Proceedings of the 14th International Middle East Power Systems Conference (MEPCON’10), Cairo University, Egypt, December 19-21,

2010, Paper ID 166.

Speed Control of Switched Reluctance Motor Based

on Fuzzy Logic Controller
Gamal M. Hashem Hany M. Hasanien
gamalhashem@eng.asu.edu.eg hanyhasanien@ieee.org
Department of Electrical Power and Machines, Ain Sham University, Cairo, Egypt

Abstract—This paper presents a Fuzzy Logic Controller for

Switched Reluctance Motors (SRM). The fuzzy logic controller is
utilized to control the SRM speed. The dynamic response of the
SRM controlled by the proposed controller is studied during the
starting period and under different loading conditions. The
effectiveness of the proposed fuzzy logic controller is then
compared with that of the conventional PI controller. The
dynamic simulation is performed on C program. Selected
experimental results are reported to verify and to validate the
effectiveness of the proposed system.

Index Terms — Fuzzy logic controller, Switched reluctance

motor, PWM Inverter.
Fig. 1. Construction of three-phase, 6/4 SRM.
I. INTRODUCTION

T HE Switched Reluctance Motor is a doubly salient

machine in which torque is produced by the tendency of
II. DYNAMIC MODEL OF SRM
The mathematical model of the SRM consists of three basic
the rotor to move to a position where the inductance of the
excited windings is maximized (the aligned position) [1]-[2]. sets of equations:-
The SRMs are considered to be attractive solutions for 1- The motor phase equations.
variable speed applications with high power density. The 2- The mechanical equation.
SRM possesses many inherent advantages such as simplicity, 3- The angular speed equation.
robustness, low manufacturing cost, high starting torque, high The motor phase equations which describe the electrical
speed, and high efficiency [3]-[4]. On the other hand, the behavior of the SRM are defined as follows:
stator windings are concentrated and no windings, no brushes
on the rotor, as shown in Fig. 1. In addition to this, only dψ k (θ k , ik )
= ±V − R ik (1)
simple converter circuit with reduced number of switches due dt
to unidirectional current requirements are needed [5]-[6]. Where, ψK is the phase k flux linkage as a function of the
These advantages make this type of motors, a competitive current and the rotor position, V is the supply voltage, R is the
choice to both the dc series motor and the squirrel cage winding resistance per phase, and i is the phase current.
induction motor [7]. The SRM can be used for general The mechanical equation which describes the mechanical
purpose industrial drives. The motor ability to operate in the motion of the motor is defined as follows:
four quadrants and its suitability for hazardous areas open a
wide range of applications for switched reluctance motor
dω 1 q
drives including mining, explosion proof machinery, traction = (∑ Tk (θ k , ik ) − Tl ) (2)
and domestic applications. Recently, a good deal of the dt J k =1
research work focus on SRM control and torque smoothness
in order to make it a competitor to both fully controlled dc and Where:
ac drives [8]. ω is the rotor speed, J is the moment of inertia of both the
The paper proposes an improved controller based on fuzzy rotor and load, Tl is the load torque, q is the number of phases,
logic technique. The controller effects on the motor dynamic Tk(θk , ik) is the torque produced by the Kth phase and θk is
response are evaluated. the rotor position as seen by the Kth phase.
The angular speed equation is defined as follows:

288
 dθ
ω= (3)
dt
These equations are representing the dynamic model of the
SRM. They are solved simultaneously using Numerical
integration technique with the aid of the motor look up tables
[5]. Fig. 3. Block diagram of FLC.

The membership functions were defined off-line, and the

III. SRM SPEED CONTROL WITH THE PROPOSED values of the variables are selected according to the behavior
CONTROLLER of the variables observed during simulations. The selected
fuzzy sets for FLC are shown in Fig. 4. The control rules of
The block diagram of the system under study is shown in the FLC are represented by a set of chosen fuzzy rules. The
Fig. 2. It consists of an SRM with its controlled and power designed fuzzy rules used in this research are given in Table 1.
supply. The motor data is given in the appendix -a. The drive The fuzzy sets have been defined as: NL, negative large, NM,
is tested under the following condition: First, the motor is negative medium, NS, negative small, ZR, zero, PS, positive
started against its full load torque until the motor reaches rated small, PM, positive medium and PL, positive large
steady state speed. Second, the motor is subjected to a severe respectively.
load disturbance where the load is suddenly decreased to 50%
from its initial value, followed by a removal of the disturbance
after the motor reaches its steady state speed.

Fig. 2. Block diagram for speed control of SRM using the FLC.

IV. THE FUZZY LOGIC CONTROLLER

The fuzzy logic controller (FLC) is proposed for SRM
controller in this study. The aim of it is to control the motor
Fig. 4. The membership functions of FLC.
speed and to enhance the speed regulation. The final output of
the fuzzy logic controller is used to regulate the switching-on
angle of the inverter to regulate the motor shaft speed. The Table 1. The rules of FLC.
inputs of the controller are the motor speed error e(t) and the
change in the error Δe(t).
The motor speed error e(t), is define as following:

e = (ωref – ω) (4)

In the FLC, The reference speed ωref is compared with the

actual speed ω to get the speed error e(t) as shown in Fig. 3.
Also this error is compared with the previous error e(t-1) to
get the change in error Δe(t). The inputs of FLC are e(t) and
Δe(t). The output of the proposed controller is Δθonref(t) which
is added to the previous state θonref(t-1) to get the θonref(t). V. PERFORMANCE EVALUATION
The SRM is tested with the same load torque variations
stated before in section III Where in this section the
descriptions of the system under study.

289 289
 A. Dynamic response during Starting (Zone one) also yields a much faster response that allows the motor to
For a good motor performance during starting, the design reach the steady state after 20 ms, while in the PI controller, it
requirements will be as follow: reaches the steady state after 60 ms.
The maximum overshooting is very small and it tends to zero,
the rise time is less than or equal to 0.01 sec, the settling time
is less than or equal to 0.04 sec, the steady state value is less
than or equal to 0.15 [9]. Therefore, according to these
requirements the damping ratio ζ equals 0.6, the undamped
natural frequency ωn is 180 rad/sec, the damped frequency ωd
is 144 rad/sec.
The new switching on angle can be expressed using the
following formula:

θon new = θon ref - θon actual (5)

Fig. 5, shows the dynamic response of the motor during

Fig. 5. The dynamic response of the motor during starting driven by the FLC
starting when provided with the FLC as compared with the PI as compared with the PI controller
controller of gains kp = 0.005 and ki = 5, where these values
represent the optimal values of the PI controller parameters.
As shown in this figure, it can be realized that the dynamic
response of the SRM when provided with the FLC is
improved compared with the motor is provided by PI
controller. The response is fast with minimum overshoots.
B. Dynamic response during a step down load torque
Disturbance (Zone two)
For a good motor performance during this zone, the design
requirements are minimum overshooting (less than or equal to
0.2), the rise time is less than or equal to 0.01 sec, the settling
time is less than or equal to 0.04 sec, the steady state value is
less than or equal to 0.15 [9]. Therefore, according to these
requirements ζ equals 0.5, ωn is 200 rad/sec, and ωd is 173 Fig. 6. The dynamic response of the motor during the second zone, driven by
rad/sec. the FLC as compared with the PI controller
Fig. 6, shows the dynamic response of the motor during the
second zone. The motor is driven by the FLC as compared
with the PI controller of gains kp = 0.005 and ki = 5. As shown
in this figure, it can be realized that the dynamic response of
the SRM in this case, has the maximum overshoot lower than
that experienced when a PI controller is used. Also, it will be
of better damped response after the first overshoot.
Accordingly, the motor reaches its steady state speed faster. In
addition, the steady state error is smaller.
C. Dynamic response during a step up load torque
Disturbance (Zone three)
For a good motor performance during this zone, the design
requirements are the same as that of zone two. Therefore, Fig. 7. The dynamic response of the motor during the third zone, driven by the
according to these requirements ζ equals 0.48, ωn is 208 FLC as compared with the PI controller
rad/sec, and ωd is 182 rad/sec.
Fig. 7, shows the dynamic response of the motor during the VI. FPGA HARDWARE ARCHITECTURE
third zone, when provided with the FLC as compared with the
case in which a PI controller of gains kp = 0.005 and ki = 5 is In this study, FPGA hardware is implemented to drive the
used. From this figure, it can be realized that the dynamic switched reluctance motor. The driver is built on FPGA
response of the SRM when provided with the FLC has the Spartan-3E Starter from Xilinx. The Spartan-3E Starter Kit
first overshoot slightly lower than that experienced by the PI board highlights the unique features of the Spartan-3E FPGA
controller. The FLC improves the system damping after the family and provides a convenient development board for
first overshoot in comparison with that of the PI controller. It

290 290
embedded processing applications [10]. Spartan-3E specific
features are illustrated in details in Appendix-b.

