Robust Localization Based on LiDAR Inertial fusion for Plant Logistics

With the gradual development of LiDAR inertial fusion localization algorithm, the technology of autonomous tractor for plant logistics has become more mature. Due to the complex and changeable characteristics of the plant environment, it is difficult to provide stable localization for autonomous tractor through single map matching localization or LiDAR inertial odometry localization method. On the one hand, the plant environment is full of dynamic interference, and the long-term map will change significantly, which will lead to jump and instability in map matching and localization. On the other hand, the LiDAR inertial odometry localization will drift seriously for a long time, which is unacceptable for the further trajectory planning. Therefore, this paper considers the integration of map matching localization and LiDAR inertial odometry localization between key frames, so as to adapt to the impact of environmental changes and odometry drift and increase the robustness and accuracy of localization.