 Open Access
 Total Downloads : 785
 Authors : Nagendra Singh, Yogendra Kumar
 Paper ID : IJERTV1IS7454
 Volume & Issue : Volume 01, Issue 07 (September 2012)
 Published (First Online): 26092012
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Moderate Random PSO Using For Economic Load Dispatch
Nagendra Singh Yogendra Kumar
Dept. of Electrical Engineering Dept. of Electrical Engineering Mewar University MANIT
Chittorgrah, India Bhopal, India
Abstract
Economic load dispatch (ELD) is an important optimization task in power system. It is the process of allocating generation among the committed units such that the constraints imposed are satisfied and the fuel cost is minimized. Particle swarm optimization (PSO) is a population based optimization technique that can be applied to a wide range of problems but it lacks global search ability in the last stage of iterations. This paper used a novel PSO with a moderaterandomsearch strategy (MRPSO), which enhances the ability of particles to explore the solution spaces more effectively and increases their convergence rates. In this paper the usefulness of the MRPSO algorithm to solve the ELD problem is demonstrated through its application to three, six and fifteen generator systems with ramp rate limit constraints. The result shows MRPSO work efficiently and give optimal solution.
Key words: ELD, Ramp rate, PSO, MRPSO.

Introduction
Electric utility system is interconnected to achieve the benefits of minimum production cost, maximum reliability and better operating conditions. The economic scheduling is the online economic load dispatch, wherein it is required to distribute the load among the generating units, in such a way as to minimize the total operating cost of generating units while satisfying system equality and inequality constraints. For any specified load condition, ELD determines the power output of each plant (and each generating unit within the plant) which will minimize the overall cost of fuel needed to serve the system load [1]. ELD is used in realtime energy management power system control by most programs to allocate the total generation among the available units. ELD focuses upon coordinating the production cost at all power plants operating on the system.
Conventional as well as modern methods have been used for solving economic load dispatch problem employing different objective functions. Various conventional methods like lambda iteration method, gradientbased method, Bundle method [2], Nonlinear programming [3], Mixed integer linear programming [4], Dynamic programming [7], Linear programming [6], Quadratic programming [8], Lagrange relaxation method [9], Newtonbased techniques [10] and Interior point methods [5], reported in the literature are used to solve such problems.
Conventional methods have many draw back such as nonlinear programming has algorithmic complexity. Linear programming methods are fast and reliable but require linearization of objective function as well as constraints with nonnegative variables. Quadratic programming is a special form of nonlinear programming which has some disadvantages associated with piecewise quadratic cost approximation. Newtonbased method has a drawback of the convergence characteristics that are sensitive to initial conditions. The interior point method is computationally efficient but suffers from bad initial termination and optimality criteria.
Recently, different heuristic approaches have been proved to be effective with promising performance, such as evolutionary programming (EP) [11], simulated annealing (SA) [12], Tabu Search (TS) [13], pattern search (PS) [14], Genetic algorithm (GA) [15], [16], Differential evolution (DE) [17], Ant colony optimization [18], Neural network [19], particle swarm optimization (PSO) [20], [21], [22], SOHPSO[23], Modified PSO[24], classical PSO[26], MRPSO[27], WIPSO[28], MOPSO[29]. Although the
heuristic methods do not always guarantee discovering globally optimal solutions in finite time, they often provide a fast and reasonable solution. EP is rather slow converging to a near optimum for some problems. SA is very time consuming, and cannot be utilized easily to tune the control parameters of the annealing schedule. TS is difficult in
defining effective memory structures and strategies which are problem dependent. GA sometimes lacks a strong capacity of producing better offspring and causes slow convergence near global optimum, sometimes may be trapped into local optimum. DE greedy updating principle and intrinsic differential property usually lead the computing process to be trapped at local optima.
Particleswarmoptimization (PSO) method is a populationbased Evolutionary technique and it is inspired by the emergent motion of a flock of birds searching for food. In comparison with other EAs such as GAs and evolutionary programming, the PSO has comparable or even superior search performance with faster and more stable convergence rates but its lacks global search ability in the last stage of iterations. This problem can be solved by using moderate random search technique with PSO. In this paper used MRPSO to solve the ELD problem. It enhance the global search ability and gives more opportunity of the particles to explore the solution space than is standard PSO.
The proposed method focuses on solving the economic load dispatch with Generator Ramp Rate Limits constraint. The feasibility of the proposed method was demonstrated for three, six and fifteen bus system. The results are obtained through the proposed approach and compared with other PSO methods reported in recent literatures.

Economic Dispatch problem Formulation

Basic formulation of ED
ED is one of the most important problem to be solved in the operation and planning of a power system. the primary concern of an ED problem is the minimization the total cost of generation(objective function) in such a way that meets the demand and satisfies all constraints associated is selected as the objective function.
The ED problem objective function is formulated mathematically in (1) and (2).
For power balance, an equality constraint should be satisfied. The total generated power should be equal to total load demand plus the total losses,
(4)
(5)
Where, PDemand is the total system demand and PLoss is the total line loss.
ijth element of loss coefficient symmetric matrix B, ith element of the loss coefficient vector and
loss coefficient constant.

Unit Operating Limits
There is a limit on the amount of power which a unit can deliver. The power output of any unit should not exceed its rating nor should it be below that necessary for stable operation. Generation output of each unit should lie between maximum and minimum limits.
(6)
Where, Pi is the output power of ith generator ,
and are the minimum and maximum power outputs of generator i respectively.

Ramp Rate Limit
According to the operating increases and operating decreases of the generators the ramp rate limit constraints described in eq. (7) & (8).

As generation increases
(7)

As generation decreases
(8)
Where, is the objective function, ai, bi and ci are the cost coefficients.
(1)
(2)
(3)
When the generator ramp rate limits are considered, the operating limits for each unit, output is limited by time dependent ramp up/down rate at each hour as given below.
=max( ) and
= min( ).
D is power equilibrium; PD and PL represent total demand power and the total transmission loss of the transmission lines respectively.
2.2. Constraints
2.2.1 Real Power Balance Equation
t) (9)
Where, current output power of ith generating unit, Previous operating point of the ith generator,
Down ramp rate limit (MW/time period) and Up ramp rate limit (MW/time period).



Overview of Some PSO Strategies
A number of different PSO strategies are being applied by researchers for solving the economic load dispatch problem and other power system problems. Here, a short review of the significant developments is presented which will serve as a performance measure for the MRPSO technique [27] applied in this paper.

Standard particle swarm optimization (PSO)
Particle swarm optimization was first introduced by Kennedy and Eberhart in the year 1995. It is an exciting new methodology in evolutionary computation and a populationbased optimization tool. PSO is motivated from the simulation of the behavior of social systems such as fish schooling and birds flocking. It is a simple and powerful optimization tool which scatters random particles, i.e., solutions into the problem space. These particles, called swarms collect information from each array constructed by their respective positions. The particles update their positions using the velocity of articles. Position and velocity are both updated in a heuristic manner using guidance from particles own experience and the experience of its neighbors.
The position and velocity vectors of the ith particle of a ddimensional search space can be represented as Pi=(pi1,pi2,pid) and Vi=(vi1,vi2,vid,) respectively. On the basis of the value of the evaluation function, the best previous position of a particle is recorded and represented as Pbesti=( pi1,pi2,pid), If the gth particle is the best among all particles in the group so far, it is represented as Pgbest=gbest= (pg1,pg2,pgd).
The particle updates its velocity and position using (10) and (11)
or cognition part which represents the private thinking of the itself and the term c2Rand2( )Ã—(gbest Sk ) is called swarm influence or the social part which represents the collaboration among the particles.
1
In the procedure of the particle swarm paradigm, the value of maximum allowed particle velocity Vmax determines the resolution or fitness, with which regions are to be searched between the present position and the target position. If Vmax is too high, particles may fly past good solutions. If Vmax is too small, particles may not explore sufficiently beyond local solutions. Thus, the system parameter Vmax has the beneficial effect of preventing explosion and scales the exploration of the particle search. The choice of a value for Vmax is often set at 1020% of the dynamic range of the variable for each problem.
W is the inertia weight parameter which provides a balance between global and local explorations, thus requiring less iteration on an average to find a sufficiently optimal solution. Since W decreases linearly from about
0.9 to 0.4 quite often during a run, the following weighing function is used in (10)
(12)
Where, Wmax is the initial weight, Wmin is the final weight, Iter max is the maximum iteration number and iter is the current iteration position.

