基于三维激光雷达的智能车辆SLAM技术研究  

Research on 3D Lidar Based SLAM for Intelligent Vehicles

作　　者：巴浩麟 马晓录[1] 吴立辉[1] BA Haolin;MA Xiaolu;WU Lihui(School of Mechanical and Electrical Engineering,Henan University of Technology,Zhengzhou 450001,China)

机构地区：[1]河南工业大学机电工程学院,河南郑州450001

出　　处：《汽车实用技术》2023年第9期197-204,共8页Automobile Applied Technology

基　　金：国家自然科学基金资助项目(U1704156)。

摘　　要：智能车辆实现自主导航的必要条件之一是定位技术,目前基于三维激光雷达的同步定位与建图(SLAM)技术是定位技术的主流方案。文章从三维激光SLAM的算法框架和关键模块进行总结,具体讨论了扫描匹配、后端优化、闭环检测、地图构建等关键模块的常用算法及改进;综述了几种开源的三维激光SLAM,并对比了其优缺点;对在应用中激光雷达局部点云稀疏、Z轴的漂移以及动态对象引发的噪声影响等关键性问题进行了分析阐述;指出了基于三维激光雷达的SLAM与深度学习相结合、多传感器融合是未来三维激光SLAM的发展和研究方向。Localization technology is one of the necessary conditions for intelligent vehicles to achieve autonomous navigation.At present,simultaneous localization and mapping(SLAM)based on 3D lidar is the mainstream solution of Localization technology.This paper summarizes the algorithm framework and key modules of 3D laser SLAM,and discusses the common algorithms of key modules such as scanning matching,back-end optimization,closed-loop detection,map construction,etc.Several 3D SLAM methods are reviewed,and their advantages and disadvantages are compared.The key problems in the application of lidar,such as local point cloud sparse,Z-axis drift and the influence of noise caused by dynamic objects,are analyzed and expounded.It is pointed out that the combination of 3D lidar based SLAM with deep learning and multi-sensor fusion will be the development and research direction of 3D laser SLAM in the future.

关 键 词：SLAM技术 扫描匹配 智能车辆 三维激光雷达 

分 类 号：U463.6[机械工程—车辆工程]