Ain Shams Engineering Journal (2018) 9, 311–318

Ain Shams University

Ain Shams Engineering Journal

www.elsevier.com/locate/asej
www.sciencedirect.com

ELECTRICAL ENGINEERING

Power system stabilizer design based on optimal

model reference adaptive system
Reza Hemmati *

Department of Electrical Engineering, Kermanshah University of Technology, P.O. Box: 63766-67178, Kermanshah, Iran

Received 6 October 2015; revised 20 January 2016; accepted 1 March 2016

Available online 15 March 2016

KEYWORDS Abstract This paper addresses an optimal model reference adaptive system (MRAS) to design
Adaptive control; power system stabilizer (PSS) in multi-machine electric power systems. Weighting factors of the
Model reference adaptive proposed MRAS are adjusted by particle swarm optimization (PSO) as well as its input signal is
system; limited by a normalization technique to assure network stability. The proposed modified-optimal
Multi machine electric power MRAS-PSS is evaluated against conventional PSS to demonstrate its advantages. In order to inves-
system; tigate the performance of the proposed MRAS-PSS under parametric uncertainties, three operating
Particle swarm optimization; conditions are defined and simulated. Several nonlinear and time-domain simulations are carried
Power system stabilizer; out to validate the viability and effectiveness of the proposed MRAS-PSS under network uncertain-
Uncertainty ties.
Ó 2016 Ain Shams University. Production and hosting by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction to the excitation system of the generators and consequently

increases the network damping torque [1].
In large-scale and interconnected electric power systems, the However, because of nonlinearity of the network as well as
nonlinearity and complication of the network are very high parametric uncertainties, conventional PSSs cannot appropri-
and this issue significantly reduces damping torque of the net- ately damp out low frequency oscillations and they are inca-
work. As a result, low frequency electromechanical oscillations pable to provide a robust performance under a wide range of
cannot be suitably damp out in the network. These un-damped operating conditions and disturbances [1]. In order to over-
oscillations greatly reduce small-signal stability of the network come this drawback, numerous advanced techniques have been
as well as limit power transfer capability of the transmission proposed to design high performance PSSs such as Fuzzy
lines. PSS is known as the best auxiliary device to damp out wavelet neural network [2], optimal fuzzy logic [3], indirect
such oscillations. PSS injects a supplementary stabilizing signal adaptive fuzzy sliding mode [4], participation factor and
genetic algorithm [5], multi-band PSS using culture-PSO-co
evolutionary (CPCE) algorithm [6], quantitative feedback the-
* Tel.: +98 831 7259980x2; fax: +98 831 7244201.
ory [7], particle swarm optimization [8], and cultural algorithm
E-mail addresses: [email protected], [email protected]
[9].
Peer review under responsibility of Ain Shams University.
As it was stated, conventional PSSs cannot guarantee the
network stability under parametric uncertainties and it is
required to design high performance PSSs based on advanced
Production and hosting by Elsevier methods. In this regard, this paper addresses PSS design using

http://dx.doi.org/10.1016/j.asej.2016.03.002
2090-4479 Ó 2016 Ain Shams University. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
312 R. Hemmati

Nomenclature

Symbols X0 d transient reactance of d axis (pu)

D deviation from the nominal value Xq steady state reactance of q axis (pu)
XT reactance of transmission line (pu) Xd steady state reactance of d axis (pu)
P active power of load (pu) Te electrical torque (pu)
Q reactive power of load (pu) T0 do time constant of excitation circuit (s)
M system inertia (MJ/MVA) Ka regulator gain
d rotor angle Ta regulator time constant (s)
x rotor speed (pu)
Dx frequency deviation (pu) Abbreviations
Pm mechanical input power (pu) GA genetic algorithms
Tm mechanical torque (pu) ITAE integral of the time multiplied absolute value of
Pe electrical output power (pu) the error
E0 q internal voltage behind x0 d (pu) MRAS model reference adaptive system
Efd equivalent excitation voltage (pu) PSO particle swarm optimization
Vref reference voltage (pu) PSS power system stabilizer

model reference adaptive system (MRAS) as a strong and second order system is completely presented in the next subsec-
advanced adaptive control technique [10]. MRAS is an effec- tion. In the proposed method, Lyapunov stability theory guar-
tive adaptive control approach to tackle the network uncer- antees the stability of system under the proposed MRAS
tainties and deal with various operating conditions and controller [10].
disturbances [11]. MRAS has been successfully utilized to
design adaptive controllers in several electrical components 2.1. MRAS design for a typical system
such as five-phase IPM Motors [12], induction motor drives
[13], active power filters [14], photovoltaic inverters [15], and In this section, the MRAS design for a typical plant is pre-
permanent magnet DC motors [16]. sented. It is assumed that in Fig. 1, the plant model be as
This paper utilizes MRAS technique to design high perfor- follows:
mance PSSs in multi-machine electric power systems. In this
paper, MRAS method is modified through several changes kp ðs þ b0 Þ
y¼ u ð1Þ
to provide a better performance, where weighting factors of s2 þ a1 s þ a0
MRAS technique are tuned by PSO as well as the input signal where u is a control variable, y is the measured output, a0, a1,
is limited by a normalization technique. The proposed modifi- b0 > 0 are some real constants and kp represents the gain. Also
cations not only improve the MRAS performance, but also the reference model is considered as follows:
guarantee the network stability. The proposed MRAS is simu-
km
lated on a multi-machine electric power system and compared ym ¼ uc ð2Þ
to the convention PSSs under various operating conditions. s þ am
Simulation results emphasize on the efficiency and feasibility where km shows the gain and am represents the pole of refer-
of the proposed MRAS in electric power systems. ence model. In this case, the MRAS controller is given by
the following:
2. Model reference adaptive system
up ðtÞ ¼ h1 ðtÞw1 ðtÞ þ h2 ðtÞw2 ðtÞ þ h3 ðtÞy þ h4 ðtÞuc ð3Þ

The model reference adaptive system (MRAS) is a control

method in which the desired performance is defined as a refer-
ence model. Then, the system with MRAS controller tracks the
reference model following any command in input [10]. In fact,
the idea behind MRAS is to create a closed loop controller
with parameters that can be updated to change the response
of the system to follow a desired model. Fig. 1 shows a general
block diagram of MRAS method. Two feedback loops are
seen, the ordinary feedback loop (inner loop) is used to com-
pose the process and the controller and another feedback loop
(outer loop) are used to adjust the controller parameters. The
parameters’ adjustment is carried out based on the tracking
error which is defined as the difference between the output
of the system and the output of the reference model. The
methodology of parameters’ adjustment in MRAS is obtained
by different methods such as gradient method or Lyapunov Figure 1 Block diagram of a model reference adaptive system
stability theory [10]. The MRAS methodology for a typical (MRAS).
Power system stabilizer design 313

where C ¼ diagðci Þ ð14Þ

dw1
¼ Fw1 þ gu ð4Þ dhi
dt ¼ ci e wi ð15Þ
dt
dw2 where e is the difference between the output of the system and
¼ Fw2 þ gy ð5Þ
dt the output of the reference model. In addition, the parameter c
is weighing factor and should be appropriately chosen [11].
u ¼ hT w ð6Þ Fig. 2 shows the structure of proposed MRAS which has been

 made in MATLAB software.
h ¼ hT1 hT2 h3 h4 ð7Þ

 2.2. Normalization of input single
w ¼ wT1 wT2 y uc ð8Þ
In order to obtain F and g, up(t) is calculated as (9) by using In adaptive control, input signal is mainly assumed as a
the network transfer function (1) and the defined reference bounded or unbounded signal. When input signal is regarded
model (2). The details for calculating up(t) as (9) are given by as bounded, it is limited to a bound and network stability is
[11]. With regard to (3) and (9), a new relationship can be writ- passed. But, when input signal is unbounded, input signal
ten as (10) and (11)–(12) are driven from (10). By solving (11) can be bounded by a normalization single to assure the net-
and (12), F and g are obtained [11]. work stability. A normalized system as (16) can ensure that
input signal is bounded. Where y(t) and u(t) indicate output
up ðtÞ ¼ h1 ðtÞf1 ðsÞu þ h2 ðtÞf2 ðsÞy þ h3 ðtÞy þ h4 ðtÞuc ð9Þ and input signals, respectively, h specifies transfer function of
the system, and m(t) represents the normalization signal. In
h1 ðtÞf1 ðsÞu þ h2 ðtÞf2 ðsÞy þ h3 ðtÞyh4 ðtÞuc order to ensure a bounded-input signal, m(t) can be considered
¼ h1 ðtÞw1 ðtÞ þ h2 ðtÞw2 ðtÞ þ h3 ðtÞy þ h4 ðtÞuc ð10Þ as proportional to the input signal such as (17). In such case,
the input signal (u/(1 + u2)) will always be bounded (i.e., the
f1 ðsÞu ¼ w1 ðtÞ ð11Þ denominator is positive and greater than numerator). In this
paper, the proposed normalization signal is utilized to limit
f2 ðsÞy ¼ w2 ðtÞ ð12Þ the input–output signals and as a result, the system stability
is always passed [11].
Based on the Lyapunov’s stability theory, the adaption rule
h is obtained as follows: yðtÞ uðtÞ
¼ h ð16Þ
dh mðtÞ mðtÞ
¼ C e w ð13Þ
dt
mðtÞ2 ¼ 1 þ uðtÞ2 ð17Þ
C is chosen as (14) and adaption rule h is obtained as (15).

