动态未知环境下基于相对坐标系的移动机器人实时运动规划

提出了一种简单、新颖的在动态未知环境下的移动机器人运动规划方法．此方法基于相对坐标系，通过传感器信息实时调整机器人的行为来实现规划．在规划过程中，机器人有两种行为：向目标运动和避碰，且避碰行为具有优先权．机器人两种行为的切换是基于加速度空问的，首先解决的是避碰问题，而向目标运动是作为避碰的反问题来考虑的．仿真研究验证了此规划方法的有效性。

A simple and novel motion planning method for mobile robot is presented in this paper. Motion planning is achieved by adjusting the robot behavior according to information from sensors based on the relative coordinates. The robot has two behaviors in the course of planning: moving to the goal and collision avoidance, and the collision avoidance behavior has the higher priority. The switch of two behaviors is based on the acceleration space coordinates. The problem of collision avoidance is solved, and the behavior of moving to the goal is considered as a reverse problem of the avoidance. Simulation results verify the validity.

国家863计划资助项目(2001AA422140)