船舶自主避碰的慎思型轨迹规划

Deliberative Trajectory Planning for Shipborne Autonomous Collision Avoidance System

基于快速扩展随机树(RRT)算法,提出一种适用于船舶自主避碰的慎思型轨迹规划(DTP)算法,能够耦合处理静态障碍物约束、船舶操纵性约束、轨迹最优性等限制条件,在两个长程路径点之间完成全局轨迹规划,以保证规划轨迹的可行性、完备性和最优性.通过一条案例三体船的自主避碰仿真及航行试验,从多方面验证了所提DTP算法的有效性、优越性和稳定性,对研究船舶自主避碰系统及其应用前景具有重要的意义.

Based on rapidly-exploring random tree(RRT) algorithm, a deliberative trajectory planner DTP was developed to search for global trajectory between the adjacent waypoints of the long voyage. The static obstacle constraints, ship maneuvering constraints, and trajectory optimality requirements are satisfied simultaneously to ensure the feasibility, completeness and optimality in planning. Through simulations and on-water tests of a trimaran model, the proposed DTP method was demonstrated to be effective, superior and stable for the autonomous collision avoidance. The present work provides an important basis for subsequent research and future application.