A Particle Swarm Optimization with stagnation detection and dispersion

Particles or candidate solutions in the standard particle swarm optimization (PSO) algorithms often face the problems of being trapped into local optima. To solve such a problem, this paper proposes a modified PSO algorithm with the stagnation detection and dispersion (PSO-DD) mechanism, which can detect a probable stagnation and is able to disperse particles. This mechanism will be described and its performance is evaluated using eight well-known 30-dimensional benchmark functions that are widely used in literature. The results show a promising alternative path for solving the common problem of local optima in PSO algorithms.