Professional Documents
Culture Documents
Optimal Power Flow 2014-TLBO
Optimal Power Flow 2014-TLBO
Optimal Power Flow 2014-TLBO
a r t i c l e i n f o a b s t r a c t
Article history: Teaching-Learning-Based Optimization is a rising star among metaheuristic techniques with highly com-
Received 13 September 2013 petitive performances. This technique is based on the influence of a teacher on learners. In this paper, the
Received in revised form 14 March 2014 Teaching-Learning-Based Optimization technique is used to solve the optimal power flow problem. In
Accepted 30 March 2014
order to show the effectiveness of the proposed method, it has been applied to the standard IEEE 30-bus
and IEEE 118-bus test systems for different objectives that reflect the performances of the power system.
Keywords:
Furthermore, the obtained results using the proposed technique have been compared to those obtained
Teaching-Learning-Based Optimization
using other techniques reported in the literature. The obtained results and the comparison with other
Metaheuristic
Optimal power flow
techniques indicate that the Teaching-Learning-Based Optimization technique provides effective and
Power system optimization robust high-quality solution when solving the optimal power flow problem with different complexities.
Voltage profile © 2014 Elsevier B.V. All rights reserved.
Voltage stability
1. Introduction among other things, which may not be suitable for the actual OPF
conditions [5,6].
The optimal power flow (OPF) problem is the backbone tool for Furthermore, the rapid development of recent computa-
power system operation [1,2]. The objective of the OPF problem tional intelligence tools have motivated significant research in
is to determine the optimal operating state of a power system by the area of non-deterministic that is, heuristic, optimization
optimizing a particular objective while satisfying certain operating methods to solve the OPF problem in the past two decades
constraints [3]. [6]. Some of these techniques are: Ant Colony Optimization
The OPF has been studied for over half a century since the (ACO), Artificial Neural Networks (ANN), Bacterial Foraging
pioneering work of Carpentier [1,2]. Earlier, many traditional Algorithms (BFA), Biogeography-Based Optimization (BBO), Black-
(deterministic) optimization techniques have been successfully Hole-Based Optimization (BHBO), Chaos Optimization Algorithms
used, the most popular were: gradient based methods, Newton- (COA), Differential Evolution (DE), Evolutionary Algorithms (EAs),
based method, simplex method, sequential linear programming, Electromagnetism-Like Mechanism (EM), Evolutionary Program-
sequential quadratic programming, and interior point methods. A ming (EP), Evolutionary Strategies (ES), Fuzzy Set Theory (FST),
survey of the most commonly used conventional optimization algo- Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Sim-
rithms applied to solve the OPF problem is given in [4,5]. Although, ulated Annealing (SA), Tabu Search (TS), Gravitational Search
some of these deterministic techniques have excellent convergence Algorithm (GSA). These methods are known for: their capabilities
characteristics and many of them are widely used in the industry of finding global solutions and avoid to be trapped with local ones,
however, they suffer from some shortcomings. Some of their draw- their ability of fast search of large solution spaces and their ability
backs are: they cannot guarantee global optimality i.e. they may to account for uncertainty in some parts of the power system. A
converge to local optima, they cannot readily handle binary or inte- review of many of these optimization techniques applied to solve
ger variables and finally they are developed with some theoretical the OPF problem is given in [6,7].
assumptions, such as convexity, differentiability, and continuity, One of the recently developed optimization techniques is the
Teaching-Learning-Based Optimization (TLBO), which is a pop-
ulation based optimization technique inspired by passing on
∗ Corresponding author. Tel.: +213 666605628; fax: +213 31908113. knowledge within a classroom environment, where learners first
E-mail address: bouchekara.houssem@gmail.com (H.R.E.H. Bouchekara). acquire knowledge from teacher and then from classmates [8,9].
http://dx.doi.org/10.1016/j.epsr.2014.03.032
0378-7796/© 2014 Elsevier B.V. All rights reserved.
