基于LiDAR/INS的野外移动机器人组合导航方法

移动机器人在地形复杂等野外环境跨区域运动时，机器人运动特性和环境特征变化更为明显，由此引起的点云畸变和特征点稀疏等问题尤为突出，有必要结合传感器标定误差、车轮打滑和车体颠簸等因素进一步改进机器人的位姿估计精度。本文对基于LiDAR/INS的移动机器人环境建模和自主导航方法展开研究，针对LeGO-LOAM等在处理车体姿态快速变化时的性能退化问题，提出一种适用于野外移动机器人运动特性的点云特征分析和多传感融合方法，利用IMU的预积分与LiDAR的scan-to-map构成优化函数，进而迭代更新机器人的位姿。野外环境实验结果表明，当机器人以较高速度做转弯运动或在短时间内多次转向时，本文所提方法仍可以为优化提供良好的初值估计，相比LeGO-LOAM等方法具有更高的位姿估计精度。

Integrated navigation approach for the field mobile robot based on LiDAR/INS

When the mobile robot moves in the large scale field with complex terrain, the motion and environmental characteristics changes dramatically, and the problems of motion distortion in point clouds and the sparse of feature points become prominently. Hence, it is necessary to improve the estimation accuracy of position and states from aspects of the calibration of sensors’ error, wheel-slip and bumpy terrain. The environmental modeling and autonomous navigation of the field mobile robot based on LiDAR/INS (Inertial Navigation System) are studied in this paper. Aiming at the degradation problem of LeGO-LOAM when the vehicle state changes dramatically, the strategy of point cloud feature extraction and sensor fusion are proposed to adapt the motion characteristics of the field mobile robot, the optimization function is composed of IMU pre-integration and scan-to-map of LiDAR, then the state of robot is updated. Finally, the field experiments are implemented when the robot makes a fast turning or steering in limited time, the proposed method are validated to provide more accurate estimation of initial values and states compared to LeGO-LOAM related approaches.