Robot repositioning method and system based on matching of positioning information and patch features

The invention provides a robot repositioning method and system based on matching of positioning information and patch features, and the method comprises the steps: obtaining the GPS positioning information of a robot when the position of the robot is lost, carrying out the longitude and latitude conversion of the GPS positioning information, obtaining a coordinate value under a laser radar coordinate system, and determining a voxel feature matching range; based on the laser radar data, in a pre-constructed voxel map with a surface patch as a unit element, point and surface matching is carried out in a voxel feature matching range; according to the matching result, state estimation is carried out, the current pose of the robot is calculated, repositioning is rapidly carried out, the repositioning accuracy of the robot is improved, and the reliability of positioning navigation is improved. 本发明提供了一种定位信息和面片特征匹配的机器人重定位方法及系统，本发明可以在机器人丢失位置时，获取机器人GPS定位信息，并将其进行经纬度转换，得到激光雷达坐标系下的坐标值，确定体素特征匹配范围；基于激光雷达数据，在预先构建的以面片为单位元素的体素