An Improved Diversity Guided Particle Swarm Optimization

Abstract

Particle swarm optimization (PSO) is a new population based stochastic search algorithm, which has shown good performance on well-known numerical test problems. However, on strongly multimodal test problems the PSO easily suffers from premature convergence. In this paper, an improved diversity guided PSO is proposed, namely IARPSO, which combines a diversity guided PSO (ARPSO) and a Cauchy mutation operator. The purpose of IARPSO is to enhance the global search ability of ARPSO by Conducting a Cauchy mutation on the global best particle. Experimental results on 6 multimodal functions with many local minima show that the IARPSO outperforms the standard PSO, ARPSO and ATRE-PSO on all test functions.