LiDAR-IMU-UWB-Based Collaborative Localization

Abstract

This article introduced a positioning system composed of different sensors, such as LiDAR, IMU, and ultra-wideband (UWB), for the positioning method in autonomous driving technology under closed coal mine tunnels. First, we processed the LiDAR data, extracted its feature points and merged the extracted feature point clouds to generate a skewed combined feature point cloud. Then, we used the skew combined feature point clouds for feature matching, performed pre-integration processing on the IMU sensor data, and completed the LiDAR-IMU odometer with the LiDAR. Finally, we added UWB data to IMU pose node as a one-dimensional over-edge constraint. By updating the sliding window, the positioning accuracy was further improved. Moreover, we have conducted experiments to verify the proposed positioning system in a simulated roadway. The experimental results showed that the method proposed in this paper is superior to the single LiDAR method and the single UWB method in terms of positioning accuracy.

Keywords

• feature extraction
• feature matching
• pose constrain
• localization