基于Kalmam滤波和Kalman-RTS平滑的高铁轨道平顺性数据融合算法  

High-speed railway track regularity data fusion algorithm based on Kalman filter and Kalman-RTS smoother

作　　者：郭锦萍 白征东[1] 辛浩浩 GUO Jinping;BAI Zhengdong;XIN Haohao(School of Civil Engineering,Tsinghua University,Beijing 100084,China)

机构地区：[1]清华大学土木工程系,北京100084

出　　处：《测绘工程》2023年第2期7-12,共6页Engineering of Surveying and Mapping

摘　　要：针对传统轨道平顺性测量方法存在的依赖于CPⅢ控制网、维护成本高、测量技术效率低等问题,文中基于高铁轨道平顺性测量系统,采用Kalman滤波和Kalman-RTS平滑算法对包括GNSS接收机、里程计、IMU在内的多种传感器的数据进行融合处理。实验表明,多传感器数据先通过Kalman滤波处理后,轨道测量绝对坐标横向偏差均值从纯GNSS的4.7 mm降低至2.2 mm,精度提升幅度达53.2%;再进行Kalman-RTS平滑处理后,绝对坐标横向偏差均值再度降低到1.6 mm,总的精度提升幅度达66.3%,相对坐标横向偏差均值精度提升幅度达10.1%,可以有效提升轨道测量作业效率。In view of the problems existing in the traditional track regularity measurement methods, such as relying on the CPⅢ control network, high maintenance cost, and low measurement technology efficiency, this paper, which is based on the high-speed rail track regularity measurement system, uses Kalman filter and Kalman-RTS smoother to process the data from various sensors including GNSS receivers, odometers, and IMUs. Experiments show that after the multi-sensor data is processed by the Kalman filter, the mean value of the absolute coordinate lateral deviation of track measurement is reduced from 4.7 mm purely in GNSS to 2.2 mm, the accuracy is improved by 53.2%. After Kalman-RTS smoother, the mean value of the absolute coordinate lateral deviation is reduced again to 1.6 mm, and the overall accuracy is improved by 66.3%, also the accuracy of the mean value of the relative coordinate lateral deviation of track measurement is improved by 10.1%, which can effectively improve the efficiency of track measurement operations.

关 键 词：工程测量 高铁轨道平顺性 KALMAN滤波 Kalman-RTS平滑 

分 类 号：P228.4[天文地球—大地测量学与测量工程] U212[天文地球—测绘科学与技术]