基于参考线的预测策略求解动态多目标优化问题
作者:
作者单位:

(兰州理工大学电气工程与信息工程学院,兰州730050)

作者简介:

通讯作者:

E-mail: lecstarr@163.com.

中图分类号:

TP273

基金项目:

国家自然科学基金项目(61763026).


Prediction strategy based on reference line for dynamic multi-objective optimization
Author:
Affiliation:

(College of Electrical Engineering and Information Engineering,Lanzhou University of Technology,Lanzhou730050,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:

    为了快速且准确地跟踪动态多目标优化问题变化的Pareto前沿与Pareto解集,在可以不依靠历史信息的前提下,提出一种基于参考线预测策略的求解动态多目标优化问题的算法(RLPS).该算法通过记录每个参考线关联的种群个体在环境变化初始时和个体自主进化一小段时间后个体位置的变化,预测最优个体所在方向,同时在该方向上均匀分布若干延伸个体,选出每个参考线关联的非支配个体作为当前环境下的引导个体,在选出的引导个体邻域内随机产生若干伴随个体增加种群多样性.通过5个标准动态测试函数对该算法测试,并与两个现有算法作对比分析,实验结果表明所提出的算法具有更快地响应环境变化的能力.

    Abstract:

    In order to track the Pareto front and Pareto solution set of dynamic multi-objective optimization problems quickly and accurately, an algorithm based on a reference line prediction strategy is proposed to solve dynamic multi-objective optimization problems without relying on historical information. The algorithm predicts the direction of the optimal individual by recording the changes of individual position at the beginning of the environmental change and a short period of time after the individual's self-evolution. At the same time, non-dominated individuals associated with each reference line are selected as the guided individuals according to the uniform distribution of several extend individuals in this direction. The results of 7 benchmark problems and the comparison with other two existing dynamicmulti-objective algorithms indicate that the proposed algorithm can maintain better performance in dealing with dynamic multi-objective problems.

    参考文献
    相似文献
    引证文献
引用本文

李二超,赵雨萌.基于参考线的预测策略求解动态多目标优化问题[J].控制与决策,2020,35(7):1547-1560

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: 2020-05-15
  • 出版日期: