为了保证无人船安全行驶，设计动态人工势场法的无人船避障路径规划方法。该方法的全局路径规划部分依据航行环境信息构建栅格地图后，采用A*算法规划无人船的全局路径，并获取无人船的全局路径节点；局部避障路径规划部分将该路径节点作为局部路径的开始节点，利用动态人工势场法规划无人船避障路径。为保证路径规划效果，采用长短记忆循环神经网络和强化学习算法相结合的方式改进动态人工势场法，确保在障碍物运动速度较快时，依旧能够可靠完成无人船避障路径规划。测试结果显示：该方法可完成无人船全局最优路径规划，并有效完成局部避障路径规划，无人船和障碍物的会遇距离和会遇时间分别大于4.26 m和大于15 s，碰撞危险度极低，满足应用标准。
The obstacle avoidance path planning method of unmanned ship based on dynamic artificial potential field method is studied to ensure the safe operation of unmanned ship. The global path planning part of the method uses A * algorithm to plan the global path of the unmanned ship and obtain the global path nodes of the unmanned ship after building a grid map based on the navigation environment information. The local obstacle avoidance path planning part takes this path node as the starting node of the local path, and uses the dynamic artificial potential field method to plan the obstacle avoidance path of the unmanned ship. In order to ensure the effect of the path planning, the dynamic artificial potential field method is improved by combining the long and short memory recurrent neural network and the reinforcement learning algorithm, so as to ensure that the obstacle avoidance path planning of the unmanned ship can still be completed reliably when the obstacle movement speed is fast. The test results show that this method can obtain the global optimal path planning results of unmanned ships, and effectively complete the local obstacle avoidance path planning. The distance and time of encounter of unmanned ships and obstacles are greater than 4.26 m and greater than 15 s, respectively. The collision risk is extremely low, which meets the application standards.
2023,45(5): 89-92 收稿日期：2022-11-04
DOI：10.3404/j.issn.1672-7649.2023.05.016
分类号：TP242
基金项目：江苏省高等学校自然科学研究面上项目（20KJD580006）
作者简介：刘涛(1988-),男,硕士,讲师,研究方向为海上智能交通、航海仿真及船舶导航技术
