动态环境下矢量化路径规划算法

Vectorization Path Planning in Dynamic Environment

对全方位机器人在完全未知动态环境下实时路径规划和针对移动障碍物的避障问题进行研究。提出矢量化路径描述方法,并将其与Bug算法思想相结合来解决机器人的路径规划问题。机器人的初始路径由其初始位置和目标位置生成,其运动过程是:首先沿初始路径行进,以规定间隔扫描当前环境,判断是否有障碍物阻挡当前路径,并检测障碍物的位置、移动方向和速度等信息;然后根据障碍物信息和机器人安全距离计算路径中间点,并插入中间点更新路径以实现避障。本文的机器人路径规划结果是以矢量形式进行描述及保存的,降低了对路径存储空间的需求,且按规划结果行进时只需要考虑直线移动距离和转动方向,简化了全方位机器人的控制。仿真结果说明本文方法的可行性及有效性。

This paper presented path planning and obstacle avoidance algorithm for omni-directional autonomous mobile agent in completely unknown dynamic environment with mobile obstacles. The vector ization path description and the thinking of Bug algorithm were combined to resolve the problem of ro bots＇ path planning. The initial path of the robot was generated by the start position and goal position. Its movement process was as follows. Firstly, the agent sensed the environment at intervals and determines whether there is obstacle blocking the current path while it moved along the initial path. The to cation, moving direction and moving speed etc. information of obstacle were detected. And then the agent calculated the intermediate point according to the obstacle information and safety radius. The free collision path was regenerated by inserting the intermediate point into the current path. The path planning result was described and stored in the form of vectors. When the robot moved along the planed path, it only needed to consider the straight distance and steering angle. Simulation results show the feasibility and efficiency of the proposed algorithm.