首页|基于纯惯性导航系统的智慧铁路人员定位系统研究

基于纯惯性导航系统的智慧铁路人员定位系统研究

扫码查看
针对GPS与IMU组合导航定位系统在不同应用场景下,存在GPS无线信号失锁和惯性传感器导航解算过程误差累积等问题,提出了一种纯惯性导航系统的智慧铁路人员定位方法.利用惯性传感器采集铁路巡检员的足部数据,并分析其运动特征;提出机器学习智能分类算法,检测行人在行走过程中的静止阶段和运动阶段.在初始状态下进行初始对准,以确定初始方位;在零速状态下,通过扩展卡尔曼滤波估计航向误差、角速度误差、速度误差,实现导航解算的姿态、速度和位置的零速修正.对比巡检员导航实验与GPS RTK差分定位实验发现,提出的基于纯惯性导航系统的智慧铁路人员定位系统可以准确跟踪巡检员的实时位置,并且平均误差为2.674%,最大误差不超过3.3%.
Research on Intelligent Railway Personnel Positioning System Based on Pure Inertial Navigation System
A pure inertial navigation system based intelligent railway personnel positioning method is proposed to address the issues of GPS wireless signal loss and accumulation of errors in inertial sensor navigation calculation process in different application scenarios of GPS and IMU integrated navigation positioning systems.Using inertial sensors to collect foot data of railway inspectors and analyze their motion characteristics.Propose a machine learning intelligent classification algorithm to detect the stationary and moving phases of pedestrians during walking.In the initial state,perform initial alignment to determine the initial orientation.At zero speed,the extended Kalman filter is used to estimate heading error,angular velocity error,and velocity error,achieving zero speed correction of attitude,velocity,and position in navigation calculation.Comparing the navigation experiment of inspectors with the GPS RTK differential positioning experiment,it was found that the proposed intelligent railway personnel positioning system based on pure inertial navigation system can accurately track the real-time position of inspectors,with an average error of 2.674%and a maximum error of no more than 3.3%.

machine learninginitial alignmentnavigation solutionextended kalman filterzero velocity update

闫双建、闫永刚、邢渊博、陈召阳、段祥玉、张开法、孙满盈

展开 >

河南思维轨道交通技术研究院有限公司,河南 郑州 450001

上海麒德机电设备工程有限公司,上海 201114

北京思维鑫科信息技术有限公司,北京 100071

机器学习 初始对准 导航解算 扩展卡尔曼滤波 零速修正

2024

自动化应用
重庆西南信息有限公司

自动化应用

影响因子:0.156
ISSN:1674-778X
年,卷(期):2024.65(13)