Relative Position Estimation Method of Multi-robot Based on 2D Lidar

In the multi-robot application scenario, the robot needs to obtain the relative position of the target robot and itself in a completely unfamiliar environment. This paper proposes a method to estimate the relative position of the target robot and the observation robot by using 2D lidar combined with the velocity information of the target robot. Observation robot with a laser radar data, in turn, produces a static map, by comparing the static map and current scanning frame, it is concluded that the dynamic map, and then through the study of the clustering algorithm of dynamic point cloud information derived its unique movement characteristics, according to the goal of the robot motion information, combined with the dynamic characteristics of point cloud data of target robot motion cluster matching. In this paper, ROS platform is used to build an experimental environment and simulation is carried out, which proves that the proposed method can effectively obtain the relative position of the target robot relative to the observation robot in real time.