q g
r
RSC GSC
Figure 3.14: Back-to-back power converter [126].
In order to derive the sub-system model of the DC link, the DC bus voltage must be
calculated which is dependent on the current flowing through the capacitor as follows:
� = ∫� (3.93)
The DC voltage control is adjusted by the control of the power exchanged by the
converter. The power and current equations of the DC-link are [7, 126, 129]:
� = =� − �� (3.94)
.� . = − � (3.95)
Let:
� =� + ∆� (3.96)
57
where � is the DC component of � (i.e. =� ), and ∆� is the ripple component

of the DC link voltage. Then by substituting equation 3.95 in equation 3.96 and
applying Laplace transformation, the following is given:
− � = � � + ∆� � (3.97)
Since ∆� ≪� ; therefore, equation 3.97 can be written as follows:
= (3.98)
Because ( � = � � ≈ � � ) then equation 3.98 then represents the transfer
function of the DC bus closed-loop control block diagram shown in figure 3.15:
= = (3.99)
Therefore, it is possible to design a feedback loop and PI controller to generate

reference value of � as follows:
�
� = = � + (� −� ) (3.100)
Then the closed loop transfer function is:
( . + � )
=� (3.101)
+ + �
� �
Similarly, as in the current control plant design, the DC-link plant in the discrete
domain characterized by equation 3.99 transformed to the z-domain and substitute
sample time equal to 5ms and capacitance value, hence DC-link plant expressed as:
.
� = �−
(3.102)
The closed-loop block diagram of the DC link is shown in figure 3.16, in which is
the DC link current of the grid-side converter and is the modulation index of the
PWM. The root locus of the DC-bus control loop is represented in figure 3.17 which
shows that the design has a stable loop with following closed loop poles. The first has
values of pole= -3.75, damping =1, and frequency=3.75Hz. The second pole is =-42.8,
at damping=1, and frequency =42.8 Hz. The step response characteristic are a peak
amplitude of 1.05 dB with an overshot of 5.49% at 0.0124 sec, rise time of 0.041 sec,
and settling time of 0.148 sec with the value of K in the design equal to 0.61076.
58
rsc
i dc

V
ref
dc i dg
 1
Ʃ k Ʃ
k pdc  sidc Cs
  V dc
Figure 3.15: DClink loop control of the GSC in S-domain.

Reference Out put DC
DC voltage k p  ki
Ts I dgref 3m I os
0.0136 voltage
+ 1
1
z 2 2 z 1
-
PI Plant
Figure 3.16: The closed loop DC link voltage control in z-domain.

4
0 0 0 0.004 0.217 4.6
80
3
Magnitude (dB)
0
60
2 40
0
1 20
Imag Axis
G.M.: -Inf dB
Freq: 0 rad/sec
45 40 35 30 25 20 15 10 5 0
0 Stable loop
-90 P.M.: 85.8 deg

Freq: 46.7 rad/sec
-1
0
Phase (deg)
-120
-2
0 -150
-3
0 0 0 0.004 0.217 4.6

-4 -180
-1 0 1 2
-45 -40 -35 -30 -25 -20 -15 -10 -5 0 10 10 10 10
Real Axis Frequency (rad/sec)
(a)
Step Response
1.4
1.2
1
Amplitude
0.8
0.6
0.4
0.2
0
0 0.2 0.4 Time0.6
(sec) 0.8 1 1.2
(b)
Figure 3.17: DC-Link controller tuning the PI parameters: (a) root locus, open loop
Bode plot; (b) step response
59
3-4 Drive Train Model

The drive train system could be approximated by a two-mass mechanical spring and
damper model connected by a flexible shaft characterized by stiffness and damping
coefficients that modeled on the low speed shaft, while the high speed shaft is assumed
to be stiff. This gives more accurate responses of the wind turbines during fluctuating
wind conditions, as shown in figure 3.18 [130]. This dynamical model is widely
accepted as expressing the dynamical behavior of the drive train in this area of
research.
The inertia of the low speed shaft comes mainly from the rotating blades and the inertia
of the high speed shaft from the generator. The mass of the gearbox itself is
insignificant and neglected. Stiffness and damping of the shaft are combined in one
equivalent for stiffness and damping placed at the low speed side The aerodynamic
torque and the generator reaction torque represent the input quantities for the
model while changes in turbine‘s rotor speed and generator speed are the output. From
figure 3.18 the variation in the angular generator speed and angular rotor speed
respectively expressed as [38, 126]:
�
− = (3.103)
�
− ℎ = (3.104)
where:
ℎ
= , and ℎ = ℎ Δ� + ℎ Δ� (3.105)
�
=� (3.106)
�
=� (3.107)
Therefore, after substitution the above equations are simplified to derive the torque
equations given by the following:
� � �
= + ℎ (� − )+ ℎ (� − ) (3.108)
� ℎ � ℎ �
− = − (� − )− (� − ) (3.109)
Equations 3.108 and 3.109 are used to describe the drive train dynamics as follows:
� � �
= − ℎ (� − )− ℎ (� − ) (3.110)
60
� � �
= − + ℎ (� − )+ ℎ (� − ) (3.111)
Figure 3.18: Two-mass-model for the drive train [38].
3-5 Wind Turbine Power Characteristics

The wind turbines convert aerodynamic power into electrical energy. In a wind turbine,
two conversion processes take place. The first converts the aerodynamic power that is
available in the wind into mechanical power. The next process converts the mechanical
power into electrical power as in figure 3.19 [130].
Aerodynamic Mechanical Power Electrical Power
Induction Electrical
Wind Rotor Model Gear box
Generator Grid
Figure 3.19: Block diagram of wind energy conversion system.

3-5-1 Annual Wind Distribution
Since wind velocities are never the same throughout the year, therefore an important
factor that characterizes the output power of a wind turbine is the annual wind speed
associated with wind distribution as shown in figure 3.20. The latter can be formed
using the statistical concept of the Weibull probability density function [131, 132]:
61
− � �
−
� = ∗ (3.112)
where v is the wind speed range where the wind distribution is evaluated, and and �
are parameters of the distribution. These values are taken from the classification given
by the International Electro-technical Commission (IEC). The average wind speed can
be obtained as follows:
∞
� = ∫ �. � . � (3.113)
700
600
500
Time (hourse/year)
400
300
200
100
0
0 5 10 15 20 25
Wind speed (m/s)
Figure 3.20: Annual wind distribution ( =11.38, � =2) [Appendix E-3].
3-5-2 Aerodynamic Characteristics of the Wind Turbine
Wind turbines convert the kinetic energy present in the wind into mechanical energy
by means of producing torque. Since the energy contained by the wind is in the form of
kinetic energy, its magnitude depends on the air density and the wind velocity. The
wind power developed by the turbine is given by the following equation [130]:
= � (3.114)
It is not possible to extract all of the kinetic energy of the wind, since this would mean
that the air would be standing still directly behind the wind turbine [86]. Therefore, the
power captured from the wind turbine is obtained as follows [98, 124]:
= �, � � (3.115)
The power coefficient ( ) gives the fraction of the kinetic energy that is converted
62
into mechanical energy by the wind turbine, while the air density ( ) depends on
factors such as plane altitude and air temperature and may vary between 1.07 � /
in hot and high altitude region to 1.395 � / in cold and low-lying regions. The
power coefficient is the ratio between the power extracted by turbine relative to the
variable wind power and is a function of tip speed ratio (TSR) and the blade pitch
angle which is the angle between the chord of the aerofoil section and the plane of
rotation and is also called the setting angle. Broadly, different mathematical formulae
have been used in previous studies to calculate the power coefficient and one of these
is shown below [86, 132, 133]:
8.
. −
�, � = . − . �− . � − . ��
(3.116)
��
where:
� = . (3.117)
�− .
−
+
Theoretically the maximum achievable power factor coefficient is 59.26% (Betz limit)
while in practice it can reach to 45% which is below the theoretical limit due to
inefficiencies and losses related to different configurations of airfoil profile, rotor tip
end losses, whirlpool losses and rotor blade number losses as shown in the typical
characteristic of the power coefficient with respect to the tip speed ratio (TSR λ) and
blade pitch control in figure 3.21 [134].
0.5
0.4
Power coefficent (%)
0.3
0.2
0.1
0
30
25
20
15
10
5 15
10
0 5
Blade pitch angle (deg) 0
Tip speed ratio

Figure 3.21: Power coefficient curve verses tip speed ratio and pitch blade angle
[see Appendix E-2].
63
3-5-3 Tip Speed Ratio (TSR) of the Wind Turbine
The relationship between the wind speed and the rate of rotation of the rotor is non-
dimensional factor named the tip speed ratio (TSR) which is described as follows
[132]:
�
�= (3.118)
The number of rotor blades of the wind turbine is proportional to the optimal tip
speed ratio, which is approximated empirically determined as follows [134]:
� ≈ (3.119)
It is clear from equation 3.115 that the maximum available power from the wind occurs
with adjusting the power coefficient which is a function of the λ. Hence, from the
curves shown in figures 3.22, explain that the maximum efficiency can be achieved at
all wind velocities with the optimal tip speed ratio. It is desirable to adapt the rotational
angular shaft speed with varying values of wind velocity to keep tip speed ratio satisfy
the optimal value associated with the maximum power capture.
Figure 3.22: Power characteristics of wind turbine verses rotational speed [134].
For each wind turbine, its ( − � − �) curves depend on the design of the blade as
shown in the 3-D curve in figure 3.21. At any specific wind speed, there is a unique
rotational speed of the wind turbine that satisfies the maximum power coefficient
( ) associated with maximum mechanical power which can be expressed as:
64
= �
� (3.120)
Thus the maximum mechanical power that can be extracted from wind is
proportional to the cube of the rotor speed, as follows [7]:
∝� (3.121)
If the wind speed is below the rated value, the pitch angle is deactivated (� is fixed)
while the wind turbine operates in variable speed mode with rotational speed adjust by
DFIG speed control or active power control. On the other hand when the wind speed is
increased above the rated value, at this condition the pitch angle control is activated
with an increase in � to reduce the mechanical power extracted from the wind [135].
3-6 Test System
The simulation configuration was set up to execute a variable-step solution using an

ode45s (dormand-prince) type solver, with a relative tolerance of 0.001. The solver
was chosen to best deal with the combination of discrete and non-discrete variables in
conjunction with the moderately stiff problem of grid fault application.
A screen-shoot for the DFIG system in the simulink model is shown in figure 3.23.
This includes: the generator, a bidirectional power electronic converter, a digital
controller with cascaded feedback control schemes, crowbar, a two-mass drive-shaft
model to represent the wind turbine mechanical dynamics, a simplified wind turbine
controller and a simplified network model. The wind turbine in this study is on
1.5MW, with a horizontal axis, and three-bladed with blad radius 35.25 m, upwind
wind turbine with pitch control. A two-pair pole DFIG using back-to-back PWM
voltage source converters in the rotor winding circuit, is adopted in the wind turbine
with carrier frequency of 5 KHz and average voltage 398.74 V, and 0.9 is the value of
the setting factor which calculated from the equation (setting factor = � � ×√
).
The parameters used in the DFIG based wind turbine model are close to that of a
commercial wind turbine as illustrated in appendix A.
The rotor-circuit crowbar model is connected by the rotor winding of the DFIG. When
activated, the DFIG rotor voltage was rendered zero, since its short circuit the rotor
winding to isolate the RSC. When disengaged, the crowbar had no effect on the rotor
65
circuit and the RSC voltage was passed unchanged to the rotor of the DFIG. In the
simulation assume an infinitely stiff grid is assumed which entails an ideal voltage
source. Specific voltage dips or swells can be applied to the DFIG system with specific
period time during system operation. The modelled voltages were applied directly to
the stator connection of the DFIG model.
Drive train model
Aerodynamic model
PLL
Grid fillter
DC link
PWM
Generator
PWM
Crowbar
GSC
RSC
Figure 3.23: Screen-shot of the DFIG system in the MATLAB/SIMULINK® model.
66
3-7 Phase Locked Loop (PLL) System Performance

Phase locked loop (PLL) performance testing is essential to demonstrate the ability of
the PLL in the DFIG-based wind turbine system to track changes in grid frequency in
different operation conditions. Figure 3.24 shows the response of the PLL after a 40Hz
step frequency reduction during period 0.5-0.6 seconds in the stator voltage signal. In
this case, the output angle is shown that the PLL ideally fast control to return tracking
the desired angle after the frequency variation end. It is noted that the system behaved
as a nearly linear system for solely frequency inputs, and also it can be illustrate the
response of the PLL output with higher frequency input produces transients of a similar
shape to this changes. As shown in figure 3.25 when the frequency of the grid increase
to 40Hz, the measured values of the angle is effected and tracking ability is lost during
this scenario.
The PLL responses were also investigated in condition where the input voltage to the
DFIG is polluted with harmonics. These harmonics are produced using
MATLAB/SIMULINK® programmable voltage source during 100ms. If the input
voltages have harmonics, the PLL responses might be influenced in terms of transient
performance or steady-state error and the output signal of the PLL may include an
undesirably high harmonics level. The PLL system is investigated here with the
injection two harmonics components superimposed on the fundamental components
during the no fault condition in order to show the system robustness. Figure 3.26
illustrates the PLL responses for voltage depression with an input voltage containing
3rd order harmonic parasitic with the 2nd order. It can be concluded from figure 3.26
that the responses are only slightly influence if input harmonics are present. The PLL
has a reasonable error output to track the desired angle which is also observed in the
second scenario as shown in figure 3.27 when 3rd order and 7th order harmonic were
injected. These results highlighted to less reversible effects with this harmonics which
will produce a DC error signal and this will induce a steady-state error in the voltage
controller and also in the phase angle controller.
It is desirable that a PLL has the ability to follow phase angle and frequency during
periods of voltage sags and harmonics and to exhibit a fast response during the fault
recovery stage transient which would demonstrate the PLL’s robustness. Figure 3.28
shows the results when PLL was tested with reduced voltage conditions. As the voltage
67
was reduced to 0.2pu for 100ms between 0.5-0.6 sec and harmonics components 3rd
and 7th were injected, the PLL was able to track the desired angle in less than 5ms.
Figure 3.29 shows the responses with a 40% voltage swell for 100ms with 3rd and 2nd
harmonic components superimposed with the fundamental voltage signal. Figure 3.30
shows the PLL system test for simultaneous application of the DFIG when a severe
single phase fault subjected with inject 3rd and 2nd harmonics components. It is seen
that the PLL system is able to follow the reference signal even there is drift error
during the fault and it took 20ms after fault clearance to return to tracking the angle
correctly. It is concluded that the results for the PLL are acceptable in the DFIG-based
wind turbine system application in this study. In addition, it is clear that this PLL is
robust and has the ability to track the desired values during different operating
conditions of harmonics, voltage sag, and voltage swell.
800 a b c
600
400
200
Vsabc (V)
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
7
(a)
Demand Measured
6
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
Figure 3.24: Instantaneous measurements of (a) grid voltage and (b) PLL
response after 40Hz frequency step change.
68
800 a b c
600
400
200
Vsabc (V)
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
Demand Measured
6
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
Figure 3.25: Instantaneous measurements of: (a) grid voltage and (b) PLL
response after +40 Hz frequency step change.
800 a b c
600
400
200
Vsabc (V)
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
6
Demand Measured
5
angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
Figure 3.26: Instantaneous measurements of: (a) grid voltage; (b) PLL
response during 3rd order harmonic (+ve sequence, 20% amplitude, and -
25degree phase angle step ) parasitic with the 2nd order harmonic ( +ve
sequence, 10% ampl itude, and 0 degree phase angl e step).
69
800 a b c
600
400
200
Vsabc (V)
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
Demand Measured
6
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
Figure 3.27: Instantaneous measurements of: (a) grid voltage and; (b) PLL
response during 3rd order harmonic (+ve sequence, 20% amplitude, and -
25degree phase angle step) and; 7th order harmonic ( +ve sequence, 2%
amplitude, and 0 degree phase angle step).
800 a b c
600
400
200
Vsabc (V)
0
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
6 Demand Measured
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
response during 80% voltage dip, having harmonics ( 3rd (+ve sequence,
20% amplitude, and -25degree phase angle step) and 7th ( +ve sequence,
2% amplitude and 0 degree phase angle step)).
70
800
a b c
600
400
200
Vsabc (V)
0
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
Demand Measured
6
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
response during 40% voltage swell, having harmonics (3rd (+ve sequence,
20% amplitude, and 25 degree phase angle step) and 2nd order harmonic
(+ve sequence, 10% amplitude, and 0 degree phase angle step)).
800 a b c
600
400
200
Vsabc (V)
0
-200
-400
-600
-800
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(a)
7
Demand Measured
6
5
Field angle (rad)
0
0.4 0.45 0.5 0.55 0.6 0.65
time (seconds)
(b)
Figure 3.30: Instantaneous measurements of (a) grid voltage and (b) PLL
response single phase fault, having harmonics ( 3rd (+ve sequence, 20%
amplitude, and -25 degree phase angle step) and 2nd order harmonic ( +ve
sequence, 10% amplitude, and 0 degree phase angle step)).
71
3-8 DFIG System Model Confirmation

