Scholarlink

<p>In this work, a sensitivity analysis of the parameters for the brainstorm optimization (BSO) algorithm was carried out using mathematical benchmark functions. This algorithm is inspired by the collective human behavior in problem-solving—specifically, brainstorming is based on swarm intelligence, which is characteristic of evolutionary algorithms. Simulation results highlight which parameter of BSO has more impact on the results for this kind of problem. This information is essential for outlining methods that could enhance BSO performance. In particular, parameter adaptation would be more beneficial for the parameter with more impact. In this way, a strategy for adapting the parameter could provide improved performance. In this regard, parameter adaptation employing fuzzy logic was undertaken with idea of enhancing BSO performance in optimization applications. Simulation results with 12 benchmark functions show that fuzzy BSO outperforms the original BSO. The study was performed using 30 executions for each function and with the mean and standard deviation metrics. In particular, for 30 dimensions fuzzy BSO outperforms BSO in 10 of 12 functions. In addition, the optimization for a real control problem was considered with good results.</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

Fuzzy logic for dynamic parameter adaptation in the brainstorm optimization algorithm

  • 期刊论文

  • 发表日期：

    2026年5月22日

查看全文
  • Norma Robles Rosario,
  • Oscar Castillo,
  • Fevrier Valdez,
  • Patricia Melin

摘要

In this work, a sensitivity analysis of the parameters for the brainstorm optimization (BSO) algorithm was carried out using mathematical benchmark functions. This algorithm is inspired by the collective human behavior in problem-solving—specifically, brainstorming is based on swarm intelligence, which is characteristic of evolutionary algorithms. Simulation results highlight which parameter of BSO has more impact on the results for this kind of problem. This information is essential for outlining methods that could enhance BSO performance. In particular, parameter adaptation would be more beneficial for the parameter with more impact. In this way, a strategy for adapting the parameter could provide improved performance. In this regard, parameter adaptation employing fuzzy logic was undertaken with idea of enhancing BSO performance in optimization applications. Simulation results with 12 benchmark functions show that fuzzy BSO outperforms the original BSO. The study was performed using 30 executions for each function and with the mean and standard deviation metrics. In particular, for 30 dimensions fuzzy BSO outperforms BSO in 10 of 12 functions. In addition, the optimization for a real control problem was considered with good results.