A Particle Swarm Optimization Algorithm with Partial Mutation Strategy and its Application

摘要

A particle swarm optimization algorithm with partial mutation strategy (PMPSO) is developed. During the searching process, only premature convergence particles are mutated to escape from local optimum, and they are no more mutated in next several generations in order to search efficiently in other areas; other non-premature particles go on their evolutions normally. Several parameters of the PMPSO algorithm are studied in this paper. Experimental results show the feasibility and validity of PMPSO. Furthermore, application result demonstrates that PMPSO is more feasible and efficient for optimal design of three-dimensional hydrofoil.