The 16th CSI International Symposium on Artificial Intelligence and Signal Processing (AISP 2012)

Load Frequency Control of Two Area Interconnected

Power System (Diesel Generator and Solar PV) with
PI and FGSPI Controller

Seyed Ali Jeddi#1,Seyed Hamidreza Abbasi#2,Fereydoon Shabaninia#3

School of Electrical and Computer Engineering
Shiraz University
s.ali.jeddi@gmail.com#1, Hamid_abs62@yahoo.com#2,Shabaninia@yahoo.com#3

Abstract—In this paper,we aim to design and control the DC voltage generated to AC, we can created and form a power
power frequency which is produced by a hybrid generator grid network and utilize it individually or through a wide
network(Diesel Generator and Solar PV Panels) while network connection if it be applicable.[3,4] Since the sun is
there are sudden changes in the connected load to the the primary source of energy in our network, there are
network. We call this type of control as the “Load multiple factors which are influencing the amount of produced
Frequency Control” and in abbreviation LFC. We will
consider the changes as a step to the solar power network.
electricity, these factors could be pointed like the change in
At first, generator network and PV are simulated by the intensity of sunlight, the amount of dust, weather condition
transfer functions and then they will be connected to each and etc [5,6]. Regarding what was went above, Frequency and
other (Tie-Line). Then the produced disturbance will be voltage control in the electricity PV grid is being discussed
controlled through two different types of PI controller and here and therefore, lots of different control methods like
FGSPI controller (Fuzzy Gain Scheduled P-I) and the Intelligent Fuzzy and Neural controllers and MPPT method
corresponding responses in each networks will be assessed. has been suggested so far. [7,8]
Also the connection location of networks (Tie-Line) will be It is observed that after connecting two networks together,
checked and the operations of both controllers are the changed and produced disturbance in each of the areas
comprised. It is obvious from the response analysis of
suggested FGSPI controller that it works more properly
cause to effect on both produced voltage and frequency of
and accurately than the PI controller.
other regions. Therefore the employed controllers must be
capable of quick and correct responses in each area and also
have this capability among the connection path.
Keywords- PV; Load Frequency control(LFC); Power System There are different suggested methods to control the
;Fuzzy Gain Scheduli produced electricity frequency of two generator networks [9,
10]. In all these studies, due to the steam valve availability in
I. INTRODUCTION the steam generator considering a regulator, the amount of
Nowadays, employing the renewable energy resources steam entering the turbine will be controlled and as a result,
such as wind and solar energy is highly regarded to provide the frequency can be controlled consequently. But in a PV
the power which is needed for the human life. electricity grid, we will employ the MPPT method to control
Heretofore, the electricity generators supply the required the produced DC voltage. Then, an inverter should be
power and there are lots of researches around to control the considered to convert the DC voltage to AC. In this case, a
voltage and frequency and eliminate the systems confusion DC-DC Boost Converter makes the MPPT output always
effects [1,2]. under control and subsequently it leads to gain a stable value
On the other hand, the PV solar panels are also due to low for reference voltage even with having changes in the
cost energy source are highly regarded throughout the world radiation levels and climatic conditions.
by Connecting multiple solar panels together and convert their

978-1-4673-1479-4/12/$31.00 ©2012 IEEE 526

 In this paper, a power generator and a PV solar network After PV Panels we use MPPT controller for regulating PV
have been connected to each other and then the effect of a step output voltage and after that we utilize BOOST CONVERTOR
disturbance has been explored on the output frequencies of the to achieve AC regulating voltage .The relevant circuit shown
networks and either along the Tie-Line route. Finally, the below:
output responses of the two different types of PI controller and
FGSPI controller are illustrated and the corrsponding results
are compared
II. MODELING THE HAYBRID NETWORK

A. PV network
At First, we model the solar cell as the smallest part of PV
Panel. Solar cell as the basic block is introduced in PV Panels
and each panel will be made of connecting these small cells in
Figure.2 Boost Convertor equivalent circuit
series – parallel method .Each cell acts as an p-n diode which
allows current to pass from one hand to the other hand .Circuit U switch shown in this figure is composed of an IGBT and
model of a PV is as follows: Diode .When U=0 , Diode is ON and IGBT is OFF and vice
versa.
Switch operation can be divided into two different time
periods [11] .First period occurs when the switch is turned ON,
it means ,the relevant equations are as follows:

(3)

(4)

Another time period is belong to turned OFF switch, it

means :

(5)
Figure.1 Equivalent circuit of solar cell
(6)
In this diagram, is used as the reference current and
is equivalent to the total resistance .So that the equations are as
follows: TS is a switching period time. But designing a Boost
Convertor for PV need to consider many factors, because
its input is affected by solar radiation and output always is
(1) affected by load changes. Also the output voltage to follow
a reference voltage MPPT controller is constantly
(2)
changing. Profile for the Boost Convertor is intended to
follow:
Each of the above values is:
Short Circuit Current ( A )
Current ( A )
Current ( A )
Current ( A )
Current ( A )
(C)
( )
olar Irradiance ( ) = 1000 Based on the data and information presented so far,
o
( C ) the following transfer function can be used for PV
Transfer Function:
)7)

527
B. Diesel Generator Network System response after after inserting disturbance to the
Frequency control in power diesel generator is be possible network load for each area and Tie-Line is given in Part C.
with employing of control valve to adjust the input steam to B. FGSPI
turbine .After the Governor and then Reheated , Generator Gain scheduling is a technique commonly used in designing
located at the end of each with a unique transfer function that controllers for systems whose dynamics change nonlinearly
each of them have. We consider the first order model for all of with operating conditions [12].In this paper, the fuzzy controller
them [10] .So, we will have the below transfer functions: output is used for determining K in PI controller .The below
figure show the system response for different Ki:
(8)
Generator

Turbine (9)

Governor (10)

Reheater (11)

Values as parameters in the above transfer functions, is given

in the Appendix. A simple form of diesel generator network is
as follows:
Figure.5 System response to different Ki
Based on the above figure, although increasing of Ki
value will cause fluctuation in response, the minimum
deviation from the initial value will achieve. Unlike the
previous case, decreasing of K value will be caused the
large deviation from set point, but with lower volatility
leads to the set point.
Figure.3 Diesel Generator model With using of fuzzy controller we can select the best value
for Ki in each time. Fuzzy controller will follow any change
and try to return set point to initial defined value. Our fuzzy
III. PROPOSED CONTROLLER controller will has tow Inputs. They are considered as ACEi
In this section two different controllers, PI and Fuzzy Gain and dACEi/dt that will obtained from the difference between
Scheduling for PI (FGSPI) for Load Frequency Control has set point and feedback signal. Output of fuzzy controller is Ki.
been suggested. The PI controller is a simple controller and The Mamdani fuzzy inference engine was selected and
FGSPI is a intelligent controller. realized by five triangular membership functions for each of
A. PI Controller the three linguistic variables (ACEi and dACEi/dt,Ki) with
suitable choice of intervals of the membership functions as
In this part, a simple PI controller is used to control of load shown in Fig. 6 and Table. 1. We have the below
frequency in two interconnection hybrid power system (Diesel
abbreviations for each of them:
Generator and PV Panels). Block diagram of the system would
look like the below figure:
NB : Negative Big , NS : Negative Small , Z : Zero
PB: Positive Big , PS : Positive Small

Figure. 4 block Diagram of System with PI controller

528
 Figure7. FGSPI schematic Diagram
IV. SIMULATION RESULT
Result of the hybrid power network (Diesel Generator and
PV) load frequency control with two type of controller (PI and
FGSPI) in each area and Tie-Line connection is as follows:

Figure.6 Membership functions for linguistic variables

ACEi
NB NS Z PS PB
dACEi/dt Figure.8 PV Network Output with PI Controller
NB PB PB PB PS PS
NS PB PB PS PS PS
Z PS PS Z NS NS
PS NS NS NS NB NB
PB NS NS NB NB NB

TABLE 1. FUZZY RULE BASE FOR FGSPI CONTROLLER

The Centriod method has been used for Defuzzification.

