Modified particle swarm optimization with switching update strategy

Abstract

This article aims at improving the Particle Swarm Optimization, by uniquely reshaping its update strategy for generating new solutions with a switching strategy that transits between exploration and convergence, a time-varying inertia weight to control particles' movement and an aging mechanism to avoid stagnation in local basins of attraction. The algorithm addressed as MPSO-SUS has been compared with eight other state-of-artEAs on a standard benchmark of sixteen functions. The results of such comparison indicate that MPSO-SUS clearly and statistically outperform the other well-known approaches, justifying its distinctive feature which makes it a successful optimizer. © 2012 Springer-Verlag.