A study on the minimal number of particles for a simplified particle swarm optimization algorithm

Abstract

Recent research on particle swarm optimization (PSO) emphasizes the need to simply this algorithm. This paper is a short study on finding the minimal number of particles in a simplified PSO. We have taken into consideration a social-only variant, Pedersen's simplified PSO, and tested it with four popular optimization benchmark functions in order to discover which is the minimal number of particles that can be used in most optimization problems.