The objective of this section is to confirm the model of the doubly fed induction
generator (DFIG)-based wind turbine in this study using MATLAB/ SIMULINK®.
This process of double checking is conducted by comparing the performance of DFIG
against the base model in the PLECS® platform. The DFIG model in PLECS® is built
and already exists in standalone simulation platform.
There are several different simulation software products available for electrical
system and power production simulations. Some of these simulation tools have similar
qualities and were developed and delivered by different companies, but on the other
hand, different software packages may have different levels of modeling and
simulation precision, and may serve different purposes. PLECS® (Piecewise Linear
Electrical Circuit Simulation) is a Simulink toolbox developed by PLEXIM for system-
level simulations of electrical circuits [136]. It is especially designed for power
electronics but it can be used for any electrical network. The SIMULINK model is
ideally suited for the simulation of control system. Therefore, SIMULINK is also a
convenient tool for the design of closed loop controlled electrical systems. PLECS®
enhances Simulink with the capability to simulate electrical circuits directly. There is
also a standalone version of PLECS® that allows the simulation of electrical circuits
and control systems directly within the PLECS® package. The complete PLECS®
schematic of the DFIG plant that employed in model confirmation is shown in Figure
3.31. The system components are available in platform’s library for different physical
domains, including electrical, magnetic, mechanical, and control systems. The
electrical and mechanical parameters of the DFIG are listed in table 3.2 where the rotor
parameters have been converted to the stator-side using the turn’s ratio. To evaluate
and compare the proposed model of DFIG-based wind turbine theoretically, several
measurements have been implemented to the system during normal operating
conditions with the MATLAB® and PLECS® simulation platforms. The purpose of
double checked between these two SIMULINKs are to assess the provided
example/default parameters, parameterized generic model operation/response in
behavior comparison and to evaluate in which level of complexity specific wind
turbines are necessary to be modelled for certain simulation purposes, e.g. what kind of
influence this would have in the overall wind turbine model accuracy under different
simulation circumstances.
72
Table 3.2: Electrical and mechanical parameters of the doubly-fed induction machine
used in the PLECS platform [136].
The electrical parameters The mechanical parameters
Figure 3.31: The DFIG wind turbine model test system in PLECS®.
By running the two models of the DFIG based wind turbine in MATLAB® and
PLECS®) at a constant wind speed of 5 m/sec by fixed the outer loop control of the
rotor-side converter in normal grid condition during 2 seconds. It will easy to illustrate
that DC-link regulator has good agreement results when compared the PLECS® with
73
the MATLAB®. The ability to tracking the reference signal is shown in figure 3.32.
Figure 3.33 presents the behaviour of the inner loop control of the rotor-side converter
of the DFIG system in terms of direct and quadrture component. Figure 3.33a shows
that the direct current component have fast and closely regulate around the set value in
MATLAB® than PLECS®. On the other hand the q- component of rotor current in
PLECS® shows better of tracking the reference value than MATLAB® as shown in
figure 3.33-b. The generator’s speed in DFIG MATLAB® model has acceptable
outcomes which approximately match the measurements in the PLECS® model as
shown in figure 3.34a. It can be observed in figure 3.34b that the stator reactive power
regulation exhibits more linear behaviour in MATLAB® when compared with the
PLECS® system and its tracking around the setting values. It can be concluded from
the comparison results that the control concepts implemented for normal operation is
as expected in the MATLAB® model. The simulation measurements for the DFIG
wind turbine model in MATLAB® and PLECS® show good agreement. In addition, it
be noticed that the results for PLECS® have less oscillation because in the demo
model the designer used filter for the stator current, rotor current and stator voltage as
shown in figure 3.35 to eliminate parasitic harmonic components.
1210
Meas MATLAB
Meas PLECS
1205
Ref
DC link voltage regulator (v)
1200
1195
1190
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 3.32: Instantaneous comparison measurements of DC link voltage

regulator in MATLAB® and PLECS®.
74
4000
Meas MATLAB
3000
Meas PLECS
2000 Ref
Idr ref and meas (A)
1000
-1000
-2000
-3000
-4000
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
time (seconds)
(a)
4000
3000
2000
1000
Iqr ref and meas (A)
Meas MATLAB
-1000
Meas PLECS
-2000 Ref
-3000
-4000
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
time (seconds)
(b)
Figure 3.33: Instantaneous measurements of: (a) d-component rotor current in
controller MATLAB® and PLECS®; (b) q-component rotor current controller
in MATLAB® and PLECS®
75
220
200 MATLAB
PLECS
180
160
wgen (rad/sec)
140
120
100
80
60
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
5
(a)
2 x 10
1.5 Meas MATLAB

Meas PLECS
1
Ref
Qsref, Qs meas(VAR)
0.5
-0.5
-1
-1.5
-2
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
(b)
Figure 3.34: Instantaneous measurements of: (a) Generator speed in MATLAB®

and PLECS®; (b) Reactive power control in MATLAB® and PLECS®.
76
Inside the
block
Inside the
block
Figure 3.35: Inside the block of RSC and GSC control of PLECS® DFIG model.
3-9 Simulation Results of Modeling the DFIG System

This section includes the results for the simulation model of the DFIG-based wind
turbine which has been built in the MATLAB/SIMULINK® platform from the
theoretical analysis presented earlier with the grid-connected case. The test is applied to
the DFIG system with sensored vector control and subject to constant wind speed. The
wind profile signal which has a rated wind speed of 8 m/s is applied to drive the
mechanical parts of the DFIG wind turbine system; this wind speed signal is used for a
realistic, reliable and accurate simulation analysis.
Step changes occurred in the active component currents in the period between 0.2-1
seconds as shown in figures 3.36-3.39. This corresponds to the stator and rotor three-
phase currents in dq coordination. This is clear that the step responses of the q axis
rotor current component together variation of stator active reference power. This
change in current clearly displays the fundamental principle of field orientation, which
is a sudden step change from one steady state condition to another.
Figure 3.40 illustrates the grid voltage where the stator of the DFIG as well as the grid-
side converter is connected. Under the circumstances of power exchange between the
DFIG and the grid, the three-phase grid-side converter current is exchanged with the
grid as shown in figure 3.41 with zoom the scope to the period of time between 0.4-
0.45 sec. This current is controlled according to the vector control principles of GSC as
77
presented earlier in this chapter. The direct and quadrture rotor voltage output-fed from
the RSC are illustrated in figures 3.42 and 3.43. The dynamic performance of the DFIG
is depicted in figures 3.44 and 3.45 were both active and reactive power values
accurately track their command signals, which means that the controllers have
successfully mitigated the impact of the coupling effect in the tracking of commands
signals. This confirms the robustness of the PI stator active power. The output voltage
of the sinusoidal pulse width modulation PWM of carrier frequency 5 kHz in the RSC
and GSC is shown in figures 3.46 and 3.47. The simulation results in figures 3.48 show
the tracking performance of the voltage regulator in the outer loop control of the GSC.
The quadrature and direct current control in the inner loop of the GSC is illustrated in
figure 3.49. The q-axis current reference of the GSC controller is set to zero, as
mention before in the modeling part of the GSC so that the DFIG-based wind turbine
reactive power demand is met by the RSC control and its obviously clear that the
measured values of q-axis current succeed to track the setting value as shown in figure
3.49a. Figure 3.49b shows the superior performance of the direct-current vector control
of the GSC.
2500
Idr
Iqr
2000
1500
Idqr (A)
1000
500
-500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.36: Instantaneous measurements of the rotor current in

synchronous reference frame � .
78
2500
a b c
2000
1500
Three phase rotor current (A)
1000
500
-500
-1000
-1500
-2000
-2500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.37: Instantaneous measurements of the three phase rotor current.
500
Ids
Iqs
-500
Idqs (A)
-1000
-1500
-2000
-2500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.38: Instantaneous measurements of the stator current in

synchronous reference frame � .
79
2500
a b c
2000
1500
three phase stator current (A)
1000
500
-500
-1000
-1500
-2000
-2500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.39: Instantaneous measurements of the three-phase stator current.
800
a b c
600
400
200
Vs abc (V)
-200
-400
-600
-800
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time (seconds)
(a)
800
Vds Vqs
700
600
500
Vdqs (V)
400
300
200
100
-100
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time (seconds)
(b)
Figure 3.40: Instantaneous measurements of the grid voltage: (a) three phase; (b)
in synchronous reference frame coordination.
80
20
a b c
15
10
5
IGSC abc (A)
-5
-10
-15
-20
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time(seconds)
Figure 3.41: Instantaneous measurements of three phase GSC current.

1000
800
600
400
Vqr RSC (V)
200
-200
-400
-600
-800
-1000
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time(seconds)
Figure 3.42: Instantaneous measurements of quadrature voltage component of the RSC.
1000
800
600
400
Vdr RSC (V)
200
-200
-400
-600
-800
-1000
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time(seconds)
Figure 3.43: Instantaneous measurements of direct voltage component of the RSC.
81
5
x 10
2
Ref
Meas
0
-2
Ps ref Ps meas (w)
-4
-6
-8
-10
-12
-14
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.44: Instantaneous measurements comparison reference and measured stator

active power.
6
x 10
5
Ref
4 Meas
3
Qsref, Qs meas(VAR)
-1
-2
-3
-4
-5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 3.45: Instantaneous comparison reference and measured stator reactive power.
82
1500
1000
500
Vab RSC (V)
-500
-1000
-1500
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time (seconds)
Figure 3.46: Instantaneous measurements phase to phase voltage of rotor voltage.
1500
1000
500
Vab GSC(v)
-500
-1000
-1500
0.4 0.405 0.41 0.415 0.42 0.425 0.43 0.435 0.44 0.445 0.45
time (seconds)
Figure 3.47: Instantaneous measurements phase to phase grid side converter output
voltage.
1215
Ref
Meas
1210
DC link voltage regulator (v)
1205
1200
1195
1190
1185
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time(sec)
Figure 3.48: Instantaneous measurements DC-link voltage regulator in the GSC.
83
-3
8 x 10
Ref
6 Meas
4
Iqg ref meas (A)
-2
-4
-6
-8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
(a)
5
Ref
Meas
Idg ref and meas(A)
-5
-10
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
(b)
Figure 3.49: Instantaneous comparison reference and measured current regulator in the
inner loop GSC control with reference frame oriented along the grid voltage vector
(a) � (b) � .
3-10 Conclusions
The modelling of the DFIG-based wind turbine system is important for the present
research work on sensorless vector control, improvement of the fault ride-through
capability of the system and its interaction with the grid in normal and fault conditions.
This chapter describes a model of wind turbines with the DFIG, developed in the
dedicated power system analysis tool Matlab/Simulink, which includes the wind
model, the aerodynamic model, the mechanical model, the DFIG model, and the PWM
voltage source converter model. The conventional vector control of the DFIG based
84
wind turbine system is built and tested to illustrate the excellent performance of the
system during normal grid condition. The decoupled control of the active and reactive
power is easily obtained to improve the dynamic behaviour of the system and the
quality of power injected to the grid. Simulation results have proven the precision of
the obtained model and the control strategy.
85
Chapter 4: Rotor Position Estimation using MRAS
CHAPTER FOUR
86
CHAPTER FOUR
ROTOR POSITION ESTIMATION USING THE MODEL

REFERENCE ADAPTIVE SYSTEM (MRAS)
4-1 Introduction
The transformation of voltages and currents from the a-b-c reference frame to the d-q
reference frame is a basic requirement of the vector control. Critical information in the
transformations includes the position of the rotor winding axes with respect to the stator
winding axes and the stator voltage angle. The rotor position is obtained by an encoder
incremental or absolute position. The stator angle is obtained by a hardware or software
phase lock loop (PLL). The rotor position signal is essential to give accurate vector control
behaviour of DFIG-based variable speed wind turbines. Incorrect rotor position information
will not only harm control performance but will also affect system stability. Several
computation blocks in the rotor-side converter control utilize the rotor position signal in the
vector control scheme to calculate � and � using the Park transformation, � ∗ and � ∗
using the inverse Park transformation, and rotor speed. The rotor position information
needs to be known at every point in time in order to decouple control to be performed.
Therefore, one of two solutions can be used to provide accurate information about the
rotor’s position. Firstly is a mechanical encoder can be used to measure the rotor position,
and this signal is continuously sent to the rotor-side converter control loops. Secondly, in
the sensorless approach rotor position is calculated from the electric quantities associated
with rotor and stator.
The latter rotor position detection method is an attractive technique which contributes to the
enhanced reliability of DFIG-based wind turbines. This is use of the mechanical
speed/position sensor coupled with the machine rotor shaft has several drawbacks in terms
of robustness and reliability since an encoder is sensitive to external perturbations such as
electromagnetic interference. Moreover, sensors are often subject to high failure rates in
harsh environments such as extreme ambient temperature, super high-speed operation, and
other adverse or heavy load conditions. Furthermore, especially in small to medium size
machines, the sensor cost is considerable and cabling and maintenance represents extra
87
costs. On other hand, the computation algorithm for rotor position using sensorless method
must be set to be stable at a wide range of operating conditions and should not require
initial condition knowledge. In addition, the flux equations are dependent upon the
generator parameters, and variations in winding temperature can lead to changes in rotor
and stator resistance and inductance.
This chapter presents the modelling of two model reference adaptive system (MRAS)
schemes adopted to conduct the rotor angle estimation in the vector controller. These are
based on rotor current and rotor reactive power with more complement analysis where the
first scheme has been used in many investigations for position detection while the second is
utilized in previews study as speed estimator with different orientation analysis.
4-2 Sensorless Control of DFIG system Using MRAS Observer

The model reference adaptive system (MRAS) represents one of the most attractive
solutions for the sensorless control of AC drives due to its simplicity and lower
computational complexity [137]. MRAS schemes have been employed for position
detection [79, 81, 84, 138] or speed estimation [78, 85, 139] which are investigated using
a PI or Hysterias controller in DFIG systems which implemented in different stator and
rotor reference frames. The MRAS method is utilized to estimate the rotor position signal
that implement to the orientation control of the RSC of the DFIG. In an MRAS, as
depicted in Figure 2.4, an adaptive (adjustable) model (AM) and a reference model (RM)
are connected together in parallel. The output of the adaptive model is expected to
converge with the output of the reference model in the presence of a proper adaptation
mechanism. The usual MRAS estimates speed or position using the outputs of these two
different AM and RM models. The AM is position-dependent while the RM is not related
to the estimated position [80]. Based on Popov’s hyperstability criterion, the difference
between the RM and the AM outputs is used to derive a suitable adaptive mechanism to
generate the estimated speed used so as to minimize the error between the reference and
estimated values and therefore the stability of the MRAS. The convergence of position
estimation can be guaranteed according to this theory [83, 140, 141]. Thus, if the tracking
errors between the output states of the AM and RM are close to zero. Then the rotor
position can be obtained by using an integrator for the output signal of the PI controller.
88
4-2-1 Rotor Current-Based MRAS (RCMRAS)

Rotor current-based MRAS represents one of the best types of MRAS models for grid-
connected or stand-alone DFIG operation [79, 81, 142]. In this method, the output of the
RM is the rotor current measured by transducers in the rotor reference frame, while the
estimated rotor current is determined using the stator voltage and stator current in the
stationary reference frame to compute the stator flux derived from equations 3.8 and 3.9
equations:
� =∫ � −�� (4.1)
� = ∫(� −�� ) � (4.2)
From the substitution of equations 4.1 and 4.2 in equations 3.16 and 3.17 respectively and
rearrangement, the formula for rotor current are estimated in the rotor ( ) frame:
∫� − −� − �
� = (4.3)
��
∫� − −�
− �
� = ��
(4.4)
Hence (� ) is the rotor’s angular position. The cross-product between the output of RM
and that of the AM is driven to zero by the adaptation mechanism in order to estimate the
rotor speed, which is integrated to get the rotor angle that will be fed back to the AM as
shown in figure 4.1. The error between the RM and AM is the cross-product between the
measured rotor current and the estimated current from equations 4.3 and 4.4:
� = � ́ .� − � ́ .� (4.5)
The adaptive mechanism is then given by:
�́ = � . � + ∫ � . �. � (4.6)
Minimizing the error between the rotor current output from the AM and RM is conducted
by adjusting the rotor position. The rotor’s electrical position is determined by the free
integral of equation 4.6 as follows:
�́ = ∫ �́ . � (4.7)
89
Irαβ
×
+ Theta
PI 1/s mod
-
Vsαβ
Isαβ AM ×
2π
 r
Figure 4.1: Simulink model of rotor current based MRAS (RCMRAS).
4-2-2 Rotor Reactive Power Based MRAS (QRMRAS)

Figure 4.2 depicts the basic structure of the QRMRAS observer. The rotor reactive power
error is computed from the difference between the outcomes of the RM and AM used to
generate the rotor quantity from the adaptation mechanism, and the output is subjected to
an integrator to estimate the rotor position [139].
The expression for instantaneous rotor reactive power is:
� = �̅ ⊗ �̅ = � .� − � .� (4.8)
It can be assumed that the inverter output voltage (� �� ) exactly follows the
reference signals ( �́ �� ́ ), and so equation 4.8 can be expressed as:
� = �́ .� − �́ . � (4.9)
By substituting equations 3.10 and 3.11 in equation 4.8 following results:
� = (� . �� − � . �� ) − � ́ . (� � −� � ) (4.10)
It should be noted that the above expressions of � are free from stator and rotor
resistance, which is a notable merit of this MRAS scheme.
The difference between the outputs of the RM and AM will form an error given as:
�=� −� (4.11)
The error signal is passed through the adaption mechanism (PI controller). The output of
the PI controller is the estimated rotor speed, which is then used to tune the adjustable
model such that the error converges to zero. The error signal is fed to the adaptation
mechanism block, which yields the rotor mechanical speed, as follows:
90
�́ = � . � + ∫ � . �. � (4.12)
In the MRAS method, it is possible to reduce the reactive power error by adjusting the
rotor position. The rotor electrical position is determined by the free integral of equation
4.12:
�́ = ∫ �́ . � (4.13)
Vrαβ
Reference
Qr
Irαβ RM
Theta
+
PI 1/s mod
-
Isαβ

AM Adaptive
r
Qr
2π
ώ
Figure 4.2: Simulink model of rotor reactive power based MRAS (QRMRAS).
4-3 Simulation Results

The performance of the estimation of the rotor position using the MRAS schemes has
been investigated in terms of its response against wind speed variations and changes in
machine parameters to test which of the two schemes is more robustness and less sensitive
to parameter changes during operation of decoupling active and reactive power control.
The set of tests conducted with the DFIG system, constant and ramp variations in
magnetizing inductance, stator resistance, and rotor resistance of the machine were
increased to 30% of the nominal values after 2 seconds of running. The simulation results
shown in figure 4.3 indicate the tracking performance in detecting the rotor angle position
using the QRMRAS, and RCMRAS as compared with the actual values of steady state
with constant value of wind speed of 10 m/sec and constant generator parameters. In this
condition, Figure 4.4 shows the comparison between the rotor reactive power that
computed in the reference and adaptive model inside the QRMRAS subsystem where the
difference between them subjected in the controller of the MRAS to detect the angular
position of the rotor. To validate the accuracy of the RCMRAS and the QRMRAS, figure
4.5 shows the difference between the actual and estimated rotor angles in each estimation
91
scheme. It’s clear that the QRMRAS gives less pulsating error as compared with
RCMRAS.
The second scenario used to test the robustness of the estimation of the DFIG rotor angle
position is implemented by applying step changes to wind speed as shown in figure 4.6.
The variation of the wind speed will increase the generator speed from 146 rad/sec to 156
rad/sec after 1.5 sec as illustrate in figure 4.7. The ability of QRMRAS and RCMRAS to
track the measured rotor angle during this case is highlighted in figure 4.8.
7
QRMRAS
Meas
6 RCMRAS
QRMRAS
5 RCMRAS
Rotor angle (rad)
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.3: Comparison results of the measured and estimated rotor position computed in
normal operating conditions.
5
x 10
Ref
Qr Ref rnd meas in QRMRAS (VAR)
1.5
Meas
1
0.5
-0.5
-1
-1.5
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2

time (seconds)
Figure 4.4: Instantaneous comparison of reference and measured values of rotor reactive
power in the QRMRAS sub-system computation.
92
6 RCMRAS
QRMRAS
4
and adaptive model in MRAS
Difference between reference
-2
-4
-6
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.5: Instantaneous error between reference and adaptive model output
in rotor RCMRAS and QRMRAS.
15
14
13
12
Wind speed (m/sec)
11
10
5
1.0 1.2 1.4 1.6 1.8 2.0
time (seconds)
Figure 4.6: Variable wind speed signal.

