Scholarlink

<p>For Multimodal Multi-objective Optimization Problem (MMOP), they consist of multiple equivalent Pareto-optimal sets (PS) within the decision space, corresponding to the same Pareto front (PF) in the objective space. However, simultaneously locating multiple equivalent PS remains a challenge. Therefore, this paper introduces a Scoring-based Multi-Neighborhood Selection Mechanism in a Multimodal Multi-objective Particle Swarm Optimization (MMOPSO-SMNSM) to address MMOPS. The algorithm introduces the use of multiple neighborhood techniques to establish diverse candidate neighborhoods within the population. Individuals make selections of their neighborhoods based on Neighborhood Evaluation Function. Additionally, the algorithm utilizes an environment selection method based on individual neighborhood convergence quality and the crowding of the decision-objective space, maintaining population diversity to enhance algorithm performance. The algorithm was compared with 7 competing algorithms on the CEC 2020 MMOPs benchmark. Experimental results demonstrated that, compared to other algorithms, MMOPSO-SMNSM exhibited superior performance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A particle swarm optimizer with a scoring-based multi-neighborhood selection mechanism for multimodal multi-objective problems

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Zhennan Wang,
  • Lei Wang,
  • Qiaoyong Jiang,
  • Zhaoqi Wang,
  • Wenqian Zhu,
  • Liangliang Wang

摘要

For Multimodal Multi-objective Optimization Problem (MMOP), they consist of multiple equivalent Pareto-optimal sets (PS) within the decision space, corresponding to the same Pareto front (PF) in the objective space. However, simultaneously locating multiple equivalent PS remains a challenge. Therefore, this paper introduces a Scoring-based Multi-Neighborhood Selection Mechanism in a Multimodal Multi-objective Particle Swarm Optimization (MMOPSO-SMNSM) to address MMOPS. The algorithm introduces the use of multiple neighborhood techniques to establish diverse candidate neighborhoods within the population. Individuals make selections of their neighborhoods based on Neighborhood Evaluation Function. Additionally, the algorithm utilizes an environment selection method based on individual neighborhood convergence quality and the crowding of the decision-objective space, maintaining population diversity to enhance algorithm performance. The algorithm was compared with 7 competing algorithms on the CEC 2020 MMOPs benchmark. Experimental results demonstrated that, compared to other algorithms, MMOPSO-SMNSM exhibited superior performance.