基于3D LiDAR SLAM的定位与制图方法研究

Research on Positioningand Mapping Method Based on 3D LiDAR SLAM

针对GNSS在室内环境中无法准确定位的问题,本文采用3D LiDAR SLAM算法实现在上述环境中的高精度定位,同时构建高质量三维点云地图。本文主要针对传感器数据预处理、LiDAR SLAM前端匹配、LiDAR SLAM后端优化、三维点云地图构建4个部分进行系统的研究。通过搭建实验平台并设计真实场景下的定位与建图实验,验证3D LiDAR SLAM算法的有效性。实验结果表明,在复杂的室内场景中,3D LiDAR SLAM的定位精度可到达分米级,建图精度可达到厘米级。

Addressing the challenge of inaccurate GNSS positioning within indoor environments,this study employs a 3D LiDAR SLAM algorithm to achieve high-accuracy positioning and simultaneously generate meticulous 3D point cloud maps.The paper′s primary focus involves four key aspects:preprocessing of sensor data,front-end matching for LiDAR SLAM,back-end optimization for LiDAR SLAM,and the construction of 3D point cloud maps.The efficiencyof the 3D LiDAR SLAM algorithm is verified through the creation of an experimental platform and the implementation of positioning and map production experiments within real-world scenarios.The outcomes of these experiments demonstrate that within intricate indoor scenes,the 3D LiDAR SLAM achieves decimeter levelpositioning accuracywhile map production accuracy reaches the centimeter level.