Tightly-Coupled LiDAR-Inertial-Range Odometry for Reducing Trajectory Drift

The LiDAR-inertial odometry (LIO) is prone to drift over long-distance trajectories due to cumulative errors and induce tracking loss during sharp or rapid turns, reducing localization accuracy and robustness. This study proposes a two-stage LiDAR-range-inertial tightly coupled system that utilizes a static UWB anchor without prior position information and employs factor graph optimization to correct trajectory drift. In the first stage, it fuses the LiDAR and IMU data using a sliding window position estimator to compute the position of the single static anchor within the line-of-sight, based on short-term LIO outputs and UWB ranging data. In the second stage, it enhances mapping and localization accuracy through a tightly coupled LiDAR-inertial-range fusion approach leveraging the estimated position of the static anchor. The effectiveness of the proposed system is validated using the simulated UWB measurements based on the public datasets and further tested in real-world scenarios. Experimental results demonstrate better accuracy than state-of-the-art LiDAR-inertial fusion methods, proving the feasibility and effectiveness of the proposed approach.