Particle Swarm Optimization with Adaptive Mutation in Local Best of Particles

Particle Swarm Optimization (PSO) has shown its good search ability in many optimization problems. But PSO easily gets trapped into local optima while dealing with complex problems due to lacks in diversity. In this work, we proposed an improved PSO, namely PSO-APMLB, in which adaptive polynomial mutation strategy is employed in local best of particles to introduce diversity in the swarm space. In the first version of this method (PSO-APMLB1), each local best is perturbed in the current search space instead of entire search space. In the second version of this method (PSO-APMLB2), each local best is perturbed in terms of the entire search space. We also proposed another local best mutation method, namely, PSO-AMLB, in which mutation size is controlled dynamically in terms of current search space. In this work, we carried out our experiments on 8 well-known benchmark functions. Finally the results are compared with PSO. From the experimental results, it is found that the proposed algorithms performed better than PSO.