Scholarlink

<p>With the development of marine science and technology, the research of unmanned surface vehicles (USVs) has received extensive attentions. Designing effective collision avoidance algorithms in complex and crowded marine environments is necessary for USV navigation. In order to improve the autonomous navigation capability of USVs, a COLREGS-based hybrid path planning algorithm for multiple USVs is proposed. The algorithm integrates an improved A-Star algorithm and an improved Dynamic Window Approach (DWA) algorithm. The improved A-Star algorithm introduces a dynamic path parameter to optimize the cost function to improve the search efficiency, and optimizes the path smoothness to remove redundant nodes. The improved DWA algorithm optimizes the evaluation function, introduces the COLREGS rule to prune the DWA speed search space, proposes a local collision avoidance strategy for multiple USVs, and uses the Deep Q-learning Network (DQN) algorithm to train the DWA objective function weights in order to improve the adaptability to the complex marine environment. In order to verify the effectiveness of proposed path planning algorithm, a large number of simulations are carried out. Extensive simulation results show that the algorithm can adequately handle the complex and congested marine environment, and multiple USVs can simultaneously travel along their respective planned paths to avoid collisions. In addition, the proposed method has good adaptability to various USVs encountered in unknown complex environments.</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

Colregs-compliant collision avoidance for multiple USVs in congested sea area

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Meiyan Zhang,
  • Wenyu Cai,
  • Hengmei Wang,
  • Qingpeng Kong

摘要

With the development of marine science and technology, the research of unmanned surface vehicles (USVs) has received extensive attentions. Designing effective collision avoidance algorithms in complex and crowded marine environments is necessary for USV navigation. In order to improve the autonomous navigation capability of USVs, a COLREGS-based hybrid path planning algorithm for multiple USVs is proposed. The algorithm integrates an improved A-Star algorithm and an improved Dynamic Window Approach (DWA) algorithm. The improved A-Star algorithm introduces a dynamic path parameter to optimize the cost function to improve the search efficiency, and optimizes the path smoothness to remove redundant nodes. The improved DWA algorithm optimizes the evaluation function, introduces the COLREGS rule to prune the DWA speed search space, proposes a local collision avoidance strategy for multiple USVs, and uses the Deep Q-learning Network (DQN) algorithm to train the DWA objective function weights in order to improve the adaptability to the complex marine environment. In order to verify the effectiveness of proposed path planning algorithm, a large number of simulations are carried out. Extensive simulation results show that the algorithm can adequately handle the complex and congested marine environment, and multiple USVs can simultaneously travel along their respective planned paths to avoid collisions. In addition, the proposed method has good adaptability to various USVs encountered in unknown complex environments.