Multi-objective Particle Swarm Optimization Algorithm Based on Self-update Strategy

In multi-objective particle swarm optimization (MOPSO) algorithms, improving the diversity of solutions is very difficult yet an important problem. In this paper, a new MOPSO algorithm of searching the Pareto-optimal solution is introduced, called multi-objective particle swarm optimization algorithm based on self-update strategy (SU-MOPSO). The mainly strategy of SU-MOPSO is that improving the diversity of each particle local best position (usually called pbest) to satisfy the swarm update´s needs, and fundamentally enhances the diversity of Pareto set by rising the candidate quantity. The proposed SU-MOPSO algorithm has been compared with ES-MOPSO algorithm. The results demonstrate that the SU-MOPSO algorithm has gained better convergence with even distributing and diversity of Pareto set.