159
158
157
156
155
154
153
152
wgen (rad/sec)
151
150
149
148
147
146
145
144
143
142
141
140
1.0 1.2 1.4 1.6 1.8 2.0
time (seconds)
Figure 4.7: Instantaneous results for generator speed (rad/sec).
93
7
Meas
6
QRMRAS
5 RCMRAS
Rotor angle (rad)
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.8: Comparison results of the measured and estimated rotor position in step
change wind speed operation condition.
The simulation results shown in figure 4.9 indicate the tracking performance in detecting
the rotor angle position using the QRMRAS, and RCMRAS as compared with the actual
values when the generator is running at 1720 rpm with a step change in the value of � of
a 30% increase after 1.5 second of operation. The QRMRAS as noted is less sensitive to
changes in the generator magnetizing inductance as illustrated in the zoom scope of figure
4.9 where the rotor detection in RCMRAS suffers deviate during this condition.
QRMRAS is more accurate in tracking in actual rotor position than the RCMRAS as a
results of absent the integration in the formula of computation of the reactive power in
AM and RM and it independent on the rotor or stator resistance.
7
Meas
6
QRMRAS
RCMRAS
5
Rotor angle (rad)
4 4
3 0
1.485 1.49 1.495 1.5 1.505 1.51 1.515 1.52 1.525 1.53 1.535
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.9: Comparison results of the measured and estimated rotor position in condition
of 30% increase of. � .
94
Figure 4.10 illustrates the results for a scenario of voltage dip subjected to the DFIG system
with sensorless control without employing any protection system to the DFIG. The three-
phase voltage drop with high incremental oscillation of the three-phase stator current is
shown in figure 4.11. Figures 4.12 and 4.13 depict the instantaneous measurements of the
RM and AM outputs in the RCMRAS and QRMRAS sub-systems respectively, and it’s
clear that the RM output in RCMRAS suffers from phase shift and high oscillation when
compare to the RM output meanwhile in QRMRAS the controller succeed to reducing the
error between the RM and AM shown in figure 4.14. This condition will lead to a minor
cumulative drift in the rotor angle estimation RCMRAS than in QRMRAS in the voltage
dip duration as illustrated in figure 4.15.
800
a b c
600
400
200
Vs abc (v)
-200
-400
-600
-800
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.10: Instantaneous measurement of three-phase voltage.

6000
a b c
4000
2000
is abc (A)
-2000
-4000
-6000
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.11: Instantaneous measurement of three-phase current during voltage dip

condition without protection.
95
x 10
1
alpha
Refrence rotor current (A)

0.5 beta
-0.5
-1
1 4 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
x 10
1
alpha
Adaptive rotor current (A)
0.5 beta
-0.5
-1
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.12: Instantaneous comparison between outputs of reference and adaptive models
in RCMRAS sub-system computation in fault condition.
6
x 10
1.5
Ref
Qr ref and meas in QRMRAS (VAR)
Meas
1
0.5
-0.5
-1
-1.5
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.13: Instantaneous comparison of reference and measured values of rotor

reactive power in the QRMRAS sub-system computation in fault condition.
40
RCMRAS
QRMRAS
20
Differnce between refrence and
adaptive model in MRAS
-20
-40
-60
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time(seconds)
Figure 4.14: Instantaneous error between reference and adaptive model output in
RCMRAS and QRMRAS in fault condition.
96
6 Meas
QRMRAS
5 RCMRAS
Rotor angle (rad)
0
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
time (seconds)
Figure 4.15: Comparison results of the measured and estimated rotor position in
voltage dip condition.
The integrated error index which is the integral of time multiplied by absolute error
(ITAE=∫ � | � | �) and the integral absolute error (IAE=∫| � | �)) are used criteria to
measure the quality of the system’s response in order to assess which of these methods
gives response which corresponds to the minimum amount [143]. The ITAE performance
index has the advantage of producing smaller overshoots and oscillations than the IAE
performance index. In addition, the former is more sensitive since it has the best
selectivity [144]. A systematic evaluation of rotor estimation performance has been
conducted for different operating conditions by means of the use of ITAE and IAE index
as shown in table 4.1 when machine parameters, which indicates that the QRMRAS gives
the lower values of ITAE and IAE as compared with RCMRAS which mean that the
accuracy of QRMRAS is better.
Table 4.1: Results of performance index of rotor position estimation at t = 1.5 second with
ramp variation applied to � �� , � .
ITAE IAE
QRMRAS RCMRAS QRMRAS RCMRAS
30% 6.5459*10-3 0.066 0.0545 0.325
30% 6.5459*10-3 0.036 0.0545 0.202
30% �� , 6.5459*10-3 0.046 0.0545 0.1259
97
4-4 Conclusions
In this chapter two different observers utilized to estimate the rotor position using currents
and stator voltages were presented. The results of simulations are presented to show the
performance of the proposed techniques. The results show that they are well-suited for
vector control applications without needing a mechanical sensor in the DFIG. The results
demonstrate the feasibility of position tracking performance using the MRAS method.
Different operating conditions of constant wind speed, step change wind speed, voltage
dips of different severity, and changing in machine parameters to highlight the robustness,
fast response and insensitivity of the methods.
The MRAS algorithms which described for rotor position computation will not conduct
any integration operation in the equations of adaptive model and reference model when
the DFIG operate at region near or equal to synchronous speed. Thus in the mathematical
computation, saturation problem of the integrator terms in MRAS scheme cannot take
place when the where the slip frequency is low. Therefore, near synchronous speed
operation, the reference model and adjustable does not create any error in the said EMF
variable calculation due to integrator saturation problem.
98
Chapter 5: Behaviour Analysis of DFIG System in Grid Disturbances
CHAPTER five
99
CHAPTER FIVE
BEHAVIOUR ANALYSIS OF DFIG SYSTEM WITH GRID

DISTURBANCES
5-1 Introduction
In this chapter, the fault response of the doubly-fed induction generator (DFIG)-based
wind turbine is analyzed and summarized with and without the modified vector control
scheme. This shows its influence on the dynamic behaviour of the wind turbine system
when faults occur. In addition a comparison is conducted with different operating
condition.
5-2 Grid Code Requirements of DFIG system

Grid code requirements vary considerably from region to region and from system to
system. The grid codes in a certain country may only cover some specific grid code
requirements. The differences in requirements, as well as local traditional practices, are
caused by different levels of wind power penetration and different degrees of power
network robustness [11, 145]. Grid codes are put in place for all power generation
plants transferring power to the grid, in order to increase the efficiency of power
transfer and to reduce the damage caused to grid-connected devices [146, 147].
According to recent grid code requirements, one of the conditions for wind turbines
that are they must remain connected to the grid during grid disturbances. Moreover,
they must also contribute to voltage support during and after grid faults. The major
typical grid code requirements for the operation and grid connection of wind turbines
are listed as following [148, 149]:
1) Voltage operating range: Wind turbines are required to operate within typical grid
voltage variations.
2) Frequency operating range: Wind turbine generation is required to operate within

typical grid frequency variations.
100
3) Active power control: Several grid codes require wind farms to provide active
power control in order to ensure a stable frequency in the system and to prevent
overloading of lines. Also, wind turbines are required to respond with a ramp rate in
the desired range.
4) Frequency control: Several grid codes require wind farms to provide a frequency
regulation capability to help maintain the desired network frequency.
5) Voltage control: Grid codes require that individual wind turbines control their own
terminal voltage to a constant value by means of an automatic voltage regulator.
6) Reactive power control: Wind farms are required to provide dynamic a reactive
power control capability to maintain the reactive power balance and the power factor in
the desired range.
7) Low voltage ride-through (LVRT): In the event of voltage sags, wind turbines are
required to remain connected for a specific period of time before being allowed to
disconnect. In addition, some utilities require that wind turbines help to support grid
voltage during faults.
8) High voltage ride-through (HVRT): In the event that the value of voltage goes
above its upper limit, wind turbines should be capable of staying on line for a given
length of time.
9) Power quality: Wind farms are required to provide electrical power of a desired
quality; for example, maintaining constant voltage or voltage fluctuations in the
desired range and maintaining voltage/current harmonics in the desired range.
10) Wind farm modelling and verification: Some grid codes require wind farm owners
or developers to provide models and system data so as to enable the system operator to
investigate by simulation the interaction between the wind farm and the power system.
They also require installation of monitoring equipment to verify the actual behaviour
of the wind farm during faults and to check the model.
11) Communications and external control: Wind farm operators are required to provide
signals corresponding to a number of parameters important for system operator to
101
enable proper operation of the power system. Moreover, it must be possible to connect
and disconnect the wind turbines remotely.
5-3 Fault Ride-Through (FRT) of DFIG System

Generally; faults are classified into as an open circuit and short circuit which occurring
on the grid side terminal. Short circuit in the power system is referring to the term fault
in this study which categorize to a temporary and permanent faults. Temporary type
results in momentarily interruption such as an insulator flashover which is cleared by
the immediate tripping of one or more circuit breakers to isolate the fault. Whereas the
permanent faults are caused by sustained disruption such as falling debris on the line
such as a tree, or snowfall causing short circuit among the phases and this type
essentially need protection devices in order to disconnect the grid from system entirely.
There are mainly two types of faults in the electrical power system. Those are
symmetrical and asymmetrical faults. But when fault occurs, it causes excessively high
currents to flow which causes the damage to equipments and devices. Fault detection
and analysis is necessary to select or design suitable switchgear
equipments, electromechanical relays, circuit breakers and other protection devices [150,
151].
Symmetrical fault refer to power system faults where all three phases are influenced
equally. These can occur when insulation fails due to system over-voltage resulting
from lighting, or switching surges, insulation contamination, or due to other physical
reasons and as well a result of adverse weather including falling trees on transmission
lines. On other hand Asymmetrical faults are unbalanced faults between one or two of
the three-phases usually caused by mechanical contact as a result of weathering on the
transmission line. These faults occur more often than the symmetrical type [152].
Basically; the configuration of the DFIG has the stator circuit directly connected to the
grid, while the rotor winding is connected to the grid via the back-to-back converter,
and this is the reason for all efforts made to develop the FRT capability. In the case of a
power grid system face the condition of a voltage drop at the PCC, the stator voltage of
the DFIG will immediately change and if it not suddenly varies the rotor voltage will be
compensate the dropping in stator voltage and this will drive to oscillate the stator and
rotor currents with high incremental in its values [106]. This disturbance in the stator
and rotor currents will cause damage to the rotor converter and increase the mechanical
102
stress on the wind turbine as a result of the transient in the electromagnetic torque of the
generator. A wind energy conversion system (WECS) can only be disconnected from
the grid during a fault condition which causes the voltage drop to exceed permitted
limit, figure 5.1 depicted the ratio as a percentage of the actual voltage and normal
voltage of the grid against time, during LVRT the systems requires reactive power from
the WECS this reduces the instabilities to the to the grid voltage. Therefore wind farms
are required to remain on line during voltage disturbances for up to specified time
periods and associated voltage levels.
A key issue of concern regarding DFIG performance is that of fault ride-through.
Network faults produce rapid voltage dips in any of the three phases connecting to a
generating unit. The standard DFIG system is sensitive to such severe dips, inducing
large transient currents in the stator and rotor circuits and risking over-current damage
to the power electronic devices in the converters. Traditionally, to protect these
devices, the rotor circuit is typically shorted or ‘crowbarred’, resulting in a
considerable demand for reactive power from the grid, exacerbating the voltage
problem [153]. The crowbar protection system is one type of the equipment used to
avoid the disconnection of doubly-fed induction wind generators from the network
during faults. The crowbar is inserted in the rotor circuits for a short period of time to
isolate the RSC. As a general rule, the activation and deactivation of the crowbar
system is based only on the DC-link voltage level of the back-to-back converters [146,
147]. However, wind power plants based on the DFIG are very sensitive to grid
disturbances, especially voltage dips. When faults occur and cause voltage dips, the
current flowing through the power converter may subsequently represent a very high
over-current. During this situation, it is common to block the converter to avoid any
risk of damage, and then to disconnect the generator from the grid [148, 149]. The
disadvantage of utilizing the crowbar is that the RSC has to be disabled when the
crowbar is activated and this means that the generator starts to consume reactive power
leading to further deterioration of the grid voltage. Some studies employ an active
crowbar control scheme to enhance the FRT capability, but this modification dose not
avoids the consumption of reactive power [154, 155].
103
Start of
100
Voltage/Vn fault
80 Wind Plant
Must
(%)
remain May
60
connected disconnect
40
Limit line
20
0 150 700 1500 3000 Time(sec)
Figure 5.1: Typical voltage dip limit for LVRT [6, 145].
In the normal mode of operation, the RSC decouple controls the injection of active and
reactive power. The optimum electric power reference is then calculated taking into
account the optimal rotor speed for the incoming wind by the maximum value of the
Cp curves. An encoder can pass the generator’s rotor position (� ) to the abc-dq0 and
to the dq0-abc transformations. The direct axis component is used to maintain the
generator power factor of 1 p.u., and thus the absorbed reactive power reference ( )
is equal to zero. The quadrature axis component is controlled in a similar way the
direct axis; however, it regulates the electrical power to the optimal value. The (� ∗ )
and (� ∗ ) reference signals are sent to the dq0-to-abc transformation and, then to the
signal generator based on the PWM methodology. Finally � ∗ represents the three
phase-voltages desired at the RSC output. In fault mode operation, the crowbar system
is activated, where the three-phase series resistance (crowbar system) is connected to
the rotor windings, and therefore the RSC is blocked.
The GSC control regulates the voltage of the DC link between the RSC and GSC.
The controller employs a PLL to provide the angle (� ) to the abc-to-dq0 and dq0-to-
abc transformations. This angle gives the reference for the synchronization of the
three-phase voltages of the converter output with the terminal voltage. The direct axis
component is used to regulate the DC link voltage (� ). The quadrature axis
component of the reference current is set to zero ( = 0), since power factor control
has already been accomplished by the GSC. (� ∗ ) and the (� ∗ ) reference signals are sent
to the dq0-to-abc transformation, and then to the PWM signal generator. Finally,
(� ∗ ) is the three-phase voltage desired at the grid-side converter output. In fault
mode operation, the DC chopper is activated to dissipate the unbalanced active power
104
between the RSC and GSC. Meanwhile the GSC still maintains control of the DC link
voltage. The reference quadreture current is switched from zero to the loop that
generates this current in comparison voltage reference and terminal voltage to inject
reactive power into the network. Tables 5.1 summarize the behavior of the RSC and
GSC in normal and fault modes of operation.
Although the main task of GSC is to keep the DC-link voltage constant, it may also be
used to compensate for the reactive power and magnetizing the DFIG or in some cases
to remove the reactive power pulsation during unbalanced conditions. The RSC would
provide the required magnetization current waveforms in rotor windings to generate
the required active and reactive power at stator terminals. Voltage can be controlled by
either RSC or GSC independently. If both, RSC and GSC are used for voltage control
but operated in an uncoordinated mode there is a capability that one of them may
dominate the response and take over the control that ultimately could lead to the
absorption of reactive power by the other. In large disturbances, the excitation of rotor
circuit through RSC may be limited or may even be completely isolated. using the
reactive power generation limit of the GSC in a coordinated manner with the stator
reactive power output contributes to terminal voltage support and reduces the
requirements imposed on the RSC [164].
Table 5.1: Summery of the conventional modes of DFIG based WT operation in
normal and fault conditions.
RSC Tracking the Max Power
V <nominal GSC Regulate the �
Normal Mode β(pitch blade angle) Set β = 0
Operation RSC Limiting the Power.
V >nominal GSC Regulate the �
β(pitch blade angle) Activated to limit power
Activated by inserted in series with the rotor
Crowbar
windings
RSC blocked
Fault Mode GSC control reactive power
Operation set to diminish the energy capture from the
β
wind
crowbar Removed after 100 ms
RSC restarted
105
5-4 Performance Analysis of the DFIG System with Voltage Dips
Over-voltages and over-current conditions are generated in the rotor winding of the
DFIG-based wind turbine when voltage dip conditions occur on the PCC. These can be
understood using the following analysis based on the voltage and flux equations in the
stationary reference frame [7]:
� = + � (5.1)
� = + � − � � (5.2)
� =� + �� (5.3)
� = �� +� (5.4)
From equation 5.3; the stator current is:
= � � − �� (5.5)
Then equation 5.5 is substituted in equation 5.4 to eliminate the stator current from the
rotor flux as follows:
��
� = �
� +� � (5.6)
where (� = − �� /� � ), and the rotor voltage in equation 5.2 will have the following
expression:
��
� = �
− � � + +� � − � (5.7)
��
where the term �
− � � =� is the rotor voltage at open circuit and it due
to the stator flux of magnetic coupling. The final term in equation 5.7 is the voltage
resulting from rotor-side circuit (RSC) and crowbar. In normal conditions, the space
vector of the stator voltage is:
�
� =� (5.8)
where � is the constant amplitude. Now equation 5.8 is substituted in equation 5.1, and
therefore:
�
� = + � (5.9)
106
If the stator resistance is neglected and integration conducted on both sides of equation
5.9, then:
� �
=� (5.10)
�
Substituting equation 5.10 in the term for the rotor open circuit:
�.��
� = � (5.11)
�
The term ( ) in equation 5.11 is the slip. Therefore, the maximum value of the rotor
open circuit is:
�.��
� = � (5.12)
�
Then the total value of rotor voltage will be:
.��
� = �
� + +� � − � (5.13)
In equation 5.13, the first term is related to stator voltage and the second term is related
to rotor current. If the rotor is open-circuit, the rotor voltage, which is (� ), is
proportional to the slip frequency and stator voltage. Even if the rotor is connected and
rotor current is present, since ( ) and � � are usually very small, the total rotor
voltage � is still dominated by (� ). Therefore, considering | | . in the normal
operation of a typical DFIG, the amplitude of rotor voltage usually does not exceed
30% of the nominal stator voltage.
In conditions where three-phase voltage dips occur due to a symmetrical voltage
disturbance on the stator’s DFIG system terminal, DFIG performance can be
investigated by characterize the limits and values of the voltage varying in the range
(� > � ). Since � = � and � = − � where ) is the voltage dip ratio [11],
thus, if it is assumed that the DFIG is to operate in the normal condition when at a
specific time a voltage dip of depth occurs, the stator voltage is given by:
�
� <
� ={ �
(5.14)
�
By substituting equation 5.5 in equation 5.1 with the assume that the condition for the
open rotor circuit ( =0) is met the following is given:
107
� =� � + � (5.15)
Solving the differential equation, the stator flux in conditions of voltage dip is then:
� � � −� − /�
� = �
+ �
(5.16)
where:
� �
 �
(forced flux) is the rotation flux space vector corresponding to the
grid during voltage dip.

� −� − /� − � − /�
 �
= �
(natural flux) is the fixed flux of the stator which
�
decreases exponentially to zero by the stator time constant (� = ).
Substituting the stator flux in equation 5.16 in the rotor open voltage equation gives:
�� −� − /�
� = �
.� − +� (5.17)
� �
Canceling ( , since it is very small, gives:

�
�� − /�
� ≈ �
.� − − � −� (5.18)
Equation 5.18 can be expressed in the rotor reference frame by multiplying it

− �
by to give:
� �� − � − /�
� ≈ . − − − (5.19)
�
And then the maximum rotor voltage is:
��
� = | |� + − � −� ) (5.20)
�
5-5 DFIG Analysis During Voltage Swell

Of the different grid faults, voltage swells are less common than voltage sags, but are
also usually associated with system fault conditions. A swell can occur due to a single
line-to-ground fault in the system, which can result in a temporary voltage rise in the
not fault-free phases. Swells can also be generated by sudden drops in load. The abrupt
interruption of the current can generate a large voltage. Switching on large capacitor
108
banks may also cause a swell [32]. High voltage ride-through means that the DFIG-
based wind turbine needs to stay connected to the grid when over-voltage occurs with
continuous and stable performance. It is necessary to reverse the power flow across the
GSC, and thus the current will flow from the grid-side into the DC-link. The DFIG’s
performance under symmetrical voltage swell conditions can be investigated by
characterizing the limits and values of the voltage variation in the range (� > � ),
knowing that � = � and � = + � where ) is the voltage incremental ratio
[156].
� �
� =� <
� ={ � �
(5.21)
� = + �
By solving equation 5.15 with the condition of equation 5.21, hence the stator flux will
be formulated as:
+ � � � − − /�
� = �
− �
(5.22)
where:
+ � �
 �
represents the forced flux which is proportional to grid voltage
+ � and which rotates at synchronous speed.

