基于动态交换策略的快速多目标粒子群优化算法研究

A fast multi-objective particle swarm optimization based on dynamic exchange strategy

本文提出了一种基于动态交换策略的快速多目标粒子群优化算法,通过把初始种群分割成Pareto和Non_Pareto集合,并在迭代过程中对Pareto解集进行动态调整,从而较好地完成了多目标优化算法对Pareto解集的搜索和逼近。实验和应用实例均表明了该算法的有效性和快速性,并通过性能指标ER的计算验证了本算法优于某些同类的多目标优化算法。

A Particle Swarm Optimization （PSO） algorithm based on a dynamic exchange strategy for multi-objective optimization is proposed, where the particles is divided into two sets, one is called by Pareto sets, the other is non-Pareto set. The number of particles in these two set is adjusted dynamically under the optimization period. Several benchmarks are tested, the results proves that dynamic exchange strategy particle swarm optimization （DE-MOPSO） could efficiently and quickly approach the Pareto front of the multi-objective optimization problems, and using ER to validate the accuracy of this algorithm.