Real-Time Autonomous Localization in Multi-Robot Scene

In the multi-robot scene, the rapidly changing environment makes it hard for robots to localize. ICP (Iterative Closest Point) is a robust point cloud matching algorithm that can be used for real-time autonomous localization. However, the occlusion of lidar by other robots or humans leads to the failure of traditional ICP localization methods. We propose a novel real-time autonomous localization method combining the improved ICP method with the lidar-odometry mixed localization mode. This method can realize high-precision and fast real-time autonomous localization in complex environments. We provide experimental results and comparisons to other well-known methods to show the competitiveness of the proposed method.