2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

MATLAB/Simulink based Modelling of TCSC-

Controller in a Power System using GSA Technique
Sthitapa Panda, *Rabindra Kumar Sahu, Sidhartha Panda
Department of Electrical Engineering
Veer Surendra Sai University of Technology (VSSUT), Burla-768018, Odisha, India
E-mail: *rksahu123@gmail.com *Corresponding author Phone no.: +91-9439702316

Abstract—Transmission networks of modern power systems results [4,7,8,9]. However, linear models cannot properly
are becoming increasingly stressed because of growing demand capture complex dynamics of the system, especially during
and restrictions on building new lines. This has led to the facing major disturbances. So by using generator equations a
of power transmission limitation crisis by energy transmission Simulink model is developed which will characterize a SMIB
systems. Better utilization of existing transmission line is system.
achieved by the use of Flexible AC Transmission Systems
(FACTS) devices without sacrificing the desired stability margin. In [6], Differential evolution (DE) algorithm is
In the present work, a MATLAB/SIMULINK based modeling employed to tune parameter of TCSC controller. In this paper,
and simulation of a power system installed with a Thyristor a comprehensive assessment of the effects of TCSC-based
Controlled Series Compensator (TCSC) based controller which is damping controller has been carried out using gravitational
suitable for carrying out power system stability studies. Lead-Lag search algorithm (GSA). In the proposed algorithm, the
(LL) controller structure based TCSC controller in Single searcher agents are a collection of masses which interact with
Machine Infinite Bus (SMIB) system has been proposed. The each other based on the newtonian gravity and the laws of
controller parameters are tuned using Gravitational Search motion. The design objective is to improve the stability of a
Algorithm (GSA) employing ITAE objective function. The SMIB power system, subjected to severe disturbances. As
control parameters of GSA algorithm are tuned by large-disturbance stability improvement is of main concern for
carrying out multiple runs of algorithm for each control the present work, tuning of the control parameters is performed
parameter variation. The superiority of the proposed approach via simulations of severe fault conditions. The dynamic
is demonstrated by comparing the results with recently published performance of proposed TCSC controller under various
technique such as Differential Evolution (DE) based TCSC loading and disturbance conditions are analyzed and compared
controller for the same power system. The dynamic performance for different fault and loading conditions when remote input
of TCSC controller under various loading and disturbance signals are provided.
conditions are analyzed and compared.
II. MODELLING OF POWER SYSTEM
Keywords—Thyristor controlled series compensator(TCSC);
MATLAB/Simulink;Gravitational Search Algorithm (GSA); Power The SMIB power system with TCSC is considered in
system stability; Transient stability the present study as shown in Fig.1 [6]. The generator has a
local load of admittance Y= G+ jB and a double circuit
I. INTRODUCTION transmission line of total impedance Z= R+ jX. Where VT and
Electric energy is an essential ingredient for the VB are the generator terminal and the infinite bus voltage
industrial and all round development of any country. Modern respectively and XT is the reactance of the transformer. The
power systems are designed to operate efficiently to generator is represented by one field circuit and one equivalent
supply power on demand to various load centers with high damper on q-axis. The synchronous machine dynamic
reliability. Recently, with the deregulation of the electricity equations can be written as:
market, power systems are expected to perform functions well
beyond its original design capacity with new loading and
power flow conditions. The increase in utilization of existing dδ
power systems may get the system operating closer to stability = ω B ( S m − S mo ) (1)
boundaries making it subject to the risk of collapse of power dt
systems. Recent development of power electronics dSm 1
introduces the employ of FACTS controllers in power = [ − D (Sm − Smo ) + Tm − Te ] (2)
dt 2H
systems and variable series compensation is highly effective in
both controlling power flow in the lines and in improving dE 'q
stability [1-3]. In stability studies, the overall accuracy is
dt
1
T do
[
= ' − E ' q + ( X d − X ' d ) id + E fd ] (3)
primarily decided by how correctly the system is modelled.
Traditionally, for the small signal stability studies of a single-
dE ' d
machine infinite-bus (SMIB) power system, the linear model of
Phillips Heffron has been used for years, providing reliable dt
1
= '
T qo
[
− E ' d + ( X q − X ' q ) iq ] (4)

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 174

 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

Max.
Tr. Line-1 σ0 σ0 + ∆σ 1 XTCSC (α)
VB ∑ 1 + sTTCSC
VT + Output
XT Z = R + jX +
Min.
TCSC ∆σ 1 + sT3T 1 + sT1T sTWT
KT SN
Infinite-bus 1 + sT4T 1 + sT2T 1 + sTWT Input
Generator Tr. Line-2
signal
Y = G + jB Two stage Washout Gain Sensor
lead-lag Block Block Block

Fig.1 SMIB power system with TCSC Fig.2 Structure of TCSC-based controller
The electrical torque Te is expressed in terms of variables E’d,
E’q, id and iq as: The firing angles of the thyristor are controlled to
adjust the TCSC reactance in accordance with a system control
algorithm, normally in response to some system parameter
Te = E 'd id + E 'q iq + ( X ' d − X 'q ) id iq (5) variations. There exists a steady-state relationship between α
and the reactance XTCSC and this relationship can be described
For a lossless network, the stator algebraic equations and the by the following equation:
network equations are expressed as:
X C2 (σ + sin σ )
E 'q + X ' d id = vq (6) X TCSC (α ) = X C − +
(XC − X L ) π (12)
E '
d − X ' q iq = v d (7) 4 X C2 2
cos (σ / 2) [k tan(kσ / 2) − tan(σ / 2)
( X C − X L ) (k 2 − 1) π
vq = − X e id + E b cos δ (8)
A TCSC is modeled as a variable capacitive reactance
(9) within the operating region defined by the limits imposed by
vd = X e iq − E b sin δ
α.
Solving the above equations, the variables id and iq can be B. Modelling of the TCSC-based controller
obtained as: Over past two decades there is a drastic change in
control structures for a better and an accurate result. The
Eb cos δ − E ' q conventional and commonly used lead–lag structure is chosen
id = (10) in many industrial applications because of its structural
X e + X 'd
simplicity, reliability and the favorable ratio between
performance and cost. Beyond these benefits, these controllers
Eb sin δ + E 'q also offer simplified dynamic modeling, lower user-skill
iq = (11)
X e + X 'q requirements and minimal development effort. TCSC is a
controller specifically designed to be connected in series with
tie lines to control their impedance. The structure of TCSC
A. Dynamic Modelling of the TCSC
controller used in the present study is shown in Fig. 2. The
In series capacitors, the reactive power increases as value of washout function TWT = 3.0s and sensor time constant
the square of the line current where as in shunt reactive power TSN = 10ms are considered. The phase compensation block
generated is proportional to square of bus voltage. For (time constants T1T, T2T, T3Tand T4T) provides the phase lag
achieving the same system benefits as that of series capacitors, between input and the output signals. In Fig. 2.σ0 represents the
shunt capacitors are there to six times more reactive power initial conduction angle as desired by the power flow control
rated than series capacitors need to employ. Reactance loop. The steady state power flow loop acts quite slowly in
practice and hence, in the present study, σ0 is assumed to be
adjusting of TCSC is a complex dynamic process[10]. The
constant during large disturbance transient period. The desired
basic conceptual TCSC module comprises a series capacitor value of line reactance is obtained according to the change in
C, in parallel with a thyristor-controlled reactor. In TCSC the conduction angle ∆σ. This signal is put through a first order
reactance adjustment is done in both inductive and capacitive lag representing the natural response of the controller and the
mode. But capacitive mode is mostly preferred due to less delay introduced by the internal control which yields the
injection of harmonics and easy controllability. Thyristor reactance offered by the TCSC, XTCSC(α). The effective
valves are fired with a phase angle α ranging between 90 0 and reactance is given by (13):
1800 with respect to the capacitor voltage [3]. According to the
 =
−
 (
) (13)
variation of the Thyristor firing angle (α) or conduction angle
(σ), this process can be modelled as a fast switch between where, XTCSC(α) is the reactance of TCSC at firing angle α.
corresponding reactance offered to the power system.

