An IMU Aided Robust LiDAR SLAM Method for UAVs in Pipeline Environment

A novel SLAM (simultaneous localization and mapping) method is proposed for 2D LiDAR used in pipeline environment. The traditional method is based on 2.5D assumption, which means that all the walls are vertical. The proposed method breaks the assumption by introducing an IMU (inertial measurement unit). The step changes of the non-flat walls can be detected, avoiding the mismatches and the consequent pose errors. The SLAM algorithm is designed, which consists of five procedures: state prediction, scan matching, environment change detection, filter update and mapping. A complex indoor test environment is built and flight experiments were carried out. The results show that the proposed method performs better than the traditional 2D LiDAR SLAM, in the respect of pose estimation and map construction.