Scholarlink

<p>Metaheuristic algorithms play a crucial role in solving complex engineering optimization problems characterized by nonlinearity, multimodality, and stringent constraints. Among recent nature-inspired methods, the cheetah optimizer (CO) has shown promising performance; however, it still faces several limitations. To further improve the performance of CO, this study proposes an improved cheetah optimizer (iCO) that integrates fast random opposition-based learning (FROBL) for population initialization and variable neighborhood search (VNS) for exploitation enhancement. FROBL improves early-stage exploration by generating a well-distributed and diverse initial population, while VNS strengthens local search through systematic neighborhood transformations, thereby mitigating stagnation around local optima. The proposed iCO is evaluated on the CEC 2022 benchmark suite and several classical constrained engineering design problems, including welded beam, tension spring, pressure vessel, and cantilever beam optimization. Extensive comparative experiments against well-established algorithms demonstrate that iCO consistently achieves superior solution quality, faster convergence, and higher robustness. The results confirm that iCO significantly enhances the exploration-exploitation balance and demonstrates strong potential as an effective and reliable tool for solving complex constrained engineering optimization problems.</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

Advancing engineering design optimization through an improved cheetah optimizer

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • Vu Hong Son Pham,
  • Thuy Dung Dau,
  • Van Nam Nguyen,
  • Nghiep Trinh Nguyen Dang

摘要

Metaheuristic algorithms play a crucial role in solving complex engineering optimization problems characterized by nonlinearity, multimodality, and stringent constraints. Among recent nature-inspired methods, the cheetah optimizer (CO) has shown promising performance; however, it still faces several limitations. To further improve the performance of CO, this study proposes an improved cheetah optimizer (iCO) that integrates fast random opposition-based learning (FROBL) for population initialization and variable neighborhood search (VNS) for exploitation enhancement. FROBL improves early-stage exploration by generating a well-distributed and diverse initial population, while VNS strengthens local search through systematic neighborhood transformations, thereby mitigating stagnation around local optima. The proposed iCO is evaluated on the CEC 2022 benchmark suite and several classical constrained engineering design problems, including welded beam, tension spring, pressure vessel, and cantilever beam optimization. Extensive comparative experiments against well-established algorithms demonstrate that iCO consistently achieves superior solution quality, faster convergence, and higher robustness. The results confirm that iCO significantly enhances the exploration-exploitation balance and demonstrates strong potential as an effective and reliable tool for solving complex constrained engineering optimization problems.