An Investigation into Mutation Operators for Particle Swarm Optimization

The Particle Swarm Optimization (PSO) technique can be augmented with an additional mutation operator that helps prevent premature convergence on local optima. In this paper, different mutation operators for PSO are empirically investigated and compared. A review of previous mutation approaches is given and key factors concerning how mutation operators can be applied to PSO are identified. A PSO algorithm incorporating different mutation operators is applied to both mathematical and constrained optimization problems. Results shown that the addition of a mutation operator to PSO can enhance optimisation performance and insight is gained into how to design mutation operators dependent on the nature of the problem being optimized.