Scholarlink

<p>Meta-heuristic algorithms (MAs) have gained significant attention as important tools for solving complex optimization problems due to their self-organization, robustness, and parallelism. However, existing algorithms generally suffer from common limitations including insufficient convergence accuracy and susceptibility to local optima. To address these issues, this paper proposes a novel hybrid algorithm named CGGWO, which integrates the global search mechanism of the chaos game optimization (CGO) with the social hierarchy structure of the grey wolf optimizer (GWO) to form a collaborative optimization framework. The main contributions of this study are as follows: (1) A deep hybrid strategy of CGO and GWO is proposed, enhancing population initialization diversity through chaotic mapping and balancing exploration-exploitation capabilities via dynamic weights; (2) Comparative experiments with 19 state-of-the-art methods on CEC2017 benchmark functions demonstrate the superiority of CGGWO at dimensions of 10, 30, 50, and 100. Statistical results from Wilcoxon signed-rank test and Friedman test confirm significant advantages (<i>p</i> &lt; 0.05) in most problems; (3) Comparative experiments with 3 champion methods on CEC2017 benchmark functions demonstrate the superiority of CGGWO at dimensions of 10, 30, 50, and 100; (4) The algorithm achieves superior results when applied to five classical engineering optimization problems (e.g., pressure vessel design, cantilever beam design); 4) Successful resolution of CEC2011 real-world optimization problems further validates its practical value. A The CGGWO achieves convergence accuracy that is 1.2 to 22 times higher than that of the standard GWO, along with significantly improved stability. The outstanding performance of CGGWO across benchmark tests, engineering applications, and real-world problems demonstrates its effectiveness as a new solution for complex optimization challenges.</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

Chaotic games driven grey Wolf optimization: optimal performance design and comprehensive analysis

  • 期刊论文

  • 发表日期：

    2026年1月6日

查看全文
  • Chenhua Tang,
  • Changcheng Huang,
  • Yi Chen,
  • Ali Asghar Heidari,
  • Huiling Chen,
  • Guoxi Liang

摘要

Meta-heuristic algorithms (MAs) have gained significant attention as important tools for solving complex optimization problems due to their self-organization, robustness, and parallelism. However, existing algorithms generally suffer from common limitations including insufficient convergence accuracy and susceptibility to local optima. To address these issues, this paper proposes a novel hybrid algorithm named CGGWO, which integrates the global search mechanism of the chaos game optimization (CGO) with the social hierarchy structure of the grey wolf optimizer (GWO) to form a collaborative optimization framework. The main contributions of this study are as follows: (1) A deep hybrid strategy of CGO and GWO is proposed, enhancing population initialization diversity through chaotic mapping and balancing exploration-exploitation capabilities via dynamic weights; (2) Comparative experiments with 19 state-of-the-art methods on CEC2017 benchmark functions demonstrate the superiority of CGGWO at dimensions of 10, 30, 50, and 100. Statistical results from Wilcoxon signed-rank test and Friedman test confirm significant advantages (p < 0.05) in most problems; (3) Comparative experiments with 3 champion methods on CEC2017 benchmark functions demonstrate the superiority of CGGWO at dimensions of 10, 30, 50, and 100; (4) The algorithm achieves superior results when applied to five classical engineering optimization problems (e.g., pressure vessel design, cantilever beam design); 4) Successful resolution of CEC2011 real-world optimization problems further validates its practical value. A The CGGWO achieves convergence accuracy that is 1.2 to 22 times higher than that of the standard GWO, along with significantly improved stability. The outstanding performance of CGGWO across benchmark tests, engineering applications, and real-world problems demonstrates its effectiveness as a new solution for complex optimization challenges.