借鉴预测控制滚动优化原理 ,研究了全局环境未知且存在动态障碍物情况下的移动机器人路径规划问题。提出的基于滚动窗口的移动机器人路径规划方法充分利用机器人实时测得的局部环境信息 ,并通过有效的场景预测 ,以滚动方式进行在线规划 ,结合了优化和反馈机制 ,具有计算量小、反应迅速的特点。大量仿真结果表明 ,该方法能很好地适应动态不确定环境。 With reference to the principle of predictive control rolling optimization, the path planning problem of mobile robot with unknown global environment and dynamic obstacles is studied. The path planning method of mobile robot based on rolling window makes full use of the local environment information measured by robots in real time and carries out on-line planning in rolling mode through effective scene prediction. Combining the optimization and feedback mechanism, it has the advantages of small calculation and quick reaction specialty. A large number of simulation results show that this method can well adapt to dynamic and uncertain environment.