合作环境下基于惯性/里程计/Lidar融合的UGV导航方法

近年来，无人车在巡检、探测等方面的应用愈发广泛,且应用环境愈发复杂。在这些应用中，无人车必须对自身的位姿进行准确估计，以确保作业安全、高效完成。其中，可在复杂环境下适用的自主导航能力是核心关键技术。提出了一种基于惯性/里程计/激光雷达的地面无人车导航方法，区别于传统的激光雷达SLAM方法，该方法根据已知的几何结构特征进行定位，避免了因有效点数量稀少而导致的匹配误差。同时对惯性/里程计/激光雷达的融合算法进行了研究，提高了自主导航系统的鲁棒性和准确性。最后，在Gazebo中搭建了相应的仿真环境，并进行了算法验证。仿真结果表明，该方法能够实现无人车在巡检过程中实时可靠的自主导航，具有较好的工程应用价值。

A UGV Navigation Method Based on Inertia/Odometer/Lidar Fusion in Cooperative Environment

In recent years, the application of unmanned ground vehicle in patrol, detection and other fields has become more and more extensive and the application environment has become more complex. In these applications, it is necessary for unmanned vehicle to accurately estimate its position and posture to ensure safe and efficient operation. Among them, the autonomous navigation technology that can be applied in complex environments is the key technology. A UGV navigation method based on inertial/odometer/lidar is proposed. Unlike the previous SLAM method, the proposed method can solve the position information based on the known geometric features, avoiding the matching error caused by the scarcity of effective points. At the same time, the fusion algorithm of inertia/odometer/lidar is studied, which improves the robustness and accuracy of the autonomous navigation system. At last, the corresponding simulation environment is built in Gazebo, and the algorithm is verified. The experimental results show that the method can realize the real-time and reliable autonomous navigation of UGV in the process of patrol inspection, and has good engineering application value.