Particle Swarm Optimization Using Adaptive Mutation

Abstract

Two new variants of particle swarm optimization (PSO) called AMPSO1 and AMPSO2 are proposed for global optimization problems. Both the algorithms use adaptive mutation using beta distribution. AMPSO1 mutates the personal best position of the swarm and AMPSO2, mutates the global best swarm position. The performance of proposed algorithms is evaluated on twelve unconstrained test problems and three real life constrained problems taken from the field of electrical engineering. The numerical results show the competence of the proposed algorithms with respect some other contemporary techniques.