Two Modified NichePSO Algorithms for Multimodal Optimization

Multimodal function optimization (MMO) has seen a lot of interest and research over the past several years due to its many real world applications, and its complexity as an optimization problem. Several niching techniques proposed in past research have been combined with popular meta heuristic search algorithms such as evolutionary algorithms (EA), genetic algorithms (GA) and particle swarm optimization (PSO). The NichePSO algorithm was one of the first PSO algorithms proposed for utilizing niching methods and parallel swarms to apply PSO to MMO problems effectively. In this paper, two modified versions of the NichePSO algorithm are proposed, the NichePSO-R and NichePSO-S algorithms, in an attempt to improve its performance. Experimental results show that both proposed algorithms are able to locate more global optima on average than the NichePSO algorithm across several popular MMO benchmark functions.