Optimal Location and Parameter Setting of TCSC for ... - ScienceDirect

Available online at www.sciencedirect.com

Physics Procedia 33 (2012) 1864 – 1878

2012 International Conference on Medical Physics and Biomedical Engineering

Optimal Location and Parameter Setting of TCSC for Loss Minimization Based on Differential Evolution and Genetic Algorithm Ghamgeen I. Rashed1 and Yuanzhang Sun1 (senior member IEEE), H. I. Shaheen2 1

School of Electrical Engineering, Wuhan University, Wuhan, China [email protected], [email protected] 2 Department of Electrical Engineering, Faculty of Mechanical and Electrical Engineering, Tishreen University, Lattakia, Syria [email protected]

Abstract Flexible Alternating Current Transmission Systems (FACTS) devices have been proposed as an effective solution for controlling power flow and regulating bus voltage in electrical power systems, resulting in an increased transfer capability, low system losses, and improve stability. However to what extent the performance of FACTS devices can be brought out highly depends upon the location and the parameters of these devices. In this paper, we propose two Evolutionary Optimization Techniques, namely Differential Evolution (DE) and Genetic Algorithm (GA) to select the optimal location and the optimal parameter setting of TCSC which minimize the active power losses in the power network, and compare their performances. To show the validity of the proposed techniques and for comparison purposes, simulations are carried out on several power systems, a three-bus power system, a five-bus power system, and an IEEE-14 bus power system. The results, we have obtained, indicate that DE is an easy to use, fast, robust and powerful optimization technique compared with genetic algorithm (GA). The results are presented in the paper together with appropriate discussion.

2011Published Published by Elsevier Ltd. Selection and/or peer-review under responsibility of [name Committee. organizer] ©©2012 by Elsevier B.V. Selection and/or peer review under responsibility of ICMPBE International Open access under CC BY-NC-ND license. Keywords: Thyristor Controlled Series Capacitor (TCSC), Differential Evolution (DE), Genetic Algorithm (GA), Power flow.

1. Introduction With ever-increasing demand for electricity, the power transfer grows, consequently the power system becomes increasingly more difficult to operate, and more insecure with unscheduled power flows and higher losses. With the rapid development of self-commutated semiconductor devices, it is possible to

1875-3892 © 2012 Published by Elsevier B.V. Selection and/or peer review under responsibility of ICMPBE International Committee. Open access under CC BY-NC-ND license. doi:10.1016/j.phpro.2012.05.296

1865

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878

design power electronic equipments known as the Flexible AC Transmission Systems (FACTS)-devices. Such kinds of devices are introduced in 1988 by Hingorani [1]. The objective of using FACTS devices in power system is to bring systems under control and to transmit power according to the ordered of the control centers. These devices also allow the increasing of the usable transmission capacity to its maximum thermal limits. By using FACTS devices, it is also possible to control the phase angle, the voltage magnitude at chosen buses and /or line impedances of a transmission system. Among the FACTS devices, TCSC is one of the most effective measures for increasing the transfer capability of the transmission system, enhancing the stability and ameliorating the dynamic characteristics of power system. However, to achieve the over mentioned benefits, the TCSC should be properly installed in the network with appropriate parameters. For this reason, some performance index must be satisfied. Following factors can be considered in the optimal installation and the optimal parameter of TCSC, the active power loss reduction, the stability margin improvement, the power transmission capacity increasing and the power blackout prevention. Therefore, conventional power flow algorithm [2], should incorporate with TCSC considering one or all of the above mentioned factors. In the last two decades, new algorithms have been developed for the optimal power flow incorporating with TCSC device as well as for the optimal placement of TCSC. A Newton-Raphson load flow algorithm to solve power flow problems in power system with thyristor controlled series capacitor (TCSC) was proposed in [3], [4]. Reference [5] suggested a real power flow performance index sensitivity to obtain the optimal placement of TCSC. Genetic Algorithm was proposed for solving the optimal location of FACTS in [6]-[10]. Differential Evolution Optimization technique for optimal location of FACTS devices was also proposed in [11], [12]. This paper deals with the application of Differential Evolution (DE) and Genetic Algorithm (GA) for the optimal location and parameter setting of the TCSC with the consideration of active power loss reduction in the power system. 2. Problem Formulation As the main objective of this work is to determine the optimal location and the optimal parameter setting of the TCSC in the power network to minimize the loss of the power system, the following performance index is selected: ntl min F PLK (1) k 1 Subject to the following equality constraints: N (2) ) cos ( ij Pgi Pdi V V Y (x j i) 0 j 1 i j ij t csc Qgi

Qdi

N

V V Y ( xt csc ) sin ( ij j 1 i j ij

i)

j

0

(3)

and following inequality constraints: min Pgi m in Q gi

Vi

m in

m in ij

max Pgi

Pgi Q gi

Vi ij

i

m ax

m ax ij

(4)

i

NG

(5)

i

N

(6)

i

N

(7)

m ax Q gi

Vi

NG

1866


m in x t csc

x t csc

min

m ax x t csc

max

(8) (9)

Where: F is the objective function. th PLk is the active power loss in the K line. ntl is the number of lines in the system. N is the set of buses indices. NG is the set of generation bus indices.

Yij and ij are the magnitude and phase angle of element in admittance matrix. Pgi and Qgi are the active and reactive power generation at bus i . Qgi is the reactive power generation at bus i . Pdi and Qdi are the active and reactive power load at bus i . Vi is the voltage magnitude at bus i . ij is the power angle. xt csc is the reactance of TCSC as a function of is the thyristor firing angle.

.

3. Methodologies for optimal location of TCSC A.

3.1 Overview of Differential Evolution (DE) DE is a parallel direct search method proposed by Storn and Price in 1995 [13]. Similar to other EAs techniques, DE is a heuristic, population-based optimization method that uses a population of points to search for a global minimum of a function over continuous search space. Basically, DE generates new vectors of parameters by adding the weighted difference between two population vectors to a third one. If the resulting individual provides a smaller objective function value than a predetermined population individual, in the next generation the new individual replaces the one with which it is compared; otherwise, the old individual is retained. There are several variants of DE, [14], [15]. The general notation of DE variants can be expressed as follows: DE / x / y / z Where x denotes the mutated vector, y is the number of difference vectors, and z is the crossover scheme. The advantage of DE can be summarized as follows [14],[16]: DE is an effective, fast, simple, robust, inherently parallel, and has few control parameters need little tuning. It can be used to minimize non-continuous, non-linear, non-differentiable space functions, also it can work with noisy, flat, multi-dimensional, and time dependent objective functions and constraint optimization in conjunction with penalty functions. 1)

DE-based optimal location and parameter setting of TCSC The step by step implementation of DE algorithm can be described as follows: Step I. initialize power flow data, and DE related parameters such as the size of population ( NP ) , the maximum number of iteration or generation (Gmax ) , the number of variables to be optimized ( D ) ,

1867


CR , and F . Step II. randomly generate the initial population of NP individuals in the feasible space by: X i ,G o

(L ) X i

rand i

(H ) 0 , 1 .( X i

(10)

(L ) X i )

Considering the variables that should be optimized (i.e., the location and the parameter setting of TCSC). These parameters are randomly initialized within feasible ranges. Therefore, all the solutions are feasible solutions and the goal is to find the optimal one. For example, the location of TCSC can be any line in the network except where the transformers are installed. Step III. evaluate the fitness for each individual in the population according to the objective function in (1). Step IV. create a new population by: 1- Mutation: randomly choose three different vectors from the current population and generate a trial vector by: (11) V X r 1, G F .( X r 2 , G X r 3 ,G ) 2- Crossover: for each X i ,G to get new vector ui ,G use: u j ,G

V j ,G ( X i ,G ) j

for j

n D , n

for all other j

1 D , 0, D

, n

L

1 D

(12)

1

3- Selection: for each X i ,G and corresponding ui ,G to select vectors for the next generation, G Use: u i ,( G 1) if f ( u i ,( G 1) ) f ( X i ,G ) X i ,( G 1) X i ,G o th erw ise

G 1.

(13)

Step V: stop the process and print the best individual (optimal location and optimal parameter setting of TCSC) if the stopping criterion is satisfied, else go back to step IV. B.

3.2 Overview of GA Genetic algorithm is a kind of stochastic search techniques based on the mechanism of natural selection and survival of the fittest [17], [18]. Further, it combines the function evaluation with the randomized and/or well-structured exchange of information among the solutions to arrive at a global optimum. More importantly, GA appears attractive because of its superior robust behavior in nonlinear environments over the other optimization techniques. The architecture of the GA implementation can be segregated into the following three constituent phases namely: initial population generation, fitness evaluation and genetic operations. 1)

A Brief Outline of GA Computational Tasks The GA control parameters, such as population size, crossover probability and mutation probability are selected, and an initial population of floating strings of finite length is randomly generated. Each of these individuals, comprising a number of chromosomes, represents a feasible solution to the search problem. Basically, average minimum and maximum fitness of all individuals within a generation are computed. If a pre-defined convergence criterion is not satisfied, then the genetic operations comprising selection and reproduction, crossover and mutation are carried out. Fundamentally, the selection and the reproduction mechanism attempt to apply pressure upon the population in a manner similar to that of the natural selection found in biological systems. A new population is created with worse performing individuals eliminated whilst the most highly fit members in a population are selected to pass on information to the next generation. In this work, the selection function called deterministic sampling selection is adopted. The method ensures that the bigger fitness individuals are remaindered into the next generation. Conceptually, pairs of individuals are chosen at random from the

1868


population and the fit of each pair is allowed to mate. Each pair of mates creates a child having some mix of the two parents. 2)

An Advanced Computational Refinement of GA:

The crossover previously mentioned is the kernel of genetic operations. It promotes the exploration of new regions in the search space using randomized mechanism of exchanging information between strings. Two individuals previously placed in the mating pool during reproduction are randomly selected. A crossover point is then randomly selected and information from one parent up to the crossover point is exchanged with the other parent. This is specifically illustrated below for the used simple crossover technique, which was adopted in this work. Parent 1:

1011

1110

offspring 1:

10111011

offspring 2: 1010 1110 Parent 2: 1010 1011 Another process also considered in this work is the mutation process of randomly changing encoded bit information for a newly created population individual. Mutation is generally considered as a secondary operator to extend the search space and cause escape from a local optimum when used prudently with the selection and crossover schemes. 4. Simulation results

For the validation of the proposed techniques, both of GA and DE algorithms have been tested on the following three test systems, a 3-Bus system (shown in Fig.1), a 5-Bus system (shown in Fig.2) and an IEEE-14 bus system (shown in Fig.3). The data for above mentioned systems is taken from [19], [20], and [21] respectively. A Matlab Code for both techniques, and modified power flow algorithm to include TCSC was developed for simulation purposes.

Figure 1. The three bus test system

Figure 2. The five bus test system


Figure 3. The IEEE-14 bus system

A.

4.1 Implementation of GA The proposed algorithm was implemented to find the optimal location and proper setting of the TCSC. Simple Crossover, deterministic sampling selection, and non uniform mutation have been adopted for the used GA. The other GA parameters are presented in Table I.