CLASSICAL PSO
In this section, for getting better solution the standard PSO algorithm, used classical PSO [26],The constriction factor is used in this algorithm given as
(13)
i
Where, V k is velocity of individual i at iteration k, k is pointer of iteration, W is the weighing factor,
(10)
(11)
Where, Ã˜ is define as 4.1Ã˜4.2
As increases, the factor c decreases and convergence becomes slower because population diversity is reduced.
Now the update its velocity using (14).
C1, C2 are the acceleration coefficients, Rand1( ), Rand2( ) are the random numbers between 0 & 1,
Sik is the current position of individual i at iteration k, Pbesti is the best position of individual i and
Gbest is the best position of the group.
1
The coefficients c1 and c2 pull each particle towards pbest and gbest positions. Low values of acceleration coefficients allow particles to roam far from the target regions, before being tugged back. on the other hand, high values result in abrupt movement towards or past the target regions. Hence, the acceleration coefficients cl and c2 are often set to be 2 according to past experiences. The term c1rand1 () x (pbest, Sk ) is called particle memory influence
(14)

WEIGHT IMPROVED PSO (WIPSO)
In this section, for getting the better global solution, the traditional PSO algorithm is improved by adjusting the weight parameter, cognitive and social factors. Based on the velocity of individual i of WIPSO algorithm [28] is rewritten as,
(15)
Where,
Where , wmin, wmax: initial and final weight, c1min, c1max: initial and final cognitive factors and c2min, c2max: initial and final social factors.

MRPSO

(16)
(17)
(18)
(19)
Minimum operating limits of the generators.
Step2:Initialize velocity and position for all particles by Randomly set to within their legal rang.
Step3:Set generation counter t=1.
Step4: Evaluate the fitness for each particle according to the objective function.
Step5:Compare particles fitness evaluation with its Pbest and Gbest.
Step6:Update position by using (20). Step7: Apply stopping criteria.

Case Study

Test Case I
The first test results are obtained for 3generator Systems in which all units with their ramprate limits. The
MRPSO was first introduced by Hao Gao and Wenbo in the year 2011[27], In order to enhance the global search ability of the PSO but not slow down its convergence rate, we used a new PSO algorithm with an MRS strategy. In this algorithm used only position update and no need of updating velocity .
The position of the ith particle at the (K + 1)th iteration can be calculated using (20), (21).
unit characteristics data are given in Table 1 The load demand is 850 MW. The B loss coefficients are given in Table 2. The best solutions of the proposed MRPSO, PSO, CPSO & WIPSO methods are shown in Table 6.
Table 1
Capacity, cost coefficients and ramp rate limits of 3 generator systems.
Unit
Pi
1
0.004820
7.97
78
200
50
170
50
90
2
0.001940
7.85
310
400
100
350
80
120
3
0.001562
7.92
562
600
100
440
80
120
(20)
Where, S denotes the population size in the MRPSO.
(21)
Table 2
0.0006760
0.0000953
0.0000507
0.0000953
0.0005210
0.0000901
0.0000507
0.0000901
0.0002940
B coefficient (in mw1 ) for 3 generator system
The parameter is obtained by changing from 0.45 to
0.35 with the lineardecreasing method during iteration,
Pd is the attractor moving direction of particles, it is given as (22).
(22)
Where, rand0 is a uniformly distributed random variable within [0, 1].
(23)
Where, rand1 and rand2 are two random variables within [0, 1] , and rand3 is a random variable within [1, 1].
4. Algorithm for ED Problem Using MRPSO
The algorithm for ED problem with ramp rate generation limits employing MRPSO for practical power system operation is given in following steps:/p>
Step1: Initialization of the swarm: For a population size the particles are randomly generated in the Range 01 and located between the maximum and the
Bio = [0.007660 0.00342 0.01890] and Boo=0.40357.

Test Case II
The second test results are obtained for sixgenerating unit system in which all units with their ramprate limits. This system supplies a 1263MW load demand.
Table 3
Capacity, cost coefficients and ramp rate limits of 6 generator systems.
Unit
Pi
1
240
7
0.0070
100
500
440
80
120
2
200
10
0.0095
50
200
170
50
90
3
220
8.5
0.0090
80
300
200
65
100
4
200
11
0.0090
50
150
150
50
90
5
220
10.5
0.0080
50
200
190
50
90
6
190
12.0
0.0075
50
120
110
50
90
The data for the individual units are given in Table 3. The B matrix of the transmission loss coefficient is given in table 4. The best solutions of the proposed MRPSO, PSO, CPSO and WIPSO methods are shown in Table 7.
Table 4
B(104) coefficients (in mw1) for six generator systems
0.17
0.12
0. 7
0.1
0.5
0.02
0.12
0.14
0.09
0.01
0.06
0.01
0.07
0.09
0.31
.000000
0.10
0.06
0.01
0.01
0.0000
2.4
0.06
0.08
0.05
0.06
0.10
0.06
1.29
0.02
0.02
0.01
0.06
0.8
0.2
1.50
Bio=104[0.3908 1.297 7.047 0.5910 2.161
6.635] Boo=0.0056.

Test Case III
The third test results are obtained for fifteengenerating unit system in which all units with their ramprate limits. This system supplies a 2650 MW load demand.
Table 5
Capacity, cost coefficients and ramp rate limits of 6 generator systems.
Unit
Pi
1
671
10.1
0.000299
150
455
400
80
120
2
574
10.2
0.000183
150
455
300
80
120
3
374
8.8
0.001126
20
130
105
130
130
4
374
8.8
0.001126
20
130
100
130
130
5
461
10.4
0.000205
150
470
90
80
120
6
630
10.1
0.000301
135
460
400
80
120
7
548
9.8
0.000364
135
465
350
80
120
8
227
11.2
0.000338
60
300
95
65
100
9
173
11.2
0.000807
25
162
105
60
100
10
175
10.7
0.001203
25
160
110
60
100
11
186
10.2
0.003586
20
80
60
80
80
12
230
9.9
0.005513
20
80
40
80
80
13
225
13.1
0.000371
25
85
30
80
80
14
309
12.1
0.001929
15
55
20
55
55
15
323
12.4
0.004447
15
55
20
55
55
The data for the individual units are given in Table 6. The best solutions of the proposed MRPSO, PSO, CPSO and WIPSO methods are shown in Table 8.
Table 6
Results of Three generator system (100 trails)
Unit Power
Output
PSO
CPSO
WIPSO
MRPSO
P1(MW)
145.73
144.8978
146.408
143.34
P2(MW)
338.45
340.9597
343.45
346.45
P3(MW)
549.7817
547.8717
543.563
534.565
Power
loss(MW)
183.043
183.7293
183.689
183.645
Total Power Output
1033.958
1033.7
1033.421
1033.355
Total
Cost($/h)
9842.228
9839.228
9834.781
9833.605
Computation
time (sec.)
0.368939
0.356130
0.479264
0.350648
Table 7
Generator output for six generator system (100 trails)
Unit Power
Output
PSO
CPSO
WIPSO
MRPSO
P1(MW)
493.24
471.66
454.39
462.6651
P2(MW)
114.63
140.03
164.279
195.36
P3(MW)
263.41
240.06
246.223
237.2409
P4(MW)
139.71
149.97
123.21
98.00
P5(MW)
179.65
173.78
167.22
197.7415
P6(MW)
84.83
99.97
120.00
83.4235
Loss
12.22
12.38
12.24
12.11
Total Power
Output
1275.46
1275.31
1275.3
1275.116
Total
Cost($/h)
15489
15481.87
15453.13
15441.9
Computation
Time(sec)
0.524359
0.479387
0.459492
0.464212


Result and Analysis
The Economic load dispatch problem solved by using the MRPSO and its performance is compared with PSO, CPSO and WIPSO. Data given for different generating units in Table 1, Table 3 and Table 5. The result obtained for these data by PSO, CPSO, WPSO and MRPSO. The program for these pso to solve ELD problem are developed in Matlab 7.5 on a 1.4GHz, core2 solo processor with 2GB DDR of RAM.
The constants used in this study was, acceleration coefficient c1=c2=2, Wmax=0.9 and Wmin=0.4.
The performance of MRPSO in this study the value of taken 3.5.
The convergence behavior of MRPSO was tested for Economic load dispatch with ramp rate constraint on different cases. The first test case is taken for three generating units, the data for first test case given in table 1, with ramp rate limit constraints. The Bcoefficients are given in table 2 for calculation of power loss of the considered system. For testing of this case a total load of 850 MW was taken. The result obtained by PSO, CPSO, WIPSO and MRPSO is given in table 6. The result of test data shows the best value of cost in this test case calculated
by MRPSO is $ 9833.605/h and its computation time is 0.350648 sec., total power loss calculated by MRPSO in this test case is 183.645 MW and obtained total generated output power is 1033.355 MW. All these result obtained for test case shows that MRPSO take less computing time and obtain least value of cost and loss of the 3 generating system.
The second test case is taken for sixgenerating units, the data for second test case given in table 3, with ramp rate limit constraints. The Bcoefficients are given in table 4 for calculation of power loss of the considered system. For testing of this case a total load of 1250 MW was taken. The result obtained by PSO, CPSO, WIPSO and MRPSO is given in table 7. The result of test data shows the best value of cost in this test case calculated by MRPSO is $ 15441.9/h and its computation time is 0.464212 sec., total power loss calculated by MRPSO in this test case is 12.11 MW and obtained total generated output power is 1275.116 MW. All these result obtained for second test case shows that MRPSO take less computing time and obtain least value of cost and loss of the 6 generating system.
The third test case is taken for fifteen generating units, the data for second test case given in table 5, with ramp rate limit constraints. In this case not considered the loss of the system. For testing of this case a total load of 2650 MW was taken. The result obtained by PSO, CPSO, WIPSO and MRPSO is given in table 8. The result of test data shows the best value of cost in this test case calculated by MRPSO is $ 32462.15/h and its computation time is 0.46114 sec., obtained total generated output power is 2650 MW. All these result obtained for thired test case shows that MRPSO take less computing time and obtain least value of cost.
Table 8
Generator output for 15 generator system (100 trails)
8000
6000
objf inal
4000
2000
0
0 10 20 30 40 50 60 70 80 90 100
itermax
Figure.1. Fitness function of the conversion system for three generator system
4
x 10
5
4
obj inal
f
3
2
1
0
0 10 20 30 40 50 60 70 80 90 100
itermax
Figure.2. Fitness function of the conversion system for six generator system
6
x 10
4
3
obj inal
f
2
1
Unit Power Output
PSO
CPSO
WIPSO
MRPSO
P1(MW)
454.3
454.98
455
422
P2(MW)
452.8
455
448.3
455
P3(MW)
132
130
130
131
P4(MW)
129
130
130
131.6
P5(MW)
336.9
335.02
265.02
341
P6(MW)
423
424.25
460
460
P7(MW)
462.5
464.98
465
465
P8(MW)
61.7
60
62
70
P9(MW)
24.9
25
25
21.6
P10(MW)
20.98
20
20
20
P11(MW)
19.08
20
59
20
P12(MW)
73.5
75
75
63.2
P13(MW)
25.08
25
25
20.6
P14(MW)
16.5
15
15
13.89
P15(MW)
17.06
15
15
15
Total Power
Output
2649.30
2649.23
2649.32
2650.0
Total
Cost($/h)
32476.7
32467.77
32464.03
32462.15
Computation time (sec.)
04821058
0.422924
0.613154
0.461147
0
0 10 20 30 40 50 60 70 80 90 100
itermax
Figure.3. Fitness function of the conversion system for fifteen generator system
Figure.1, figure.2 and figure.3 show the graph between object final V/s itermax in eah iteration for the 3,6 and 15 generation unit system respectively.

Conclusion
In This paper MRPSO is used to solve the economic dispatch with ramp rate limit constraints. The test results obtained by MRPSO clearly demonstrated that it is capable of achieving global solution, it is computationally efficient and give better optimal results (minimum cost) than other PSO methods. Overall, the MRPSO algorithms have been shown to be very helpful in studying optimization problems in economic load dispatch problem.
REFERENCES

M.E. EIhawary & G.S. Christensen, Optimal economic operatiom of Electrical power system, New York, Academic, 1979.

Mezger Alfredo J & Katia de Almeida C, Short term hydro thermal scheduling with bilateral traction via bundle method, International Journal of Electrical power & Energy system 2007, 29(5), pp387396.

Martinez Luis Jose, Lora Trancoso Alicia & Santos Riquelme Jesus, Short term hydrothermal coordination based on interior point nonlinear programming and genetic Algorithm,IEEE porto power Tech Confrence, 2001.

M. Gar CW, Aganagic JG,Tony Meding Jose B & Reeves S,
Experience with mixed integer linear programming based approach on short term hydrothermal scheduling, IEEE transaction on power system 2001, 16(4), pp.743749.

G.Torres and V. Quintana, On a non linear multiple centrality corrections interiorpoint method for optimal power flow, IEEE. transaction on power system, vol.16, no2, 2001, pp.222228.

K.Ng and G.Shelbe, Direct load control a profitbased load management using linear programming, IEEE transaction on power system,vol.13,no.2, 1998, pp.688694.

Shi CC, Chun HC, Fomg IK & Lah PB., Hydroelectric generation scheduling with an effective differential dynamic programming algorithm, IEEE transaction on power system 1990, 5(3), pp.737743.

Erion Finardi C, silva Edson LD, & Laudia sagastizabal CV., Solving the unit commitment problem of hydropower plants via Lagrangian relaxation and sequential quadratic programming, Computaional & Applied Mathematics 2005, 24(3).

Tkayuki S & Kamu W., Lagrangian relaxation method for price based unit commitment problem, Engineering optimization taylor Francis 2004, pp. 3641.

D.I. sun, B.Ashley,B.Brewer,A.Hughes and W.F. Tinney, Optimal power flow by Newton Aproach, IEEE transaction on power system, vol.103, 1984, pp.28642880.

Nidhul Sinha, R.Chakrabarti & P.K. Chattopadhyay, Evolutionary programming techniques for Economic load Dispatch, IEEE transactions on Evolutionary Computation, Vol.7, No1, 2003, pp.8394.

K.P. Wong & C.C. Fung, Simulated annealing based economic dispatch algorithm, proc. Inst. Elect. Eng. C., Gen., transm.,Distrib.,vol.140, no.6, Nov.1993, pp.505519.

W.M. Lin, F.S. Cheng & M.T. Tsay, An improved Tabu search for economic dispatch with multiple minima, IEEE transaction on power system , vol.17, no.2, 2002, pp.108 112.

J.S. AlSumait, A.K. AlOthman & J.K. Sykulski, Application of pattern search method to power system valve point economic load dispatch, Elect. Power energy system, vol.29, no.10, 2007, pp.720730.

D.C.Walter & G.B.Sheble, genetic algorithm solution of economic dispatch with valve point loading, IEEE transaction on power system, vol.8, no.3, 1993, pp.1325 1332.

Tarek Bouktir, Linda Slimani & M.Belkacemi, A genetic algorithm for solving for the optimal power flow problem, Leonardo journal of sciences, Issue4, 2004, pp.4458.

K. Vaisakh & L.R. Srinivas, Differential Approach for optimal power flow solutions, Journals of theoretical and applied information Technology, 200508, pp. 261268.

Boumediene Allaoua & Abedellah Laoufi, Optimal power flow solution Unsing ant manners for electrical network, Advance in Electrical & Computer engg., Vol.9, 2009, pp.3440.

L.L. Lai & Mata Prasad, Application of ANN to economic load dispatch, proceeding of 4th international conference on Advance in power system control, Operation and management, APSCOM97, HongKong, nov1997, pp.707 711.

J.Kennedy & R.C. Eberhart, Particle Swarm Optimization, proceeding of IEEE international conference on Neural networks , Vol.4, 1995, pp. 19421948.

C.H. Chen & S.N. Yeh, PSO for Economic power dispatch with valve point effects, IEEE PES transmission & Distribution conference and Exposition Latin America, Venezuela, 2006.

K.S. Swarup, Swarm intelligence Approach to the solution of optimal power flow, Indian Institute of science, Oct 2006, pp. 439455.

K.T. Chaturvedi, Manjaree pandit & Laxmi Srivastava, Self Organizing Hierachical PSO for nonconvex economic load dispatch, IEEE transaction on power system, vol.23, no.3, Aug. 2008, pp.10791087.

Adel Ali Abou ELEla & Ragab AbdelAziz EISehiemy, Optimized Generation costs using modified particle Swarm optimization version, Wseas transactions on power systems, Oct2007, pp.225232.

A.PereiraNeto, C. Unsihuay & O.R. Saavedra, Efficient evolutionary strategy optimization procedure to solve the nonconvex economic dispatch problem with generator constraints, IEE proc. Gener. Transm. Distributed, vol.152, no.5, Sept 2005, pp.653660.

R.C. Eberhart & Y. Shi, Comparing inertia weights and constriction factor in PSO, in proc. Congr. Evolutionary computation, 2000, vol.1, pp.8488.

Hao Gao & Wenbo Xu, A new particle swarm algorithm and its globally convergent modifications, IEEE transactions on systems, man, and cyberneticsPart B: cybernetics, vol. 41, no. 5, october 2011, pp.13341351.

Phan Tu Vu, Dinhlungle & Joef, A Novel weight Improved Particle swarm optimization algorithm for optimal power flow and economic load dispatch problem, IEEE confrence, 2010, pp.17.

H. Shayeghi & A. Ghasemi, Application of MOPSO for economic load dispatch solution with transmission losses, International journal on technical and physical problems of Engineering, Issue 10, vol. 4, 2012, pp. 2734.