11 V May 2023

https://doi.org/10.22214/ijraset.2023.52161
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

Solving Economic Load Dispatch Problem Using

Grey Wolf Optimization Algorithm
Ramamoorthi R1, Balamurugan R2
1
Department of Electrical and Electronics Engineering, Government College of Engineering, Dharmapuri, Tamilnadu-636704,
India
2
Department of Electrical and Electronics Engineering, Government College of Engineering. Thanjavur, Tamilnadu-613402, India

Abstract: This article presents a new evolutionary optimization approach named grey wolf optimization (GWO), which is based
on the behavior of grey wolves, for the optimal operating strategy of economic load dispatch (ELD). Nonlinear characteristics of
generators like ramp rate limits, valve point discontinuities and prohibited operating zones are considered in the problem. GWO
method does not require any information about the gradient of the objective function, while searching for an optimum solution.
The GWO algorithm concept appears to be a robust and reliable optimization algorithm is applied to the nonlinear ELD
problems. The proposed algorithm is implemented and tested on two test systems having 40 Thermal generators. The results
confirm the potential and effectiveness of the proposed algorithm compared to various other methods available in the literature.
The outcome is very encouraging and proves that the GWO is a very effective optimization technique for solving various ELD
problems.
Keywords: Economic load dispatch, Valve point effects, Meta-heuristic algorithms, Grey wolf optimization algorithm, and non-
convex optimization.

I. INTRODUCTION
Economic load dispatch (ELD) is a process used in power systems to determine the optimal distribution of power output from
various generating units to meet the system demand while minimizing the operating costs. This is achieved by optimizing the output
of each generator in the system, subject to the operating constraints of the system such as transmission capacity, voltage limits, and
stability limits. The ELD problem is a nonlinear optimization problem that requires solving a set of equations to determine the
optimal operating point for each generator. The solution is typically obtained using mathematical optimization techniques such as
linear programming; nonlinear programming but these mathematical methods are having various disadvantages such as the
increased computational complexity of techniques with the size and complexity of the problem. As a result, it may be difficult or
even impossible to solve large-scale non convex optimization problems using these techniques. On the other hand, heuristic
optimization methods can handle nonlinear and non-convex problems effectively by searching the solution space using a population
of solutions and applying evolutionary operators such as mutation, crossover, and selection etc., also ELD problems involve
multiple constraints such as power balance, generator output limits, transmission capacity limits, and voltage limits.
Evolutionary algorithm can handle multiple constraints by incorporating penalty functions or constraints handling techniques into
the fitness function. In recent literature various heuristic algorithms are reported to solve the ELD problems. Gaing proposed
particle swarm optimization (PSO) to solve the ELD problem in power systems and compared with Genetic Algorithm. Several
nonlinear characteristics of the generator such as ramp rate limits, POZs and non-smooth cost functions were considered [1]. In [2],
Firefly algorithm [FA] was used to determine optimal solution for the ELD problems FA emulates social conduct of fire flies
dependent on their blazing quality. Dubey, pandit and panigrahi [3] presented modified flower pollination algorithm [MFPA] to
deal the ELD problems. In the MFPA neighborhood fertilization of FPA was constrained by a scaling component and a
concentrated exploitation stage was added to determine the best solutions. A continuous version of quick group search optimizer
(QGSO) algorithm was proposed to realize me ELD formulation with VPL effect POZs, transformation losses and ramp rate limits
[4]. Cuckoo search algorithm (CSA) was developed for solving both convex and non-convex ELD problems [5]. It was inspired
from the obligate brood parasitism of some cuckoo species by laying their eggs in the nests of other host birds of different species.
Simulated annealing technique has been applied to determine the optimal generation schedule for economic dispatch problems in a
power system [6]. Hybrid evolutionary programming (HEP) [14] is used to obtain best optimal solution.[7]. In [8] authors used
Evolutionary programming method for solving ELD problems .

