华东师范大学学报(自然科学版) ›› 2019, Vol. 2019 ›› Issue (1): 66-75.doi: 10.3969/j.issn.1000-5641.2019.01.08

• 计算机科学 • 上一篇    下一篇

基于动态分组的多策略引力搜索算法

张强, 王梅   

  1. 东北石油大学 计算机与信息技术学院, 黑龙江 大庆 163318
  • 收稿日期:2018-01-12 出版日期:2019-01-25 发布日期:2019-01-24
  • 作者简介:张强,男,博士,研究方向为进化算法、神经网络和智能优化.E-mail:zhangqiang@nepu.edu.cn.
  • 基金资助:
    国家自然科学基金（61702093）；黑龙江省自然科学基金（F2015020）

Multi-strategy gravitational search algorithm based on dynamic grouping

ZHANG Qiang, WANG Mei   

  1. School of Computer and Information Technology, Northeast Petroleum University, Daqing Heilongjiang 163318, China
  • Received:2018-01-12 Online:2019-01-25 Published:2019-01-24

RichHTML

16

PDF

407

摘要： 给出了一种基于动态分组的多策略引力搜索算法.算法迭代初期利用自适应分组策略对种群进行分组寻优，每个分组内只更新最差个体，采用云模型理论来改进最优个体的进化行为；迭代后期将种群分为优势子群和拓展子群，采用差分变异算子更新优势子群提高寻优精度和速度，利用Tent混沌理论进化拓展子群完成个体变异.典型复杂函数测试表明，该算法具有很好的收敛精度和计算速度.

关键词: 引力搜索算法, 云模型, 佳点集, 混沌, 连续空间优化

Abstract: A multi-strategy gravitational search algorithm based on dynamic grouping is proposed in this paper. At the initial stage of the algorithm iteration, adaptive grouping strategies are used to optimize populations. Only the least-optimal individuals are updated in each group. The cloud model theory is used to improve the evolutionary behavior of the optimal individuals. In the later part of the algorithm iteration, the populations are divided into dominant and extension subgroups. The differential mutation operator is subsequently used to update the dominant subgroups to improve the precision and speed of the optimization. Tent chaos theory is used to update the extension subgroups to complete the individual variation. Typical complex function tests show that the algorithm has good convergence accuracy and computational speed.

Key words: gravitational search algorithm, cloud model, good point set, chaos, continuous space optimization

中图分类号: