基于IMU与激光雷达融合的无人弹药补给车SLAM系统研究
Research on SLAM system of unmanned ammunition supply vehicle based on the fusion of lidar and IMU
樊宏丽 1李郁峰 1郭荣 1陈晓锋1
作者信息
- 1. 中北大学 智能武器研究院,太原 030051
- 折叠
摘要
针对单一传感器建图精度低、实时性不足的问题,将IMU融合到激光雷达SLAM算法中.首先,采用手眼标定方法对2 种传感器坐标系外参进行标定,实现传感器在时间与空间上的对齐.然后,结合因子图优化模型,解决在建图过程中产生的漂移现象,并将IMU融合到激光雷达LeGO-LOAM算法中.最后,在室外场景下搭建了无人弹药补给车SLAM实验平台,分别进行了LeGO-LOAM算法融合IMU前后的建图和定位试验.结果表明,融合IMU后的SLAM算法建图和定位精度都明显提高,满足了在未知环境下无人弹药补给车建图和定位的性能要求.
Abstract
To address the issues of low accuracy and slow efficiency in single sensor mapping,IMU is integrated into the laser radar SLAM algorithm.Firstly,the hand-eye calibration method is used to calibrate the external parameters of the two sensor coordinate systems,achieving the alignment of the sensors in time and space.Then,combined with graph optimization model,the drift phenomenon generated during the mapping process was solved,and IMU was integrated into the LiDAR LEGO-LOAM algorithm.Finally,an unmanned ammunition supply vehicle SLAM experimental platform was built in an outdoor scene,and mapping and positioning experiments were conducted before and after the LeGO-LOAM fusion IMU algorithm.The results show that the SLAM algorithm fused with IMU significantly improves the accuracy of mapping and positioning,meeting the performance requirements of unmanned ammunition supply vehicle mapping and positioning in unknown environments.
关键词
激光SLAM/无人驾驶/多传感器融合/惯性测量单元/位姿优化Key words
laser SLAM/unmanned driving/multi-sensor fusion/inertial measurement unit/pose optimi-zation引用本文复制引用
基金项目
山西省基础研究计划联合资助项目(太重)(TZLH20230818005)
出版年
2024
NETL
NSTL
万方数据