VII. HARDWARE IMPLEMENTATION

Fig. 8 presents the structure of the switched reluctance
motor drive system. The modules of the system are: a four
phase switched reluctance motor, FPGA Spartan-3E Starter
Kit, a driving circuit, dc generator and resistive loads. The
rated current is 1.5 A per phase, the rated torque is 3 N.m, the
dc supply voltage is 24 V, and the rated speed is 100 rpm. The
motor is equipped with an optical encoder to measure position
and speed of rotor shaft. The encoder is optical type with 2048
lines. The driving circuit is asymmetric half bridge circuit,
which consists of two IGBTs and two freewheeling diodes for
each phase. The input to the driving circuit is a pulse train Fig. 9. The measured speed ripple of the motor during both of the starting
process and load torque disturbances, when it was driven by the conventional
voltage signal with magnitude of 0 or 5 V. The output of the PI controller. (5 rpm/division vertically and 0.5 m sec/division horizontally).
driving circuit is switch, with a 24 V nominal dc voltage. A dc
generator is coupled to the shaft of the SRM to simulate the
load torque disturbance. The dc generator is loaded by
different resistive loads. VHDL is used for programming the
FPGA kit.

Fig. 10. The measured speed ripple of the motor during both of the starting
process and load torque disturbance, when it was driven by the adaptive ANN
controller. (5 rpm/division vertically and 0.5 msec/division horizontally).

IX. CONCLUSION
The paper presents a new fuzzy logic controller to ensure
excellent reference tracking of switched reluctance motor
drives. The fuzzy logic controller is enhanced the speed
Fig. 8. Snapshot of the laboratory setup. regulation of this type of drives over both starting and load
disturbance periods. The SRM response when controlled by
FLC is superior to that corresponding to the conventional PI
VIII. EXPERIMENTAL RESULTS controller. Experimental results have verified the validity and
The proposed control technique was tested in the laboratory effectiveness of the proposed control scheme.
when the motor was operated under full load torque at the
starting process and under load torque disturbances. X. APPENDIX
Fig. 9 and Fig. 10 demonstrate the measurements of speed a) The motor under study is three phases, SRM, the rated
under full load torque and load torque disturbances. It can be power is 4 kW at 3000 rpm. The phase resistance is 0.1
observed that the speed response is very smooth and the ripple Ω, the motor inertia is 0.0012 kgm2. DC supply voltage
is reduced to a very small value using the fuzzy logic = 360 V.
controller in comparison with the conventional PI controller.
These good practical results agree with the simulation results. b) Spartan-3E kit features.
In addition, the measured speed has a fast and better damped • Spartan-3E specific features
response than that of the conventional PI controller. 1. Parallel NOR Flash configuration
2. MultiBoot FPGA configuration from
Parallel NOR Flash PROM

