The impact of parameter adjustment strategies on the performance of particle swarm optimization algorithm

Abstract

To determine the reasonable parameter settings of particle swarm optimization (PSO) algorithm, this paper discusses the impact of the time-varying inertia weight and velocity-based mutation strategies on the performance of PSO algorithm. The performance of the PSO algorithm with these two kinds of parameters adjustment strategies are tested through four well-known benchmark functions. The simulation results show that the PSO algorithm has better convergence performance with the quickly decreasing inertia weight. Also, the velocity-based mutation strategy will slow down the convergence speed of PSO algorithm if the global solutions over the adjacent generations are close to each other.