面向无人艇集群避障的速度障碍算法研究

Research on velocity obstacle algorithm for unmanned surface vehicle swarm avoidance

无人艇集群在海上公共区域航行需要考虑互相避障以及水面其他障碍的规避问题,故提出一种无人艇集群避障算法解决该问题。算法改进了速度障碍法,考虑障碍物位置和速度以估计未来时间窗口下可能发生碰撞的锥形区域,并规划出可行域半平面。之后引入海事避碰规则进一步约束无人艇的避碰方向和可行域范围。多障碍场景下,通过在若干个可行域的交集上寻找最优的避碰速度矢量,可引导无人艇避障。算法中各无人艇相互独立,不存在中心节点。进行了仿真验证,将数十条无人艇划分为多个集群进行相互避障测试。仿真结果表明,该方法能够有效地避免无人艇集群之间的碰撞,具有较好的鲁棒性和实时性。

In this paper,an Unmanned Surface Vehicle(USV)cluster avoidance algorithm is proposed to address the problem of mutual avoidance and obstacle avoidance in the public area of the sea.The algorithm improves the speed obstacle method by taking into account the position and speed of obstacles to estimate the cone region where a collision may occur in the future time window,and plans out the feasible domain half-plane.Moreover,the International Regulations for Preventing Collisions at Sea(COLREGs)are introduced to constrain the avoidance direction and feasible domain range of USVs.In the case of multiple obstacles,by finding the optimal avoidance speed vector on the intersection of several feasible domains,it can guide USVs to avoid obstacles.This algorithm is independent for each USV,without a central node.The effectiveness of the proposed algorithm is verified by simulation,which divides dozens of USVs into multiple clusters for mutual avoidance tests.The simulation results demonstrate that this method can effectively prevent collisions between UsV clusters,and has good robustness and real-time performance.