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 175

 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

III. PROBLEM FORMULATION Mpi(t)*Maj(t)

The controller gain KT and the time constants T1T, T2T, Fijd (t) = G(t) (xdj (t)−xid (t)) (17)
T3T and T4T are to be determined. During steady state Rij(t)+∈
conditions ∆σ and σ0 are constant. In dynamic conditions, where
conduction angle (σ) and XTCSC (α) is modulated to improve M aj is the active gravitational mass related to agent j
power system stability. The desired value of compensation is
obtained through the change in the conduction angle (∆σ), M pi
is the passive gravitational mass related to agent i
according to the variation in ∆ω. The effective conduction G(t ) is the gravitational constant at time t ∈ is small constant
angle σ during dynamic conditions is given by (14):
Rij (t )
σ =σ 0+ ∆σ (14) is the Euclidian distance between two agents i and j
given by:
The TCSC-based controller is designed to minimize
the power system oscillations after a disturbance as a result it Rij (t ) = X i (t ), X j (t ) 2
(18)
improves the stability. These oscillations are reflected in the The stochastic characteristic in GSA algorithm is
deviation in the generator rotor speed (∆ω). In the present incorporated by assuming that the total force that act on agent
study, an integral time absolute error (ITAE) of the speed ‘i’ in a dimension‘d’ be a randomly weight sum of d-th
deviations is taken as the objective function J, expressed as: components of the forces exerted from other agents as:

 =   |()| (15) n
(19)
Fi d (t ) = ∑
j =1, j ≠ i
rand j F ijd ( t )
Where ‘e’ is the error signal (∆ω) and t1 is the time range of where rand j is a random number in the interval [0, 1]
simulation. Here t1 is considered as 10 Sec.
The acceleration of the agent ‘i' at the time t and in
d
IV. OVERVIEW OF GRAVITATIONAL SEARCH ALGORITHM the direction d-th, a i (t ) is given by the law of the motion as:
In GSA, agents are considered as objects and their Fid (t)
d
performance is measured by their masses. All these objects a i (t ) = (20)
attract each other by the force of gravity and this force causes M ii ( t )
a global movement of all objects towards the objects with a Where M ii (t ) is the inertia mass of i-th agent. The velocity
heavier mass. Hence masses co-operate using a direct form of of an agent is updated depending on the current velocity and
communication through gravitational force. The heavy masses acceleration. The velocity and position are updated as:
which correspond to good solution move more slowly than
lighter ones, this guarantees the exploitation step of the vid (t + 1) = rand i *vid (t ) + a id (t ) (21)
algorithm. x id (t + 1) = xid (t ) + v id (t + 1) (22)
A. Law of gravity where rand i is a uniform random variable in the interval
Each particle attracts every other particle and the (0,1).The random number is used to give a randomized
gravitational force between the two particle is directly characteristic to the search process.
proportional to the product of their masses and inversely
The gravitational constant G is initialized at the
proportional to the distance between them R. It has been
beginning. To control the search accuracy it is reduced with
reported in literature that R provides better results than R2 in
all experiment cases [5]. time and expressed as function of the initial value ( G 0 ) and
time t as:
B. Law of motion:
G (t ) = G0 e ( −αt / T ) (23)
The current velocity of any mass is equal the sum of
where α is a constant and T is the number of iteration.
the fraction of its previous velocity and the variation in the
velocity. Variation in the velocity or acceleration of any mass The masses (gravitational and inertia) are evaluated
is equal to the force acted on the system divided by mass of by the fitness function. Efficient agents are characterized by
inertia. For a system with ‘n’ agent (masses), the i-th position heavier masses. Assuming the equal gravitational and inertia
of an agent Xi is defined by: mass, the values of masse are calculated using the map of
fitness. The gravitational and inertial masses are updated as:
M ai = M pi = M ii = M i , i = 1,2, .. n. (24)
X i = ( xi1 ,........, x id ,........xin ) for i =1,2,…..n (16)
fiti (t ) − worst (t )
Where mi (t ) = (25)
d best (t ) − worst (t )
x represents the position of i-th agent in the d-th dimension.
i
At a specific time ‘t’, the force acting on mass ‘i ' from mass
mi (t )
M i (t ) = N
(26)
‘j’ is defined as:
∑ j =1 m j (t)

