Scholarlink

Unmanned surface vehicles (USV) have gained a significant position in ocean resource development due to their cost-effectiveness and safety. Collision avoidance is the crucial technology determining whether a USV can complete resource exploration missions. This paper proposes a novel collision avoidance algorithm incorporating designed clustering algorithm and improved deep reinforcement learning (DRL) consisting of two parts. The improved collision avoidance algorithm involves collision risk assessment based on the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) Algorithm used to classify obstacle ships. Secondly, Proximal policy optimization. Finally, multi-ship collision avoidance simulation platform is established to verify the reliability of the proposed algorithm. The simulation results show the higher collision avoidance success rate than traditional algorithms.

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

Collision Avoidance for Unmanned Surface Vehicles Based on the PPO with DBSCAN Algorithm

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Hui Yu,
  • Defeng Wu,
  • Guoqiang Li,
  • Yuqin Li

摘要

Unmanned surface vehicles (USV) have gained a significant position in ocean resource development due to their cost-effectiveness and safety. Collision avoidance is the crucial technology determining whether a USV can complete resource exploration missions. This paper proposes a novel collision avoidance algorithm incorporating designed clustering algorithm and improved deep reinforcement learning (DRL) consisting of two parts. The improved collision avoidance algorithm involves collision risk assessment based on the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) Algorithm used to classify obstacle ships. Secondly, Proximal policy optimization. Finally, multi-ship collision avoidance simulation platform is established to verify the reliability of the proposed algorithm. The simulation results show the higher collision avoidance success rate than traditional algorithms.