50 H.R.E.H. Bouchekara et al. / Electric Power Systems Research 114 (2014) 49–59
The main objective of this paper is to apply the TLBO to solve the 2.3. Objective constraints
OPF problem. The performance of the proposed technique is sought
and tested on the standard IEEE 30-bus and IEEE 118-bus test sys- OPF constraints can be classified into equality and inequality
tems where the objective functions are: minimization of generation constraints, as detailed in the following sections.
fuel cost, voltage profile improvement, voltage stability enhance-
ment, voltage stability enhancement during contingency condition, 2.3.1. Equality constraints
piecewise quadratic fuel cost curve and fuel cost minimization of The equality constraints of the OPF reflect the physics of the
generators with valve-point loading. power system. The physics of the power system are represented by
The remainder of the paper is organized as follows. First, the the typical power flow equations. These equality constraints are as
OPF is mathematically formulated. Then, the TLBO is presented. follows.
Next, we apply the proposed TLBO to solve the OPF problem in
order to optimize the power system operating conditions. Finally, (a) Real power constraints
we conclude our paper with some remarks and points.
NB
PGi − PDi − Vi Vj [Gij cos (ij ) + Bij sin (ij )] = 0 (6)
2. Optimal Power Flow formulation
j=1
As mentioned earlier, OPF is a power flow problem which gives (b) Reactive power constraints
the optimal settings of control variables for a given settings of load
by minimizing a predefined objective function such as the cost of
NB
QGi − QDi − Vi Vj [Gij sin (ij ) + Bij cos (ij )] = 0 (7)
active power generation and under the consideration of operating
j=1
limits of the system. The OPF problem can be formulated as a non-
linear constrained optimization problem as follows:
where ij = i − j , NB is the number of buses, PG is the active power
Minimize J(x, u) (1) generation, QG is the reactive power generation, PD is the active
load demand, QD is the reactive load demand, Gij and Bij are the
Subject to g(x, u) = 0 (2) elements of the admittance matrix (Yij − Gij + j Bij ) representing the
conductance and susceptance between bus i and bus j, respectively.
and h(x, u) ≤ 0 (3)
2.3.2. Inequality constraints
where u, vector of independent variables or control variables; x,
The inequality constraints of the OPF reflect the limits on physi-
vector of dependent variables or state variables; J(x,u), objective
cal devices present in the power system as well as the limits created
function; g(x,u), set of equality constraints; h(x,u), set of inequality
to guarantee system security. These inequality constraints are as
constraints.
follows.
The control variables u and the state variables x of the OPF
problem are stated in (4) and (5), respectively.
(c) Generator constraints
For all generators including the slack: voltage, active and reac-
2.1. Control variables tive outputs ought to be restricted by their lower and upper
limits as follows:
These are the set of variables which can be modified to satisfy the
load flow equations. The set of control variables in the OPF problem VGmin ≤ VGi ≤ VGmax , i = 1, . . ., NG (8)
i i
formulation are:
PG , active power generation at the PV buses except at the slack PGmin ≤ PGi ≤ PGmax , i = 1, . . ., NG (9)
i i
bus; VG , voltage magnitude at PV buses; T, tap settings of trans-
QGmin ≤ QGi ≤ QGmax , i = 1, . . ., NG (10)
former; QC , shunt VAR compensation. i i
It is worth mentioning that control variables are self- the students is assessed through the mean value of the population.
constrained. The inequality constraints of dependent variables Moreover, the teacher puts effort to increase the mean of students
which contain load bus voltage magnitude; real power generation to a higher level, at which students will require another teacher of
output at slack bus, reactive power generation output and line load- better quality to teach them [8].
ing can be included into an objective function as quadratic penalty Remarkable results have been reported about TLBO outper-
terms. In these terms, a penalty factor multiplied with the square of forming ES, PSO, ABC, DE, and GEM on a number of constrained
the disregard value of dependent variable is added to the objective benchmark functions and constrained mechanical design prob-
function and any unfeasible solution obtained is declined. Mathe- lems, TLBO discovers better or equal solutions much faster than
matically, penalty function can be expressed as follows: the abovementioned techniques as reported in [9].
2
NL
2
Jaug = J + P (PG1 = PGlim ) + v (VLi − VLlim ) + Q 3.2. TLBO algorithm
1 i
i=1
860
In order to illustrate the effectiveness of the proposed TLBO tech-
nique, it has been tested on: the IEEE 30- and the IEEE 118-bus test
850
systems.
The developed program is written in MATLAB computing envi-
840
Cost ($/hr)
ronment and applied on a 2.20 GHz i7 personal computer with
8.00 GB-RAM. 830
The IEEE 30-bus test system has the following characteristics: 810
six generators at buses 1, 2, 5, 8, 11 and 13, four transformers with
off-nominal tap ratio at lines 6–9, 6–10, 4–12, and 28–27, 3) and 800
nine shunt VAR compensation buses at buses 10, 12, 15, 17, 20, 21,
790
23, 24 and 29 [10,11]. 0 50 100 150 200 250 300 350 400 450 500
The line, bus data, generator data and the minimum and max- Iterations
imum limits for the control variables are given in [12] for Case 1
through Case 4 and in [13] for Cases 5 and Case 6. The proposed Fig. 1. Fuel cost variation for Case 1.
technique has been applied to solve the OPF problem for several
cases with different objective functions.
total fuel generation cost is highly reduced compared to Case 0
(the initial case). Quantitatively, it is reduced from 901.9516 $/h to
4.1.1. Case 1: minimization of generation fuel cost 799.0715 $/h which represents a percent cost reduction of 11.41%.
The minimization of the generation fuel cost is studied in this Furthermore, it is noteworthy to mention that Case 0 has some
case. Generally, the generation fuel cost curve is expressed by a voltage violations at bus 19 through bus 30. However, there are
quadratic function, that is: no longer violations in the results obtained using TLBO. The trend
fi = ai + bi PGi + ci PG2 ($/h) (17) of minimizing the total generation fuel cost is sketched in Fig. 1. It
i
appears from this figure that the proposed method has an excellent
where ai , bi and ci are the basic cost coefficient, the linear cost convergence speed. Moreover, the execution time of one iteration
coefficient and the quadratic cost coefficient of the ith generator, is 0.92 s. The power flow through transmission lines of the consid-
respectively. The values of these coefficients are given in [12]. ered system for this case is tabulated in Table 2 and a sketch of this
Therefore, the total fuel cost of all generator units, which repre- flow is given in Figs. 2 and 3.
sents the objective function for this case is expressed as follows:
4.1.2. Case 2: voltage profile improvement
NG
J= fi ($/h) (18) One of the most important and significant safety and service
quality indices is bus voltage [12]. Though the minimization of total
i=1
generation fuel cost may give a feasible solution, the voltage profile
where fi is the fuel cost of the ith generator given by (17). may not be acceptable. Therefore, the goal of this second case is the
The proposed technique has been run for Case 1 and the opti- minimization of the fuel cost along with the improvement of the
mal settings obtained are tabulated in Table 1. It appears that the voltage profile by minimizing the voltage deviation of PQ buses
Power i-->k
100 Power k-->i
50
Power (MW)
-50
-100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
3
Losses
Losses (MW)
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Fig. 2. Active power flow and losses through transmission lines for Case 1.
H.R.E.H. Bouchekara et al. / Electric Power Systems Research 114 (2014) 49–59 53
Table 1
Optimal settings of control variables.
from the unity 1.0. Thus, the objective function can be expressed as
NL
follows: JVoltage deviations = |Vi − 1.0| (21)
i=1
J = Jcost + wJVoltage deviations (19)
and w is a weighting factor which has to be selected carefully. By
this section, a weight (an importance) is given to each one of the
where two terms of the objective function. In this study w is chosen as
100.
NG
The TLBO technique has been applied to search for the optimal
Jcost = fi (20) solution of Case 2. The variations of total fuel cost and voltage devia-
i=1 tions over the iterations are sketched in Fig. 4. These optimal results
20
Power i-->k
15 Power k-->i
10
5
Power (Mvar)
-5
-10
-15
-20
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
5
Losses
Losses (Mvar)
-5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Fig. 3. Reactive power flow and losses through transmission lines for Case 1.
54 H.R.E.H. Bouchekara et al. / Electric Power Systems Research 114 (2014) 49–59
Table 2
Power flow through transmission lines for Case 1.
are given in Table 1. The total generation fuel cost and voltage one of Case 1 as shown in Fig. 5. We can notice from this figure that
deviations are 803.7871 $/h and 0.0945 p.u. for this case compared the voltage profile has been greatly improved.
to 799.0715 $/h and 1.8925 p.u. for Case 1. Although the cost has
been increased by 0.59%, the voltage profile has been improved by 4.1.3. Case 3: voltage stability enhancement
95.01%. The system voltage profile in this case is compared to the Transmission systems are forced to operate very close to their
security limits, due to economical and environmental reasons.
One important characteristic of the power system is its ability to
840 1
Cost
1.1 Case 1
Voltage deviations
Case 2
1 p.u.
Voltage deviations (p.u.)
Voltage Profile (p.u.)
1.05
820 0.5
800 0 0.95
50 100 150 200 250 300 350 400 450 500 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
PQ Bus Number PQ Bus Number
Fig. 4. Variations of fuel cost and voltage deviations for Case 2. Fig. 5. System voltage profile improvement.
H.R.E.H. Bouchekara et al. / Electric Power Systems Research 114 (2014) 49–59 55
conserve constantly acceptable bus voltage at each node under 920 0.145
Cost
normal operating conditions, after load increase, following system L
900 max 0.14
configuration changes or when the system is being subjected to a
disturbance. The non-optimized control variables may lead to pro-
880 0.135
gressive and uncontrollable drop in voltage resulting in an eventual
L max
is attracting more and more attention. Out of the different meth- 840 0.125
ods to assess the voltage stability margin, Kessel and Glavitch [15]
have developed a voltage stability index (L-index) based on the fea- 820 0.12
sibility of power flow equations for each node. The L-index of a bus
indicates the proximity of voltage collapse condition of that bus. 800 0.115
Table 3
Comparison of the simulation results for Case 1.
760 and
NG
720 Fuel cost coefficients of generators 1 and 2 are given in [17] and
those of the remaining generators have the same values as in Case
Cost ($/hr)
1.
700
The obtained optimal settings of control variables when the
proposed technique is run, are given in Table 1 and the fuel cost
680 variation during the optimization process is sketched in Fig. 8.
1040 Table 7
Comparison of the simulation results for Case 6.
TLBO 923.4147
DSA 923.4573
1000
BBO 919.7647
Cost ($/hr)
EP 919.8900
FPSO 923.5400
980
PSO 923.7200
GSA 929.7240
960 MDE 930.7930
IEP 953.5730
940
problems because the results obtained using TLBO are either bet-
ter or comparable to those obtained using other techniques. This
920
0 50 100 150 200 250 300 350 400 450 500 highlights its ability to find better quality solution.
Iterations Moreover, the proposed technique has been compared with
three mathematical non-linear programming methods for the solu-
Fig. 8. Fuel cost variation for Case 6.
tion of OPF [34]. Specifically, the Interior Point (IP), the Sequential
Quadratic Programming (SQP) and the Active Set (AS) methods
Table 4 have been applied to Case 1 through Case 6. The results of the pro-
Comparison of the simulation results for Case 2 where VD stands for voltage posed TLBO are compared with those of IP, SQP and AS techniques
deviation. as given in Table 8. The superiority of the proposed TLBO method
Methods Cost VD Objective function is clearly shown as it gives better solutions. Furthermore, in Case
5 the three mathematical non-linear programming methods have
TLBO 803.7871 0.0945 813.2355
DSA 803.8274 0.0977 813.5961 converged to local optima i.e. they are trapped on local optima.
GSA 804.3148 0.0933 813.6417
BBO 804.9982 0.1020 815.1982 4.1.8. Robustness study
PSO 806.3800 0.0891 815.2900
In order assess the robustness and the effectiveness of the pro-
EM 804.2600 0.1270 816.9600
BHBO 804.5975 0.1262 817.2140 posed technique a statistical study has been carried out. In this
DE 805.2619 0.1357 818.8319 study the TLBO has been run 50 times with different initial solu-
tions for three selected cases namely: Case 1, Case 5 and Case 6. The
best, the mean, the median, the worst and the standard deviation
Table 5 (noted as SD) obtained are displayed in Table 9. We can notice that
Comparison of the simulation results for Case 3. the best, the mean, the median and the worst values of objective
Method Cost Lmax Objective function functions after 50 trials are very close. This is also shown by the
low values of the standard deviations calculated. This study reveals
TLBO 799.9780 0.1131 1478.6611
BBO 805.7252 0.1104 1468.1252
the effectiveness of the proposed technique and its ability to reach
DSA 799.8895 0.1140 1484.1108 either to optimum value or very near to it in every trial. It is worth
GSA 806.6013 0.1162 1504.0833 mentioning that similar statistical analysis can be done for the other
BHBO 803.5193 0.1168 1504.4869 cases.
EM 805.0580 0.1167 1505.0420
DE 807.5272 0.1219 1538.9272
PSO 801.1600 0.1246 1548.7600 4.2. IEEE 118-bus test system
Table 8
Comparison of the obtained results with those obtained using mathematical programming.
TLBO 799.0715 Cost = 803.7871VD = 0.0945 Cost = 799.9780Lmax = 0.1131 Cost = 804.7369Lmax = 0.1150 647.9202 923.4147
IP 799.0718 Cost = 803.8182VD = 0.1025 800.0225 Lmax = 0.1131 Cost = 804.7376Lmax = 0.1150 777.2738 923.4231
SQP 799.0718 Cost = 803.8106VD = 0.0942 800.0225 Lmax = 0.1131 Cost = 804.7377Lmax = 0.1150 737.6695 923.4160
AS 799.0718 Cost = 803.8118VD = 0.0942 800.0224 Lmax = 0.1131 Cost = 804.7376Lmax = 0.1150 736.3832 923.4160
58 H.R.E.H. Bouchekara et al. / Electric Power Systems Research 114 (2014) 49–59
quadratic fuel cost curve and fuel cost minimization of generators [13] O. Alsac, B. Stot, Optimal load flow with steady-state security, IEEE Trans. Power
with valve-point loading. App. Syst. PAS-93 (3) (1974) 745–751.
[14] C.A. Belhadj, M.A. Abido, An optimized fast voltage stability indicator, in: Elec-
This paper is significant for the following reasons among oth- tric Power International Conference on Engineering, PowerTech, Budapest,
ers: 1) the high ranking of the proposed technique among other 1999, pp. 79–83.
techniques which has been proven when compared with other [15] P. Kessel, H. Glavitsch, Estimating the voltage stability of a power system, IEEE
Trans. Power Deliv. 1 (3) (1986) 346–354.
techniques, 2) the efficiency of the TLBO which has been proven by [16] A. Bhattacharya, P.K. Chattopadhyay, Application of biogeography-based opti-
carrying out a statistical analysis that has revealed its strength by misation to solve different optimal power flow problems, IET Gener. Transm.
converging every time the optimum value or very near to it 3) unlike Distrib. 5 (1) (2011) 70–80.
[17] J. Yuryevich, K.P. Wong, Evolutionary programming based optimal power flow
other optimization techniques TLBO has no internal parameter to
algorithm, IEEE Trans. Power Syst. 14 (4) (1999) 1245–1250.
tune. [18] S. Duman, U. Güvenç, Y. Sönmez, N. Yörükeren, Optimal power flow using
gravitational search algorithm, Energy Convers. Manag. 59 (2012) 86–95.
[19] H.R.E.H. Bouchekara, M.A. Abido, Optimal power flow using differential search
Acknowledgement
algorithm, Electric Power Compon. Syst. (2013) (under review).
[20] C.A. Roa-Sepulveda, B.J. Pavez-Lazo, A solution to the optimal power flow using
Dr. M.A. Abido would like to acknowledge the support of Dean- simulated annealing, Int. J. Electr. Power Energy Syst. 25 (1) (2003) 47–57.
[21] A.F. Attia, Y.A. Al-Turki, A.M. Abusorrah, Optimal power flow using adapted
ship of Scientific Research, King Fahd University of Petroleum
genetic algorithm with adjusting population size, Electric Power Compon. Syst.
and Minerals, through the Electrical Power and Energy Systems 40 (11) (2012).
Research Group. [22] H.R.E.H. Bouchekara, Optimal power flow using black-hole-based optimization
approach, Appl. Soft Comput. (2013) (under review).
[23] H.R.E.H. Bouchekara, M.A. Abido, Optimal power flow using electromagnetism-
References like mechanism, Electr. Energy (2013) (under review).
[24] K. Vaisakh, L.R. Srinivas, Genetic evolving ant direction HDE for OPF with non-
[1] R.B. Squires, Economic dispatch of generation directly from power sys- smooth cost functions and statistical analysis, Expert Syst. Appl. 38 (2011)
tem voltages and admittances, IEEE Trans. Power App. Syst. 79 (3) (1961) 2046–2062.
1235–1245. [25] K. Vaisakh, L.R. Srinivas, Evolving ant direction differential evolution for OPF
[2] J. Carpentier, Contribution à l’étude du Dispatching Economique, Bull. Soc. Fran- with non-smooth cost functions, Eng. Appl. Artif. Intell. 24 (2011) 426–436.
caise Electriciens (1962) 431–447. [26] R.H. Liang, S.R. Tsai, Y.T. Chen, T. Wan-Tsun, Optimal power flow by a fuzzy
[3] T. Niknam, M.R. Narimani, M. Jabbari, A.R. Malekpour, A modified shuffle frog based hybrid particle swarm optimization approach, Electr. Power Syst. Res.
leaping algorithm for multi-objective optimal power flow, Energy 36 (11) 81 (7) (2011) 1466–1474.
(2011) 6420–6432. [27] L.L. Lai, J.T. Ma, R. Yokoyama, M. Zhao, Improved genetic algorithms for optimal
[4] M. Huneault, F.D. Galiana, A survey of the optimal power flow literature, IEEE power flow under both normal and contingent operation states, Int. J. Electr.
Trans. Power Syst. 6 (2) (1991) 762–770. Power Energy Syst. 19 (5) (1997) 287–292.
[5] S. Frank, I. Steponavice, S. Rebennack, Optimal power flow: a bibliographic [28] A.J. Ghanizadeh, G. Mokhtari1, M. Abedi, G.B. Gharehpetian, Optimal power
survey I, formulations and deterministic methods, Energy Syst. 3 (3) (2012) flow based on imperialist competitive algorithm, Int. Rev. Electr. Eng. 6 (June
221–258. (4)) (2011), Papers Part B, 2011.
[6] M.R. AlRashidi, M.E. El-Hawary, Applications of computational intelligence [29] A.G. Bakirtzis, P.N. Biskas, C.E. Zoumas, V. Petridis, Optimal power flow by
techniques for solving the revived optimal power flow problem, Electr. Power enhanced genetic algorithm, IEEE Trans. Power Syst. 17 (2) (2002) 229–236.
Syst. Res. 79 (4) (2009) 694–702. [30] M.A. Abido, Optimal power flow using Tabu search algorithm, Electric Power
[7] S. Frank, I. Steponavice, S. Rebennack, Optimal power flow: a bibliographic Compon. Syst. 30 (2002) 469–483.
survey II, non-deterministic and hybrid methods, Energy Syst. 3 (3) (2012) [31] S. Sayah, K. Zehar, Modified differential evolution algorithm for optimal power
259–289. flow with non-smooth cost functions, Energy Convers. Manag. 49 (11) (2008)
[8] R.V. Rao, V.J. Savsani, D.P. Vakharia, Teaching-learning-based optimization: a 3036–3042.
novel method for constrained mechanical design optimization problems, Com- [32] W. Ongsakul, T. Tantimaporn, Optimal power flow by improved evolutionary
put. Aided Des. 43 (3) (2011) 303–315. programming, Electric Power Compon. Syst. 34 (1) (2006) 79–95.
[9] M. Črepinšek, S.H. Liu, L. Mernik, A note on teaching-learning-based optimiza- [33] S.R. Paranjothi, K. Anburaja, Optimal power flow using refined genetic algo-
tion algorithm, Inform. Sci. 212 (2012) 79–93. rithm, Electric Power Compon. Syst. 30 (2002) 1055–1063.
[10] M.A. Abido, Optimal power flow using particle swarm optimization, Int. J. Electr. [34] P.N. Biskas, N.P. Ziogos, A. Tellidou, C.E. Zoumas, A.G. Bakirtzis, V. Petridis,
Power Energy Syst. 24 (7) (2002) 563–571. Comparison of two metaheuristics with mathematical programming methods
[11] K. Lee, Y. Park, J. Ortiz, A united approach to optimal real and reactive power for the solution of OPF, Proc. Gener. Transm. Distrib. 153 (January (1)) (2006)
dispatch, IEEE Trans. Power App. Syst. 104 (5) (1985) 1147–1153. 16–24.
[12] A.A. Abou El Ela, M.A. Abido, Optimal power flow using differential evolution [35] Matpower Matlab toolbox http://www.pserc.cornell.edu/matpower/
algorithm, Electr. Power Syst. Res. 80 (7) (2010) 878–885. matpower.html