Table I PARAMETER VALUES FOR GA Parameter of GA No. of variables Length of individual Population size (np) of individuals

2 3 30

Maximum number of generations g m a x

30, 50

Number of offspring per pair of parents

1

In this study, the location and the reactance of the TCSC were considered as variables optimized to reduce the system losses. Therefore, the TCSC is modeled for the load flow computation like a controllable reactance inserted in the system branch, which can increase or decrease the line reactance. The working reactance range of the TCSC was considered to be [-0.05, 0.05]. For the considered three-bus, five-bus and IEEE-14 bus systems, the GA was applied in two simulation cases. The simulation was done with an initial population having 30 individuals with a maximum generation number equal to 30 and 50 respectively, the results are shown as follows: 1)

Three-Bus Test System From the simulation result, it is found that the optimal location to use the TCSC for this system is in line one (from bus 1 to bus 3) and the TCSC reactance xt csc 0.0220 . Power loss minimization for 30

generation case is shown in Fig.4. Simulation results for voltage magnitude, phase angles, active and reactive power losses with the TCSC for this system are shown in Table II.

1869

1870


0.1915 0.191 0.1905

PLoss (pu)

0.19 0.1895 0.189 0.1885 0.188 0.1875

0

5

10

15 Generation

20

25

30

Figure 4. Power losses-30 generation

Table II VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES WITH TCSC To Bus From Bus line number 1 2 3

1 2 2

3 3 3

Ptlloss (PU) 0.0419 0.0728 0.0728

(PU)

Bus numb er

VM (pu)

VA (deg)

1.035 0.311 0.311

1 2 3

1.05 1.0 0.96

-0.1574 0.0000 -5.9276

Qtlloss

2)

Five-Bus Test System For the five-bus test system, simulation results showed that the optimal placement of the TCSC is in line number one (from bus 1 to bus 2) and the TCSC reactance xt csc 0.001 .

Fig.5. shows the power loss minimization for 50 generation case. Simulation results for voltage magnitude, phase angles, active and reactive power losses with the TCSC for this system are shown in Table III. 0.0469 0.0468 0.0467

PLoss (pu)

0.0466 0.0465 0.0464 0.0463 0.0462 0.0461


0

5

10

15

20

25 30 Generation

35

40

45

50

1871

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878 Table III VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES

2 3 3 4 5 4 5

0.0121 0.012 0.0043 0.0034 0.0114 0.00039 0.0026

(PU)

Bus number

1 1 2 2 2 3 4

To Bus

From Bus line number 1 2 3 4 5 6 7

Ptlloss

VM (pu)

0.0238 0.0181 0.0319 0.0291 0.0025 0.0197 0.0506

1 2 3 4 5 -

1.060 1.041 1.019 1.018 1.012 -

Qtlloss

(PU)

VA (deg) 0.000 -2.676 -4.922 -5.251 -6.065 -

WITH TCSC

3)

IEEE-14 Bus Test System Simulation result shown that the optimal location of TCSC in this system is in line number three (from 0.042 . bus 2 to bus 3) and the TCSC reactance xt csc

Power loss minimization is shown in Fig.6. Table IV shows simulation results for voltage magnitude, phase angles, active and reactive power losses with the TCSC for this system. 0.2176 0.2175 0.2174

PLoss (pu)

0.2173 0.2172 0.2171 0.217 0.2169 0.2168

0

5

10

15

20

25 30 Generation

35

40

45

50


Table IV VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES

From Bus

To Bus

1 2 3 45 6 7 8 9 10

1 1 2 2 2 3 4 4 4 5

2 5 3 4 5 4 5 7 9 6

0.064 0.0437 0.0375 0.026 0.02 0.00188 0.001 0 0 0

(PU)

(PU)

Bus number

line number

Ptlloss

VM (pu)

0.166 0.169 0.113 0.041 0.027 0.033 0.007 0.020 0.013 0.0616

1 2 3 4 5 6 7 8 9 10

1.060 1.015 1.010 1.005 1.008 1.070 1.036 1.090 1.019 1.020

Qtlloss

VA (deg) 0.0 -6.202 -14.394 -13.281 -12.376 -18.356 -16.518 -16.518 -18.197 -18.507

