A Hybrid Multi-phased Particle Swarm Optimization with Sub Swarms

As is known, multi-phase particle swarm optimization (MPPSO) outperforms standard PSO and it is fitful for searching with small swarm size but does not converge as quickly as expected. A hybrid MPPSO with sub swarms called HMPPSO is developed in this paper. In HMPPSO, following the MPPSO updating rules, the sub swarms work cooperatively and alternatively with the optimal swarm, which is formed by the global best particle of each sub swarm, to enhance the global exploration and local exploration. The reinitiating operator is applied to maintain the diversity of the swarms; the feedback operator is used to improve the information exchange among the sub swarms and the optimal swarm. In addition, the size of each sub swarm and the size of the optimal swarm are kept small to save the computation time of fitness function. Numerical simulation results on some benchmarks show that the HMPPSO outperforms MPPSO in both the optimization ability and computation efficiency.