Multi-strategy comprehensive article swarm optimization algorithm based on population partition
- 分享到:
摘要：
针对标准粒子群优化算法容易陷入局部最优收敛精度不高的问题,提出一种基于种群分区的多策略综合学习粒子群优化算法(MSPSO).该算法利用竞争机制将种群分为两个子种群:潜力子群与普通子群,对这两个子群实行不同的进化策略,潜力子群中的粒子主要负责全局探索,普通子群中的粒子则侧重于局部勘探.为验证算法的性能,在不同类型的基准函数上与其他粒子群算法及其他群智能算法进行对比,所提算法都能取到最优的平均结果,证明所提算法具有更优异的算法性能.
In order to solve the problem that the canonical particle swarm algorithm is prone to fall into the local optima and is of low convergence accuracy,a multi-strategy comprehensive learning particle swarm optimization algorithm(MSPSO)is proposed based on population partition.In MSPSO the population is divided into potential subgroup and common subgroup based on the competition mechanism.Both subgroups use different evolutionary strategies.The particles in potential subgroup are mainly responsible for global exploration,while particles in common subgroup focus on local exploitation.In order to verify the performance of the proposed algorithm,MSPSO is compared with other PSO-based competitors and other swarm intelligence algorithms.It gets the best average results on different benchmark functions,and it is proved that the proposed algorithm has better performance.
作者：
李冰晓 万睿之 朱永杰 赵新超
Li Bingxiao;Wan Ruizhi;Zhu Yongjie;Zhao Xinchao(School of Science,Beijing University of Posts and Telecommunications,Beijing 100876,China)
机构地区：
北京邮电大学理学院
出处：
《betway官方app 学报：自然科学版》 CAS 北大核心 2022年第3期85-94,共10页
Journal of Henan Normal University(Natural Science Edition)
基金：
国家自然科学基金(61973042) 北京市自然科学基金(1202020).
关键词：
粒子群优化 竞争机制 多策略学习 种群分区 综合学习
particle swarm optimization competition mechanism multi-strategy learning population partition comprehensive learning
分类号：
TP301.6 [自动化与计算机技术—计算机系统结构]
