Scholarlink

<p>To address the limitations of the Salp Swarm Algorithm (SSA), such as low precision and susceptibility to local optima, this paper proposes an Improved Salp Swarm Algorithm based on Multi-strategy Fusion (ISSA). The ISSA incorporates three key enhancements: a good point set initialization for uniform population distribution, a golden section chaotic search to bolster global exploration efficiency, and a dynamic double-region optimization strategy that balances rapid convergence with the ability to escape local optima via distinct guide and learning mechanisms. Performance evaluations using CEC2020 and CEC2022 benchmarks, alongside six engineering problems, demonstrate that ISSA significantly outperforms mainstream optimization algorithms in both convergence speed and accuracy. Furthermore, application to 2D and 3D path planning tasks confirms the algorithm’s robustness; notably, in 3D environments, ISSA achieved a path length 7.9% shorter than the standard SSA with superior stability and shorter execution times. These results validate ISSA as an efficient and practicable tool for complex real-world optimization and autonomous navigation scenarios.</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

Improved salp swarm algorithm based on multi-strategy fusion for engineering design problems and path planning

  • 期刊论文

  • 发表日期：

    2026年7月14日

查看全文
  • Fengtao Wei,
  • Yue Feng,
  • Tao Zhao,
  • Jianwei Zhao,
  • Xin Shi

摘要

To address the limitations of the Salp Swarm Algorithm (SSA), such as low precision and susceptibility to local optima, this paper proposes an Improved Salp Swarm Algorithm based on Multi-strategy Fusion (ISSA). The ISSA incorporates three key enhancements: a good point set initialization for uniform population distribution, a golden section chaotic search to bolster global exploration efficiency, and a dynamic double-region optimization strategy that balances rapid convergence with the ability to escape local optima via distinct guide and learning mechanisms. Performance evaluations using CEC2020 and CEC2022 benchmarks, alongside six engineering problems, demonstrate that ISSA significantly outperforms mainstream optimization algorithms in both convergence speed and accuracy. Furthermore, application to 2D and 3D path planning tasks confirms the algorithm’s robustness; notably, in 3D environments, ISSA achieved a path length 7.9% shorter than the standard SSA with superior stability and shorter execution times. These results validate ISSA as an efficient and practicable tool for complex real-world optimization and autonomous navigation scenarios.