控制与信息技术Issue(4)：74-81,8.DOI:10.13889/j.issn.2096-5427.2024.04.010
轨道交通高实时性高精度定位方法研究
Research on Highly Real-Time and High Precision Localization Method for Rail Transit
黄文宇 1潘文波 1李源征宇 1陈志伟 1杨振宇 1袁超1
作者信息
- 1. 中车株洲电力机车研究所有限公司,湖南 株洲 412001
- 折叠
摘要
Abstract
In order to address the real-time high-precision localization needs in open-air and tunnel scenarios within the rail transit sector,this paper proposes a real-time matching localization method based on LiDAR and inertial measurement sensors. This approach effectively combines the short-term high-frequency advantages of inertial sensors with the long-range stability and accuracy of point cloud-map matching. As a result,it effectively eliminates motion distortion and generates data with enhanced quality. First,the results from inertial computation are directly used as predictive input values for the filter,to fully leverage the dynamic response characteristics of inertial sensors. The filter is then updated relying on the precise matching results between the point cloud and the pre-built map. Second,a matching algorithm based on normal distributions transform (NDT) is employed to obtain the accurate positional offsets of point cloud data in actual environments. The results of subsequent experiments conducted in open-air and tunnel scenarios of rail transit environments showed an average localization error of 0.1 m and average processing time of 98.13 ms,without trajectory non-convergence and degradation observed. These results demonstrate the method's alignment with the localization requirements for autonomous perception systems in trains.关键词
轨道交通/匹配定位/正态分布变换/无迹卡尔曼滤波器Key words
rail transit/matching localization/normal distributions transform(NDT)/unscented Kalman filter(UKF)分类
计算机与自动化引用本文复制引用
黄文宇,潘文波,李源征宇,陈志伟,杨振宇,袁超..轨道交通高实时性高精度定位方法研究[J].控制与信息技术,2024,(4):74-81,8.