WITH TCSC

1872

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878 11 12 13 14 15 16 17 18 19 20

6 6 6 7 7 9 9 10 12 13

11 12 13 8 9 10 14 11 13 14

0.001 0.001 0.0045 0 0 0.002 0.001 0.001 0.009 0.0033

0.003 0.0017 0.0052 0.016 0.010 0.003 0.0022 0.0019 0.0101 0.0027

11 12 13 14 -

1.041 1.052 1.045 1.012 -

-18.542 -19.179 -19.166 -19.664 -

B.

4.2 Implementation of DE Differential Evolution Optimization technique is also implemented in two numbers of generation 30, and 50, to find out the optimal location and parameter setting of the TCSC. The DE parameters utilized in this simulation are shown in Table V. The results are shown as follows:

Table V PARAMETER VALUES FOR DE

Population size NP

30

Maximum number of generation Gmax Number of variables (NV) Length of individual

( LI )

DE-step size, F Crossover probability constant DE strategy

CR

DE / rand / 1/ bin

Termination criteria

1

e

6

or G m ax

1)

Three-Bus Test System The optimal placement of TCSC for this system was found to be in line one (from bus 1 to bus 3) and the TCSC reactance xt csc 0.0226 . Power loss minimization for 30 generation case is shown in Fig.7. 0.197 0.196 0.195

PLtotal (pu)

0.194 0.193 0.192 0.191 0.19 0.189 0.188 0.187


0

5

10

15 Generations

20

25

30


Voltage magnitudes, phase angles, active and reactive power losses with the TCSC for the system are shown in Table VI. Table VI VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES WITH TCSC To Bus From Bus line number 1 2 3

1 2 2

Ptlloss Qtl loss

3 3 3

(PU)

(PU)

0.0489 0.0694 0.0694

1.2233 0.2774 0.2774

Bus number

VM (pu)

VA (deg)

1 2 3

1.05 0.98 0.952

-0.24208 0.0 -4.6268

Table VII shows the average, worst, and best values of the objective function in this system using TCSC in the optimized location obtained by DE after 30 trials for different initial parameter values of DE. Table VII BEST, WORST, AND AVERAGE OF OBJECTIVE FUNCTION FOR 3-BUS SYSTEM

Objective function value

F

0 .5

CR

average worst best

F

0 .5

0.189385 0.205934 0.18783

CR

0 .5 0 .8

0.189738 0.204401 0.18781

F CR

0 .8

F

0 .5

CR

0.187726 0.188748 0.18770

0 .8 0 .8

0.187769 0.188863 0.18774

2)

Five-Bus Test System Line number one (from bus 1 to bus 2) was found to be the optimal placement of the TCSC for the system and the TCSC reactance xt csc 0.0016 .Power loss minimization is shown in Fig.8, voltage magnitude, phase angles, active and reactive power losses with the TCSC for 50 generation case are shown in table VIII. 0.0470 0.0469 0.0468

Ploss (pu)

0.0467 0.0466 0.0465 0.0464 0.0463 0.0462 0.0461


0

5

10

15

20

25 30 Generations

35

40

45

50

1873

1874

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878 Table VIII VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES WITH TCSC

To Bus

From Bus line number

Ptlloss

1

1

2

0.0132

0.027

1

1.04

0.0

2

1

3

0.0129

0.0168

2

0.99

-2.644

3

2

3

0.0025

0.0415

3

1.03

-4.899

4-

2

4

0.0029

0.0283

4

0.961

-5.226

5

2

5

0.0122

0.0019

5

0.986

-6.0359

6

3

4

0.00029

0.02

-

-

-

7

4

5

0.0022

0.0493

-

-

-

(PU)

Qtlloss

Bus number

(PU)

VM (pu)

VA (deg)

Table IX shows the average, worst, and best solutions for DE technique in this system. Table IX BEST, WORST, AND AVERAGE OF OBJECTIVE FUNCTION FOR 5-BUS SYSTEM

Objective function value

F CR

0 .5

F

0 .5

CR

0 .5 0 .8

F CR

0 .8

F

0 .5

CR

0 .8 0 .8

average

0.053149

0.052077

0.046153

0.046186

worst

0.061099

0.052269

0.046268

0.046283

best

0.051147

0.051955

0.046190

0.046247

IEEE- 14 Bus Test System The optimal placement of the TCSC for this system is in line number three (from bus 2 to bus 3) and the TCSC reactance xt csc 0.05 .

3)

Table X shows the voltage magnitude, phase angles, active and reactive power losses with TCSC for 50 generation, and Fig. 9 shows the Power loss minimization for 50 generation case. 0.2184 0.2182 0.218

Ploss (pu)

0.2178 0.2176 0.2174 0.2172 0.217 0.2168


0

5

10

15

20

25 30 Generations

35

40

45

50

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878 Table X VOLTAGE MAGNITUDE, PHASE ANGLES, ACTIVE AND REACTIVE POWER LOSSES WITH TCSC To Bus From Bus line number 1 2 3 45 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

1 1 2 2 2 3 4 4 4 5 6 6 6 7 7 9 9 10 12 13

2 5 3 4 5 4 5 7 9 6 11 12 13 8 9 10 14 11 13 14

Ptlloss (PU) 0.07336 0.0524 0.0363 0.025 0.019 0.001 0.0012 0 0 0 0.0011 0.001 0.0026 0 0 0.002 0.0013 0.0004 0.0001 0.0001

Qtlloss (PU) 0.0166 0.1644 0.110 0.0377 0.0235 0.0316 0.0074 0.0182 0.0167 0.0482 0.0023 0.0018 0.0053 0.0024 0.0119 0.0003 0.0027 0.0010 0.0001 0.0019

Bus number

VM (pu)

VA (deg)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 -

1.060 1.0279 1.0 0.987 0.991 1.0 0.979 1.0 0.964 0.961 0.977 0.982 0.975 0.949 -

0.0 -6.42 -14.044 -12.947 -12.063 -18.387 -16.511 -16.511 -18.433 -18.751 -18.705 -19.334 -19.351 -20.017 -

Table XI shows the average, worst, and best solutions for DE technique in this system. Table XI BEST, WORST, AND AVERAGE OF OBJECTIVE FUNCTION FOR 14-BUS SYSTEM

Objective function value average worst best

F CR

0 .5 0 .5

0.221072 0.232443 0.217019

F CR

0 .5 0 .8

0.220226 0.232684 0.217120

F CR

0 .8 0 .5

0.216894 0.217032 0.216860

F CR

0 .8 0 .8

0.216919 0.217513 0.216891

From the obtained results it was observed that the unique advantages of these techniques (GA and DE) are, on one hand, their capability of finding the global optimal solution to the optimal location and parameter settings of TCSC problem, on the other hand, they don’t suffer from the extant computational complexity and other limiting mathematical assumptions that the traditional optimization techniques suffered from. 5. Comparisons between the performance of DE, and GA techniques

From the simulation results, we found that the applied DE technique is outstanding, suitable, and efficient for the considered optimization problem. In the following, we summarize the main observations that we have noticed from the implementation results: (a) In general, DE is simple, accurate, and robust. It convergence fast and finds the global optimum almost in every run. In addition, it has few parameters to tune, and the same settings can be used for many different problems. (b) From the convergence perspective, it is observed that DE outperforms GA. DE is able to achieve good results consistently with fast performance. It finds the smallest value of the objective function in all

