返回
A novel multi-ship collaborative collision avoidance method based on key risk ship identification
DOI:10.1016/j.oceaneng.2026.128356.png)
摘要
En 中文
• 开发了一种多层图注意力网络,用于识别演化中的海上交通网络中的关键风险船舶。
• 提出了一种集成模拟退火和外部档案策略的IMOSSA,用于多船协同避碰。
• 一种两阶段框架使用三阶段筛选,将IMOSSA聚焦于GAT识别的关键风险船舶,减少优化集合。
• 该框架产生符合COLREGS的角色分配,并在所有测试场景中保持安全的船舶间距。
Keyword:
Graph attention network
Multi-ship collaborative collision avoidance
Multi-objective optimization
Ship collision risk
期刊
IF:
5.5
论文数:
7.7K
被引数:
7.6W
机构
引用论文
Identification of Complex Multi-Vessel Encounter Scenarios and Collision Avoidance Decision Modeling for MASSs复杂多船交汇场景的识别与岸基自动船舶操作系统的避碰决策建模
A novel method for identifying key focus ships in a complex network based on ship collision risks一种基于船舶碰撞风险的复杂网络关键焦点船舶识别新方法
Incorporating graph theory and time series analysis for fine-grained traffic flow prediction in port areasYu, H.; Cui, X.; Bai, X.; Chen, C.; Xu, L. 将图论和时间序列分析应用于港区精细交通流量预测。Ocean. Eng. 2025, 335, 121693. [Google Scholar] [CrossRef]
Ship behavior pattern recognition method based on hybrid graph neural networksMa, L.; Cao, H.; Shi, G.Y. 基于混合图神经网络的船舶行为模式识别方法. Front. Mar. Sci. 2025, 12, 1605216. [Google Scholar] [CrossRef]