� − − /�
 �
represents the natural flux which is a transient flux that
guarantees no discontinuities appear in the machine. Its amplitude decreases

exponentially to zero and it does not rotate.
By substituting equation 5.22 in first term of equation 5.7, the rotor voltage:
− −
�� − � � � − �
� = .� + + − � � + (5.23)
� � �
Neglect term (� ) in the open circuit rotor voltage induced by stator flux:
−
�� −
� ≈ .� + + − � � (5.24)
�
109
5-6 Reactive Current Constraint of the GSC During Grid Voltage Dips and
Swells
The stability of the DC bus voltage is the premise of uninterrupted operation of DFIG
systems when grid faults occur. The steady-state voltage equation of the GSC in the
synchronously rotating coordinate can be derived from equations 3.70 and 3.71, hence:
� = +� −� � (5.25)
� = +� � +� (5.26)
Figure 5.2: Spatial relationship of GSC steady-state voltage vectors [157].
DC link voltage should obey the power balance relationship shown in equation 3.95.
The steady-state voltage space vector diagram of the GSC can be depicted as in figure
5.2 with (�) being the power factor angle. From figure 5.2, it can be concluded that the
terminal of the output voltage vector should always fall on the hypotenuse of the
impedance triangle, with its magnitude restricted by the rated operational voltage
across the DC-bus capacitor. This is actually based on voltage modulation theory,
which states that without over-modulation the modulation ratio m needs to satisfy the
following equation [157]:
�
� = √� + � /( ⁄ ) /√ (5.27)
110
If grid voltage orientation is adopted, � is the magnitude of the grid phase voltage
vector. Meanwhile, if small voltage drops across are ignored, equations 5.25 and
5.26 can be further simplified as:
� =� +� � (5.28)
� = −� � (5.29)
Substituting equations 5.28 and 5.29 into equation 5.27, it is found that [158]:
� √ √(� +� � ) + (−� � ) (5.30)
Equation 5.30 gives the operational constrains of the GSC with the DC link voltage,
grid voltage, and inductor and load currents being the main parameters. It can be
determined that the minimum value of the DC link voltage should be no less than the
grid line voltage (i.e. � √ � ) when the GSC operates in unity power factor
mode, i.e., = . Actually, this is caused by the natural attributes of the GSC boost
circuit. Based on the above discussion, the GSC reactive current limit can then be
analyzed and acquired for voltage sags and swells. In voltage sag conditions the
maximum reactive current of the GSC depends on the overall active power of the
system and the slip ratio while in the voltage swell conditions the minimum reactive
current depends on both the grid voltage and the active current [157, 158]. The
stable control of DC-bus voltage value is the mark for deciding whether the wind
power generators can detect voltage swell. In order to ensure the stable control of the
normal operation of GSC and DC voltage during the voltage swell of the power grid,
DC-bus voltage needs to meet the limits in equation 5.30. The right side of formula
5.30 is expressed as minimum value of DC voltage. When GSC absorbs certain
inductive reactive current, therefore the minimum DC voltage limit is reduced. Then,
when the voltage swell of the power grid causes the voltage rise of the wind power
generator port, the normal operation of GSC can be realized through the control of
quadrture current component. If the voltage continuously rises, then the output range of
GSC will be exceeded. Therefore, GSC is based on the reactive power outputted which
is cannot meet the high voltage ride-through demand for the DFIG system. The GSC
reactive current usually possesses an inductive behaviour, which is useful for reducing
the required DC-bus voltage. Due to the fact that the current capacity of the RSC is
111
usually designed a approximately larger than that of the GSC, the output active power
of the RSC during the voltage swell cases should be set so that it is not more than that
of the GSC [157].
5-7 Improved FRT Control Scheme for the DFIG Wind Turbine
Different methods have been investigated to monitor the improvement of the DFIG-
based wind turbine’s ride-through ability during severe voltage dip conditions. These
techniques of control have several advantages, since they fully utilize existing
electrical resources in DFIG wind turbine system without the need for additional
hardware components. In addition these strategies will keep the back-to-back converter
connected to the DFIG even when the grid suffers from voltage drops, and also the
DFIG can be equipped with a crowbar to protect against severe grid faults but to reduce
the activation of the crowbar during the brief periods of voltage dips. It is necessary to
formulate the stator and rotor currents in terms of stator voltage in order to
approximately describe the influence of voltage on these currents. The stator flux
equations 3.20 and 3.21 are used along with the stator voltage in the synchronous
reference frame as formulated in equations 3.12 and 3.13 and by apply the Laplace
transformation to equations 3.12 and 3.13, and after substitute into equations 3.20 and
3.21 in the stator voltage equations, the stator currents can be characterized in terms of
stator voltages and rotor currents as follows [120, 121]:
� �+ �� +� � � (� � + �+� � )�� −� ��
= − (5.31)
� � + � �+ +� � � � + � �+ +� �
� + � −� � �� (� + +� � )�� +� ��
= − (5.32)
� + � + +� � � + � + +� �
By simplifying equations 5.31 and 5.32 by assuming that (� = � ), ( � = ),

(� = ), (� = � ), and ( �� ≪ � , � ), then [155]:
� � ��
= − (5.33)
� + +� �
�
+ �
� ��
= − (5.34)
� + +� �
�
112
As can be noticed from equations 5.33 and 5.34, the rotor current and the stator voltage
appear as input variables, as the former is fixed by the RSC while � depends on the
grid behaviour. Any variation in stator voltage introduces oscillations in the dq
components of the stator currents in the synchronous reference frame. The frequency
of such oscillations is equal to the grid frequency and its damping is very poor, due to
the low value of the stator resistance ( ).
The configuration set the reference values set for active and reactive power in the outer
RSC loop control, as soon as the voltage dips detect by the fault relies that activate the
values until the fault clear to limit the fluctuating of stator and rotor currents. This
scheme is already existing strategy for reducing the current peaks during voltage sag.
The condition of this case considering that the control of the power delivery is based
on the field-oriented control (FOC) philosophy where the active and reactive power
magnitudes can be calculated from equations 5.35 and 5.36 [120]:
= � (5.35)
=− � (5.36)
The second term of equations 5.35 and 5.36 represents the stator currents, and these
can be substituted with the expressions found in equations 5.33 and 5.34 and condition
of null active and reactive power:
+ �
� ��
= � − = (5.37)
�
� + +�
�
� � ��
=− � − =0 (5.38)
�
� + +�
�
Now the rotor currents equations can be deduced from equations 5.37 and 5.38. Given
that the magnitude of the stator voltage can be different from zero, the only way to
guarantee a null delivery of active and reactive power is to set the rotor current
reference equal to zero. This yield the rotor current components shown in equations
5.39 and 5.40 which will regulate in the inner control loop will take in consideration
the drop of in stator voltage.
113
� �
= � (5.39)
�� + +�
�
�+ �
�
= � (5.40)
�� + +�
�
From equations 5.39 and 5.40, it can be concluded that the feedback of the rotor
currents produce an oscillation with an attenuation factor equal to (1/�� ) and the rotor
currents will never be equal to zero since there is always a small residual voltage at the
PCC. Thus, there is a small magnetizing current. The proposed control strategy will
permit to reduction the over-currents in the generator's windings during the fault. The
soft damping of the oscillations in the currents is due to the relationship between the
stator resistance and the magnetizing inductance which yields a low damping factor.
Figure 5.3 depicts the RSC outer loop control modification to improve the FRT ability
of the DFIG system using the PQ null method.
 
    L 
e r
m
ds
  i dr Lr 

L s 
 
ref
1

ref
P s i qr
Ʃ Ʃ Ʃ
*
PI PI v
0
  v qr1
 qr
2  
ref
Q 1  i
ref

Ʃ Ʃ Ʃ
s dr *
PI PI
  v dr1  v dr
0
2  
cal cal
Q s P
cal
s i
cal
dr i qr
Computation
the stator active
Computation
rotor current in
    i L 
e r qr r
and reactive synchronise
power reference frame
Figure 5.3: Diagram of null the active and reactive power method.
5-8 Detection of Disturbances in the DFIG System

The over-current relay has an inverse time-current characteristic where the tripping
time becomes longer as the fault current decreases, and the tripping time is at a
maximum for a fault current equal to the pick-up current �� of the relay. The
114
following equation represents the relationship between the fault current and the
tripping time [159]:
�
�� = ∝ +� (5.41)
(�⁄� )
��
where is the time for pick-up, � is a time constant depending on the trip relay
design parameters, � is a time constant that accounts for saturation in the magnetic
circuit, ∝ changes according to the type of relay (inverse, extremely inverse ..and so
on), and finally � is the relay set pick-up current [159]. Equation 5.41 implies that
the relay responds to the true rms value of the input current. Figure 5.4 shows the
relay’s current-time characteristics which were employed to the delay time of the over-
current protection system in this study. This is 0.03 second and because at the instant
of an 80% voltage dip test the maximum over-current in the system reached 3500A as
shown in figure 5.5 where the relay characteristics are superimposed on the over-
current seen by the relay.
As shown in figure 5.6a, an over-current trip model is used in the grid-connected

DFIG-based wind turbine system. In the model, amplitude phase currents are compared
with a defined current in p.u, which represents the protection activation trip. The three-
phase current is split into its three phase counterparts which are inputted to the
amplitude operator. If the current of any phase exceeds the set tolerances then the
system will be assume that the over-currents are a direct result of a fault occurrence. If
the input current has amplitude greater than the stipulated limits, the output will then
become logic “1”. In other words if any of the phase currents exceed the threshold, the
protection is activated.
115
Current (A)
time(seconds)
Figure 5.4: Inverse time relay characteristics [159].
overcurrent current
Current (A)
time(seconds)
Figure 5.5: Relay characteristic and over-current.
The model for wind turbine voltage trip protection used in this study is the same as that
in 1.5 MW DFIG Demo/ MATLAB® in which the positive sequence voltage is given
as an input, and for its corresponding value trip data is used to see whether it should be
116
tripped or not. The different reasons for tripping may be AC over-voltage, under-
voltage. Figure (5.6b) illustrates the under/over voltage trip model that compares of the
positive sequence voltage component magnitude calculated from the input three-phase
voltage with the upper and lower limits of the normal operation voltage as well as
detecting any disturbance in the voltage profile and sending the logic signal to the
crowbar system in the DFIG to active if with an appropriate delay. The under-voltage
setting shall be less than or equal to the lowest operation voltage of the grid.
U Y
detection Ac overcurrent
1 U Y |u| max
> 1
abc
MinMax
U Y
Max_Iinst
(a)
t
< 1
Min_V1 0.1 s
under voltage
U Y
1 U Y [1 a a2] 1/3 In1 Out1
Vabc U Y magnitude
Gain1
t
> 2
Max_V1 0.1 s
overvoltage
(b)
Figure 5.6: MATLAB/SIMULINK® subsystem: (a) over-current trip model; (b)

over/under voltage trip model
5-9 Performance of Grid Connected DFIG System Without Protection
5-9-1 The Case of (80%) Voltage Dips
The transfer function of the DFIG machine have two poorly damped poles which
oscillated frequency approach to the line frequency and these poles will leads to a
fluctuation in the flux if a grid disturbance occur on the PCC of the DFIG. Therefore,
117
after the system is subjected to grid disturbance, it will be important to increase the
rotor voltage to control the rotor currents if this needs the voltage of exceed the limits
of the power converter, thus it will be incapable of control the current as required. This
reveals that the voltage dip will cause highly induced voltage and currents if there is no
longer a protection strategy adapted to the system. In this investigation the DFIG is
subjected to a severe voltage dip with drop value of 80% depth for a period of 0.5-0.6
second. The RSC and GSC have both been controlled using the vector control with
assume the DFIG system produce zero reactive power and the nominal rated active
power is injected to the grid during a constant wind speed of 4 m/sec during this
simulation. When reductions in stator flux result from the voltage dip, the direct and
quadrature flux components will fluctuate during the voltage dip and after the
clearance of the fault. In addition, the q-axis stator flux cannot be maintained at zero
due to the voltage dip.
As shown in figures 5.7 and 5.8 the voltage dip of 80% at the terminal of the DFIG
system leads to transient over-currents in both the stator and rotor windings as
illustrated in figures 5.9-5.12 with increments of about 2 p.u. Then it decays
exponentially subject to the RSC which causes a large stress upon it. In addition, the
step increment is clearly depicted in the total rms grid current in figure 5.13 as a result
of the sag condition. On the other hand the DC-link voltage fluctuated during the
voltage dip until it reaches limit and then returns to regulate it around the set value
after fault removal as depicted in figure 5.14.
118
1.4
1.2
1
Vgrid rms(pu)
0.8
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.7: Instantaneous rms grid voltage.
1500
a b c
1000
500
Vs abc (v)
-500
-1000
-1500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.8: Instantaneous three-phase stator voltage.
4000
a b c
3000
2000
1000
Is abc (A)
-1000
-2000
-3000
-4000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.9: Instantaneous three-phase stator current.
119
2000
Ids
Iqs
1000
0
Idqs (A)
-1000
-2000
-3000
-4000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time(second)
Figure 5.10: Instantaneous dq stator current.

80% voltage dip
4000
a b c
3000
2000
1000
Ir abc (A)
-1000
-2000
-3000
-4000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.11: Instantaneous three-phase rotor current.

4000
Idr
Iqr
3000
2000
Idqr (A)
1000
-1000
-2000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.12: Instantaneous dq rotor current.
120
1.4
1.2
1
Igrid rms (pu)
0.8
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.13: Instantaneous measurements of rms grid current.
1204
Ref
1203 Meas
1202
1201
Vdc ref meas (v)
1200
1199
1198
1197
1196
1195
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.14: DC-link voltage regulator in the GSC.
The stator active and reactive power tracking of the reference set points is shown in
figure 5.15 and figure 5.16, the results indicate severity of the system simulation
during the voltage dip period where both the active and reactive power vary and lose
controllability. The active power decreases nearly to zero while the reactive power is
absorbed by the machine. At voltage dip initiation, the vector control of the rotor-side
converter is provisionally lost. The brief surge in power output is associated with the
rapid demagnetization of the DFIG associated with large, oscillatory active and
reactive power as illustrated in figures 5.15 and 5.16. After fault clearance, the power
controller requires 20ms to settle the highly oscillatory output, and its ability for
recovery depends on the magnitude of the recovery voltage and the stabilize the
generator speed. The reactive power rises quickly after fault initiation to the same peak
level as the active power, of 0.1 MVAR, before oscillatory decay. After fault
121
clearance, the DFIG start to re-magnetize. However, this effect lasts for less than one
system cycle.
The generator speed shown in figure 5.17 represent as increase of 0.19% associated
with the brief increase in power output following the fault. The DFIG rotor speed
confirmed the relatively small impact after a voltage dip duration. Although the speed
rises by over 6% during the voltage dip, this was partially offset by a 2% decrease in
speed associated with the demagnetization energy dissipated from the machine. The
currents of the grid-side converter (GSC) shown in figure 5.18 appear thicker with
high frequency noise due to the relatively low-impedance low-pass filter fitted to the
converter. Note that the pre-fault currents are far smaller than the rotor current because
only active power is delivered to the grid. In normal conditions, the rotor converter
also provides a magnetisation current to the generator. The GSC inject only enough
current to balance the rotor exchang power input. The grid fault caused an immediate
increase in converter current. After a small spike in current, the active current was
brought downto -40 A by the controller in an attempt to reduce the DC-link voltage.
6
x 10
0
Ref
-0.2 Meas
-0.4
-0.6
Ps ref Ps meas (w)
-0.8
-1
-1.2
-1.4
-1.6
-1.8
-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.15: Instantaneous regulation of the stator active power in the RSC.
122
6
x 10
1.5
Ref
Meas
1
0.5
Qsref, Qs meas(VAR)
-0.5
-1
-1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.16: Instantaneous regulation of the stator reactive power in the RSC
165
160
Wgen (rad/sec)
155
150
145
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.17: Instantaneous measurement of generator speed.
800
600
400
200
Iabc GSC (A)
-200
-400
-600
-800
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

time (seconds)
Figure 5.18: Instantaneous measurement of three-phase current of the GSC.
123
To demonstrate the performance of the mechanical drive train 2-mass model during
voltage dips, The electromagnetic and aerodynamic torque behaviour and rotational
speed generator and rotor are illustrated in figures 5.19 and 5.20 respectively. Here the
speed and torque values of the low-side shift have been converted to the high-speed
side of the gearbox which is the induction machine shaft side to allow a simpler
evaluation of the system during voltage dip conditions.
Aerodynamic and electromagnetic torque represent the input quantities for the model,
while changes in the turbine‘s rotor speed and generator speed are the output. During a
worst case grid-side voltage dip condition, where the grid voltage falls to 20% at 0.5
sec, the stator flux decreases to an extremely small value, and the induction machine is
no longer able to generate electrical torque. When this happens, the power absorbed by
the blades from the wind will be completely stored in the rotating mechanical
components in the form of kinetic energy, and the generator will accelerate. After the
voltage starts to recover due to the clearing of the fault after 0.6 sec, the stator flux
gradually recovers such that electrical torque can be produced again to counteract the
driving torque from the wind. As a result, the electromagnetic torque will be restored
back to the reference value after approximately 200 msec, as shown in figure 5.19.
It should be noted that, during the voltage sag, the active power delivered to the grid is
not balanced (voltage dip increases active and reactive powers continue to swing as
rotor speed varies) with the mechanical power obtained from the wind turbine. This
causes an acceleration of the generator that in any case will be dangerous due to the
exceptional moment of inertia of the drive train itself to compensate for the power drop
due to the voltage drop. In figure 5.20 it can be seen that the expected acceleration is
only 17 rad/sec during the dip’s duration 0.5-0.6 sec which is less than a 2% increase.
The 17 rad/sec increase in the machine rotor speed represents a 0.4 rad/sec increase in
the turbine’s rotating speed and this can be considered to be negligible as shown in
figure 5.20.
124
4
x 10
3 The aerodynamic torque

Generator and Aerodynamic torque(Nm) Generator torque
-1
-2
-3
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1
time(seconds)
Figure 5.19: Electromagnetic and aerodynamic torque during voltage dip.