1875

1876


studied cases. If the computational time is not important, GA always achieves good solutions. When the computational time is considered as the priority, DE is the good choice. (c) From the perspective of convergence speed, it is observed that DE is always faster than GA. (d) DE and GA are robust techniques: the performed 30, 50 trials show that DE and GA can achieve the same results consistently over many trials. (e) DE uses real number representation while the conventional GA uses binary, although it sometimes uses integer or real number representation as well. (f) In GA, two parents are selected for crossover and the child is a recombination of the parents. In DE, three parents are selected for crossover and the child is a perturbation of one of them. (g) The new child in DE replaces a randomly selected vector from the population only if it is better than it. In conventional GA, children replace the parents with some probability regardless of their fitness. (h) In DE, the difference vector is created on the basis of randomly extracted individuals that partially present the population state. (i) A random choice of the vector of differentiation produces many useless steps in the global neighborhood, and a local search is needed at the end of optimization. (j) Both DE and GA are sensitive to the control parameters. 6. Conclusion

In this paper, the effectiveness of the optimal installation of TCSC for minimizing the active power losses in power system has been investigated. One of the newest computational intelligence techniques, namely: DE has been successfully applied to the problem under consideration. The performance of DE is compared with that of GA. With the above proposed algorithms, it is possible for utility to place TCSC in the transmission line such that proper power planning can be achieved with minimum system losses. The result for the three tested power systems showed that both DE and GA techniques can easily find out the optimal location and the best parameters of the TCSC, but DE technique has superior features including high-quality solution, stable convergence characteristics, and good computation efficiency. Acknowledgments

The work was supported by Wuhan University, School of Electrical Engineering. The authors would like to thank the organization for their support. References [1]

N. G. Hingorani, “Power electronics in electrical utilities: role of power electronics in future power systems,” Proceedings of the IEEE Vol. 76 No. 4, pp.481-482, April 1988.

[2]

C. R. Puerle-Esquivel and E. Acha, “A Newton-type algorithm for the control of power flow in electrical power networks,”

[3]

C. R. Fuerte-Esquivel, E. Acha, and H. Ambriz-Perez, “A thyristor controlled series compensator model for the power flow

IEEE Trans. Power System, Vol. 12, no. 4, pp. 1474-1480, November 1997. solution of practical power networks”, IEEE Trans. power system, Vol. 15, No. 1,pp.58-64, February 2000. [4]

M. A. Abdel Moamen, P. P. Narayana, “Newton-Raphson TCSC model for the power flow solution of practical power networks,” IEEE Power Engineering Society Summer Meeting, Volume 3, pp.1488 – 1493, 2002.

[5]

K.S. Verma, S.N. Singh and H. O. Gupta, “ FACTS devices location for enhancement of total transfer capability”, IEEE Power Engineering Society Winter Meeting, Vol. 2, pp. 522 – 527, 28 Jan.-1 Feb. 2001.

[6]

P. P. Narayana, M. A. Abdel Moamen, and B.J. Praveen Kumar “ Optimal location and initial parameter settings of multiple TCSCs for reactive power planning using genetic algorithm”, IEEE Power Engineering Society General Meeting, Vol.1, pp.

1877


1110 – 1114, 6-10 June 2004. [7]

L. Ippolito and P. Siano, “ selection of optimal number and location of thyristor-controlled phase shifters using genetic

[8]

D. Radu, and Y. Besanger, “ A multi-objective genetic algorithm approach to optimal allocation of multi-type FACTS

based algorithms,” IEE proc.-Gener. Trasm. Distrib., Vol. 151, No. 5, September 2004. devices for power system security,” IEEE Power Engineering Society General Meeting, pp. 8, 18-22 June 2006 [9]

G. Stephane, C. Rachid and J. G. Alain, “optimal location of multi-type FACTS devices in a power system by means of genetic algorithms,” IEEE Trans. Power systems, Vol. 16, No.3, pp. 537-544, August 2001..

