| 数据驱动选择策略的多目标差分进化算法 |
| |
| 引用本文： | 侯莹,吴毅琳,白星,韩红桂.数据驱动选择策略的多目标差分进化算法[J].控制与决策,2023,38(7):1816-1824. |
| |
| 作者姓名： | 侯莹 吴毅琳 白星 韩红桂 |
| |
| 作者单位： | 北京工业大学 信息学部,北京 100124;北京工业大学 教育部数字社区工程研究中心, 北京 100124;北京工业大学 信息学部,北京 100124;北京工业大学 计算智能与智能系统北京市重点实验室,北京 100124;北京工业大学 信息学部,北京 100124;北京工业大学 教育部数字社区工程研究中心, 北京 100124;北京工业大学 计算智能与智能系统北京市重点实验室,北京 100124 |
| |
| 基金项目： | 国家自然科学基金青年项目(61903010)；国家自然科学基金杰出青年基金项目(62125301)；国家重点研发计划项目(2018YFC1900800)；北京高校卓越青年科学家项目(BJJWZYJH01201910005020)；国家自然科学基金重大项目(61890931)；国家自然科学基金创新研究群体项目(62021003). |
| |
| 摘 要： | 针对多目标差分进化算法求解复杂多目标优化问题时,最优解选择策略中非支配排序计算复杂度高的问题,提出一种数据驱动选择策略的多目标差分进化(MODE-DDSS)算法.首先,设计多目标差分进化算法的优化解排序等级评估准则,建立基于评估准则的优化解排序等级评估库;其次,设计基于优化解双向搜索机制和无重复比较机制的数据驱动选择策略,实现优化解的高效搜索和快速排序;最后,构建数据驱动选择策略的多目标差分进化算法,降低算法在最优解选择操作中的时间复杂度,提高算法的寻优效率.实验结果表明,所提出的MODE-DDSS算法能够有效减少最优解在选择过程中的比较次数,提升多目标差分进化算法解决复杂多目标优化问题的寻优效率.
|
| 关 键 词： | 数据驱动 选择策略 非支配排序 多目标优化 差分进化算法 寻优效率 |
Multi-objective differential evolution algorithm with data-driven selection strategy |
| |
| HOU Ying,WU Yi-lin,BAI Xing,HAN Hong-gui.Multi-objective differential evolution algorithm with data-driven selection strategy[J].Control and Decision,2023,38(7):1816-1824. |
| |
| Authors: | HOU Ying WU Yi-lin BAI Xing HAN Hong-gui |
| |
| Affiliation: | Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Engineering Research Center of Digital Community of Ministry of Education,Beijing University of Technology,Beijing 100124,China;Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Beijing Key Laboratory of Computational Intelligence and Intelligent System,Beijing University of Technology,Beijing 100124,China; Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;Engineering Research Center of Digital Community of Ministry of Education,Beijing University of Technology,Beijing 100124,China;Beijing Key Laboratory of Computational Intelligence and Intelligent System,Beijing University of Technology,Beijing 100124,China |
| |
| Abstract: | The multi-objective differential evolution(MODE) algorithm has high computational complexity of the selection strategy in solving complex multi-objective optimization problems. To address this issue, a multi-objective differential evolution with data-driven selection strategy(MODE-DDSS) is proposed. First, the ranking evaluation criteria of optimization solutions is designed, and the ranking evaluation database of optimization solutions based on evaluation criteria is established. Then, a data-driven selection strategy, based on a two-way search mechanism and a non-repeated comparison mechanism, is designed to search and compare the optimal solutions efficiently, and select the optimal solutions. Finally, a multi-objective differential evolution algorithm with the data-driven selection strategy is constructed, which reduces the complexity of optimal solution selection operation and improves the optimization efficiency of the algorithm. Experimental results show that the proposed MODE-DDSS algorithm can effectively reduce the number of comparison operations in the selection strategy, and improve the efficiency of the multi-objective differential evolution algorithm in solving complex multi-objective optimization problems. |
| |
| Keywords: | |
|
| 点击此处可从《控制与决策》浏览原始摘要信息 |
|
点击此处可从《控制与决策》下载全文 |
|
