A Multi-population Coevolution Multi-objective Particle Swarm Optimization Algorithm

Abstract

In this paper, a multi-population coevolution multi-objective particle swarm optimization(MPCMOPSO) algorithm is proposed. In the proposed algorithm, multiple subpopulations and an inferior population are used to search jointly, which strengthens the utilization of inferior particles and not only improves the efficiency of the algorithm, but also effectively improves the convergence and searching ability of the algorithm. Archive set is used to retain non-dominated paticles to realize information exchange between subpopulations and the inferior population, where non-dominated particles in the archive set are used to replace the inferior particles. Moreover, a strategy to prevent particles from falling into local optima is proposed. Finally, experiment results on ZDTs series functions and comparison with other algorithms are given to validate the proposed algorithm.