中国邮电高校学报(英文) ›› 2020, Vol. 27 ›› Issue (3): 42-52.doi: 10.19682/j.cnki.1005-8885.2020.0015

• Artificial Intelligence • 上一篇    下一篇

Meta-heuristic optimization inspired by proton-electron swarm

刘永利,刘燊   

  1. 河南理工大学
  • 收稿日期:2019-09-17 修回日期:2020-04-13 出版日期:2020-06-24 发布日期:2020-08-30
  • 通讯作者: 刘永利 E-mail:liuyongli@hpu.edu.cn
  • 基金资助:
    国家自然科学基金

Meta-heuristic optimization inspired by proton-electron swarm

  • Received:2019-09-17 Revised:2020-04-13 Online:2020-06-24 Published:2020-08-30
  • Supported by:
    National Natural Science Foundation of China

摘要: While solving unimodal function problems, conventional meta-heuristic algorithms often suffer from low accuracy and slow convergence. Therefore, in this paper, a novel meta-heuristic optimization algorithm, named proton-electron swarm (PES), is proposed based on physical rules. This algorithm simulates the physical phenomena of like-charges repelling each other while opposite charges attracting in protons and electrons, and establishes a mathematical model to realize the optimization process. By balancing the global exploration and local exploitation ability, this algorithm achieves high accuracy and avoids falling into local optimum when solving target problem. In order to evaluate the effectiveness of this algorithm, 23 classical benchmark functions were selected for comparative experiments. Experimental results show that, compared with the contrast algorithms, the proposed algorithm cannot only obtain higher accuracy and convergence speed in solving unimodal function problems, but also maintain strong optimization ability in solving multimodal function problems.

关键词: meta-heuristic, proton, electron swarm

Abstract: While solving unimodal function problems, conventional meta-heuristic algorithms often suffer from low accuracy and slow convergence. Therefore, in this paper, a novel meta-heuristic optimization algorithm, named proton-electron swarm (PES), is proposed based on physical rules. This algorithm simulates the physical phenomena of like-charges repelling each other while opposite charges attracting in protons and electrons, and establishes a mathematical model to realize the optimization process. By balancing the global exploration and local exploitation ability, this algorithm achieves high accuracy and avoids falling into local optimum when solving target problem. In order to evaluate the effectiveness of this algorithm, 23 classical benchmark functions were selected for comparative experiments. Experimental results show that, compared with the contrast algorithms, the proposed algorithm cannot only obtain higher accuracy and convergence speed in solving unimodal function problems, but also maintain strong optimization ability in solving multimodal function problems.

Key words: meta-heuristic, proton, electron swarm

中图分类号: