《计算机应用研究》|Application Research of Computers

曲线递增策略的自适应粒子群算法研究

Research on adaptive particle swarm optimization algorithm with curve increasing strategy

免费全文下载 (已被下载 次)  
获取PDF全文
作者 吴凡,洪思,杨冰,胡贤夫
机构 1.安徽工业大学 管理科学与工程学院,安徽 马鞍山 243032;2.丘钛微电子科技有限公司,江苏 昆山 215300
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2021)06-009-1653-04
DOI 10.19734/j.issn.1001-3695.2020.09.0235
摘要 群智能算法以其动态寻优能力强、实现途径简单等特点不断成为进化算法领域的研究热点。控参的选择对算法寻优性能有着极大影响,首先从数学推导角度对粒子群参数进行深入研究,接着提出一种契合粒子本身进化公式的,且具有反向思维的曲线递增策略的改进算法。最后验证该算法具备以下两点突出优势:a)有效避免早熟问题,在处理维度灾难问题上,寻优性能更强,且具备良好的平衡全局与局部寻优性能;b)算法控参简单,可有效解决鲁棒性低且繁琐的人工调参问题。
关键词 控参; 数学推导; 曲线递增; 粒子群算法
基金项目 国家自然科学基金资助项目(61702006)
本文URL http://www.arocmag.com/article/01-2021-06-009.html
英文标题 Research on adaptive particle swarm optimization algorithm with curve increasing strategy
作者英文名 Wu Fan, Hong Si, Yang Bing, Hu Xianfu
机构英文名 1.School of Management Science & Engineering,Anhui University of Technology,Ma'anshan Anhui 243032,China;2.Qiutai Microelectronics Technology Co. Ltd. ,Kunshan Jiangsu 215300,China
英文摘要 Swarm intelligence algorithm has become a hotspot in the field of evolutionary algorithm because of its strong dynamic optimization ability and simple implementation approach. The choice of control parameters has a great influence on the optimization performance of the algorithm. Firstly, this paper studied the particle swarm parameters from the point of mathematical derivation, then proposed an improved algorithm with inverse thinking curve increment strategy, which fitted the evolution formula of the particle itself. Finally, it verified that the algorithm has the following two outstanding advantages. It can effectively avoid the problem of precocity. In dealing with the problem of dimensional disaster, the optimization performance is stronger and has a good balance between global and local optimization performance. The algorithm is simple and can effectively solve the problem of low robustness and tedious manual parameter adjustment.
英文关键词 control parameters; mathematical deduction; increasing curve; particle swarm optimization(PSO)
参考文献 查看稿件参考文献
 
收稿日期 2020/9/8
修回日期 2020/10/26
页码 1653-1656,1661
中图分类号 TP301
文献标志码 A