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 176

 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

Dam ping
1 2*pi *f Power angl e
Mechani cal
Sm o 1
Electri cal power power i nput -K- [delta]
s
after faul t clearence 1
Peo -K-
s Speed
[pe] si go
1/2H sig

Tw.s T1.s+1 T3.s+1 dsig

Clock Kp
Tw.s+1 T2.s+1 T4.s+1
Pe Swi tch
sig+dsig
Electri cal power xt csc sig
before & duri ng faul t
Reactance Xe, before, duri ng 0 Sat uration.
& after fault cl earence
sigma to xtcsc
Clock.
[Xe] Xtcsco
Switch.
Xe
0 T i me
Reactance
Cl ock' T i me before & after faul t

Fig.3 MATLAB/Simulink model of SMIB with TCSC controller

Where fiti (t ) represents the fitness value of the agent ‘i' at TABLE I. SIMULATION RESULTS OVER 50 INDEPENDENT RUNS
OF TUNING GSA PARAMETERS
time t and best (t ) is defined for a minimization problem
Para Min. Ave. Max. St.dev. Other
as: meter x 10-4 x 10-4 x 10-4 X 10-4 parameters
Best ( t ) = min fit j (t ) (27) α =5 9768.88 1129.87 13039.36 626.82
j∈{1 ... n } α =10 9338.60 1123.19 12701.15 590.41 G0 =100
(28) α =20 10196.85 1193.63 13694.53 785.04 NP=30
Worst (t ) = max fit j ( t ) T=30
j∈(1 ... n ) α =30 10260.35 1189.77 14143.29 718.61
α =40 10489.84 1157.02 12848.00 411.11
To achieve a good compromise between α =50 10013.45 1163.22 14316.07 485.61
exploration and exploitation, the number of agents is α =60 10207.27 1159.00 13141.37 392.79
reduced with lapse of Eq. (20) and therefore a set of agents G0=80 9218.75 1130.49 12594.85 610.55
with bigger mass are used for applying their force to the G0=90 9132.36 1140.66 12817.94 595.55 α =10
other. The performance of GSA is improved by controlling G0=100 9338.60 1123.81 12701.15 590.41 NP=30
G0=110 9815.69 1125.67 11789.03 550.96 T=30
exploration and exploitation. To avoid trapping in a local
G0=120 9618.13 1143.42 13177.77 544.83
optimum GSA must use the exploration at beginning. G0=130 8680.06 1130.45 13297.15 804.31
G0=140 8688.75 1129.20 13185. 87 746. 43
The different steps of the GSA are the followings: NP=20 8720.06 1138.23 12818.09 860.96
NP=30 8680.06 1130.45 13297.15 804.31 α =10
(a) Identify the search space of parameters to be searched NP=40 8678.48 1128.33 12995.41 736.06 G0=130
NP=50 8677.30 1123.89 12094.65 735.10 T=30
(b) Initialize the variables.
T=30 8680.06 1130.45 13297.15 804.31
(c) Evaluate the fitness of each agent. α=10
T=50 8451.08 1116.47 12597.46 803.73
(d) Update G (t ) , best (t ) , worst (t ) and T=100 8333.29 1105.88 12169.93 954.70 G0 =130
(e) Calculate the total force in various directions. NP=30
T=150 8332.30 1104.78 12163.57 954.52
(f) Calculate the acceleration and velocity.
B. GSA parameters tuning
(g) Update the position of the agents
(h) Repeat steps (c) to (g) until the stop criteria is reached. The success of GSA is heavily dependent on setting
(i) End of control parameters namely; constant α, initial gravitational
constant G0, population size NP and number of iteration T.
V. RESULTS AND D ISCUSSIONS While applying GSA these control parameters should be
carefully chosen for the successful implementation of the
A. Application of GSA algorithm. A series of experiments were conducted to
In order to determine the TCSC controller properly tune the GSA control parameters in order to
parameters, the simulink model of power system under study optimize the TCSC controller parameters employing ITAE
in MATLAB environment is developed which is shown in objective function. Table I shows the GSA outcomes as a
Fig. 3. GSA program is written (in .mfile). The objective result of varying its control parameters. To quantify the
function (ITAE) is calculated in the .m file and used in the results, 50 independent runs were executed for each
optimization algorithm. Simulations were conducted on an parameter variation. It is clear from results shown in Table I
Intel, core i-3core CPU, of 2.4 GHz and 4 GB RAM that the best settings for constant α, gravitational constant G0,
computer in the MATLAB 7.10.0.499 (R2010a) population size NP and number of iteration T are α=10,
environment. G0=130, NP=30 and T=100 respectively. Note that
increasing the population size NP beyond 30 and iterations T

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 177

 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

beyond 100 will improve the average, maximum and TABLE II. OPTIMALPARAMETERS FOR REMOTE INPUT
standard deviation values slightly (with same minimum CONTROL SIGNALS
value) at the expense of increasing the computation time Cases KT T1T T2T T3T T4T
significantly. Nominal loading 14.8977 1.0000 0.0701 0.0100 0.2363
Light loading 46.0355 0.0100 0.5770 0.9987 0.0258
C. Analysis of results Heavy loading 34.1753 0.0188 0.0547 0.6950 0.5255
To assess the effectiveness and robustness of the -3
x 10
proposed controller, different loading conditions and 10
W ith out control DE with TCSC [6] GSA with TCSC
parameters variations as given in Table II are considered.
The performance of the proposed approach is compared 5
with a published [6] results for the same operating

∆ ω (pu)
conditions.
0
Case I: Three-phase fault disturbance at nominal loading
A three phase fault is applied at the generator -5
terminal bus bar at t= 1.0 sec and cleared after 5 cycles.
Figs. 4-7 show the time response of the speed deviation, line 0 2 4 6 8 10
current, active power and power angle for three different Time (sec)
conditions. It is clear from the Figs. 4-7 that without control
Fig.4. Response of speed deviation for Case I
the system is oscillatory and becomes unstable. Stability of
the system is maintained and power system oscillations are
effectively damped out with TCSC based controller. It can
also be seen from Figs (4-7) that the performance of the 1.2

Line current (pu)

system is better with the proposed GSA optimized TCSC
1
controller compared to the DE optimized TCSC controller
[6]. The minimum ITAE value is obtained with GSA 0.8

(ITAE=0.8333) compared to ITAE values with DE 0.6

(ITAE=1.4988) techniques. So it can be concluded that for
0.4
the similar system and same objective function (ITAE) GSA
outperforms DE technique. 0.2 W ith out control DE with TCSC [6] GSA with TCSC
0 2 4 6 8 10
Case II: Line outage disturbance at light loading Time (sec)
To test the robustness of the proposed approach, Fig.5. Line current variation for Case I
another disturbance is considered. The operating point is
changed to light loading condition. One of the parallel 1.4
transmission line is permanently tripped off at t= 1.0 sec.
Reactance of the transmission line increases in the post-fault 1.2
steady state period due to permanent line outage. The system
1
response for the above contingency is shown in Fig. 8–10. It
P (pu)

is clear from Figs.8-10 that for stability criterion with GSA 0.8
based TCSC controller response is improved compared to
e

DE optimized TCSC controller. In this case also minimum 0.6

ITAE value is obtained with GSA (ITAE=0.7413) compared 0.4

to ITAE values with DE (ITAE=1.5416) techniques.
W ith out control DE with TCSC [6] GSA with TCSC
0.2
Case III: Small disturbance at heavy loading 0 2 4 6 8 10
Time (sec)
To complete the analysis, the performance of the Fig.6. Electrical power variation response for Case I
TCSC controller is verified at heavy loading conditions
under small disturbance. The input mechanical power is 70
Power angle (degree)

decreased by 10% at t= 1.0 sec and the system response is 65

shown in Fig. 11. It is clear from Fig. 11 that stability of the 60
system is maintained with TCSC controller by applying 55
speed deviation as input signal and it significantly suppresses 50
the low frequency oscillations by stabilizing the system 45
quickly. 40
35
30 W ith out control DE with TCSC [6] GSA with TCSC
0 2 4 6 8 10
Time (sec)
Fig.7. Power angle variation response for Case I

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 178

 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT)

