基于指标和自适应边界选择的高维多目标优化算法
作者:
作者单位:

兰州理工大学

作者简介:

通讯作者:

中图分类号:

TP273

基金项目:


An Indicator-Based Many-Objective Evolutionary Algorithm with Adaptive Boundary Selection
Author:
Affiliation:

Lanzhou University of Technology

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:

    多目标优化算法的主要目标是实现好的多样性和收敛性。传统的高维多目标优化算法,在目标维数增加的时候,选择方式难以平衡种群的收敛性和多样性。针对这个问题,本文提出了一个基于指标和自适应边界选择的高维多目标优化算法。在环境选择中,首先计算种群中两两个体的指标 作为第一选择标准,然后提出一种自适应边界选择策略,利用种群进化信息对超平面系数进行模糊预测,后近似计算待选个体到超平面的范式距离作为第二选择标准。最后将所提算法与4种代表性的高维多目标算法进行比较,实验结果表明,算法处理复杂Pareto前沿高维多目标优化问题时,能在平衡收敛性和多样性的同时,更好的维护多样性。

    Abstract:

    The main goal of the multi-objective optimization algorithm is to achieve good diversity and convergence. In traditional high-dimensional multi-objective optimization algorithms, the selection operator is difficult to balance the convergence and diversity of the population. when the dimensionality of the objective increases. To solve this problem, this paper proposes a high-dimensional multi-objective algorithm named an indicator-based many-objective evolutionary algorithm with adaptive boundary selection. In environmental selection, first calculate the index of the two bodies in the population as the first selection criterion, and then propose an adaptive boundary selection strategy, which uses population evolution information to make fuzzy predictions of hyperplane coefficients, and then approximately Calculate the paradigm distance from the candidate individual to the hyperplane as the second selection criterion. Finally, the proposed algorithm is compared with four representative high-dimensional multi-objective algorithms. The experimental results show that when the algorithm handles the complex Pareto frontier high-dimensional multi-objective optimization problem, it can balance convergence and diversity while achieving better Maintain diversity.

    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:2020-11-03
  • 最后修改日期:2021-02-02
  • 录用日期:2021-02-10
  • 在线发布日期:
  • 出版日期: