Scholarlink

<p>The Snake Optimizer (SO) is a widely utilized metaheuristic algorithm that has demonstrated considerable success across various practical applications. Nevertheless, SO exhibits slow convergence when applied to nonconvex and high-dimensional complex problems, often resulting in entrapment within local optima. To address these limitations, an enhanced variant of SO, termed NSO, is introduced. NSO incorporates a two-population competition strategy to augment population diversity, four adaptive inertia weight factors to enhance adaptability, and a Lévy flight perturbation strategy to bolster global search capabilities. Additionally, a novel development strategy is proposed to prevent NSO from succumbing to local optima. To assess the efficacy of NSO, it was applied to a comprehensive suite of 41 global optimization problems, 12 multi-constraint engineering design problems, and Unmanned Aerial Vehicle (UAV) path planning scenarios. The experimental results show that for the 29 optimization problems in CEC2017, NSO obtains 5, 12, 15, and 17 best results in 4 different dimensions, respectively; for the 12 optimization problems in CEC2022, NSO obtains 4 and 6 best results in 2 different dimensions, respectively. NSO also obtains the best paths in the UAV path planning problem as compared to the competing algorithms.</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

Multi-strategy collaborative improved Snake Optimizer for complex optimization problems and 3D UAV path planning

  • 期刊论文

  • 发表日期：

    2025年10月19日

查看全文
  • Heng Wang,
  • Kai Yang,
  • Jiadui Chen,
  • Haisong Huang,
  • Jingwei Yang

摘要

The Snake Optimizer (SO) is a widely utilized metaheuristic algorithm that has demonstrated considerable success across various practical applications. Nevertheless, SO exhibits slow convergence when applied to nonconvex and high-dimensional complex problems, often resulting in entrapment within local optima. To address these limitations, an enhanced variant of SO, termed NSO, is introduced. NSO incorporates a two-population competition strategy to augment population diversity, four adaptive inertia weight factors to enhance adaptability, and a Lévy flight perturbation strategy to bolster global search capabilities. Additionally, a novel development strategy is proposed to prevent NSO from succumbing to local optima. To assess the efficacy of NSO, it was applied to a comprehensive suite of 41 global optimization problems, 12 multi-constraint engineering design problems, and Unmanned Aerial Vehicle (UAV) path planning scenarios. The experimental results show that for the 29 optimization problems in CEC2017, NSO obtains 5, 12, 15, and 17 best results in 4 different dimensions, respectively; for the 12 optimization problems in CEC2022, NSO obtains 4 and 6 best results in 2 different dimensions, respectively. NSO also obtains the best paths in the UAV path planning problem as compared to the competing algorithms.