Modified Particle Swarm Optimization Algorithms for Solving Economic Load Dispatch

Abstract

Electrical power systems consist of many generation units with each unit is limited to its characteristics. These units must be operated such that their total output power meets the total system demand and system losses at the minimum operation cost. This problem is known as the economic load dispatch (ELD) problem. The cost function of generation units is non-smooth and non-linear. Therefore, metaheuristic techniques are employed to solve this non-convex optimization problem. In this paper, the particle swarm optimization (PSO) algorithm and three other modified versions of the PSO are used to solve this highly non-linear and constrained optimization problem. The modified versions of the PSO are weight enhanced particle swarm optimization (WEPSO), chaotic particle swarm optimization (CPSO), and time-varying acceleration coefficients particle swarm optimization (TVACPSO). These algorithms are applied to solve the ELD problem for IEEE 15-unit test system. Results show that the WEPSO algorithm gives the minimum system operation cost and has the highest convergence rate.