基于种群关联策略和强化解集准则的高维多目标进化算法
DOI:
作者:
作者单位:

南昌航空大学

作者简介:

通讯作者:

中图分类号:

TP18

基金项目:

国家自然科学基金(No.62066031,No.61866025,No.61866026); 江西省自然科学基金(No.2018BAB202025); 江西省优势科技创新团队计划(No.2018BCB24008);基于自适应参考点策略和降维技术的高维多目标进化优化研究(YC2020030)


Many-objective Evolutionary Algorithm based on Population Association Strategy and Enhanced Solution set Criterion
Author:
Affiliation:

Nanchang Hangkong University

Fund Project:

National Natural Science Foundation of China(No.62066031,No.61866025,No.61866026);Jiangxi Natural Science Foundation(No.2018BAB202025);The plan of Jiangxi superior science and technology innovation team(No.2018BCB24008);Many-objective evolutionary optimization based on adaptive reference point strategy and dimension reduction technology(YC2020030)

  • 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:

    研究表明一般的高维多目标进化算法无法有效处理不同类型的Pareto前沿,针对上述情况,本文提出了一种基于种群关联策略和强化解集准则的高维多目标进化算法(Many-objective evolutionary algorithm based on population association strategy and enhanced solution set criterion,MaOEA/PAS-ESC)。该算法在环境选择中采用种群关联策略(Population association strategy, PAS)和强化解集准则(Enhanced solution set criterion, ESC)协同指导种群进化。其中,PAS利用解与参考向量的角度和欧氏距离以及种群中解之间的距离来选择多样性良好的解,ESC利用参考点与种群间的联系选择来收敛性良好的解,以共同达到有效平衡多样性和收敛性的目的。实验结果证明MaOEA/PAS-ESC在处理高维多目标优化问题不仅具有更强的竞争性能,而且提高了处理不同类型Pareto前沿的能力。

    Abstract:

    Related research shows that general many-objective optimization evolutionary algorithms cannot effectively process different types of Pareto fronts (PFs). To address this problem, we propose a many-objective evolutionary algorithm based on population association strategy and enhanced solution set criterion,named MaOEA/PAS-ESC. In the environment selection, the population association strategy (PAS) and enhanced solution set criterion (ESC) are used to cooperatively guide population evolution. For the purpose of trade-off between diversity and convergence, PAS uses the angles, Euclidean distances between solutions and reference vectors and the distances between solutions in current population to maintain diversity. And ESC uses the association between reference points and current population to select solutions with good convergence. The experimental results prove that MaOEA/PAS-ESC not only has more competitive performance in dealing with many-objective optimization problems, but also improves the ability to handle different types of PFs.

    参考文献
    相似文献
    引证文献
引用本文
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:2021-01-22
  • 最后修改日期:2021-07-19
  • 录用日期:2021-07-29
  • 在线发布日期:
  • 出版日期: