Family Particle Swarm Optimization with piecewise mutation probability

Abstract

Family Particle Swarm Optimization (FPSO), which combines the sociological conception with particle swarm optimization (PSO), has been proposed. In this paper, to improve the convergence accuracy of FPSO, the mutation strategy was introduced into the FPSO. To study the effect of mutation probability on the performance of the algorithm, piecewise functions were used to determine the mutation probability. The experimental results showed that the FPSO with mutation could improve the performance of FPSO and the selected mutation probability depended on the complexity of problem being optimized. For high dimensional problem, piecewise mutation had high evolution velocity in an early stage. In low dimensional space, piecewise decreasing mutation had obvious advantage at the convergence accuracy and evolution velocity.