180
Generator speed
175 Rotor speed
Generator and rotor speed(rad/sec)
170
165
160
155
150
145
140
135
130
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1
time(seconds)
Figure 5.20: Generator speed and rotor speed during the voltage dip.
5-9-2 Case of Voltage Swell of 1.3 p.u.
A voltage swell can result from either switching off a large load or switching on a large
capacitor bank across the PCC. The behaviour of the DFIG during voltage swell is
investigated by increasing the PCC voltage level at the grid-side terminal to 1.3 p.u.
which tested in operating period 0.5-0.6 sec. In this case an increment in the power
generated will noted by the DFIG system in parallel to voltage swell occurrence, and
this will then be reduced after clearance the fault. Under normal conditions, active and
reactive power decoupling control is adopted for the rotor side-converter (RSC) and
grid-side converter (GSC) to adjust the active and reactive powers output by the DFIG
stator by controlling the exciting current output by the RSC, wherein GSC control aims
at stabilizing the DC-link voltage. When DC link voltage is higher than the maximum
125
continuous operating voltage, any conventional protection circuit used in a DFIG, such
as the DC chopper in parallel connection to the DC-bus, can be triggered for break-
over to restrain the possible bus voltage pumping-up at the moment of sudden voltage
change, so as to ensure the safety the DC-link.
From the simulation results shown in figures 5.23-5.30, it is clear that when the voltage
swell occurs, the DFIG-based wind turbine changes from being in an excited condition
to over-excited mode for LVRT voltage support. After voltage recovery and to reduce
of the adverse influence on grid voltage upon the DFIG, the system will dynamically
operate in excited mode again. The reactive power support is shared between the stator
winding and GSC so as to reduce the GSC output voltage in order to avoid the over
modulation of the PWM via the control of DC voltage, as shown in figure 5.22.
Therefore, during grid voltage swell conditions, from the wind turbine needs to inject
reactive current into the power system with specific proper control design of the GSC
(which is not included in this study), but of inductive property to ensure the safe
operation of the capacitor so that the GSC will contribute in support reactive power. In
addition, the incremental of total rms grid current over-shoot to 1.2 p.u. As shown in
figure 5.21, the PCC voltage rise to 1.3 p.u. and therefore the direct and quadrature
axis rotor currents in the synchronous reference frame have rise as shown in figure
5.26 at the start of voltage swell. Then after fault clearance while the stator current in
figure 5.24 increased by step and recovered at the end of the swell condition. In figure
5.26 d axis rotor current and q axis rotor current fluctuated bigger under voltage
recovery condition as compared to voltage swell since stator oscillation of voltage
recovery will based on natural flux decay which is not zero.
126
1.5
1
Vgrid rms(pu)
0.5
0
0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.21: Instantaneous rms stator voltage.
1500
a b c
1000
500
Vs abc (v)
-500
-1000
-1500
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.22: Instantaneous measurement of the three-phase stator voltage.
4000
a b c
3000
2000
1000
Is abc(A)
-1000
-2000
-3000
-4000
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.23: Instantaneous three-phase stator current.
127
2000
Ids
Iqs
1000
0
Idqs (A)
-1000
-2000
-3000
-4000
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.24: Instantaneous dq stator current.

4000
a b c
3000
2000
1000
Ir abc (A)
-1000
-2000
-3000
-4000
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.25: Instantaneous three-phase rotor current.

4000
Idr
Iqr
3000
2000
Idqr (A)
1000
-1000
-2000
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.26: Instantaneous dq rotor current.
128
1204
Ref
1203 Meas
1202
1201
Vdc Ref Meas (v)
1200
1199
1198
1197
1196
1195
0.4 0.5 0.6 0.7 0.8 0.9 1
time(sec)
Figure 5.27: DC-link voltage regulator.

6
x 10
1
Ref
Meas
0.5
0
Ps ref Ps meas (w)
-0.5
-1
-1.5
-2
-2.5
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
6
x 10
1.5
Ref
Meas
1
0.5
Qsref, Qs meas(VAR)
-0.5
-1
-1.5
-2
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.29: Instantaneous regulation of the stator reactive power in the RSC.
129
1.8
1.6
1.4
Igrid rms(A)
1.2
0.8
0.6
0.4
0.2
0
0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.30: Instantaneous rms grid current.
5-10 Performance of Sensored and Sensorless Vector Control in DFIG

System with PQ Null Modification during Disturbances
The aim of this set of tests is to illustrate some of the most representative
MATLAB/SIMULINK® dynamic behaviour results so as to evaluate and compare the
performance of sensored and sensorless vector control of the DFIG system during fault
conditions in the grid connection. In order to study the validity of enhanced FRT using
the PQ null control strategy against voltage dips, a three-phase fault is considered
which causes a voltage dip of about 80% in depth and of duration in the range 0.5-0.6
seconds at the stator terminal of the DFIG. Both RCMRAS and QRMRAS have been
employed in this investigation with a constant wind speed of 4 m/sec in the grid fault
calculation to conduct the vector control associated with the PQ null method in the
RSC. As shown in figures 5.33 –5.42, three major sets of simulink models were
conducted, including: in the first a model of the PQ null method used with a sensored
drive DFIG. In the second, RSMRAS model used to verify the PQ null with sensorless
drive. In the third, QRMRAS model was utilized to validate the PQ null with
sensorless drive based DFIG operation. Immediately after the fault occurs at 0.5
second, the voltage at the wind turbine terminal drops, as shown in figures 5.31 and
5.32. The DFIG reacts to the three-phase voltage dip with dramatic increments of the
stator currents and thus high rotor currents are induced in the rotor winding. At the
moment when the stator voltage becomes lower than the sett values in the fault
detectors, this will directly activate the PQ null method by switching off the
130
conventional power loop control and using the zero values as a reference to reduce
over-current of the rotor/stator in parallel with regulation the DC-bus voltage with an
acceptable fluctuation level, as shown in figures 5.33, 5.36 and 5.41. here the PQ null
method proves its success in employing the available resources instead of using
conventional protection devices. In addition, this method has been tested with
RCMRAS and QRMRAS, as depicted in figures 5.33-5.42, and it succeeds in tracking
the performance of sensored vector control with results based on estimation of the rotor
position. In this condition these schemes have proved their ability to eliminate the need
for conventional position detectors with less sensitivity to some types of regular grid
disturbances such as voltage dips. As shown in figure 5.33 and 5.36, the wind turbine
system is protected by the PQ null solution which is activated before the crowbar
because the voltage dip is very short-lived. The stator and rotor currents are reduced,
and the stator currents decay slowly having a DC component. The RSC can stay
running and connected to the DFIG rotor windings. When the PQ null control is
activated, the outer power control loops are disabled and thus active and reactive
power controls are not achieved. The power control can be implemented to fulfill grid
code requirements when the transients have decayed. After fault clearance, the wind
turbine system can continue with normal operation.
Figures 5.33 - 5.42 illustrate the comparison of the responses of this system in sensored
and sensorless vector control of an 80% voltage dip using RCMRAS and QRMRAS to
estimate the rotor position and employ this information in the closed loop decouple
control. Rotor and stator currents begin to drop with the activation of the PQ null loop
at the RSC without any crowbar engagement at around 4m/sec. Rotor and stator
currents in each approach show acceptable fluctuation levels around their pre-fault
values. Voltage recovery allows sufficient control of the GSC currents to return the DC
link voltage to the reference value. However, from the performance evaluation of
DFIG system based on the results, it is clear that the QRMAS succeeds to estimating
the rotor angle under these operating conditions.
The dynamic behaviour of the dq stator and rotor currents when both RCMRAS and
QRMRAS are implemented during a 80% voltage dip can be compared with the
sensored outcomes from the simulation plots depicted in figures 5.34, 5.35, 5.37 and
5.38. Figure 5.35 shows the response of the quadrature stator currents during the
131
voltage sag, and they oscillate around zero during the sage duration, while the direct
component exponentially fluctuate and then returns to zero after the end of the sag. The
reduction of the rms value of grid current during the fault, shown in figure 5.42 with no
protection condition. In figure 5.42, it is clear that the method with either sensored or
sensorless vector control succeed in reducing the currents in the generator’s windings
during the fault.
1.4
1.2
1
Vgrid rms(pu)
0.8
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.31: Instantaneous rms grid voltage.
1500
a b c
1000
500
Vs abc (v)
-500
-1000
-1500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (second)
Figure 5.32: Instantaneous three-phase stator voltage.
132
3000
Phase c in sensored vector control
Phase c in RCMRAS sensorless vector control
2000 Phase c in QRMRAS sensorless vector control
1000
Is c (A)
-1000
-2000
-3000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.33: Instantaneous measurements of phase c stator current.
1500
Ids in QRMRAS Sensorlessvector control

1000 Ids in sensored vector control
Ids in RCMRAS Sensorlessvector control
500
Ids (A)
-500
-1000
-1500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.34: Instantaneous value of the d components stator current.
3000
Iqs in sensored vector control

Iqs in RCMRAS Sensorless vector control
2000
Iqs in QRMRAS Sensorless vector control
1000
Iqs (A)
-1000
-2000
-3000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.35: Instantaneous value of the q components stator current.
133
4000
Phase c in sensored vector control

3000 Phase c in RCMRAS sensorless vector control
Phase c in QRMRAS sensorless vector control
2000
1000
Ir c (A)
-1000
-2000
-3000
-4000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.36: Instantaneous measurements of phase c rotor current.
1500
Idr in RCMRAS Sensorlessvector control

1000 Idr in sensored vector control
Idr in QRMRAS Sensorlessvector control
500
Idr (A)
-500
-1000
-1500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.37: Instantaneous d-component rotor current.
4000
Iqr in sensored vector control

3000
Iqr in RCMRAS Sensorlessvector control
Iqr in QRMRAS Sensorlessvector control
2000
1000
Iqr (A)
-1000
-2000
-3000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.38: Instantaneous q-component rotor current.
134
6
x 10
1
Ref
0.5
Meas in sensored vector control
Meas in RCMRAS sensorless vector control
0
Meas in QRMRAS sensorless vector control
Ps ref Ps meas (w)
-0.5
-1
-1.5
-2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
6
x 10
1.5
Ref
1 Meas in RCMRAS sensorless vector control
Meas in QRMRAS sensorless vector control
Qsref, Qs meas(VAR)
0.5
-0.5
-1
-1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (seconds)
Figure 5.40: Instantaneous regulation of the stator reactive power in the RSC.
1204
Ref
1203
Meas in RCMRAS sensorless vector control
1202 Meas in QRMRAS sensorless vector control
1201
Vdc ref meas (V)
1200
1199
1198
1197
1196
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time(second)
Figure 5.41: DC-link voltage regulator.
135
1.4
Sensored vector control
In RCMRAS Sensorless vector control
1.2
In RCMRAS Sensorless vector control
1
Igrid rms (pu)
0.8
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time(sec)
Figure 5.42: Instantaneous-three phase grid current.
5-11 Conclusions
Given the new grid code requirements, it is of great importance that variable-speed
wind turbines to remain connected to the grid network during severe grid voltage dips
and they must contribute to voltage recovery in these conditions. The necessities of
fault ride-through (FRT) have been explained in this chapter and the requirements
made of grid operators have been explained. The fault ride-through ability of the
DFIG-based turbine is highly dependent on the speed and the converter’s ability to
withstand heavy rotor current transients with or without crowbar protection. The FRT
behaviour could thus be improved by designing the electrical drive so that it can handle
large currents, but this is limited by commercial viability. This technology has full
employing the available resource also it can be used as a back up to the crowbar.
136
Chapter 6: Conclusions and Further Work
CHAPTER six
137
CHAPTER SIX
CONCLUSIONS AND FURTHER WORK
6-1 Conclusions
 This investigation include the principles of the modelling and design the
DFIG-based wind turbine system along with its vector control strategy of the
back-to-back converter includes the aerodynamic, mechanical and electrical
components and controllers. The system’s performance during normal
operation was conducted.
 This study includes complement analysis of the vector control DFIG-based
wind turbine in two operating conditions of sensored and sensorless vector
control. The QRMRAS algorithm has been employed with a PI controller to
detect the rotor position and its performance compared with that of the
RCMRAS. Effects of parameter variations on the dynamic rotor detection of
QRMRAS and RCMRAS have been investigated
 By employing the integrated error index; integral of time multiplied by
absolute error and integral absolute error were used as criteria to measure the
quality of system response. A systematic evaluation of rotor angle detection
has been accomplished for different operating condition, and the results show
QRMRAS gives lower value of ITAE and IAE as compared with the
RCMRAS.
 Given the new grid code requirements, it is of great importance that variable-
speed wind turbines remain connected to the grid network during severe grid
voltage dips and that they contribute to voltage recovery. When subjected to
severe grid voltage dips, this will be responsible for the deterioration of the
back-to-back converter. The whole DFIG system previously needed to be
disconnected from the grid during severe voltage sags. This is inconsistent
with the standards set in the new grid requirements, which demand the DFIG
system have FRT capability or already-used protection schemes should be
138
activated to enable the system stay connected to the network and behave as a
typical power generator.
 The results show that the system can more easily ride-through a voltage swell
of 0.3pu compared to a voltage dip of 0.8pu.
6-2 Contributions
The main contributions of this thesis are summarized as follows:-
 A complete dynamic comparative performance assessment for grid-connected

variable-speed wind turbines based on the DFIG has been built mathematically
in MATLAB/SIMULINK® platform and confirmed the system with PLECS®
model.
 The rotor reactive power-based MRAS (QRMRAS) is used to estimate the
rotor position with normal and voltage dip condition where the results has
been tested with the conventional rotor current based MRAS (RCMRAS)
observer. The QRMRAS shows it has insensitive response to the parameter
variation of the DFIG as compared with the RCMRAS.
 This study put the use of QRMRAS as a back-up to the conventional
mechanical encoder to increase the DFIG system reliability in the remote areas
and harsh environment of operation.
6-3 Further Work
There are many possibilities for future development, improvement, verification

and analysis of the present project to overcome the present research limitations. The
following recommendations can be made.
 If possible, experimental sets of tests should be conducted to verify the

simulation results documented in this study. With the verification work, the
proposed measures, sensorless control strategies and enhanced fault ride-
through (FRT) capability could be improved. Furthermore, the field validation
of the proposed system is needed. A small-scale kit DFIG system could be
used which has behaviour similar to a full-scale machine, and also a proper
grid-simulator should be used to emulate a realistic voltage drop.
139
 Voltage recovery investigation should be carried out with interaction among

wind turbines in wind-farms after grid disturbances.
 The accuracy and sensitivity of several sensorless methods should be
investigated, includes the use QRMRAS in the DFIG-based stand-alone
system with different operating conditions.
 Converter protection strategies for asymmetrical faults still need to be
developed for the DFIG system in terms of the behaviour and FRT strategies
because such faults induce permanent negative sequence components whose
influence persist throughout the duration of voltage dip.
 Further studies should focus on improving the power converter controller of
the DFIG to contribute to the FRT capability for faults of longer duration.
 Complete speed and frequency deviation protection schemes should be
improved with more attention to providing a robust DFIG system.
140
References
References
141
References
REFERENCES
[1] S. Rahman, "Going Green: The Growth of Renewable Energy", IEEE Power and
Energy Magazine, Vol. 99, No. 6, pp. 16-17, Nov. /Dec. 2003.
[2] B. Fox, D. Flynn, L. Bryans, N. Jenkins, D. Milborrow, M. O. Malley, R. Waston
and Olimpo Anaya-lara, "Wind Power Integration Connection and System
Operational Aspects", IET Power and Energy series 50, 2007.
[3] P. A. Lynn, "Onshore and Offshore Wind Energy", 2012 John Wiley & Sons Ltd.
[4] http://www.gwec.net/global-figures/graphs/
[5] A. Z. Amin, "Renewable Power Generation Costs in 2012: An Overview",
International Renewable Energy Agency, 2013.
[6] B. Wu, Y. Lang, N. Zargari, S. Kouro, and Institute of Electrical and Electronics
Engineers, Power Conversion and Control of Wind Energy Systems. Oxford:
Wiley Blackwell, 2011.
[7] G. Abad, J. Lopez, A. R. Miguel, M. Luis and G. Iwanski, Doubly Fed Induction
Machine: Modeling and Control For Wind Energy Generation, 1st edition. John
Wiley & Sons, INC., Publication, 2011.
[8] J. F. Manwell, J. G. McGowan, and A. L. Rogers, Wind Energy Explained
Theory, Design, and Application, 2nd edition, 2009, John Wiley & Sons Ltd.
[9] F. Blaabjerg and Z. Chen, Power Electronics for Modern Wind Turbines, San
Rafael, Calif.: Morgan & Claypool, 2006.
[10] C. Zhe, J. M. Guerrero, and F. Blaabjerg, "A Review of the State of the Art of
Power Electronics for Wind Turbines", IEEE Transactions on Power Electronics,
Vol. 24, No. 8, pp. 1859-1875, Aug. 2009.
[11] F. Blaabjerg, and Z. Chen, Power Electronics for Modern Wind Turbines,
Morgan & Claypool Publishers, 1st edition, 2006.
[12] G. O. Suvire, Wind Farm Technical Regulations, Potential Estimation and Siting
Assessment, InTech, June, 2011.
[13] http://www.energy.gov/eere/wind/how-does-wind-turbine-work.
[14] L. Wang, Y.-F. Yang, and S.-C. Kuo, ''Analysis of Grid-connected Induction
Generators Under Three-phase Balanced Conditions'', IEEE, Power Engineering
142
References
Society Winter Meeting, Vol. 1, pp. 413 – 417, 2002.

