Modified particle swarm optimization and its application in multimodal function optimization

Abstract

In multimodal optimization, the basic particle swarm optimization is easy to duplicate and miss points of the optimal value. To solve this problem, a modified particle swarm optimization algorithm, called BNPSO, is proposed. This modified particle swarm optimization algorithm is based on the niche particle swarm optimization (NPSO) algorithm, and implemented a Bernoulli trial. It is proved theoretically that the algorithm BNPSO is much more effective than the algorithm NPSO for multimodal function optimization problems. However, the time complexity of the new scheme is increased. Testing of the algorithm indicate that the algorithm BNPSO has better perform in stability and convergence.