BDS信号失锁下基于自适应因子图的列车组合定位模型  

Train combination positioning model based on adaptive factor graph under BDS signal loss lock

作　　者：程相 王运明 王新屏 初宪武 CHENG Xiang;WANG Yunming;WANG Xinping;CHU Xianwu(School of Automation and Electrical Engineering,Dalian Jiaotong University,Dalian 116028,China)

机构地区：[1]大连交通大学自动化与电气工程学院,辽宁大连116028

出　　处：《铁道科学与工程学报》2024年第3期1146-1155,共10页Journal of Railway Science and Engineering

基　　金：辽宁省科学技术计划项目(2021-BS-219);辽宁省教育厅基本科研项目(LJKMZ20220857);大连市高层次人才创新支持计划项目(2021RQ060)。

摘　　要：针对列车运行环境复杂,容易产生BDS信号失锁,影响BDS/IMU列车定位系统的精确性问题,提出了基于自适应因子图的BDS/IMU/OD列车组合定位模型。在BDS/IMU列车定位系统的基础上,引入里程计(Odometer,OD)定位技术,利用北斗卫星导航系统(Beidou Navigation Satellite System,BDS)与惯性测量元件(Inertial Measurement Unit,IMU)、OD 3类传感器获取列车量测信息,根据因子图理论,将多源量测信息描述为状态空间方程,抽象BDS、IMU、OD因子节点和先验因子,确定因子节点与变量节点之间的无向连接关系,建立多变量列车组合定位因子图模型,解算列车的位置信息。当BDS信号产生变化时,借助因子图的即插即用特性,提出了自适应因子算法,动态调整列车组合定位因子图模型结构。在BDS信息部分失锁时,利用BDS的部分信息,建立BDS/IMU/OD列车组合定位因子图模型,在BDS信息完全失锁时,转换为IMU/OD列车组合定位因子图模型,抑制BDS完全失锁造成的发散误差。利用卡尔曼算法、因子图算法、自适应因子图算法进行了列车定位的仿真分析,在BDS信息部分失锁时,自适应因子图模型的定位位置均方根误差比卡尔曼算法分别降低了52.3%、48.2%和42.7%,比因子图算法分别降低了34.8%、27.0%和25.2%。在BDS信息完全失锁时,自适应因子图模型的定位位置误差比卡尔曼算法分别降低了46.7%、46.7%和50%。自适应因子图算法提高了BDS信息失锁的情况下的列车定位精度,实现了不同传感器之间的即插即用,为构建高精确性、强鲁棒性、高可扩展性的列车组合定位系统提供模型支持。Aiming at the problem that the train running environment was complex and the BDS signal was easy to lose lock,which affected the accuracy of the BDS/IMU train positioning system,a combined BDS/IMU/OD train positioning model based on an adaptive factor graph was proposed.On the basis of the BDS/IMU train positioning system,the Odometer(OD)positioning technology was introduced.The Beidou Navigation Satellite System(BDS)and the Inertial Measurement Unit(IMU)and OD three types were used.The sensor obtained the train measurement information,according to the factor graph theory,described the multi-source measurement information as a state space equation,abstracted BDS,IMU,OD factor nodes and prior factors,and determined the undirected connection relationship between factor nodes and variable nodes.It established a multivariate train combination location factor graph model,and calculated the location information of the train.When the BDS signal undergoes changed,an adaptive factor algorithm was proposed using the plug and play feature of the factor graph to dynamically adjust the structure of the train combination positioning factor graph model.When the BDS information was partially unlocked,a BDS/IMU/OD train combination positioning factor graph model was established using the partial information of BDS.When the BDS information was completely unlocked,it was converted into an IMU/OD train combination positioning factor graph model to suppress the divergence error caused by the complete unlocking of BDS.Using the Kalman algorithm,factor graph algorithm,and adaptive factor graph algorithm,the simulation analysis of train positioning is carried out.When the BDS information is partially out of lock,the root mean square error of the positioning position of the adaptive factor graph model will be reduced separately 52.3%,48.2%and 42.7%than that of the Kalman algorithm which are 34.8%,27.0%and 25.2%lower than the factor graph algorithm.When the BDS information is completely out of lock,the positioning position error of the ad

关 键 词：列车组合定位 BDS信号失锁 因子图 BDS/IMU/OD 自适应因子 

分 类 号：U284[交通运输工程—交通信息工程及控制]