Abstract: The paper presents a new approach to robot path planning under dynamic environment.The whole system includes two parts，the global path planning and the local planning for obstacle avoidance.In the global path planning，an optimal route to the goal is founded by RRT algorithm；in the local planning for obstacle avoidance，while following the global path，several collision-free strategies for different situations are used after the environment detection and collision prediction based on rolling windows in order that the robot reaches the goal safety.The results of the simulation experiment indicate that the proposed method is feasible.

Key words: robot path planning, dynamic obstacle avoidance, rapidly-exploring random tree algorithm, global path planning, local planning for obstacle avoidance, grid method

摘要： 针对机器人动态路径规划问题，提出了在动态环境中移动机器人的一种路径规划方法，适用于环境中同时存在已知和未知，静止和运动障碍物的复杂情况。采用栅格法建立机器人空间模型，整个系统由全局路径规划和局部避碰规划两部分组成。在全局路径规划中，用快速搜索随机树算法规划出初步全局优化路径，局部避碰规划是在全局优化路径的同时，通过基于滚动窗口的环境探测和碰撞规则，对动态障碍物实施有效的局部避碰策略，从而使机器人安全顺利地到达目的地。仿真实验结果说明该方法具有可行性。

关键词: 机器人路径规划, 动态避障, 快速搜索随机树算法, 全局路径规划, 局部避碰规划, 栅格法