基于多区域中心点预测的动态多目标优化算法
作者:
作者单位:

燕山大学电气工程学院

作者简介:

通讯作者:

中图分类号:

TP273

基金项目:

国家自然科学基金项目(面上项目,重点项目,重大项目)No.61803327; 河北省青年基金自然科学基金No.E2018203162;河北省博士后科研项目 B2019003021;河北省自然科学基金F2020203031;河北大学科技攻关项目QN2020225;燕山大学博士基金No.BL18048


A dynamic multi-objective optimization algorithm based on multi-regional center point prediction
Author:
Affiliation:

Institute of Electrical Engineering, Yanshan University

Fund Project:

National Natural Science Foundation of China (Grant No.61803327), the Youth Fund Natural Science Foundation of Hebei (No.E2018203162), Post-Doctoral Research 24 Projects of Hebei(B2019003021), the Natural Science Foundation of Hebei (F2020203031), Science and Technology Research Projects of Hebei University (QN2020225) and Doctoral Foundation of Yanshan University (No.BL18048).

  • 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:

    现实生活中存在很多动态多目标优化问题(DMOPs),这类问题要求算法在环境变化后快速收敛到新的Pareto前沿,并保持解集的多样性,且随着Pareto前沿复杂程度的增加,这一问题更加突出. 针对此问题,提出一种基于多区域中心点预测的动态多目标优化算(MCPDMO). 首先,根据环境变化的严重程度将种群划分为多个子区域, 使个体的分配更加适应动态变化的环境; 其次,分别计算每个子区域的中心点,对不同子区域在不同时刻的中心点建立时间序列,并利用差分模型预测新环境的最优解集,以提高算法对不同环境变化的响应能力. 为 验证算法的有效性,与3种动态多目标优化算法在10个标准测试函数上进行仿真实验. 结果表明,该算法在具有复杂Pareto前沿的动态问题上表现出更优的收敛性和分布性.

    Abstract:

    In real life, there are many dynamic multi-objective optimization problems (DMOPs), which require algorithms to quickly converge to the new Pareto front and maintain the diversity of the solution set. As the complexity of the Pareto front increases, this problem is more prominent. For this problem, a dynamic multi-objective optimization algorithm based on multi-regional center point prediction (MCPDMO) is proposed. Firstly, the population is divided into multiple sub-regions according to the degree of environmental change, which makes the allocation of individuals more adaptable to the dynamic environment. Secondly, the center point of each sub-region is calculated respectively. The time series are established with the center point of different sub-regions at different times, and the difference model is used to predict the optimal solution set of the new environment. It improves the algorithm’s ability to respond to changes in different environments. To verify the effectiveness of the algorithm, the simulation experiments are conducted on 10 test functions with 3 dynamic multi-objective optimization algorithms. The results show that the algorithm exhibits better convergence and distribution in dynamic problems with complex Pareto front.

    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:2021-02-09
  • 最后修改日期:2021-06-29
  • 录用日期:2021-07-05
  • 在线发布日期: 2021-08-01
  • 出版日期: