A Diversity Guided Particle Swarm Optimization with Chaotic Mutation

Particle Swarm Optimization (PSO) as well as genetic algorithm has shown good search abilities in many optimization problems. However, PSO easily falls into local minima on complex problems because of the loss of swarm diversity. This paper presents an improved diversity guided PSO algorithm, called DCPSO, by employing a modified velocity model and a chaotic mutation operator. In order to verify the performance of DCPSO, we test it on six benchmark functions. The simulation results show that DCPSO outperforms other two variants of PSO in all test cases.