一种用于群无人艇避碰的动态分组策略

为提高水面无人艇(Umanned Surface Vehicle, USV)的避碰能力,实现群水面无人艇之间的自主避碰,针对遵守国际海上避碰规则的无人艇提出一种动态分组策略的避碰方法。在群无人艇避碰的动态分组策略中,考虑无人艇的操纵特性,分析其几何位置关系,对遵守国际海上避碰规则的运动目标进行行为预测,利用交通流的特性,实现周边多个避碰目标的动态分组,同时使用凸包扫描算法扫描障碍物边界,进而达到简化避碰态势的目的。开展多种场景下的案例仿真和对比试验,结果验证上述动态分组策略在群无人艇避碰问题中的有效性,同时为设计、实现复杂环境下的群无人艇自主避碰系统提供有益的借鉴。

Collision-Avoidance Method for a Swarm of Unmanned Surface Vehicles Using Dynamic Grouping Strategy

To enhance the collision avoidance ability of unmanned surface vehicle USV(Unmanned Surface Vehicle)and fulfill collision avoidance autonomously for a swarm of USVs, a dynamic grouping strategy is proposed, particularly for the USVs that comply with water traffic regulations. The maneuvering characteristics of the vehicles and their geometric position relations are analyzed thoroughly. The behavior of USVs facing obstacles are predicted, and the dynamic grouping are arranged based on the features of the traffic flows. Meanwhile, to simplify the collision avoidance situation, the convex hull scanning algorithm is employed to scan the obstacle boundary. The effectiveness of the dynamic grouping strategy for the USVs collision avoidance are verified through the simulations under typical scenarios.