A particle swarm optimization algorithm based on genetic selection strategy

Abstract

The standard particle swarm optimization algorithm (simply called PSO) has many advantages such as rapid convergence. However, a major disadvantage confronting the PSO algorithm is that they often converge to some local optimization. In order to avoid the occurrence of premature convergence and local optimization of the PSO algorithm, a particle swarm optimization algorithm based on genetic selection stra-tegy, simply called GSS-PSO, is singled out in this paper. GSS-PSO not only retains the rapid convergence charactering of the standard PSO algorithms, but also scales up their global search ability. At last, we experimentally tested the efficiency of our new GSS-PSO algorithm using eight classical functions. The experimental results show that our new GSS-PSO algorithm is generally better than the PSO algorithm. © 2009 Springer Berlin Heidelberg.