基于改进Voronoi图的移动机器人在线路径规划

Online path planning of mobile robots based on improved Voronoi diagram

针对移动机器人在部分环境信息已知下的路径规划问题,运用Voronoi图理论及动态路径最优算法(D*算法)理论,研究了一种基于传感器信息的移动机器人在线路径规划的方法.该方法利用现有的已知环境信息离线生成路图,并根据起点与终点的位置规划出一条无碰撞的全局最优路径,然后移动机器人沿着最优路径前进,安装在机器人上的传感器不断地探测环境新信息以在线完成路图的重构及路径的重规划,实时搜索一条全局最优路径.最后,通过在自制的小车平台上的实验证明方法的可行性.

Pertaining to partially-sensored environment information,an on-line path planning is proposed for mobile robots using the Voronoi diagram and dynamic D＊ algorithm.Accordingly,the road map is configured based on the acquired environment information.In this manner,a globally-optimal non-collision path is obtained between the origin and terminal.Afterwards,the rod map,together with the path,is reconfigured in terms of the real-time environment information updated.Eventually,this approach is proven feasible by the means of a robot car testing platform.