The system block diagram of two area interconnection
hybrid power system with FGSPI is similar to the
system block diagram with PI controller except that
instead of PI controller block , FGSPI controller block
,as shown below ,is used. Figure.9 Diesel Generator Network Output with PI Controller

529
 Figure.10 Tie-Line Output with PI Controller Figure.13 Tie-Line Output with FGSPI Controller

V. COUNCLUSION
The connection between two different type of electricity
generator,Diesel Generator and solar PV, being adaptive and
intelligent is the most important factor for controller. The
proposed intelligent controller is not only simple in design but
also easy to implement. Using of proposed FGSPI will be
accompanied with better response when we have load
disturbance in any area of power generation. It is expected that
the controller will perform effectively under different
operating conditions. Simulation results obtained demonstrate
the usefulness of the proposed controller.
VI. APPENDIX
, = 0.33 , , , = 10 ,
Figure.11 PV Network Output with FGSPI Controller ,=0.8 , a12=a22=-1 , R=0.4

REFERENCE

[1] Kroposki, B.; Vaughn, A. DG Power Quality, Protection, and

Reliability Case Studies Report, NREL/SR-560-34635. Golden, CO:
National Renewable Energy Laboratory. General Electric Corporate
R&D, 2003.
[2] Shigeyuki Yanagawa, Takeyoshi Kato, Wu Kai Akimori Tabata,
Yasunobu Yokomizu,Tatsuki Okamoto,Yasuo Suzuoki," LOAD
FREQUENCY CONTROL", IEEE Transactions on Energy Conversion,
Vol. 11, No. 1, March 1996
[3] Miller, N.; Ye, Z. Report on Distributed Generation Penetration Study.
NREL/SR-56034715. Golden, CO: National Renewable Energy
Laboratory, 2003.
[4] Manoj Datta, Tomonobu Senjyu, Atsushi Yona, , Toshihisa Funabashi,
and Chul-Hwan Kim, "A Cordinated Control Method for Leveling PV
output Power flactuatins of PV-Diesel Hybrid System connected to
Figure.12 Diesel Generator Network Output with FGSPI Isolated Power Utility" IEEE TRANSACTIONS ON ENERGY
CONNVERSION VOL.24,NO.1,MARCH2009
Controller [5] Chalmers, S.; Hitt, M.; Underhill, J.; Anderson, P.; Vogt, P.; Ingersoll,
R. “The Effect of Photovoltaic Power Generation on Utility Operation.”
IEEE Transactions on Power Apparatus and Systems; PAS-104, 1985,
pp. 524–530.
[6] EPRI report EL-6754, Photovoltaic Generation Effects on Distribution
Feeders, Volume 1: Description of the Gardner, Massachusetts, Twenty-
First Century PV Community and Research Program,” 1990.
[7] Manoj Datta, Tomonobu Senjyu, Atsushi Yona, , Toshihisa Funabashi,
and Chul-Hwan Kim, "A Cordinated Control Method for Leveling PV
output Power flactuatins of PV-Diesel Hybrid System connected to

530
 Isolated Power Utility" IEEE TRANSACTIONS ON ENERGY conference on Communication System and Neural Network
CONNVERSION VOL.24,NO.1,MARCH2009 Technologies,2011
[8] C.S.Chain,K.T.K.Teo , "Fuzzy Logic Based MPPT for Photovoltaic [11] A.J. Mahdi, W.H. Tang and Q.H. Wu " Improvement of a MPPT
Modules Influenced by Solar Irradiation and Cell Temperature" UKSim Algorithm for PV Systems and Its Experimental Validation"
13th International Conference on Modelling and Simulation,2011 International Conference on Renewable Energies and Power Quality
[9] Francoise Beaufays,Youssef Abdel-Majid , Bernard Widrow (ICREPQ’10) Granada (Spain), 23rd to 25th March, 2
"Application of Neural Networks to Load-Frequency Control in Power [12] Zhen-Yu Zhao, Masayoshi Tomizuka and Satoru Isaka, “Fuzzy gain
System"Neural Network, Vol7, No.1,pp 183-194 1994 scheduling of a PID controllers”, IEEE Trans. Syst., Man Cyber. 23 (5)
[10] Sathans,Akhilesh Swarup “Intelligent Load Frequency of Two-Area (1993)
Interconnected Power System and Comparative Analaysis” International

531