基于椭圆碰撞锥的无人艇动态避障方法

Maritime autonomous obstacle avoidance in a dynamic environment based on collision cone of ellipse

为了提高无人艇在海洋动态避障中的效率,提出一种基于椭圆聚类-碰撞锥推演的可预测型无人艇海洋动态避障控制方法。在现有的路径规划算法,大多数对障碍物进行圆形建模,进一步采取规避措施。依据无人艇在海洋环境下实际的碰撞问题,提出对动态船只进行椭圆聚类,进一步提高无人艇的避障效率。提出的避障算法结合了《国际海上避碰规则》(COLREGS),并在避障结束后回归到初始规划轨迹上,最后,对提出的动态避障算法进行了仿真分析和实验测试。

In order to improve the efficiency of obstacle avoidance of unmanned surface vehicles （USV）, this paper presents an avoidance algorithm for USV to navigate safely in dynamic, cluttered environments. Most of the existing path planning algorithms take avoidance measures by making circular modeling to obstacles. Based on the actual collision issue of USV in the marine environment, this work presents an elliptical clustering method of dynamic vessels to improve the obstacle avoidance efficiency. Furthermore, the proposed algorithm also applies the International Regulations for Preventing Collisions at Sea （ known as international regulatitns for preventing collisions at sea （COLREGS） , for collision regulations） , and USV returns the specified routes after the obstacle avoidance. Finally, simulation analysis and experimental tests of avoidance algorithm are involved.