[15] C. S. Demoulias and P. S. Dokopoulos, ''Transient behavior and self excitation of
wind-driven induction generator after its disconnection from the power grid'',
IEEE Transactions on Energy Conversion, Vol. 5, No. 2, pp. 272-278, June,
1990.
[16] M. A. Ouhrouche, D. D. Xuan, M. L. Quang, and C. Robert, ''EMTP Based
Simulation of a Self-Excited Induction Generator After its Disconnection From
the Grid'', IEEE Transactions on Energy Conversion, Vol. 13, No. 1, March
1998.
[17] H. L. Hess, N. A. Abdul Melek, and E. Muljadi, ''Power Converter for Wind
Turbine Application'', IEEE Power Engineering Society Summer Meeting, Vol. 2,
pp. 1275 -1276, 2000.
[18] S. M. Muyeen, T. Junji, and M. Toshiaki, ''Stability Augmentation of a Grid-
connected Wind Farm'', Springer, 2009, page 61-62.
[19] B. Singh, R. B Saxena, S. S. Murthy, and B. P Singh, ''A Single-Phase Self-
Excited Induction Generator for Lighting Loads in Remote Areas'', International
Journal on Electricity Engineering Education, Vol. 25, pp. 269-275, 1988.
[20] D. Rane, ''Voltage and Frequency Control of Wind generator System'', M.sc.
thesis, University of Delhi, 2008.
[21] I. Cadirei, and M. Ermis, ''Performance Evaluation of Wind Driven DOIG using a
Hybrid Model'', IEEE Transactions on Energy Conversion, Vol. 13, No. 2, June
1998.
[22] R. Pena, J. C. Clare and G. M. Asher, "Doubly Fed Induction Generator Using
Back-To-Back PWM Converters and Its Application to Variable-Speed Wind-
Energy Generation", IEE Proceedings Electric Power Applications, Vol. 143,
No. 3, pp. 231-241, May 1996.
[23] L. Refoufi, B. A. T. Al Zahawi, and A. G. Jack, "Analysis and Modeling of the
Steady State Behavior of the Static Kramer Induction Generator", IEEE
Transactions on Energy Conversion, Vol. 14, No. 3, pp. 333-339, Sept. 1999.
[24] J. W. Kolar, T. Friedli, J. Rodriguez, and P. W. Wheeler, ''Review of Three-
Phase PWM AC-AC Converter Topologies'', IEEE Transactions on Industrial
Electronics, Vol. 58, No. 11, pp. 4988-5006, Nov.2011.
[25] E. D. Basset and F. M. Potter, ''Capacitive Excitation for Induction Generators'',
143
References
Transactions of the American Institute of Electrical Engineers, Vol. 54, No. 5,

pp. 540-545, May 1935.
[26] P. K. Shadhu Khan and J. K. Chatterjee, Three-Phase Induction Generators: A
Discussion on Performance, Taylor & Francis on Electric Machines and Power
Systems, Vol. 27, No. 8, pp. 813-832, 1999.
[27] J. M. Elder, J. T. Boys and J. L. Woodward, ''Self-Excited Induction Machine as
a Small Low-cost Generator'', IEE Proceedings, Vol. 131, Pt. C, No. 2, 1984.
[28] K. K. Raj, E. Swati, Ch. Ravindra, ''Voltage Stability of Isolated Self Excited
Induction Generator (SEIG) for Variable Speed Applications using
Matlab/Simulink'', International Journal of Engineering and Advanced
Technology (IJEAT), Vol. 1, Issue-3, pp.2249 – 8958,February 2012.
[29] B. Hopfensperger and D. Atkinson, "Doubly-Fed A.C. Machines: Classification
and Comparsion", European Conference on Power Electronics and Applications
(EPE), topic 06f, Graz, 2001.
[30] B. Hopfensperger, D. Atkinson, and R. A. Lakin, "Stator Flux Oriented Control
of a Cascaded Doubly-Fed Induction Machine", IEE Proceedings: Electric
Power Applications, Vol. 146, No. 6, pp. 597–605, Nov. 1999.
[31] A. Petersson, "Analysis, Modeling and Control of Doubly-Fed Induction
Generators for Wind Turbines", Ph.D. dissertation, Chalmers University of
Technology, Sweden, 2005.
[32] A. K. Wallace, R. Spee, and G. C. Alexander, "The Brushless Doubly-Fed
Machine: Its Advantages, Applications and Design Methods", In 6th IEE
International Conference on Electrical Machines and Drives, 8–10 Sept. 1993,
No.376, pp. 511–517.
[33] A. J. S. Filho, and E. R. Filho, "Model-Based Predictive Control Applied to the
Doubly-Fed Induction Generator Direct Power Control", IEEE Transactions on
Sustainable Energy, Vol. 3, No. 3, pp. 398-406, July 2012.
[34] W. Leonhard, Control of Electrical Drives, Berlin, Heidelberg, New York, and
Tokyo: Springer-Verlag, 1985.
[35] X. Yu, Z. Jiang, and Y. Zhan, "Control of Doubly-Fed Induction Generators for
Distributed Wind Power Generation", IEEE Power and Energy Society General
Meeting Conversion and Delivery of Electrical Energy in the 21st Century, 20-24
July 2008, pp. 1-6.
144
References
[36] L. Yang, Z. Xu, J. Østergaard, Z. Y. Dong, and K. P. Wong, "Advanced Control

Strategy of DFIG Wind Turbines for Power System Fault Ride Through", IEEE
Transactions on Power Systems, Vol. 27, No. 2, pp. 713-722, May 2012.
[37] V.T. Phan and H. H. Lee, "Stationary Frame Control Scheme for A Stand-Alone
Doubly Fed Induction Generator System with Effective Harmonic Voltages
Rejection", IET Electric Power Applications, Vol. 5, No. 9, pp. 697–707, 2011.
[38] M. G. Gracia, M. P. Comech, J. Sallán, and A. Llombart, "Modelling Wind
Farms for Grid Disturbance Studies", Renewable Energy, Vol. 33, No. 9, pp.
2109–2121, 2008.
[39] A. Abbaszadeh, S. Lesan, V. Mortezapour, "Transient Response of Doubly Fed
Induction Generator under Voltage Sag Using an Accurate Model", Sustainable
Alternative Energy (SAE), 2009 IEEE PES/IAS Conference, pp. 1 – 6, 28-30 Sep.
2009.
[40] K. Elkington and M. Ghandhari, "Comparison of Reduced Order Doubly Fed
Induction Generator Models for Nonlinear Analysis", IEEE Electrical Power &
Energy Conference, pp.1-6, Oct. 2009.
[41] P. Sørensen, A. D. Hansen, T. Lund and H. Bindner, "Reduced Models of
Doubly Fed Induction Generator System for Wind Turbine Simulations", Wind
Energy, Vol. 9, No. 4, pp. 299-311, Aug. 2006.
[42] A. S. Neto, S. L. A. Ferreira, J. P. Arruda, F. A. S. Neves, P. A. C. Rosas and M.
C. Cavalcanti, "Reduced Order Model for Grid Connected Wind Turbines with
Doubly Fed Induction Generators", IEEE International Symposium on Industrial
Electronics, pp. 2655-2660, June 2007.
[43] P. Ledesma and J. Usaola, "Effect of Neglecting Stator Transients in Doubly Fed
Induction Generators Models", IEEE Transactions on Energy Conversion, Vol.
19, No. 2, pp. 459-461, June 2004.
[44] I. Erlich and F. Shewarega, "Modeling of Wind Turbines Equipped with Doubly-
Fed Induction Machines for Power System Stability Studies", 2006 IEEE Power
Systems Conference and Exposition, pp. 978-985, November 2006.
[45] W. Qiao, W. Zhou, J. M. Aller, and R. G. Harley, "Wind Speed Estimation Based
Sensorless Output Maximization Control for a Wind Turbine Driving a DFIG",
IEEE Transactions on Power Electronics, Vol. 23, No. 3, pp. 1156-1169, May
2008.
145
References
[46] A. Tapia, G. Tapia, J. X. Ostolaza and J. R. Sáenz, "Modeling and Control of a

Wind Turbine Driven Doubly Fed Induction Generator", IEEE Transactions on
Energy Conversion, Vol. 18, No. 2, pp. 194-204, June 2003.
[47] S. Muller, M. Deicke, R. De Doncker, "Doubly Fed Induction Generator Systems
for Wind Turbines", IEEE Industry Applications Magazine, Vol. 8, No. 3, pp. 26-
33, May/Jun 2002.
[48] S. Chondrogiannis and M. Barnes, "Stability of Doubly-Fed Induction Generator
Under Stator Voltage Orientated Vector Control", IET Renewable Power
Generation, Vol. 2, No. 3, pp. 170-180, Sep. 2008.
[49] C. Batlle, A. D. Cerezo and R. Ortega, "A Stator Voltage Oriented PI Controller
for the Doubly-Fed Induction Machine", Proceedings of the 2007 American
Control Conference, New York City, USA, pp. 5438-5443, July 2007.
[50] A. Petersson, L. Harnefors, and T. Thiringer, "Comparison Between Stator-Flux
and Grid-Flux-Oriented Rotor Current Control of Doubly-Fed Induction
Generators", 2004 35th Annual IEEE Power Electronics Specialisfs Conference,
1: pp.482-486.
[51] H. Akagi, and H. Sato, "Control and Performance of a Doubly-Fed Induction
Machine Intended for a flywheel Energy Storage System", IEEE Transactions
Power Electronics, Vol. 17, No. 1, pp. 109–116, Jan. 2002.
[52] M. Yamanmoto, and O. Motoyoshi, "Active and Reactive Power Control of
Doubly-Fed Wound Rotor Induction Generator", IEEE Transactions on Power
Electronics, Vol. 6, No. 4, pp.624-629, Oct. 1991.
[53] T. M. Masaud and P.K. Sen, "Modeling and Control of Doubly Fed Induction
Generator for Wind Power", IEEE Conference on North American Symposium
(NAPS), pp. 1-8, 4-6 Aug. 2011.
[54] F. Wu, X. P. Zhang, K. Godfrey, and P. Ju, "Modeling and Control of Wind
Turbine with Doubly Fed Induction Generator", IEEE PES Conference on Power
Systems Conference and Exposition, pp. 1404-1409, 2006.
[55] Y. Zhang, J. Jia, W. Li, D. Wang, and J. Liu, "Steady State Characteristic
Analysis and Stability Assessment of Doubly Fed Induction Generator Based
Wind Power Generation System", IEEE International Conference on Power
System Technology, pp. 1-5, 24-28 Oct. 2010.
[56] A. Ostadi, , A. Yazdani, and R. k. Varma, "Modeling and Stability Analysis of a
146
References
DFIG-Based Wind-Power Generator Interfaced With a Series-Compensated

Line", IEEE Transactions on Power Delivery, Vol. 24, No. 3, pp. 1504-1514,
July 2009.
[57] Y. Zhang, Z. Li, J. Hu, W. Xu and J. Zhu, "A Cascaded Brushless Doubly Fed
Induction Generator for Wind Energy Applications Based on Direct Power
Control", 2011 International Conference on Electrical Machines and Systems ,
pp.1-6, Aug. 2011.
[58] K. Protsenko and D. Xu, "Modeling and Control of Brushless Doubly-Fed
Induction Generators in Wind Energy Applications", IEEE Transactions on
Power Electronics, Vol. 23, No. 3, pp.1191- 1197, May 2008.
[59] S. Arnalte, J. C. Burgos and J. L. Rodríguez-Amenedo, "Direct Torque Control
of a Doubly-Fed Induction Generator for Variable Speed Wind Turbines",
Electric Power Components and Systems, Vol. 30, No. 2, pp. 199-216, Nov.
2002.
[60] K. C. Wong, S. L. Ho and K. W. E. Cheng, "Direct Torque Control of a Doubly-
fed Induction Generator with Space Vector Modulation", Electric Power
Components and Systems, Vol. 36, No. 12, pp. 1337-1350, Nov. 2008.
[61] F. Bonnet, P. E. Vidal and M. Pietrzak-David, "Direct Torque Control of Doubly
Fed Induction Machine", Bulletin of the Polish Academy of Sciences Technical
Sciences, Vol. 54, No. 3, pp. 307-314, 2006.
[62] J. Arbi, M. J. B. Ghorbal, I. S. Belkhodja and L. Charaabi, "Direct Virtual Torque
Control for Doubly Fed Induction Generator Grid Connection", IEEE
Transactions on Industrial Electronics, Vol. 56, No. 10, pp. 4163-4173, Oct.
2009.
[63] M. Tazil, V. Kumar, R. C. Bansal, S. Kong, Z. Y. Dong, W. Freitas and H. D.
Mathur, "Three-Phase Doubly Fed Induction Generators: an Overview", IET
Electric Power Applications, Vol. 4, No. 2, pp. 75-89, Feb. 2010.
[64] T. Noguchi, H. Tomiki, S. Kondo and I. Takahashi, "Direct Power Control of
PWM Converter Without Power-Source Voltage Sensors", IEEE Transaction on
Industry Application, Vol. 34, No. 3, pp. 473-479, May/June 1998.
[65] R. Datta and V. T. Ranganathan, "Direct Power Control of Grid-Connected
Wound Rotor Induction Machine Without Rotor Position Sensors", IEEE
Transactions on Power Electronics, Vol. 16, No. 3, pp. 390-399, May 2001.
147
References
[66] D. S. Martin, J. L. Amenedo and S. Arnalte, "Direct Power Control Applied to

Doubly Fed Induction Generator Under Unbalanced Grid Voltage Conditions",
IEEE Transactions on Power Electronics, Vol. 23, No. 5, pp. 2328-2336,
September 2008.
[67] G. Abad, M. Rodrıguez, G. Iwanskiand and J. Poza, "Direct Power Control of
Doubly-Fed-Induction-Generator-Based Wind Turbines Under Unbalanced Grid
Voltage", IEEE Transactions on Power Electronics, Vol. 25, No. 2, pp. 442-452,
Feb. 2010.
[68] L. Xu and P. Cartwright, "Direct Active and Reactive Power Control of DFIG for
Wind Energy Generation", IEEE Transactions on Energy Conversion, Vol. 21,
No. 3, pp. 750-758, Sep. 2006.
[69] E. Tremblay, S. Atayde, and A. Chandra, "Comparative Study of Control
Strategies for the Doubly Fed Induction Generator in Wind Energy Conversion
Systems: A DSP-Based Implementation Approach", IEEE Transactions on
Sustainable Energy, Vol. 2, No. 3, pp. 288-299, July 2011.
[70] F. Iov, A. D. Hansen, P. Sørensen, and F. Blaabjerg, "Wind Turbine Blockset in
Matlab/ Simulink, General Overview and Description of the Models", Aalborg
University , March, 2004.
[71] K. Rajashekara, A. Kawamura, and K. Matsuse, Sensorless Control of AC
Motors, Piscataway, NJ: IEEE Press, 1996.
[72] J. Holtz, "Sensorless Control of Induction Motor Drives", Proceedings IEEE,
Vol. 90, No. 8, pp. 1359-1394, Aug. 2002.
[73] J. Holtz, "Sensorless Control of Induction Machines With or Without Signal
Injection", IEEE Transactions on Industrial Electronics, Vol. 53, No. 1, pp. 7-30,
Feb. 2006.
[74] L. Xu and W. Cheng, "Torque and Reactive Power Control of a Doubly-Fed
Induction Machine by Position Sensorless Scheme", IEEE Transactions
Industrial Applications, Vol. 31, No. 3, pp. 636- 641, May 1995.
[75] L. Morel, H. Godfroid, A. Mirzaian, and J. M. Kauffmann, "Doubly-Fed
Induction Machine: Converter Optimization and Field Oriented Control without
Position Sensor", IEE Proceedings Electric Power Applications, Vol. 145, No. 4,
pp. 360-368, Jul. 1998.
[76] B. Hopfensperger, D. J. Atkinson, and R. A. Lakin, "Stator Flux Oriented Control
148
References
of A Doubly-Fed Induction Machine with and without Position Encoder IEE

Proceedings Electric Power Applications, Vol. 147, No. 4, pp. 241-250, Jul.
2000.
[77] R. Datta and V. T. Ranganathan, "A Simple Position Sensorless Algorithm For
Rotor Side Field Oriented Control of Would Rotor Induction Machine", IEEE
Transactions on Industrial Electronic, Vol. 48, No. 4, pp. 786-793, Aug. 2001.
[78] X. Hao, W. Change and X. Zahay, "A Comparision of Sensorless Control
Strategies of DFIG", 2009 International Conference on Energy and Environment
Technology, Oct. 2009, pp. 3-6.
[79] G. D. Marques and D. M. Sousa, "Air-gap Power Vector Based Sensorless
Method for DFIG Control without Flux Estimator", IEEE Transactions on
Industrial Electronics, Vol. 58, No. 10, Oct. 2011.
[80] J. W. Finch and D. Giaouris, "Controlled AC Electrical Drives", IEEE
Transactions on Industrial Electronics, Vol. 55, No. 2, pp. 481-491, Feb. 2008.
[81] R. Cárdenas, R. Peña, J. Clare, G. Asher, and J. Proboste, "MRAS Observers for
Sensorless Control of Doubly-Fed Induction Generators", IEEE Transactions on
Power Electronics, Vol. 23, No. 3, pp. 1075-1084, May 2008.
[82] A. A. Naassani and A. Ghazal, A. Joukhadar and A. El Shafei, "Speed Sensorless
Control of DFIG Based MRAS Observer", Proceedings of the 14th International
Middle East Power Systems Conference (MEPCON’10), December, 2010, pp.
476-481.
[83] S. M. Gadoue, D. Giaouris, and J. W. Finch, ''Sensorless Control of Induction
Motor Drives at Very Low and Zero Speeds Using Neural Network Flux
Observers'', IEEE Transaction on Industrial Electronics, Vol. 56, No. 8, Aug.
2009, pp.3029-3039.
[84] G. D. Marques, V. F. Pires, S. Sousa and D. M. Sousa, ''A DFIG Sensorless
Rotor-Position Detector Based on a Hysteresis Controller'', IEEE Transactions on
Energy Conversion, Vol. 26, No. 1, March. 2011, pp. 9-17.
[85] L. Gao, B. Guan, Y. Zhou, and Longya Xu, ''Model Reference Adaptive System
Observer Based Sensorless Control of Doubly-Fed Induction Machine'',
International Conference on Electrical Machines and Systems (ICEMS), Oct
.2010, pp. 931 – 936.
[86] T. Ackermann, Wind Power in Power Systems, John Wiley & Sons, 2006.
149
References
[87] R. Cárdenas, R. Peña, J. Proboste, G. Asher, and J. Clare, ''MRAS Observer for
Sensorless Control of Stand-Alone Doubly-Fed Induction Generators'', IEEE
Transaction Energy Conversion, Vol. 20, No. 4, Dec. 2005, pp. 710-718.
[88] B. Boukhezzar and H. Siguerdidjane, ''Nonlinear Control of a Variable-Speed
Wind Turbine Using a Two-Mass Model'', IEEE Transactions on Energy
Conversion, Vol. 26, No. 1, March 2011, pp.149-162.
[89] M. Soliman, O. P. Malik, and D. T. Westwick, ''Multiple Model Predictive
Control for Wind Turbines With Doubly Fed Induction Generators'', IEEE
Transactions on Sustainable Energy, Vol. 2, No. 3, July 2011, pp. 215- 225.
[90] D. Zhi, L. Xu and B. W. Williams, ''Model-Based Predictive Direct Power
Control of Doubly Fed Induction Generators'', IEEE Transactions on Power
Electronics, Vol. 25, No. 2, Feb. 2010, pp. 341-351.
[91] L. Xu, D. Zhi and B. W. Williams, ''Predictive Current Control of Doubly Fed
Induction Generators'', IEEE Transactions on Industrial Electronics, Vol. 56, No.
10, pp. 4143-4153, Oct. 2009.
[92] F. E. V. Taveiros, L.S. Barros, and F.B. Costa, ''Back-to-Back Converter State -
Feedback Control of DFIG (doubly-fed induction generator)-Based Wind
Turbines'', Energy, Vol. 89, issue C, pp. 896-906, 2015.
[93] S. M. Muyeen, Wind Power, INTECH, June 2010.

