基于邻域驱动的粒子群算法  

Particle swarm algorithm based on neighborhood driven

作　　者：李国森 闫李 郭倩倩 赵启凤 岳彩通 LI Guo-sen;YAN Li;GUO Qian-qian;ZHAO Qi-feng;YUE Cai-tong(School of Electronic and Information Engineering,Zhongyuan University of Technology,Zhengzhou 450007,China;School of Electrical Engineering,Zhengzhou University,Zhengzhou 450001,China;Industrial Technology Research Institute,Zhengzhou University,Zhengzhou 450001,China)

机构地区：[1]中原工学院电子信息学院,河南郑州450007 [2]郑州大学电气工程学院,河南郑州450001 [3]郑州大学产业技术研究院,河南郑州450001

出　　处：《计算机工程与设计》2022年第11期3159-3167,共9页Computer Engineering and Design

基　　金：国家自然科学基金项目(62103456、61976237、61922072、61876169、61673404);河南省高等学校重点科研基金项目(19A120014、20A120013);河南省自然科学基金项目(202300410511、212300410321);中原英才计划基金项目(ZYQR201810162);河南省高等学校青年骨干教师培养计划基金项目(2021GGJS111);河南省高校创新人才基金项目(16HASTIT033);中国纺织工业联合会科技指导性基金项目(2017054、2018104)。

摘　　要：针对算法存在多样性缺失、获得的Pareto解集不完整等不足,提出基于邻域驱动的粒子群算法(DNPSO)。采用邻域驱动策略,构造多个不同的邻域,推动种群协同搜索;采用吸收机制,充分利用邻域的搜索经验,加强对邻域的深度开采;设计3个多模态多目标测试函数。将DNPSO应用到15个测试函数和1个实际优化问题,并与8个多模态多目标算法进行对比。实验结果表明,DNPSO能够在决策空间获得更多的解,且解的多样性和收敛性较好。Aiming at the problem that the algorithm has diversity loss and incomplete Pareto set,a particle swarm algorithm based on neighborhood driven(DNPSO)was proposed.The neighborhood driven strategy was used to construct multiple neighborhoods to promote the collaborative search of the population.The absorption mechanism was used to make full use of the neighborhood search experience,strengthen the deep mining ability of the neighborhood.Three multi-modal multi-objective test functions were designed.DNPSO was applied to fifteen test functions and one practical optimization problem,and compared with eight multi-modal multi-objective algorithms.Experimental results show that DNPSO can obtain more solutions in decision space,and the diversity and convergence of solutions are better.

关 键 词：进化算法 粒子群优化 多模态优化 多目标优化 帕累托最优解集 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]