Scholarlink

<p>Three-dimensional UAV path planning necessitates strong global search capabilities due to its high-dimensional optimization nature. To address premature convergence and enhance local search efficiency in comparison to traditional DBO methods, this study proposes MSDBO, a multi-strategy fusion algorithm. The approach involves utilizing piecewise chaotic mapping to increase population diversity, integrating OOA for improved global exploration, and implementing a dynamic balance mechanism comprising Sigmoid convergence factors, adaptive t-distribution mutation, and dynamic weights. Additionally, simulated annealing is enhanced to achieve better convergence precision. Through systematic validation encompassing 21 benchmark functions, Wilcoxon tests, and CEC2021, MSDBO exhibits superior convergence accuracy and robustness when compared to seven other metaheuristic algorithms. Urban flight experiments further demonstrate MSDBO’s ability to generate smoother paths with a 7.5% lower optimal cost and a 31% reduced standard deviation than DBO. These findings confirm the efficacy of MSDBO in tackling UAV path planning challenges in complex scenarios through coordinated multi-stage optimization.</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 hybrid adaptive dung beetle optimization for UAV photogrammetric 3D path planning under complex constraints

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月29日

查看全文
  • Muyang Wu,
  • Guo Li,
  • Jinwei Liao,
  • Hang Wang,
  • Wen Liu,
  • Xiang Yan,
  • Mengdie Yang,
  • Shun Li

摘要

Three-dimensional UAV path planning necessitates strong global search capabilities due to its high-dimensional optimization nature. To address premature convergence and enhance local search efficiency in comparison to traditional DBO methods, this study proposes MSDBO, a multi-strategy fusion algorithm. The approach involves utilizing piecewise chaotic mapping to increase population diversity, integrating OOA for improved global exploration, and implementing a dynamic balance mechanism comprising Sigmoid convergence factors, adaptive t-distribution mutation, and dynamic weights. Additionally, simulated annealing is enhanced to achieve better convergence precision. Through systematic validation encompassing 21 benchmark functions, Wilcoxon tests, and CEC2021, MSDBO exhibits superior convergence accuracy and robustness when compared to seven other metaheuristic algorithms. Urban flight experiments further demonstrate MSDBO’s ability to generate smoother paths with a 7.5% lower optimal cost and a 31% reduced standard deviation than DBO. These findings confirm the efficacy of MSDBO in tackling UAV path planning challenges in complex scenarios through coordinated multi-stage optimization.