-3
4
x 10 tuned and the recommended GSA parameters are found to
be: α=10, G0=130, NP=30 and T=100 respectively. The
2 superiority of the proposed approach is demonstrated by
comparing the results with Differential Evolution (DE),
∆ ω (pu)

0 algorithm technique. Investigations show that the dynamic

performance of proposed optimal GSA based TCSC
-2
controller under various loading and disturbance conditions
are improved compared to others.
-4
W ith out control DE with TCSC [6] GSA with TCSC
0 2 4 6 8 10 APPENDIX
Time (sec)
System data: All data are in pu unless specified otherwise.
Fig.8. Response of speed deviation for Case II Generator:
H= 3.542, D= 0, X d= 1.7572, X q= 1.5845, X′ d = 0.4245,
0.8
X’ q =1.04, T′ d0 = 6.66, T’qo= 0.44, f= 60, VT= 1.0, Pe=
Line current (pu)

0.7
0.8 pu, Qe= 0.02 pu, δ0= 53.2657°.
Exciter (simplified IEEE Type):
0.6 KA1= 400, TA1= 0.025 s.
Transmission line and transformer:
0.5 X= 0.8125, XT= 0.1, Ra=0
TCSC controller:
0.4
W ith out control DE with TCSC [6] GSA with TCSC XTCSC0 = 0.6263, α0= 142°, XL= 0.8125, XC= 0.3XL, XP=
0 2 4 6 8 10 0.25 XC, k= 2, XTCSCmax = 0.8 X, XTCSCmin = 0.
Time (sec)
Fig.9. Line current variation for Case II
REFERENCES
65 [1] P. Kundur, Power System Stability and control, Tata McGraw-Hill,
Power angle (deg)

8th reprint 2009.

60 [2] N.G. Hingorani, and L.Gyugyi, Understanding FACTS: concepts and
technology of flexible AC transmission system, IEEE Press, 2000.
55
[3] R.M. Mathur and R.K. Varma, Thyristor-based FACTS controllers
for electrical transmi.sion systems , IEEE Press, John Wiley & Sons,
50
inc. publication, 2002.
45 [4] S.Panda, and N.P. Padhy, “Power system with PSS and FACTS
W ith out control DE with TCSC [6] GSA with TCSC controller: modelling, simulation and simultaneous tuning employing
0 2 4 6 8 10 genetic algorithm”, International Journal of Electrical, Computer,
Time (sec) and Systems Engineering, Vol. 1, No. 1, pp.9–18, 2007.
Fig.10. Power angle variation response for Case II [5] E. Rashedi, H. Nezamabadi and S. SaryazdiJ, “GSA: A Gravitational
x 10
-3 Search Algorithm”, Information Sciences, vol.179, pp. 2232–2248,
5 2009.
[6] S. Panda and R.K. Sahu, “Modelling, simulation and optimal design
of a TCSC-based controller in a power system”, Int. J. Engineering
Systems Modelling and Simulation, Vol. 1, No. 4, pp. 222 - 230,
∆ ω (pu)

2009.
0 [7] P. Mattavelli, G.C. Verghese, A.M. Stankovitctt, “Phasor dynamics of
thyristor-controlled series capacitor systems”, IEEE Trans. Power
Systems, vol. 12, No. 3, pp.1259-1267, 1997.
[8] Y.H. Song and T.A. Johns, Flexible AC Transmission Systems
(FACTS),IEE, London, 2000.
W ith out control DE with TCSC [6] GSA with TCSC [9] A.D. Rosso, C.A. Canizares and V.M. Dona, “A study of TCSC
-5 controller design for power system stability improvement”, vol. 18,
0 2 4 6 8 10
Time (sec) No. 4, pp. 1487–1496, 2003.
[10] Y.Chang and Z. Xu, “A novel SVC supplementary controller based
Fig.11. Response of speed deviation for Case III on wide area signals”, Electric Power System Research, vol. 62,
pp.127–137, 2007.

VI. CONCLUSION
In this study, gravitational search algorithm (GSA)
optimization technique is employed for the design of TCSC-
based damping controllers. The design problem is
transferred into an optimization problem and GSA is
employed to search for the optimal TCSC based controller
parameters. The parameters of GSA technique are properly

978-1-4799-4190-2/14/$31.00 ©2014 IEEE 179