[10] F. Wang, G. B. Shrestha, “Allocation of TCSC devices to optimize total transmission capacity in a competitive power market,” IEEE Power Engineering Society Winter Meeting, Vol. 2, pp. 587 – 593, 28 Jan.-1 Feb2001 [11] H. R. Baghaee, etal, “Power system security improvement by using differential evolution algorithm based FACTS allocation,” Power System Technology and IEEE power India Conference, POWERCON, pp. 1 - 6

2008.

[12] Husam I. Shaheen, Ghamgeen I. Rashed and S. J. Cheng, “Application of differential evolution algorithm for optimal location and parameters setting of UPFC considering power system security,” European Transactions on Electrical Power, Vol. 19, pp. 911-932, 2009. [13] Storn R, Price K., “ Differential evolution- a simple and efficient adaptive scheme for global optimization over continuous spaces”, Technical report, International computer science Institute, Berkley, 1995. [14] Price KV., An introduction to differential evolution, In New Ideas in Optimization, Corne D. Dorigo M. Glover M (eds). McGraw-Hill:London, 1999 , pp.79-108. [15] Onwubolu GC., Optimization CNC drilling machine operations: traveling salesman problem-differential evolution approach, In New Optimization Techniques in Engineering, Onwubolu GC, Babu BV (eds)., Springer-Verlag: Berlin, Heidelberg, 2004, pp.537-566. [16] Wong KP, Dong ZY., Differential evolution, an alternative approach to evolutionary algorithm. In modern Heuristics Optimization Techniques, Lee KY, El-Sharkawi M (eds), John Wiley & Sons. Hooken, New Jersey, 2008,pp.171-187. [17] D. E. Goldberg, Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley, 1989. [18] D. A. Coley, An Introduction to genetic Algorithms for Scientists and Engineers, World Scientific Publishing Co., 1999. [19] Power System Toolbox Version 2.0: Load Flow Tutorial and Functions. Cherry Tree Scientific Software, RR-5 Colborne, Ontario 1S0, 1991-1999. [20] G. W. Stagg, and A. H. El-Abiad, Computer Methods in Power System Analysis, McGraw-Hill, 1968. [21] L. L. Freris, and A. M.Sasson, “Investigation on the load flow problem,” Proceeding of IEE, Vol. 115, pp.1459-1470, 1968.

Ghamgeen I. Rashed was born in Sulaimani-Iraq, on Sept. 16, 1974. He received his B.Sc. in Electrical Engineering in Salahaadin University-Iraq at 1995, his M.Sc. in Sulaimani University-Iraq in 2003, and his ph.D degree in power system and its automation, from Huazhong University of Science and Technology (HUST), China in 2008. Now he is assistant professor in Wuhan university, school of electrical engineering, china. His special research of interest is AI and its application to power system, FACTS devices specially TCSC and its control

Yuanzhang Sun was born in Hunan Province, China, on September 21, 1954. He received the B.S. degree from Wuhan University of Hydro and Electrical Engineering, China in 1978, the M.Sc. degree from Electric Power Research Institute (EPRI), China in 1982, and Ph.D. degree from Tsinghua Univ., Beijing in 1988, all in electrical engineering. He is now a chair professor at Tsinghua University, vice director of State Key Lab. of Power System Control and Simulation, and dean of School of Electrical Engineering of Wuhan University. His main research interests are in the area of power systems and control, voltage stability and control..

1878

Ghamgeen I. Rashed et al. / Physics Procedia 33 (2012) 1864 – 1878 Husam. I. Shaheen received his Bachelor degree and his Diploma in Electrical Engineering from Techreen University, Syria, in 1999 and 2000 respectively, and his Master and Ph.D. degree in Power System Automation, from Huazhong University of Science and Technology (HUST), China in 2005 and 2008 respectively. Recently he is a lecturer in department of electrical engineering, faculty of mechanical and electrical engineering, Tishreen University, Lattakia, Syria. His main interest researches are AI and its application to power system, FACTS devices and their control,