摘要
为进一步提高多目标粒子群算法的收敛性和多样性,提出一种多策略融合改进的多目标粒子群优化算法.首先,引入分解思想以增加Pareto解集的多样性;然后,在速度和位置更新时,引入"多点"变异,即随着迭代次数的递增,根据相应判据对位置的更新作出不同的变异,避免算法早熟现象的发生;最后,将更新后种群和最优解集进行非支配排序,最优解放入精英外部存档.仿真实验结果表明,与另外4种进化算法对比,所提出算法表现出良好的整体性能.
In order to improve the convergence and diversity of the multi-objective particle swarm optimization,an improved multi-objective particle swarm optimzation algorithm integrating multiply strategies is proposed. Firstly,decompositions are introduced to increase the diversity of the pareto sets. Then, in the updating of the velocity and location, the"multi-point"variation is introduced. With the increase of the number of iterations, the updating of the location varies according to the corresponding criterion, which avoids the premature phenomenon of the algorithm.Finally, the updated population and the optimal solution set are subjected to non-dominated sorting and stored in the elite external archive. Simulation experiment results show that the proposed algorithm has better overall performance than other four multi-objective optimizers.
出处
《控制与决策》
EI
CSCD
北大核心
2018年第2期226-234,共9页
Control and Decision
基金
河北省高等学校创新团队领军人才培育计划项目(LJRC013)
国家冷轧板带及装备工程研究中心开放课题项目(2012006)
河北省自然科学基金面上项目(F2016203249)
河北省博士研究生创新资助项目(CXZZBS2017049)
关键词
多目标优化
粒子群算法
多策略改进
非支配排序
multi-objective optimization
particle swarm optimization algorithm
multiple strategies improved
non- dominated ranking