Figure 2 MRAS controller in MATLAB-SIMULINK.

314 R. Hemmati
8
2.3. Weighing factors of MRAS >
> x_ ¼ ðPm PMe DxÞ
> i
>
>
< d_i ¼ x0 ðx  1Þ
In MRAS technique, weighing factors, c, have a great impact ð20Þ
E_qi 0 ¼ Tq 0 fd
ðE þE Þ
>
>
on the results, and optimal tuning of these factors is very >
>
>
do
: _
important. These parameters are conventionally tuned by Efdi ¼ Efd þKTa ðV
a
ref Vt Þ

‘try and error’ method. In this paper, the weighing factors

are tuned using PSO technique to find the optimist perfor- Furthermore, a linear dynamic model can be realized by lin-
mance. PSO is an optimization algorithm and finds the opti- earization of the nonlinear equations around nominal operat-
mist value of the weighing factors. PSO optimizes a problem ing condition. A linear dynamic model of the power system can
by iteratively trying to enhance a candidate solution regarding be given as (21), where A is the power system state matrix, B is
the objective function (measure of quality). PSO starts with a the input matrix, C is the output matrix, and D is the feed-
random population of candidate solutions (population matrix) forward matrix [17]. It is worth mentioning that matrixes A,
and rows of this matrix are called particles. PSO optimizes the B, C, and D are large-scale matrixes and it may not be suitable
problem by moving the particles around in the search-space to present them in the paper. The details for calculating such
based on the mathematical formulation over the particle’s matrixes are completely presented in [17].

position and velocity. Each particle’s movement is influenced x_ ¼ Ax þ Bu
by its local best known position as well as the global best ð21Þ
y ¼ Cx þ Du
known position found by all particles so far. As a result, the
particles mainly move toward the best solution.
5. Controller design using MRAS
3. Power system stabilizer
The first step in the controller design using MRAS is to define
From the perspective of control, PSSs are classified as regula- a reference model. Here, PSS operates as a regulator and mit-
tory controllers. In such controllers, the aim is to damp out igates oscillations in electric power systems, and PSS design
oscillations in output signal and the output signal mainly using MRAS technique can be regarded as a regulatory con-
moves toward zero. PSS is regulatory type controller in power troller design. Therefore, in this paper, the reference model
systems which is equipped on the excitation system of the syn- of MRAS is considered as a regulatory controller such as
chronous generators. PSS is mainly utilized to damp out oscil- (22). The step response of (22) is depicted in Fig. 3. It is clear
lations in electric power systems. PSS injects a stabilizer signal that the magnitude of step response is driven back to zero after
to the system and provides an electrical damping torque in- six seconds. With respect to this reference model, MRAS-PSS
phase with speed deviation (Dx) to increase damping torque can be structured as the model given by Fig. 2, where the plant
of the network [1]. Conventional PSS structure is usually struc- model is obtained by (21), the controller is given by (3), and the
tured as (18), where Dx is the speed deviation. This type of reference model is defined by (22).
PSS comprises a washout filter and a dynamic compensator. It is worth mentioning that there is no a well-defined
The washout filter in which its time constant is Tw is used to methodology to find the reference model. The reference model
reset the steady state offset in the PSS output and is fixed as is mainly defined based on the designer experience and desired
10 s in this paper. The dynamic compensator is made up to characteristics of output (i.e., settling time, rise time, overshot).
two lead–lag stages with time constants, T1–T4 and an addi-
tional gain KDC. The proposed conventional PSS cannot sys-
-4
tematically deal with uncertainties. As a result, in this paper, x 10
a high performance PSS is designed by MRAS technique. 10
The performance of the proposed MRAS-PSS is evaluated
against conventional PSS. 8
STW 1 þ ST1 1 þ ST3
Uout ¼ KDC Dx ð18Þ
1 þ STW 1 þ ST2 1 þ ST4 6
Magnitude (pu)

4. Dynamic model of power system 4

In this paper, the generators are modeled as two-axis, three- 2

order model. For all operating conditions, the power system
with generators, PSSs, excitation systems, and turbine-
governor systems can be modeled by a set of nonlinear differ- 0
ential equations as (19) [17].
x_ ¼ fðx; uÞ ð19Þ -2
0 1 2 3 4 5 6 7 8 9 10
where x = (d, x, E0 q) and u are the vectors of state variables. Time (s)
The nonlinear dynamic model of the system can be rewritten as
(20) [17]. Figure 3 Step response of the reference model based on Eq. (22).
Power system stabilizer design 315

In this paper, the network is simulated several times to find the damped Eigen-values close to zero as well as one un-damped
desired output. Then, a reference model that provides the Eigen-value is placed on zero. As a result, small signal stability
desired characteristics is adopted. of the network is very low and damping torque of the network
should be increased through installing PSSs.
1  103 sðs þ 1Þ
y¼ u ð22Þ
s2 þ 2s þ 2
7. Test system
In order to demonstrate the effectiveness of the proposed
MRAS-PSS over conventional PSS, the traditional PSS given Fig. 4 shows a multi machine power system. In this network,
by (18) is designed using GA optimization algorithm. PSS the generators are shown with their nominal apparent power
design using GA is presented in details by [18]. In order to ratings. This system is an modified version of the IEEE
design PSS using GA, the objective function of optimization 9-bus Anderson-Farmer network [20]. The network includes
is considered as (23). This objective function is known as ITAE four synchronous generators G1, G2, G3, and G4. The
[19], where parameter ‘‘i” specifies number of generators. The synchronous machines are connected to the network via
results of GA-PSS are listed in Table 1. It should be noted that lossless step-up transformers featuring a fixed turns ratio.
system is equipped with two PSSs on G2 and G3. The GA Synchronous machines G2 and G3 represent single physical
parameters are considered as follows: number of parameters unit, whereas synchronous machines G1 and G4 denote aggre-
for optimization: 10, population size: 48, mutation rate: gated machines comprising several physical units. In order to
10%, mating rate: 50%, convergence criterion: 2% error in evaluate robustness of the proposed MRAS-PSS over a wide
the responses, objective function: ITAE as shown in (23). range of operating conditions, three loading conditions as
Xn Z t
nominal, light, and heavy are defined and listed in Table 3.
ITAE ¼ tjDxi jdt ð23Þ
i¼1 0
8. Results and discussions

6. Small signal stability analysis In this section, the designed MRAS-PSS and GA-PSS are sim-
ulated on the given test system and compared under the uncer-
For the nominal operating condition, network Eigen-values tainty presented in Table 3. Furthermore, in order to study the
are calculated using power system state matrix, A, and listed performance under different disturbances, three disturbances
in Table 2. The table indicates that there are several low- are considered as follows, where disturbances 1 and 2 are sym-
metrical faults, and disturbance 3 indicates an asymmetrical
fault.
Table 1 PSS parameters using GA.
Disturbance 1: 6-cycle three-phase short circuit on bus 6.
KDC T1 T2 T3 T4 Disturbance 2: disconnection of line between bus 5 and bus
G2 27.77 0.11 0.01 0.14 0.01 7 by breaker.
G3 40.91 0.10 0.01 0.12 0.01 Disturbance 3: 3-cycle single-phase short circuit on bus 7.

In order to demonstrate the robust performance of the

designed PSSs, ITAE index is calculated and results are listed

Table 2 Eigen-values of test case.

166.67, 166.67
0.25, 0.25
10,000, 10,000
100, 100, 100, 100
0.1, 0.1
31.466
26.915
25.788
17.46
14.459
3.9478
1.0595
1.0351
0.99016 Figure 4 Network topology of the 9-bus dynamic network.
0.18032
0
0.02336
0.05289 Table 3 Loading conditions of the network.
0.04479
Loading condition Light Nominal Heavy
0.04938
10,000, 10,000 Value Pnominal  80% Pnominal Pnominal  120%
5, 5, 5, 5, 5, 5, 5, 5 Qnominal  90% Qnominal Qnominal  110%
316 R. Hemmati

in Table 4. It is clear that MRAS-PSS indicates better perfor- 1.002

mance than GA-PSS at all operating conditions. Here, MRAS- MRAS-PSS
1.0015 GA-PSS
PSS resulted in lower ITAE index compared to GA-PSS. As a
result, MRAS-PSS can damp out oscillations more effective
1.001
than GA-PSS.

Speed G2 (pu)
Figs. 5–8 indicate the simulation results under different 1.0005
symmetrical disturbances and various operating conditions.
1

Table 4 Calculated ITAE for test system. 0.9995

Disturbance 1 Disturbance 2 0.999

MRAS-PSS GA-PSS MRAS-PSS GA-PSS
0.9985
Nominal operating 0.054 0.054 0.061 0.069
condition
0.998
Heavy operating 0.083 0.097 0.09 0.096 0 5 10 15 20 25 30
condition Time (s)
Light operating 0.064 0.070 0.072 0.081
condition Figure 7 Speed G2 in the heavy operating condition following
disturbance 2.

1.002 1.005
MRAS-PSS MRAS-PSS
1.0015 GA-PSS 1.004 GA-PSS

1.003
1.001
1.002
Speed G2 (pu)

Speed G3 (pu)

1.0005
1.001
1 1

0.9995 0.999

0.998
0.999
0.997
0.9985
0.996
0.998
0 2 4 6 8 10 12 14 16 18 20 0.995
0 5 10 15 20 25 30
Time (s) Time (s)

Figure 5 Speed G2 in the nominal operating condition following Figure 8 Speed G3 in the heavy operating condition following
disturbance 1. disturbance 2.

1.003 1.0003
MRAS-PSS MRAS-PSS
1.0025 GA-PSS
1.0003 GA-PSS
1.002
1.0002
1.0015
Speed G3 (pu)

Speed G1 (pu)

1.001 1.0002

1.0005 1.0001
1
1.0001
0.9995
1
0.999

0.9985 1

0.998 0.9999
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Time (s) Time (s)

Figure 6 Speed G3 in the nominal operating condition following Figure 9 Speed G1 in the nominal operating condition following
disturbance 1. disturbance 3.
Power system stabilizer design 317

415 conventional PSS which is tuned by GA approach, i.e. GA-

With PSS PSS. Simulation results emphasize on the viability and robust-
Without PSS
414 ness of the proposed adaptive method under various operating
conditions and disturbances. The proposed MRAS-PSS sys-
Active power of G1 (pu)

413 tematically tackles the network uncertainties and provides a

robust performance following changing parameters. In addi-
412 tion, the results indicate that MRAS-PSS is more effective than
GA-PSS to damp out oscillations and stability enhancement
411 under uncertain conditions. Eventually, it can be concluded
that MRAS-PSS ensures robust stability and robust perfor-
410 mance of the network under a wide range of operating condi-
tions, uncertainties, and disturbances.
409
References
408
0 2 4 6 8 10 12 14 16 18 20 [1] Yu Y-N. Electric power system dynamics. Academic Press, Inc.;
Time (s) 1983. p. 256 [111 Fifth ave., New York, NY 10003, USA].
[2] Tofighi M, Alizadeh M, Ganjefar S, Alizadeh M. Direct adaptive
Figure 10 Active power of G1 following 3-cycle three-phase power system stabilizer design using fuzzy wavelet neural network
short circuit on bus 3. with self-recurrent consequent part. Appl Soft Comput
2015;28:514–26.
[3] Sun Z, Wang N, Srinivasan D, Bi Y. Optimal tunning of type-2
fuzzy logic power system stabilizer based on differential evolution
It is clear that under all operating conditions and disturbances,
algorithm. Int J Electr Power Energy Syst 2014;62:19–28.
MRAS-PSS operates better than GA-PSS. The results demon- [4] Saoudi K, Harmas MN. Enhanced design of an indirect adaptive
strate that under heavy operating condition, while the perfor- fuzzy sliding mode power system stabilizer for multi-machine
mance of GA-PSS becomes poor, MRAS-PSS provides a power systems. Int J Electr Power Energy Syst 2014;54:425–31.
stable and robust performance. It can be concluded that [5] Hassan LH, Moghavvemi M, Almurib HAF, Muttaqi KM,
MRAS-PSS provides better parameter adaptation compared Ganapathy VG. Optimization of power system stabilizers using
to GA-PSS under varying operating conditions. In other participation factor and genetic algorithm. Int J Electr Power
words, together with changing operating condition from the Energy Syst 2014;55:668–79.
nominal values to the heavy or light values, adaptive PSSs alter [6] Khodabakhshian A, Hemmati R, Moazzami M. Multi-band
their parameters and tackle the uncertainty to show a robust power system stabilizer design by using CPCE algorithm for
multi-machine power system. Electr Power Syst Res
performance. Fig. 9 also demonstrates the simulation result
2013;101:36–48.
under asymmetrical disturbance. It is clear that, MRAS-PSS [7] Khodabakhshian A, Hemmati R. Robust decentralized multi-
operates better than GA-PSS following an asymmetrical machine power system stabilizer design using quantitative feed-
disturbance. back theory. Int J Electr Power Energy Syst 2012;41:112–9.
In order to investigate the effect of the proposed method on [8] Mostafa HE, El-Sharkawy MA, Emary AA, Yassin K. Design
the generator produced active power under disturbances, and allocation of power system stabilizers using the particle
active power of G1 following 2-cycle three-phase short circuit swarm optimization technique for an interconnected power
on bus 3 is depicted in Fig. 10. It is clear that improvement system. Int J Electr Power Energy Syst 2012;34:57–65.
of generator speed profile via using the suggested method does [9] Khodabakhshian A, Hemmati R. Multi-machine power system
not cause worse oscillations on the generator active power pro- stabilizer design by using cultural algorithms. Int J Electr Power
Energy Syst 2013;44:571–80.
file. The figure shows that suggested PSSs not only do not 
[10] Aström KJ, Wittenmark B. Adaptive control. Dover Publications;
cause worse oscillations, but also provide more damping tor- 2008.
que and less oscillations. This is due to this fact that the pro- [11] Ioannou PA, Sun J. Robust adaptive control. Courier Dover
posed PSSs inject a stabilizing signal to the excitation system Publications; 2012.
of the generators and increase the network damping torque. [12] Lusu G, Parsa L. Model reference adaptive control of five-phase
As a result of increasing damping torque, all network oscilla- IPM motors based on neural network. IEEE Trans Ind Electr
tions following disturbances are expected to be damped out 2012;59:1500–8.
appropriately. [13] Ravi Teja AV, Chakraborty C, Maiti S, Hori Y. A new model
reference adaptive controller for four quadrant vector con-
trolled induction motor drives. IEEE Trans Ind Electr
9. Conclusions 2012;59:3757–67.
[14] de Araujo Ribeiro RL, de Azevedo CC, de Sousa RM. A robust
adaptive control strategy of active power filters for power-factor
This paper presents an application of modified MRAS tech- correction, harmonic compensation, and balancing of nonlinear
nique to design PSS as a regulator controller in electric power loads. IEEE Trans Power Electr 2012;27:718–30.
systems. In order to investigate the performance of the pro- [15] Gonzalez-Espin F, Garcera G, Patrao I, Figueres E. An adaptive
posed MRAS-PSS under uncertainties, various operating con- control system for three-phase photovoltaic inverters working in a
ditions and disturbances are defined and simulated. polluted and variable frequency electric grid. IEEE Trans Power
Furthermore, the proposed MRAS-PSS is evaluated against Electr 2012;27:4248–61.
318 R. Hemmati

[16] Rashidi B, Esmaeilpour M, Homaeinezhad MR. Precise angular Reza Hemmati was born in Kermanshah, Iran,
speed control of permanent magnet DC motors in presence of in 1982. He received the B.S. degree in
high modeling uncertainties via sliding mode observer-based electrical engineering from Hormozgan
model reference adaptive algorithm. Mechatronics 2015;28:79–95. University, Bandarabbas, Iran, in 2005, the M.
[17] Kundur P, Balu NJ, Lauby MG. Power system stability and S. degree in electrical engineering from Kashan
control. New York: McGraw-hill; 1994. University, Kashan, Iran, in 2007, and the Ph.
[18] Abido MA. Parameter optimization of multimachine power D. degree in electrical engineering from
system stabilizers using genetic local search. Int J Electr Power Isfahan University, Isfahan, Iran, in 2013.
Energy Syst 2001;23:785–94. Presently, he is assistant professor at
[19] Shayeghi H, Shayanfar HA, Safari A, Aghmasheh R. A robust Kermanshah University of Technology,
PSSs design using PSO in a multi-machine environment. Energy Kermanshah, Iran. His research interests
Convers Manage 2010;51:696–702. include power systems planning, operation
[20] Anderson PM, Farmer RG. Series compensation of power and control.
systems. PBLSH! Incorporated; 1996.