一种基于多策略差分进化的分解多目标进化算法
CSTR:
作者:
作者单位:

1. 中国民航大学 电子信息与自动化学院,天津 300300;2. 广西民族大学 人工智能学院,南宁 530006

作者简介:

通讯作者:

E-mail: hmzhao@cauc.edu.cn.

中图分类号:

TP39

基金项目:

国家自然科学基金项目(61771087,62066005);中国民航大学科研启动基金项目(2020KYQD123).


A novel decomposition multi-objective evolutionary algorithm based on differential evolution model with multi-strategy
Author:
Affiliation:

1. College of Electronic Information and Automation,Civil Aviation University of China,Tianjin 300300,China;2. College of Artificial Intelligence,Guangxi University for Nationalities,Nanning 530006,China

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    为了提高多目标优化问题非支配解集合的分布性和收敛性,根据不同差分进化策略的特点,基于切比雪夫分解机制,提出一种基于多策略差分进化的分解多目标进化算法(MOEA/D-WMSDE).该算法首先采用切比雪夫分解机制,将多目标优化问题转化为一系列单目标优化子问题;然后引入小波基函数和正态分布实现差分进化算法的参数控制,探究一种基于5种变异策略优势互补的最优变异策略,提出一种基于参数控制和最优变异策略的多策略差分进化(WMSDE)算法;在此基础上,实现一种基于WMSDE的分解多目标进化算法.采用ZDT和DTLZ测试函数验证MOEA/D-WMSDE算法的有效性,实验结果表明:所提算法在收敛性和分布性方面获得了较大的改进与提高,能够有效求解多目标优化问题;与其他算法对比分析表明,所获得的解集整体质量更优,为多目标问题求解提供了新方法.

    Abstract:

    In order to improve the distribution and convergence of the non-dominated solution set of a multi-objective optimization problem, according to the characteristics of different differential evolution strategies, a novel decomposition multi-objective evolution algorithm based on the Chebyshev decomposition mechanism and differential evolution model with multi-strategy, namely MOEA/D-WMSDE is proposed in this paper. The MOEA/D-WMSDE uses the Chebyshev decomposition mechanism to transform the multi-objective optimization problem into a series of single objective optimization subproblems. Then the wavelet basis function and normal distribution are used to control parameters. An optimal mutation strategy based on complementary advantages of five mutation strategies is deeply studied in order to propose a new differential evolution(WMSDE) algorithm with multi-strategy. On this basis, the MOEA/D-WMSDE algorithm is realized. Finally, the ZDT and DTLZ benchmark functions are used to prove the optimization performance of the MOEA/D-WMSDE. The experimental results show that the MOEA/D-WMSDE has greatly improved the convergence and distribution, and can effectively solve the multi-objective optimization problem. Compared with the other algorithms, the overall quality of the obtained solution set is superior, which provides a new method to solve multi-objective optimization problems.

    参考文献
    相似文献
    引证文献
引用本文

邓武,蔡幸,周永权,等.一种基于多策略差分进化的分解多目标进化算法[J].控制与决策,2022,37(2):387-392

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2022-01-07
  • 出版日期: 2022-02-20
文章二维码