[94] A. B. Xavier and S. Betty, ''Energetic Macroscopic Representation and Inversion
Based Control Illustrated on a Wind-Energy-Conversion System Using
Hardware-in the Loop Simulation'', IEEE Transactions on Industrial Electronics,
Vol. 56, No. 12, pp. 4826-4835, Dec. 2009.
[95] S. Drid, A. Makouf and M. Naït-Saïd, ''Variable Speed DFIG Power Control with
Wind Turbine Maximum Power Point Tracking'', JHMER, Vol. 2, No. 1, pp. 13-
18, 2011.
[96] G. Simões, and F. Farret, ''Alternative Energy Systems: Design and Analysis with
Induction Generators'', Power Electronics and Applications Series CRC Press,
(2008).
[97] T. Pan, Z. Ji, and Z. Jiang, ''Maximum Power Point Tracking of Wind Energy
Conversion System Based on Sliding Code Extremum Seeking Control'', IEEE
Energy 2030 Conference, 17-18 Nov. 2008.
150
References
[98] B. Shen, B. Mwinyiwiwa, Y. Zhang, and B. T. Ooi, ''Sensorless Maximum

Power Point Tracking of Wind by DFIG using Rotor Position Phase Lock
Loop(PLL)'', IEEE Transactions on Power Electronics, Vol. 24, No. 4, pp. 942-
951, April 2009.
[99] K. Han and G. Chen, ''A Novel Control Strategy of Wind Turbine MPPT
Implementation for Direct-Drive PMSG Wind Generation Imitation Platform'',
Power Electronics and Motion Control Conference, 2009. IPEMC '09. IEEE 6th
International, pp. 2255 - 2259, 17-20 May 2009.
[100] G. Ahmed and A. Kalil, ''Maximum Power Point Tracking Based on Sensorless
Wind Speed Using SVR'', Proceedings of the 14th International Middle East
Power Systems Conference (MEPCON’10), Cairo University, Egypt, December
19-21, 2010, Paper ID 264, pp. 678-682.
[101] S. Bhowmik, R. Spee, and J. H. R. Enslin, ''Performance Optimization For
Double Fed Wind Power Generation System'', IEEE Transactions on Industry
Applications, Vol. 35, No. 4, pp. 949-958, July/Aug. 1999.
[102] K. Tan and S. Islam, ''Optimal Control Strategies in Energy Conversion of PMSG
Wind Turbine System without Mechanical Sensors'', IEEE Transactions on
Energy Conversion, Vol. 19, No. 2, pp. 392-399, June 2004.
[103] L. Fan, Z. Miao, X. Wang, ''Sensorless Maximum Power Point Tracking in
Multi-Type Wind Energy Conversion Systems'', Proceedings of the 48th IEEE
Conference on Decision and Control, 2009 held jointly with the 2009 28th
Chinese Control Conference. CDC/CCC 2009, pp. 6823- 6828, 15-18 Dec. 2009.
[104] S. Nouali and A. Ouali, ''Sensorless Control Strategy for Wind Energy Conversion
System Based on Cascaded Doubly Fed Induction Generator Using Artificial
Neuronal Network'', International Journal of Electrical and Power Engineering,
Vol. 5, No. 1, pp. 42-48, 2011.
[105] C. Feltes, H. Wrede, and F. Koch, ''Fault Ride-Through of DFIG-based Wind
Farms connected to the Grid through VSC-based HVDC Link'', 16th PSCC,
Glasgow, Scotland, July 14-18, 2008.
[106] F. Iov, A. D. Hansen, P. Sørensen, N. A. Cutululis, ''Mapping of Grid Faults and
Grid Codes'', Riso-R-1617(EN), January 2007.
[107] V. Ignatova, P. Granjon, S. Bacha and F. Dumas, ''Classification and
Characterization of Three Phase Voltage Dips by Space Vector Methodology'',
151
References
Future Power Systems, 2005 International Conference, pp. 1-6, 18 Nov. 2005.
[108] S. Choudhury, K. B. Mohanty, B. C. Babu, ''Performance Analysis of Doubly fed
Induction Generator for Wind Energy Conversion System'', The 5th PSU-UNS
International Conference on Engineering and Technology (ICET-2011), Phuket,
pp. 532-536, 2-3 May 2011.
[109] M. B. C. Salles, K. Hameyer, J. R. Cardoso, A. P. Grilo and C. Rahmann,
''Crowbar System in Doubly Fed Induction Wind Generators'', Energies, Vol. 3,
No. 4, pp.738-753, Feb. 2010.
[110] C. Abbey, and G. Joos, ''Supercapacitor Energy Storage for Wind Energy
Applications'', IEEE Transactions On Industry Applications, Vol. 43, No. 3, pp.
769-776, May/June 2007.
[111] A. Petersson, S. Lundberg, T. Thiringer, ''A DFIG Wind Turbine Ride-Through
System. Influence on The Energy Production'', Wind Energy, Vol. 8, No. 3, pp.
251–263, July/Sep. 2005.
[112] P. S. Flannery, and G. Venkataramanan, ''A Fault Tolerant Doubly Fed Induction
Generator Wind Turbine Using a Parallel Grid Side Rectiﬁer and Series Grid
Side Converter'', IEEE Transactions on Power Electronics, Vol. 23, No. 3, pp.
1126-1135, May 2008.
[113] A. H. Kasem, E. F. EI-Saadany, H. H. EI-Tamaly, and M. A. A. Wahab, ''An
Improved Fault Ride-Through Strategy for Doubly Fed Induction Generator-
Based Wind Turbines'', IET Renewable Power Generation, Vol. 2, No. 4,
pp.201–214, Dec. 2008.
[114] S. Foster, L. Xu, and B. Fox, ''Coordinated Reactive Power Control for
Facilitating Fault Ride Through of Doubly Fed Induction Generator- and Fixed
Speed Induction Generator-Based Wind Farms'', IET Renewable Power
Generation, Vol. 4, No. 2, pp. 128–138, Mar. 2008.
[115] L. Peng, B. Francois, and Y. Li, ''Improved Crowbar Control Strategy of DFIG
Based Wind Turbines for Grid Fault Ride-Through'', Applied Power Electronics
Conference and Exposition, 2009. APEC 2009. Twenty-Fourth Annual IEEE, pp.
1932–1938, 15-19 Feb. 2009.
[116] M. B. C. Salles, J. R. Cardoso, A. P. Grilo, C. Rahmann, and K. Hameyer,
''Control Strategies of Doubly Fed Induction Generators to Support Grid
Voltage'', IEEE International Electric Machines and Drives Conference, 2009.
152
References
IEMDC '09, pp. 1551–1556, May 2009.

[117] J. López, E. Gubía, E. Olea, J. Ruiz, and L. Marroyo, ''Ride Through of Wind
Turbines With Doubly Fed Induction Generator Under Symmetrical Voltage
Dips'', IEEE Transactions On Industrial Electronics, Vol. 56, No. 10, pp. 4246–
4254, Oct. 2009.
[118] L. G. Meegahapola, T. Littler, and D. Flynn, ''Decoupled-DFIG Fault Ride-
Through Strategy for Enhanced Stability Performance During Grid Faults'', IEEE
Transactions on Sustainable Energy, Vol. 1, No. 3, pp. 152–162, Oct. 2010.
[119] K. E. Okedu, S. M. Muyeen, R. Takahashi, and J. Tamura, ''Wind Farms Fault
Ride Through Using DFIG with New Protection Scheme'', IEEE Transactions on
Sustainable Energy, Vol. 3, No. 2, pp. 242-254, April 2012.
[120] C. Wessels, and F. W. Fuchs, ''Fault Ride Through of DFIG Wind Turbines
During Symmetrical Voltage Dip with Crowbar or Stator Current Feedback
Solution'', Energy Conversion Congress and Exposition (ECCE), 2010 IEEE, pp.
2771- 2777, 12-16 Sept. 2010.
[121] K. Lima, A. Luna, P. Rodriguez, E. Watanabe, R. Teodorescu, and F. Blaabjerg,
''Doubly-Fed Induction Generator Control Under Voltage Sags'', IEEE Energy
2030 Conference, 2008. ENERGY 2008, pp.1- 6, 17-18 Nov. 2008.
[122] A. Petersson, T. Thiringer, L. Harnefors, and T. Petru, ''Modeling and
Experimental Verification of Grid Interaction of a DFIG Wind Turbine'', IEEE
Transactions on Energy Conversion, Vol. 20, No. 4, pp. 878-886, Dec. 2005.
[123] F. K. A. Lima, A. Luna, P. Rodríguez, E. H. Watanabe, and M. Aredes, ''Study of
a Simplified Model for DFIG-Based Wind Turbines'', IEEE Transactions on
Energy Conversion Congress and Exposition, pp. 345-349, 20-24 Sept. 2009.
[124] H. Polinder, F. F. A. Pijl, G. Vilder, and P. J. Tavner, ''Comparison of Direct-
Drive and Geared Generator Concepts for Wind Turbines'', IEEE Transactions
on Energy Conversion, Vol. 21, No. 3, pp. 725-733, Sep. 2006.
[125] Y. Lei, A. Mullane, G. Lightbody, and R. Yacamini, "Modeling of The Wind
Turbine with a Doubly Fed Induction Generator for Grid Integration Studies'',
IEEE Transactions on Energy Conversion, Vol. 21, No. 1, pp. 257-264, March,
2006.
[126] M. Ali Dami, K. Jemli, M. Jemli, and M. Gossa, "Doubly Fed Induction
Generator, With Crow-Bar System, under Micro-Interruptions Fault",
153
References
International Journal on Electrical Engineering and Informatics, Vol. 2, No.3,

p.p 216 – 231, 2010.
[127] A. D. Hansen, P. Sørensen, F. Blaabjerg, and J. Bech, "Dynamic Modelling of
Wind Farm Grid Interaction", Wind Engineering, Vol. 26, No. 4, pp. 191-208,
2002.
[128] M. Liserre, A. Aquila, and F. Blaabjerg, "Design and Control of a Three-Phase
Active Rectifier Under Non-Ideal Operating Conditions", 37th IAS Annual
Meeting Conference Record of The Industry Applications Conference, Vol.2, pp.
1181–1188, 2002.
[129] E. Carrasco, A. Gómez, R. Amenedo, and J. L. EPS, ''Sensorless Control of
Doubly-fed Asynchronous Machines for WECS Applications'', Universidad
Carlos III de Madrid, España, http://www.aedie.org/9CHLIE-paper-send/279-
garcia.pdf.
[130] A. D. Hansen, C. Jauch, P. Sørensen, F. Iov, and F. Blaabjerg, ''Dynamic Wind
Turbine Models in Power System Simulation Tool Digsilent'', Risoe Report
R1400(EN), Risø National Laboratory, Denmark, 2003.
[131] G. L. Johnson, Wind Energy Systems, Prentice-Hall, 1985.
[132] J. V. Barreras, F. V. Sánchez, ''A Fast Method To Evaluate Annual Energy
Production of Different Off-Shore Wind Farm Configurations Connected to a
VSC-HVDC Line Transmission Including Evaluation of a New Variable Speed
Wind Turbine Concept Based on RCC-WRIG'', Aalborg University, Institute of
Energy Techonology, Msc. Thesis, 2009.
[133] A. M. Eltamaly, A. I. Alolah, and M. H. Rahman, ''Modified DFIG Control
Strategy for Wind Energy Applications'', SPEEDAM 2010, International
Symposium on Power Electronics, Electrical Drives, Automation and Motion, pp.
653-658, 14-16 June 2010.
[134] N. S. Çetin, M. A. Yurdusev, R. Ata, and A. Özdemir, ''Assessment of Optimum
Tip Speed Ratio of Wind Turbine'', Mathematical and Computational
Applications, Vol. 10, No. 1, pp. 147-154, 2005.
[135] W. Qiao, ''Dynamic Modelling and Control of Doubly Fed Induction Generators
Driven by Wind Turbines'', Power Systems Conference and Exposition, 2009.
PSCE '09. IEEE/PES, pp. 1-8, 15-18 March 2009.
[136] J. H. Allmeling and W. P. Hammer, "PLECS - Piece-wise Linear Electrical
154
References
Circuit Simulation for Simulink". IEEE 1999 International Conference on Power

Electronics and Drive Systems, PEDS’99, July 1999, Hong Kong.
[137] M. Rashed and A. F. Stronach, "A Stable Back-EMF MRAS-Based Sensorless
Low Speed Induction Motor Drive Insensitive to Stator Resistance Variation",
IEE Proceedings Electric Power Applications, Vol.151, pp. 685-693, 2004.
[138] G. D. Marques and D. M. Sousa, ''New Sensorless Rotor Position Estimator of a
DFIG Based on Torque Calculations—Stability Study'', IEEE Transactions on
Energy Conversion, Vol. 27, No. 1, March 2012, pp. 196-203.
[139] A. A. Naassani, A. Ghazal, A. Joukhadar and A. El Shafei, "Speed Sensorless
Control of DFIG Based MRAS Observer", International Middle East Power
Systems Conference (MEPCON’10), Cairo University, Egypt, Paper ID 210, pp.
476-481, Dec, 2010.
[140] F. Peng, and T. Fukao, "Robust Speed Identification for Speed Sensorless Vector
Control Of Induction Motors", IEEE Transactions on Industry Applications, Vol.
30, No. 5, pp. 1234–1240, Sep/Oct. 1994.
[141] F. C. Dezza, G. Foglia, M. F. Iacchetti, and R. Perini, "An MRAS Observer for
Sensorless DFIM Drives with Direct Estimation of the Torque and Flux Rotor
Current Components", IEEE Transactions on Power Electronics, Vol. 27, No. 5,
pp. 2576-2584, May 2012.
[142] R. Peña, R. Cárdenas, J. Proboste, G. Asher and J. Clare, "Sensorless Control of
Doubly-Fed Induction Generators Using a Rotor-Current-Based MRAS
Observer", IEEE Transactions on Industrial Electronics, Vol. 55, No. 1, Jan.
2008.
[143] S. M. Shinners, Modern Control System Theory and Design, John Wiley & Sons,
Inc., 2nd edition, 1998.
[144] W. C. Schultz and V. C. Rideout, ''Control System Performance Measures: Past,
Present and Future'', IRE Trans. on Automatic Control, AC-6: 22-35, 1961.
[145] W. Qiao' and R. G. Harley, "Grid Connection Requirements and Solutions for
DFIG Wind Turbines", IEEE Energy 2030 Conference, pp.1-8, 17-18 Nov. 2008.
[146] X. Zheng, and D. Guo, "A Noval Ride-Through Control Strategy of DFIG Wind
Generation Under Grid Voltage Dip", Journal of Information & Computational
Science, Vol. 8, No. 3, pp. 579-591, 2011.
[147] M. Rahimi, and M. Parniani, "Grid Fault Ride Through Analysis and Control of
155
References
Wind Turbines with Doubly Fed Induction Generators", Science direct, Electrical
Power System Research, Vol. 80, No. 2, pp. 184-195, Feb. 2010.
[148] M. Rahimi, and M. Parniani, "Efficient Control of Wind Turbines with Doubly
Fed Induction Generators for Low-Voltage Ride-Through Capability
Enhancement", IET Renewable Power Generation, Vol. 4, No. 3 , pp. 242-252,
May 2010.
[149] A. D. Hansen, and G. Michalke, "Fault Ride-Through Capability of DFIG Wind
Turbines", Renewable Energy, Vol. 32, No. 9, pp. 1594–1610, July 2007.
[150] G. C. Lampley, "Permanent and Temporary Faults", IEEE Industry Applications
Magazine, Vol. 15, No. 5, pp. 25 – 31, Sept/Oct. 2009.
[151] H. L. Zhou, G. Yang, and D.Y. Li, "Short Circuit Current Analysis of DFIG
Wind Turbines with Crowbar Protection", Electrical Machines and Systems,
2009. ICEMS 2009. International Conference, pp. 1 – 6, 15-18 Nov. 2009.
[152] J. D. Glover, M. S. Sarma, and T. J. Overbye, "Power System Analysis and
Design", Wadsworth Publishing co. Inc, 4th edition, 2007.
[153] G. Pannell, D. Atkinson, R. Kemsley, L. Holdsworth, P. Taylor, and O. Moja,
"DFIG Control Performance Under Fault Conditions For Offshore Wind
Applications", 18th International Conference and Exhibition on Electricity
Distribution, 2005, Cired 2005, pp. 1 – 5, 6-9 June 2005.
[154] J. Morren, and S. W. H. Haan, "Ridethrough of Wind Turbines with Doubly-Fed
Induction Generator During a Voltage Dip", IEEE Transactions On Energy
Conversion, Vol. 20, No. 2, pp. 435-44, June 2005.
[155] I. Erlich, J. Kretschmann, J. Fortmann, S. Engelhardt, and H. Wrede, "Modeling
of Wind Turbines Based on Doubly-Fed Induction Generators for Power System
Stability Studies", IEEE Transactions On Power Systems, Vol. 22, No. 3, pp.
909-919, Aug. 2007.
[156] X. Zhang, T. Qu, Z. Xie, and R. Cao, "Dynamic analysis of doubly fed induction
generator during symmetrical voltage swells", IEEE, 2011 Second International
Conference on Mechanic Automation and Control Engineering, pp. 1245–1248,
15-17 July 2011.
[157] H. Xu, X. Ma, and D. Sun, "Reactive Current Assignment and Control for DFIG
Based Wind Turbines during Grid Voltage Sag and Swell Conditions", Power
Electronics, Vol. 15, No. 1, pp. 235-245, January 2015.
156
References
[158] H. Zehua, "Safety Voltage Control Strategy for High Voltage Ride-Through of
Doubly Fed Induction Generator (DFIG) ", International Power, Electronics and
Materials Engineering Conference (IPEMEC 2015), pp. 562-566.
[159] P. M. Anderson, Power System Protection, John Wiley & Sons, INC.,
Publication, 1999.
[160] D. Aouzellag, K. Ghedamsi, and E. M. Berkouk, ''Network Power Flux Control
of a Wind Generator'', Renewable Energy, Vol. 34, No. 3, pp. 615–622, March
2009.
[161] J. Sundvall, M. C. P. Værens, T. Ofeigsson, and M. Swierczynski, ''Dynamic
Control of an Electrical Truck Motor'', Institute of Energy Technology, 8.
Semester Report, spring 2007.
[162] R. Carriveau, Advances in Wind Power, InTech, November, 2012
[163] J. A. Baroudi, V. Dinavahi, and A. M. Knight, " A Review of Power Converter
Topologies for Wind Generators", Renewable Energy, Vol. 32, No. 14, pp. 2369-
2385, Nov 2007
[164] M. Edrah, K. L. Lo, and O. Lara, " Impact of High Penetration of DFIG Wind
Turbines on Rotor Angle Stability of Power Systems", IEEE Transactions on
Sustainable Energy, Vol. 6, No. 3, pp. 759-766, July, 2015.
157
Appendices
Appendices
158
Appendices
APPENDIX [A]
MODEL PARAMETERS
A-1- DFIG [160]
Rated power = 1.5 MW;
Rated R.M.S. line-to-line voltage (stator) = 690 V;
Rated R.M.S. line-to-line voltage (rotor) = 389 V;
Nominal current: In=1900 A;
Frequency = 50Hz;
Rated rotational speed=1450 rev/min;
Number of pole pairs = 2;
Stator resistance = 0.012 Ω.;
Stator leakage inductance (Ls) =0.0137 H.;
Rotor resistance = 0.021Ω.;
Rotor leakage inductance (Lr) =0.0136 H;
Magnetizing inductance (Lm) =0.0135 H;
A-2 Grid Side Filter
The objective of this filter in the design is to enable fast-acting response of the current
controller of the GSC to any variation in current in a time scale of milliseconds. In
addition it is responsible for maintaining a stable controller output.
Filter series inductance (� ) = 0.0006 H;
Filter parallel resistance (� ) = 0.37 Ω;
159
Appendices
1
(1/Lf)*(u[1] -u[2]-(Rf*u[3]))
2 s
Vs_abc Integrator1
Fcn
1
(1/Lf)*(u[1] -u[2]-(Rf*u[3])) 1
s
It_abc
Fcn1 Integrator2
1 1
(1/Lf)*(u[1] -u[2]-(Rf*u[3]))
Vrgsc_abc s
Fcn2 Integrator3
Figure A.1: Simulink model of grid side filter.

A-3 DC- Link
The DC-link capacitor is usually not big enough to smooth the DC-link voltage
variation caused by the alternating rotor current [155].
2 Resistance1
idc_red 1
C vdc
1
s
capacitor xo
1
idc_inv -C- Integrator
init_cond
Figure A.2: Simulink model of the DC-link
Nominal DC Link voltage = 1200 V;

DC Link capacitor = 0.038 F;
Resistor= 0.004 Ω
A-4 Wind Turbine
Blade radius R= 35.25 m;
Gear-box ratio= 90;
160
Appendices
Air density (ρ) = 1.255 Kg/m3
2
Electric machine - Moment of inertia = 90 [kg m ]
5 2
Wind turbine rotor - Moment of inertia: 49.5 ×10 [kg m ]
6
Shaft stiffness: 114×10 [Nm/rad]
3
Damping coefficient of shaft: 755.658×10 [Nm sec/rad]
Friction = 0.0024 Nm/s.
Cut-in wind speed 3 m/s
Optimum tip-speed ratio 6.3
Maximum power coefficient 0.44
A-5 Simulation Controller Parameters
RSC PWM Switching frequency = 5 kHz;
GSC PWM Switching frequency = 5 kHz;
Table A.1: PI controller parameters.
Controller ��
Current control 2 0.016

RSC
Power control 50 5354.55
Current control 33 50
GSC
DC-link voltage control 5 350
PLL 180.03 127.2265
QRMRAS 4 70
RCMRAS 2/70 15
Pitch angle controller 200 60
161
Appendices
APPENDIX [B]
TRANSFORMATION PROCESS OF THE CURRENTS
After the state space vector model parameters of the induction machine have
been calculated, the stator and rotor currents are to be transformed from the two-phase
DC current dq –axis frame back to the three-phase AC quantities. The transformation
process required is the inverse of the two-step voltage transformation used for supply
quantities.
s s
i
e
dq i i abc
d   c a b
Figure B.1: Current Transformation Process [7].
� cos � −sin � �
�{ } [ ]=[ ]⌊ ⌋ (B.1)
� sin � cos � �
� cos � −sin � �
� { ′} [ ]=[ ]⌊ ⌋ (B.2)
� sin � cos � �
Step one is to transform the 2-phase DC currents to their 2-phase AC currents. By

applying the appropriate angular position of the stator and rotor, this will give their
true magnitude and sine wave position [6].
� √ �
�{ } [� ] = − [ ] (B.3)
�
� √
[− − ]
162
Appendices
� √ �
� { ′} [� ] = − [ ] (B.4)
�
� √
[− − ]
Step 2 transforms the current quantities from their 2-phase alpha beta magnitudes and
positions to their 3-phase magnitude and positions. The ratio applied keeps the 2-
phase to 3-phase magnitudes and positions in equal quantities; therefore, the output
for the 3-phase system is balanced providing the input is also balanced [6].
163
Appendices
APPENDIX [C]
PULSE GENERATION OF THE CONTROLLED

SWITCHED MODULATION
The design of the modulation technique should allow for the conversion of the bi-
directional power flow. The switching design has to be capable of the inversion and
rectification of the bi-directional power flow.
The switching illustrated in figure C.1 represents the rectification of the 3-phase AC
voltages to DC voltages through SPWM.
The switching algorithm is:
̂
�
� = ̂
(C.1)
�
�
� = (C.2)
�
Definition of switching states
� >� ∴ �� =�
� <� �� ∴ � � =
� >� ∴ �� =�
� <� �� ∴ � � = } (C.3)
� >� ∴ �� =�
� <� �� ∴ �� =
� = = ∗�
The triangular wave form overlapping the 3 sinusoidal waveforms are the pulses
from the switching on/off of the IGBT power devices, through a successfully designed
switching algorithm, the back to back inversion and rectification for voltage
transformation can be used for the RSC and GSC of a DFIG system.
164
Appendices
Figure C.1: Output voltages of two level converters with sinusoidal SPWM [6, 7].
165
Appendices
Te
Teta_r Vsabc
[Vsabc]
[W_gen] [Vdqs] vs_dq Teta_s Teta_s
Wrot 7
1 teta_r
Te
Goto22 Vsdq [Vdqs]
is_d A
Isd Is_abc ee_phase VabcA
described in detail.
W_gen omg_mec a 13
sensor [Isd] 2 B
b teta_s Teta_s
W_gen
Iq ref1 WR_IM is_q Isq B1 cC
Beta Trot
Vrd [Isq]
Teta_r Teta_r 3 B23
sensor vr_d Teta_sl Teta_sl
4 Vwind (m/sec)
C Vgrid
ird Ird
Subsystem s
Vrq
Pitch_control + Vs_alpha_beta
vr_q -
irq Irq
Vwind
Aerodynamic_system Wound Rotor grid voltage and
Induction Machine s slip angle computation
constant parameters1 +
Teta_sl4 -
Vrgsc_abc
It_abc
D-1 Complete DFIG system
[Vsabc] Vs_abc
From54 Filter
t Teta_sl3
vdc
Clock1
DC
DC-link circuit2
Link
Continuous
idc_inv
idc_red
From52
powergui
theta_actual_final [Irabc]
166
Teta_sl Trigonometric
Clock Start time
Function Divide1
>= atan
-C- Iqr_ref Idr_ref P_ref Q_ref
Qrmras1
Vdq*
APPENDIX [D]
GSC_Current control1
P_ref
Q_ref
Iqr_ref
Idr_ref
Vs_alpha_beta
Vdr_star
Vqr_star
SIMULINK MODEL
Power and Current control_RSC
vdc
Iq*
Id*
Igabc1
Vgabc1
theta1
[Isabc] Is_abc theta_r_est
Irmras
fernce_thet
[Irabc] Ir_abc
P_meas
Q_meas
Irq
Ird
Vdqs
W_gen
From11 Iq ref
IrMras2 [Vsabc] From23
[W_gen] 0
P_meas
[Vdqs]
Q_meas
Id*
Irq Ird GSC_Voltage control
Vdc
Vdc*
Figure D.1: Block diagram of the DFIG wind turbine model.

Vdc ref1 1200
source converters are applied in the electrical circuit as controllable voltage sources.
rotor-side PWM converter control blocks. The output voltages of the PWM voltage
The DFIG wind turbine model mainly comprises the turbine block, DFIG,
source converter and the grid and rotor-side PWM converter control blocks are
transformer, PWM voltage source converters, the DC-link block and the grid- and
In the following sections, the functions of the turbine block, DFIG, PWM voltage
Figure D.2: Diagram of the aerodynamic block in the DFIG wind turbine model.
3 u(1)^3 In1
wind_power
Vwind(m/s) radius wind_power(w)
Vwind^3
compute_wind_power
Divide
1
Wrot_speed(rad/s)
Avoid division
by zero1
Tturbin
wr
AND 1
cp
Cp
Trot(Nm)
35.25 raduis1
radius
167
wind m/s lamda
lamda
2 beta >=
3 Switch
beta(deg)
limiter_beta cut_in
Subsystem
<=
25
cut_out
D-2 Aerodynamic Model
0
null
Appendices
Appendices
D-3 Pitch Angle Controller Model
speed_limite
Vwind (m/sec)
Gene
200
1
atan(u[1]/35.25*u[2])
Sigma
alpha_optim
kp=200
ki=60
Fcn
8.1
PI
0-5
B_optim
0-90
1
s
Beta
1
Figure D.3: Diagram of the pitch angle control block in the DFIG wind turbine model.
168
[Iqr]
in the state variable form using generalized machine theory. The stator and the rotor
The model is shown in Figure D.4. The wound rotor induction generator is modelled
windings are transformed into their 2-phase equivalents using the dq0- transformation.
4 [W_gener]
3 Vds W_gen3 From8
Goto3 (3*P*u(2)*Lm*u(1))/(2*Ls*2*pi*Fs) 1
Vdqs [Vqs] Te
[Vqs]
Goto1
From
[Iqr] (Lr-(Lm^2/Ls))
From9 (Lr-(Lm^2/Ls))s+Rr
Figure D.4: Diagram of the mathematical model of DFIG.

2*pi*Fs Divide4
Constant4 2
Vdr
Add4
[W_gener] 1 [Idr] Vdqr
From5 (Lr-(Lm^2/Ls))s+Rr
Add Goto2
[Vqss] Idr 5
(Lm/Ls)s
From11
(Lr-(Lm^2/Ls))s+Rr
2*pi*Fs Divide2
Constant2
2*pi*Fs
Constant3
4
Add3 1 Iqr
[Vqs] [W_gener]
Vqr
169
From1 From6 1
[Iqr]
(Lm/Ls) (Lr-(Lm^2/Ls))s+Rr
2*pi*Fs Add1 Goto
Divide Divide3
Constant
[Idr] (Lr-(Lm^2/Ls))
From3 Add2
[Idr]
[Vqs] From4 (u(2)-(Lm*u(1)))/Ls 3
[Iqr] u(1)*Lm*(-1)/Ls 2
Isq From2 Ids
From10
D-4 Generator Model
2*pi*Fs Divide1
Constant1
Appendices
The PWM converter block, built with custom components, is shown in Figure D.5.
Figure D.5: Block diagram of the PWM.

sin((u(1)*u(2)*2*pi)-(pi/2)) asin 2/pi
carrier_frequence Gain1
1
(u(1)*cos(u(4))+u(2)*sin(u(4)))*sqrt(2/3)
d Switch2
2 Relay1
D-5 PWM Voltage Source Converter Model
Ref b
q (u(1)*cos(u(4)-((2*pi)/3))+u(2)*sin(u(4)-((2*pi)/3)))*sqrt(2/3) 2
0 vr_abc*1
Relay2 Switch1
Ref c switching
(u(1)*cos(u(4)+((2*pi)/3))+u(2)*sin(u(4)+((2*pi)/3)))*sqrt(2/3)
3
teta_sl Relay6
170
Switch
4 0.5 -1
Vdc i_ph
Gain
5
DC_link current
Idc
1 0.5*(u(1)+u(2)+u(3))
Appendices
Appendices
D-6 GSC and RSC Model
Figure D.6: Diagram of inner and outer loops control blocks in the RSC and GSC.
171
Appendices
D-7 Stator Angle Estimator Model

The stator voltage angle is determined by a PLL and a VCO as shown in Figure D.7.
sin
2/3*u(1)-(u(2)+u(3))/3 f(u)
1 vx Mux cos(gama) 1
ref out 1
s
v_abc
-(u(2)-u(3))/sqrt(3) f(u) teta_s
vy sin(gama) PI anti wind-up2

cos 2*pi*50
Figure D.7: Diagram of PLL used to compute the stator voltage angle.
D-8 Rotor Reactive Power MRAS (QRMRAS) Model
1 Vr_alpha_beta
Vr_alph_beta
Vr_abc Qr_ref _model
abc_alpha beta1 Qr_ref _model
Ir_alpha_beta1
2
Refernce_model
Ir_abc
abc_alpha beta2 PI In1Out1
mod 1
integration theta_r_est
Ir_alpha_beta3
3
Is_alpha_beta2
Is_abc
abc_alpha beta4 Qr_adap_model 2*pi
wr
angle
Adaptive_model
Figure D.8: Diagram of rotor reactive power-based MRAS (QRMRAS).
172
Appendices
D-9 Rotor Current MRAS (RCMRAS) Model
3
Ir_abc
abc_alpha beta
Product4
PI
mod 1
theta_r_est
Vs_alpha_beta Product3 2*pi

Vs_alpha_beta
1
2 Is_abc Iralpa
Is_abc
angle
Subsystem10
Figure D.9: Diagram of rotor current MRAS (RCMRAS) subsystem.
D-10 Crowbar Model
|u| 2
Product2
Product3
2 sqrt
1 emu > NOT
triger crowbar 3 emu 1
Idqr*
Relational vdqr from RSC vdqr_ref
Operator1
2
|u|
Product Product5
3900
crowbar limit currnt
Figure D.10: Diagram of Crowbar subsystem.
173
Appendices
APPENDIX [E]
MATLAB INITIALISATION CODE
E-1 Appendix Model Parameter

%Generated power=1.5 MW
%Rated voltage (stator):398/690 V.
Vs=398;
%rated voltage (rotor):225/389 V
Vr=225;
% nominal current: In=1900 A.
%Rated rotational speed :1450 rev/min
% slip: g
p=2;
Rr=0.021;
Rs=0.012;
Ls=0.0137;
Lr=0.0136;
Lm=0.0135;
g=0.03;
f=50;
ws=2*pi*f;
G=90;
R=35.25;
Tr=Lr/Rr;
Ts=Ls/Rs;
sig=1-((Lm^2)/(Ls*Lr));
k=Lm/(sig*Ls*Lr);
Y=1/(sig*Tr);
T_ond=0.01;
E-2 Plotting Tip Speed Ratio Variation with Power Coefficient and Pitch
Blade Angle
%plotting curve of lambda with power coefficient and
%pitch blade angle
clc
clear all
%Cp parameters from Thomas Ackermann, ?Wind Power in Power Systems?
c1=0.73;
c2=151;
c3=0.58;
c4=0.002;
c5=2.14;
c6=13.2;
c7=18.4;
c8=-0.02;
c9=-0.003;
[lambda,Beta]=meshgrid(1:0.5:15,1:1:30);
vwind=(3:1:25); %wind-speed [m/s];
174
Appendices
R=35.25; %radius of the rotor or the blade length [m];

%in next steps the calculation of the cp curve for different
lambda_i=1./(lambda+c8.*Beta)-c9./(Beta.^3+1);
cp = c1 * (c2*lambda_i - c3*Beta - c4.*Beta.^c5 -c6).* exp( -
c7*lambda_i);
cp=1/2*(cp+abs(cp)); %only positive values of Pref
mesh(lambda,Beta,cp)
surf(lambda,Beta,cp)
E-3 Calculation of the Annual Weibull Distribution of the Wind [132]
%Calculation of the annual WEIBULL distribution of the Wind

clc
clear all
V=(4:1:25); %Wind speed [m/s]
% constant of Weibull parameters from IEC classification
c=2;
a=11.38;
%Calculation of the distribution
P=365*24*(c./a.^c).*Vwind.^(c-1).*exp(-(Vwind./a).^c);
plot (Vwind,P)
title('annual wind distribution curve based on weibull distribution
function')
xlabel('wind speed (m/s)')
ylabel('time (hourse/year)')
175
Appendices
APPENDIX [F]
PI CONTROLLER DESIGN IN Z DOMAIN
The general equation for PI control in the time domain as follows [161]:
= � + �� ∫ (F.1)
where is the output signal, is the error signal, ‘� ’ is the proportionality

constant and ‘�� ’ is the integration constant. PI control in z-domain for a backward
transform previous equation is expressed as:
�
= � + �� (F.2)
� � −
�
where is the sampling time. In Figure (F.1) is presented the diagram of the PI in z
domain.
kp +
E (z ) +
U (z )
ki Ts +
+
1
z
Figure F.1: PI control in the z-domain using backward transform.
176

Performance Analysis of Doubly-Fed Induction Generator (DFIG) - Based Wind Turbine With Sensored and Sensorless Vector Control

Uploaded by

Copyright:

Available Formats

Performance Analysis of Doubly-Fed Induction Generator (DFIG) - Based Wind Turbine With Sensored and Sensorless Vector Control

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Performance Analysis of Doubly-Fed Induction Generator (DFIG) - Based Wind Turbine With Sensored and Sensorless Vector Control

Uploaded by

Copyright:

Available Formats

Performance Analysis of Doubly-Fed Induction Generator

(DFIG)- Based Wind Turbine with Sensored and

Amer Obaid Kareem

Student Number: 089112111

A thesis submitted for the degree of

School of Electrical and Electronic Engineering

One of the main objectives of this study is to investigate the comparative

The behavioral response of the DFIG-based wind turbine system to grid

From the depths of my heart, I would like to express my sincere gratitude to my

Of course, no acknowledgments would be complete without giving thanks to my

Special appreciation goes to the School of Electrical and Electronic Engineering,

Acronyms and Abbreviations

List of Figures xiv

List of Tabels xxi

1-1 Overview of Wind Energy Conversion System Technology 2

1-2 Wind Power Generators 5

1-3 Research Methodology and Challenges 13

1-4 Objectives and Contributions of the Study 14

1-5 Thesis Layout 16

Chapter 2 Literature Survey 18

Modeling and Control System of the DFIG

2-3 Sensorless Control of the DFIG-Based Wind Turbine 22

2-4 Maximize Power Point Tracking (MPPT) in DFIG-Based Wind 25

2-5 Fault Ride-Through (FRT) of the DFIG-Based Wind Turbine 29

2-6 Wind Turbine Topologies 34

Chapter 3 Mathematical Model of the Doubly-Fed Induction Generator System 38

3-2 Dynamic Model of Doubly-Fed Induction Generator 38

3-3 Vector Control System 42

3-3-1 Vector Control of the Rotor Side Converter (RSC) 43

3-3-2 Vector Control of the Gride Side Converter (GSC) 51

3-3-3 Control of the DC-Bus Voltage 57

3-4 Drive Train Model 60

3-5 Wind Turbine Power Characteristics 61

3-5-1 Annual Wind Distribution 61

3-5-2 Aerodynamic Characteristics of the Wind Turbine 62

3-5-3 Tip Speed Ratio (TSR) of the Wind Turbine 64

3-6 Test System 65

3-7 Phase Locked Loop (PLL) System Performance 67

3-8 DFIG System Model Confirmation 72

3-9 Simulation Results of Modelling the DFIG system 77

4-2 Sensorless Control of DFIG System Using MRAS Observer 88

4-2-1 Rotor Current-Based MRAS (RCMRAS) 89

4-2-2 Rotor Reactive Power-Based MRAS (QRMRAS) 90

4-3 Simulation Results 91

Chapter 5 Behavioral Analysis of DFIG System with Grid Disturbances 100

5-1 Introduction 100

5-2 Grid Codes Requirements of DFIG System 100

5-3 Fault Ride-Through (FRT) of DFIG System 102

5-5 DFIG Analysis During Voltage Swell 108

5-8 Detection of Disturbances in the DFIG System 114

5-9 Performance of Grid Connected DFIG System Grid Without 117

5-9-1 Case of 80% Voltage Dips 117

5-9-2 Case of Voltage Swell of 1.3 p.u. 125

5-10 Performance of Sensored and Sensorless Vector Control of DFIG 130

5-11 Conclusions 136

Chapter 6 Conclusions and Further Works 138

6-1 Research Summary 138

6-2 Contributions 139

6-3 Further Work 139