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2556
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

This paper presents a new swarm based optimization algorithm known as Gray wolf optimization (GWO) developed by proposed
by Mirjalili et al. in 2014 [9]. This algorithm is inspired by the hunting nature of the gray wolf. To assess the effectiveness of the
GWO, Two test system having different constraints are taken in to consideration and optimized for best optimal fuel cost. Obtained
results are compared with the other heuristic methods in the literature.
Organization of this research article as follows: In Section 2, the formulation of the economic load dispatch problem is discussed. In
Section 3, the applied Gray wolf optimization algorithm is explained, as well as its implementation process for the ELD problem. In
Section 4, Simulation results are presented to compare GWO’s effectiveness to that of the original algorithm and other algorithms.
Final section summarizes this research work.

II. PROBLEM FORMULATION

The goal of the ELD problem is to find an optimal power generation schedule while minimizing fuel costs and also satisfying the
operating constraints of different power systems. The fuel consumption cost of a generator is explained as the following equation.
( )= + + (1)

Where a,b,c are the fuel cost weigh constants. The objective of the ELD problem is to minimize the fuel cost while satisfying
equality and inequality constraints
(2)
( )= ( )

Where ( ) is the fuel cost function of unit, is the power output of the unit, N is the number of generating units in the
system. The sequential valve opening in the multivalve steam turbines generates ripple effect on the fuel cost curve of the generator.
To model an accurate and practical ELD problem this valve point effect should be included in the fuel cost function as shown in
figure 1. Then the fuel cost function of each generating unit is expressed in the non-convex form as follows
( )= + + + ∗ ∗ − (3)

Figure 1 : Fuel cost curve for generators

A. Power system Constraints
1) Power balance constraints
The generators complete power output must be equal to the sum of power requirement and complete transmission losses and it is
provided by the following equations
(5)
= +

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2557
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

The transmission losses are expresses as

(6)
= + +

Here is the total load demand and is the system loss. In order to calculate the system loss the method based on penalty factor
and constant loss formula coefficient or B coefficient is adopted. , , are the loss coefficient of the generators.

2) Generator capacity Constraints

Each generating unit output power needs to be restricted by limiting inequality between its limits. This constraint is represented by
≤ ≤ (7)

III. GREY WOLF OPTIMIZATION ALGORITHM

Grey wolf optimization (GWO) is a new population based meta-heuristic algorithm proposed by Mirjalili et al. in 2014 [9]. The
method imitates the hunting behavior and social hierarchy of grey wolves. On the basis of behavior of grey wolves, GWO is
implemented where a specific number of grey wolves in a pack moves through a multi-dimensional search space to look for prey. In
this optimization algorithm, the positions of grey wolves are considered as different position variables and the distances of the prey
from the grey wolves determine the fitness value of the objective function. The movement of each individual is influenced by four
processes, namely searching for prey (exploration), Encircling prey, Hunting, Attacking prey (exploitation).
These operators are briefly explained and mathematically expressed in the following subsection.

A. Searching for Prey (Exploration)

The grey wolves diverge from each other position for searching a victim. Make use of AM with random values to compel the search
agent to diverge from the victim. The component CM provides random weights for searching prey in the search space. Hence
exploration through AM and CM permits this algorithm to globally search the area. CM vector also presents the effect of obstacles
to impending the prey.

B. Encircling Prey
The alpha, beta and delta estimate the position of the three best wolves and other wolves updates their positions using the positions
of these three best wolves. Encircling behavior can be represented by DM. The expected boundary is mathematically represented by
the following equations:
= | . ( ) − ( )| (8)

( + 1) = ( )− . (9)

Here t indicates the current iteration, AM and CM are coefficient vectors, XP(t) is the position vector of prey, X(t) represents the
position vector of a grey wolf. r1 and r2 are random vectors in [0, 1].a is linearly decreased from 2 to 0.
=2 ∗ − (10)

= 2∗ (11)
C. Hunting
Conservation of regional habitat connectivity has the potential to facilitate recovery of the grey wolf. After encircling, alpha wolf
guides for hunting. Later, the delta and beta wolves join in hunting. It is tough to predict about the optimum location of prey. The
hunting behavior of grey wolf, based on the position of alpha, beta, gamma (candidate solution) wolf can be represented by
=| . ( )+ | (12)

= . ( )+ (13)

=| . ( )+ | (14)

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2558
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

Finally, the position of various categories of wolves is modified as follows:

= − . (15)

= − . (16)

= − . (17)

+ + (18)
( + 1) =
3

D. Attacking prey (exploitation)

The grey wolves stop the hunting by attacking the prey when it stop moving. It depends on the value of a* AM is a random value in
the interval [-2a, 2a]. In GWO, search agents update their positions based on the location of alpha, beta, delta wolves mentioned in
hunting phase and attack towards the prey.

E. Grey wolf optimization applied to ELD

The different steps of GWO algorithm for solving ELD problems are described below.
1) Step 1: Active power generation of all the generating units initialized randomly within their lower and upper real power
operating limits
2) Step 2: Evaluate fitness of each solution of current population using (1)–(3). Each fitness value represents the distance of the
individual wolf from the prey.
3) Step 3: Sort the population from best to worst. The best, second best and third best solutions respectively, represent the
positions of , β and δ categories of wolves.
4) Step 4: Modify the position of each search agents using the searching prey, encircling prey, hunting and attacking prey
concepts. The position of each search agent represents a potential solution comprised of active power generation of ELD
problem.
5) Step 5: Check whether the operating limits of the active power of all generating units except last unit are violated or not. If any
power generation is less than the minimum level, it is made equal to minimum value. Similarly, if it is greater than the
maximum level, it is assigned its maximum value. Subsequently, last unit of the power generation is evaluated using (5) and
whether it satisfies all the inequality constraints or not is checked. The infeasible solutions are exchanged by the best feasible
solutions.
6) Step 6: Go to Step 2 until termination criteria is met. The GWO is stopped executing when the maximum number of iterations
(generations) is reached or there is no noteworthy improvement in the solution. In this paper, the ending criterion is the
maximum number of iterations for which most of the grey wolves or search agents are idle.

IV. CASE STUDIES AND NUMERICAL RESULTS

In order to validate the feasibility of the proposed GWO method for the ELD problems, it is employed on a relatively large system
consisting of 40 generating units. The load demand used in the simulations is 10500 MW. In order to judge the efficacy of the
proposed in nonlinear environment, the valve-point effect and prohibited operating zones are considered. Data for the test system is
referred from [10]. The developed algorithm is simulated and tested in MATLAB 7.1 Software on 2 GHz Pentium IV, 1 GB RAM
personal computer. The population size and the maximum iteration number are taken as 50 and 500 respectively for the test systems
under consideration.
Test Case 1: 40-unit system without valve point loading effect and without Transmission loss.
Test Case 2: 40-unit system including valve point loading effect and Transmission Loss.

A. Test System 1
In this 40 unit test system Valve point effects and Transmission losses are not considered for the simulation. The results obtained by
applying the GWO algorithm and other heuristic method known as variable scaling hybrid differential evolution (VSHDE) [11] are
summarized in Table 1 for 40-unit power system without considering the effects of valve-point loading without transmission losses.
Analyzing the data, it can be observed that the GWO method succeeds in finding a satisfactory solution.

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2559
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

The minimum cost obtained by the proposed GWO method was given by 121244 $/hr, which is the best cost found so far. The
analysis of these comparative results demonstrates that the proposed approach shows superior performance compared to other
method reported in the literature. Convergence curve for this test system is shown in figure 3.