291 291
 3. SPI serial Flash configuration [10] Spartan-3E starter kit board user guide, UG 230, vol. 1,
• Embedded development March 2006, Xilinx.
1. MicroBlaze™ 32-bit embedded RISC
Processor
2. PicoBlaze™ 8-bit embedded controller XII. BIOGRAPHY
3. DDR memory interfaces
Gamal M. Hashem was born in September 1960, in
The Spartan-3E Starter Kit board is more advanced and Cairo, Egypt. He received the B.Sc. and the M.Sc.
degrees in Electrical Engineering from Ain Shams
complex compared to other Spartan development boards. The University, Cairo, Egypt, in 1983, and 1988,
key features of the Spartan-3E Starter Kit board are: Xilinx respectively. In 1994 he received the Ph.D. degree in
XC3S500E Spartan-3E FPGA, up to 232 user-I/O pins, 320- Electrical Engineering according to the agreed
pin FBGA package, and over 10,000 logic cells. It has 4 Mbit channel system between Ain Shams University,
Egypt, and Brunel University, U.K.
Platform Flash configurations PROM, 64 MByte (512 Mbit) From 1994 to 2006, he was an Assistant Professor, and since September
of DDR SDRAM, x16 data interface, 100+ MHz, and 16 2006 to recent, he is Associate Professor in Electrical Power and
MByte (128 Mbit) of parallel NOR Flash (Intel StrataFlash). Machines Dept., Ain Shams University, Egypt. His research activities are
Also, it contains FPGA configuration storage, 2-line, 16- in the area of Power Electronics Including PWM Techniques,
Enhancement of Induction Motors Performance, PWM AC Choppers, and
character LCD screen, VGA display port, and 10/100 Ethernet Multilevel PWM Inverter.
PHY (requires Ethernet MAC in FPGA). In addition, the Dr. Gamal is listed in Who’s Who in science and engineering, in the 9th
processor is MicroBlaze™ 32-bit embedded RISC. Moreover, edition. He authored or coauthored numerous technical papers published
it has a four-output, SPI-based Digital-to-Analog Converter in leading journals and conference proceedings.
(DAC) and two-input, SPI-based Analog-to-Digital Converter Hany. M. Hasanien (M’09). He was born in Cairo,
(ADC) with programmable-gain pre-amplifier. Egypt on 1976. He received his B.Sc., M.Sc. and
Ph.D degrees in Electrical Engineering from Ain
XI. REFERENCES Shams University, Faculty of Engineering, Cairo,
Egypt, in 1999, 2004, and 2007 respectively.
[1] I. Kioskeridis, and C. Mademlis, “Maximum Efficiency in Currently, he is an Assistant Professor at the
Single-Pulse Controlled Switched Reluctance Motor Electrical Power and Machines Dept., Ain Shams
Drives”, IEEE Transactions on Energy Conversion, VOL. University. His research interests include machine
design, modern control techniques, electrical drives,
20, NO. 4, pp. 809-817, DECEMBER 2005. artificial intelligence applications on electrical
[2] Yana Zhou, Changliang Xia, Ziming He and Ximing Xie, machines and renewable energy systems.
“Torque Ripple Minimization in a Sensorless Switched Dr. Hasanien is a member of the Institution of Electrical and Electronics
Reluctance Motor Based on Flexible Neural Networks”, Engineers (IEEE) and also of Power & Energy Society (PES). He has
published in many International Conferences and Journals. He is a reviewer of
in Proc. IEEE International Conference of Control and many International Conferences and Journals papers. He has published a book
Automation, May 2007. (Co-authored with Dr. S. M. Muyeen and Prof. J. Tamura), “Switched
[3] T. J. E. Miller, “Switched reluctance motors and their Reluctance Machine”, Praise Worthy Prize, Italy, in Feb. 2010.
control”, Oxford University Press, 1993.
[4] P. J. Lawrenson, “Variable speed switched reluctance
motors”, IEE Proc. Vol. 127, No. 4, pp. 253-265, 1980.
[5] Hany M. Hasanien, N.H. Saad, M.A. Mostfa, and M.A.
Badr, “Speed control of axial laminations switched
reluctance motor provided with digital pole placement
controller”, in Proc. of the International Conference on
Electrical Machines (ICEM), Sept. 2006.
[6] B.W. Williams and C. Pollock “Power converter circuit
for switched reluctance motors with minimum number of
switches”, IEE Proceedings, vol.137, pt. B, No. 6,
November 1990.
[7] Hassan Moghbelli, Gayle E.Adams and Richard G. Hoft,
“Performance of a 10-HP switched reluctance motor and
comparison with induction motors”, IEEE Transactions
on industrial applications, vol. 27, No. 3, May/June 1991.
[8] Nicholas J. Nagel and Robert D. Loranz, “Rotating vector
method for torque control of switched reluctance motor
drive”, IEEE Transaction of industrial Applications, vol.
36, No. 2, March/April 2000.
[9] Katsuhiko. Ogata, “Modern control engineering”,
reference book, Fourth edition, New jersey, U.S.A, 2006.

292 292