Impact of inertia weight strategies in particle swarm optimization for solving economic dispatch problem

Abstract

Particle Swarm Optimization (PSO) is a population based stochastic optimization technique inspired by the social learning of birds or fish. Some of the appealing facts of PSO are its convenience, simplicity and easiness of implementation requiring but few parameters adjustments. Inertia Weight (ω) is one of the essential parameters in PSO, which often significantly the affects convergence and the balance between the exploration and exploitation characteristics of PSO. Since the adoption of this parameter, there have been large proposals for determining the value of Inertia Weight Strategy. In order to show the efficiency of this parameter in the Economic Dispatch problem(ED), this paper presents a comprehensive review of one or more than one recent and popular inertia weight strategies reported in the related literature. Among this five recent inertia weight four were randomly chosen for application and subject to empirical studies in this research, namely, Constant (ω), Random (ω), Global-Local Best (ω), Linearly Decreasing (ω), which are then compared in term of performance within the confines of the discussed optimization problem. Morever, the results are compared to those reported in the recent literature and data from SONELGAZ. The study results are quite encouraging showing the good applicability of PSO with adaptive inertia weight for solving economic dispatch problem.