Scholarlink

<p>There are problems in optimizing structural parameters for switched reluctance motors (SRM), such as high computational costs, low optimization efficiency, and heavy decision-making burden. In order to solve these problems, this article offers a multi-objective optimization method for SRMs based on fuzzy inference multi-objective particle swarm optimization (FIMOPSO). The algorithm can search for the solution in the Pareto front and non-dominated sort according to the designer’s preference, which overcomes the difficulty of selecting the optimal solution. The initial step involves defining the optimization problem of SRM, followed by establishing a multi-objective optimization model. Secondly, the sensitivity analysis of the parameters involves utilizing the Pearson correlation coefficient analysis and the orthogonal table. This analysis leads to the division of parameters into two distinct subspaces. Then, a relative strength fuzzy inference system considering the designer’s preference is established, which is integrated with the MOPSO to guide the optimization direction of the algorithm. Finally, the Kriging surrogate model is used to replace the finite element analysis of the optimization space, and FIMOPSO is used to optimize the two subspaces in turn. The optimization results solved by FIMOPSO and MOPSO are compared and analyzed. From the discussion of results, the proposed method obtains more preference solutions and improves decision-making efficiency.</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

Application of Fuzzy Inference Multi-Objective Particle Swarm Optimization Combined with Kriging Model on Switched Reluctance Motor Optimization

  • 期刊论文

  • 发表日期：

    2025年3月20日

查看全文
  • Jie Gao,
  • Yuxiang Zhou,
  • Meng Xu,
  • Yanqing Mi

摘要

There are problems in optimizing structural parameters for switched reluctance motors (SRM), such as high computational costs, low optimization efficiency, and heavy decision-making burden. In order to solve these problems, this article offers a multi-objective optimization method for SRMs based on fuzzy inference multi-objective particle swarm optimization (FIMOPSO). The algorithm can search for the solution in the Pareto front and non-dominated sort according to the designer’s preference, which overcomes the difficulty of selecting the optimal solution. The initial step involves defining the optimization problem of SRM, followed by establishing a multi-objective optimization model. Secondly, the sensitivity analysis of the parameters involves utilizing the Pearson correlation coefficient analysis and the orthogonal table. This analysis leads to the division of parameters into two distinct subspaces. Then, a relative strength fuzzy inference system considering the designer’s preference is established, which is integrated with the MOPSO to guide the optimization direction of the algorithm. Finally, the Kriging surrogate model is used to replace the finite element analysis of the optimization space, and FIMOPSO is used to optimize the two subspaces in turn. The optimization results solved by FIMOPSO and MOPSO are compared and analyzed. From the discussion of results, the proposed method obtains more preference solutions and improves decision-making efficiency.