Table 1 : Best fuel cost Simulation test results for the Test system 1
Unit GWO VHSDE Unit GWO VHSDE
1 80 79.63 22 550 550
2 120 119.99 23 550 550
3 190 189.98 24 550 550
4 41.24206 36.27 25 550 550
5 38.08678 42 26 550 550
6 140 140 27 550 549.99
7 300 300 28 10.17545 10
8 300 299.98 29 10.17545 10
9 300 300 30 10.17545 10
10 130 131.97 31 20 20.01
11 94 94.03 32 20 20.01
12 100.7911 94 33 20 20
13 171.2593 174.03 34 20 20
14 339.2156 327.7 35 18 18.01
15 337.9593 339.51 36 18 18
16 337.9593 339.49 37 20 20
17 337.9593 350.34 38 25 25.06
18 500 500 39 25 25
19 500 500 40 25 25
20 550 550 Total 121244 121253
21 550 550 Cost
(S/hr)

170000 Test system-1

160000
Best fuel cost ($/hr)

150000

140000

130000

120000

0 100 200 300 400 500

Iteration number
Figure 2 : Convergence curve for the test system-1

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2560
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

B. Test system 2
The simulation results obtained by proposed GWO method for 40- unit test system considering transmission losses and valve point
loading effects are compared with GA-API [12],SDE [10],TLBO [13] in Table 2. From Table 2, it was observed that GWO
outperforms other optimization methods. In this case, the GWO obtained total cost 2267.8 $/hr, 560.3 $/hr, 217 $/hr lesser than the
GA-API, SDE, and TLBO algorithms respectively. Convergence curve for this test system is shown in figure 3. The simulation
results clearly suggest that GWO produces feasible solutions. To judge the superiority and robustness,

Table 2 : Best fuel cost Simulation test results for the Test system 1
Unit GA-API SDE TLBO GWO Unit GA-API SDE TLBO GWO

1 114 110.06 114 114 22 550 550 522.1852 523.2794

2 114 112.41 114 114 23 550 550 526.1804 523.2794

3 120 120 120 120 24 550 528.16 521.1967 523.2794

4 190 188.72 182.4448 179.7331 25 550 524.16 525.801 523.2794

5 97 85.91 90.6923 87.7999 26 550 539.1 526.0022 541.3818

6 140 140 140 140 27 11.44 10 13.0804 10

7 300 250.19 300 300 28 11.56 10.37 11.0397 10

8 300 290.68 296.0682 300 29 11.42 10 12.9373 10

9 300 300 288.8518 300 30 97 96.1 89.7412 87.7999

10 205.25 282.01 281.952 279.5997 31 190 185.85 190 190

11 226.3 180.82 238.1293 243.5997 32 190 189.54 190 190

12 204.72 168.74 251.012 243.5997 33 190 189.96 190 190

13 346.48 469.96 483.1175 484.0392 34 200 199.9 200 200

14 434.32 484.17 481.9042 484.0392 35 200 196.25 200 200

15 431.34 487.73 488.2883 484.0392 36 200 185.85 164.7435 164.7998

16 440.22 482.3 396.3448 484.0392 37 110 109.72 110 110

17 500 499.64 494.2577 489.2794 38 110 110 110 110

18 500 411.32 408.3826 489.2794 39 110 95.71 110 110

19 550 510.47 510.5206 511.2794 40 550 532.47 547.9677 511.2794

20 550 542.04 521.2217 511.2794 Total 139865 138157.5 137814.2 137597.2

cost
($/hr)
21 550 544.81 540.57 433.5196 Total 1045.06 974.43 1002.63 1021.504
Loss
MW

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2561
 International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Volume 11 Issue V May 2023- Available at www.ijraset.com

260000
Test system-2
240000

Best fuel cost ($/hr)

220000

200000

180000

160000

140000

120000
0 100 200 300 400 500
Iteration no
Figure 3 : Convergence curve for Test system-2

V. CONCLUSION
In this work, an efficient meta-heuristic algorithm named GWO is proposed to solve the ELD problem taking the valve point
loading effects, prohibited operating zone, ramp rate limits into consideration. Two case studies are employed to demonstrate the
applicability of the GWO method. The benefit of the proposed GWO is that it does not impose any convexity limitations on the
generating unit characteristics. Numerical results show that the GWO method has superior features, advantages over other
algorithms in terms of robustness, avoids premature convergence, simple applicability and stable convergence characteristic.
Although, the proposed algorithm is applied to solve ELD problems in the current study, it seems from its unique feature that GWO
has the potential to solve many other optimization problems in the field of power system planning and operation.

REFERENCES
[1] Z.-L. Gaing, “Particle swarm optimization to solving the economic dispatch considering the generator constraints,” IEEE Trans. Power Syst., vol. 18, no. 3, pp.
1187–1195, Aug. 2003, doi: 10.1109/TPWRS.2003.814889.
[2] X.-S. Yang, S. S. Sadat Hosseini, and A. H. Gandomi, “Firefly Algorithm for solving non-convex economic dispatch problems with valve loading effect,”
Appl. Soft Comput., vol. 12, no. 3, pp. 1180–1186, Mar. 2012, doi: 10.1016/j.asoc.2011.09.017.
[3] H. M. Dubey, M. Pandit, and B. K. Panigrahi, “A Biologically Inspired Modified Flower Pollination Algorithm for Solving Economic Dispatch Problems in
Modern Power Systems,” Cognit. Comput., vol. 7, no. 5, pp. 594–608, Oct. 2015, doi: 10.1007/s12559-015-9324-1.
[4] M. Moradi-Dalvand, B. Mohammadi-Ivatloo, A. Najafi, and A. Rabiee, “Continuous quick group search optimizer for solving non-convex economic dispatch
problems,” Electr. Power Syst. Res., vol. 93, pp. 93–105, Dec. 2012, doi: 10.1016/j.epsr.2012.07.009.
[5] M. Basu and A. Chowdhury, “Cuckoo search algorithm for economic dispatch,” Energy, vol. 60, pp. 99–108, Oct. 2013, doi: 10.1016/j.energy.2013.07.011.
[6] K. P. Wong and C. C. Fung, “Simulated annealing based economic dispatch algorithm,” IEE Proc. C Gener. Transm. Distrib., vol. 140, no. 6, p. 509, 1993, doi:
10.1049/ip-c.1993.0074.
[7] A. K. Swain and A. S. Morris, “A novel hybrid evolutionary programming method for function optimization,” in Proceedings of the 2000 Congress on
Evolutionary Computation. CEC00 (Cat. No.00TH8512), pp. 699–705, doi: 10.1109/CEC.2000.870366.
[8] Hong-Tzer Yang, Pai-Chuan Yang, and Ching-Lien Huang, “Evolutionary programming based economic dispatch for units with non-smooth fuel cost
functions,” IEEE Trans. Power Syst., vol. 11, no. 1, pp. 112–118, 1996, doi: 10.1109/59.485992.
[9] S. Mirjalili, S. M. Mirjalili, and A. Lewis, “Grey Wolf Optimizer,” Adv. Eng. Softw., vol. 69, pp. 46–61, Mar. 2014, doi: 10.1016/j.advengsoft.2013.12.007.
[10] A. Srinivasa Reddy and K. Vaisakh, “Shuffled differential evolution for large scale economic dispatch,” Electr. Power Syst. Res., vol. 96, pp. 237–245, Mar.
2013, doi: 10.1016/j.epsr.2012.11.010.
[11] J.-P. Chiou, “Variable scaling hybrid differential evolution for large-scale economic dispatch problems,” Electr. Power Syst. Res., vol. 77, no. 3–4, pp. 212–
218, Mar. 2007, doi: 10.1016/j.epsr.2006.02.013.
[12] I. Ciornei and E. Kyriakides, “A GA-API Solution for the Economic Dispatch of Generation in Power System Operation,” IEEE Trans. Power Syst., vol. 27,
no. 1, pp. 233–242, Feb. 2012, doi: 10.1109/TPWRS.2011.2168833.
[13] P. K. Roy and S. Bhui, “Multi-objective quasi-oppositional teaching learning based optimization for economic emission load dispatch problem,” Int. J. Electr.
Power Energy Syst., vol. 53, pp. 937–948, Dec. 2013, doi: 10.1016/j.ijepes.2013.06.015.

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 | 2562