Particle Swarm Optimization: Dynamic parameter adjustment using swarm activity

In this paper, swarm activity, which is a new index for assessing the diversification (global search) and intensification (local search) during particle swarm optimization (PSO) searches, is introduced. It is shown that swarm activity allows the quantitative assessment of the diversification and intensification during the PSO search. Using this concept, a new PSO called activity feedback PSO (AFPSO) is constructed, which involves feedback based on swarm activity to control diversification and intensification during the search. For each of the 5 benchmark problems, this method is used to determine the globally optimal solutions. These numerical experiments show that AFPSO has generality and effectiveness.