Scholarlink

<p>This study proposes an enhanced crested porcupine optimizer (CPO) framework by integrating a fuzzy logic-based intelligent selection mechanism to improve optimization performance and robustness. The core innovation lies in dynamically assigning global exploration or local exploitation capabilities to individuals during evolution, guided by real-time assessments of system states—such as the iterative process and population diversity metrics. Methodologically, a selection factor (SF) is introduced to balance exploration–exploitation trade-offs, while type 1 and interval type 2 fuzzy logic systems are designed to dynamically adjust key parameters (SF, trade-off factor, and convergence factor) based on fuzzy rules. The proposed fuzzy-CPO algorithms have been rigorously evaluated using the CEC2017 benchmark suite and a UAV 3D path planning problem. Experimental results demonstrated that type 1 and interval type 2 fuzzy-CPO variants significantly outperform the traditional CPO and other swarm intelligence algorithms. Type 1 fuzzy-CPO exhibits superior time efficiency, making it ideal for real-time applications, whereas interval type 2 fuzzy-CPO excels in handling problems with intricate uncertainties due to its enhanced robustness. These findings highlight the efficacy of the fuzzy-driven adaptive mechanism in strengthening optimization capabilities. This study concludes that the choice between the two variants depends on application-specific requirements: Type 1 fuzzy-CPO for high-speed scenarios and interval type 2 fuzzy-CPO for complex, uncertain environments, thereby broadening the practical applicability of the CPO framework.</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

Intelligent Selection Mechanism for Enhancing CPO Algorithm Robustness Based on Fuzzy Logic System

  • 期刊论文

  • 发表日期：

    2026年1月8日

查看全文
  • Zuqiang Long,
  • Zhiyong Hu,
  • Jiaying Gu,
  • Ke Sun

摘要

This study proposes an enhanced crested porcupine optimizer (CPO) framework by integrating a fuzzy logic-based intelligent selection mechanism to improve optimization performance and robustness. The core innovation lies in dynamically assigning global exploration or local exploitation capabilities to individuals during evolution, guided by real-time assessments of system states—such as the iterative process and population diversity metrics. Methodologically, a selection factor (SF) is introduced to balance exploration–exploitation trade-offs, while type 1 and interval type 2 fuzzy logic systems are designed to dynamically adjust key parameters (SF, trade-off factor, and convergence factor) based on fuzzy rules. The proposed fuzzy-CPO algorithms have been rigorously evaluated using the CEC2017 benchmark suite and a UAV 3D path planning problem. Experimental results demonstrated that type 1 and interval type 2 fuzzy-CPO variants significantly outperform the traditional CPO and other swarm intelligence algorithms. Type 1 fuzzy-CPO exhibits superior time efficiency, making it ideal for real-time applications, whereas interval type 2 fuzzy-CPO excels in handling problems with intricate uncertainties due to its enhanced robustness. These findings highlight the efficacy of the fuzzy-driven adaptive mechanism in strengthening optimization capabilities. This study concludes that the choice between the two variants depends on application-specific requirements: Type 1 fuzzy-CPO for high-speed scenarios and interval type 2 fuzzy-CPO for complex, uncertain environments, thereby broadening the practical applicability of the CPO framework.