EGT — PSO for organized distribution of swarm members in crowd and diverse population

A key point of Evolutionary Game Theory (EGT) deals with equilibrium called Evolutionary Stable Strategy (ESS). Three main strategies commonly used in Particles Swarm Optimization (PSO) i.e. best neighbor, personal memory, and boosted only by inertia, combined together with ESS provide an efficient payoff matrix to optimize three coefficients of search direction of PSO algorithm. Meanwhile, SPSO does not satisfy mathematical theory of convergence, thus EGPSO is employed. Evolutionary Game based Particles Swarm Optimization (EGPSO) have been tested in unimodal function, therefore in this paper, a multi-purpose EGPSO is presented. The experiment was conducted in three different neighborhood topology variation, uniform, stars, and ring topology. The result shows that the speed of the proposed algorithm outperforms its predecessor, both in single and multi-objective functions. However, it still requires further development in stars topology towards MOPSO.