Developing Niching Algorithms in Particle Swarm Optimization


Abstract

Niching as an important technique for multimodal optimization has been used widely in the Evolutionary Computation research community. This chapter aims to provide a survey of some recent efforts in developing state-of-the-art PSO niching algorithms. The chapter first discusses some common issues and difficulties faced when using niching methods, then describe several existing PSO niching algorithms and how they combat these problems by taking advantages of the unique characteristics of PSO. This chapter will also describe a recently proposed lbest ring topology based niching PSO. Our experimental results suggest that this lbest niching PSO compares favourably against sonic existing PSO niching algorithms.