Magnetometer IMU-based Trolley Mapping and Navigation Optimization on Complex Road Surfaces

AHRS-LeGO-LOAM: Based on the LeGO-LOAM algorithm (a lightweight ground-optimized Lidar ranging and mapping method), we have added a new measurement module IMU (AHRS Attitude and Heading Reference System) with magnetometer to detect data deviations caused by asymmetry in the body pose when the robot moves to a non-planar or sunken road section. LeGO-LOAM, a system for rapid mapping of automobiles to determine motion trajectory and pose estimation, can be used to collect point cloud data at nine degrees of freedom after adding a magnetometer. We started by using a point cloud segmentation algorithm to filter out the noise, and then use feature extraction to obtain distinct planar and edge features. We compare the performance of AHRS-LeGO-LOAM with the current LeGO-LOAM, using a data set collected from the variable terrain environment of ground trolleys, and the results show that AHRS-LeGO-LOAM has a higher accuracy rate on the same dataset. We also integrated AHRS-LeGO-LOAM into the SLAM framework to eliminate drift-induced pose estimation errors, and tested them using the KITTI data set.