多星座组合精密动态定位的抗差扩展Kalman滤波方法研究  

Robust Extend Kalman Filtering Method Based on Precise Relative Positioning by Using Multi-constellation Integrated System

作　　者：高晓[1,2] 戴吾蛟[1] 张超[1] 余文坤[1] 

机构地区：[1]中南大学测绘与国土信息工程系,湖南长沙410083 [2]中国电力工程顾问集团西北电力设计院有限公司,陕西西安710032

出　　处：《武汉大学学报（信息科学版）》2015年第10期1329-1333,共5页Geomatics and Information Science of Wuhan University

基　　金：湖南省国土资源厅科技资助项目(2012-41);中南大学研究生自主探索资助项目(2014zzts249)~~

摘　　要：扩展Kalman(EKF)滤波算法可有效地进行多卫星系统数据融合处理,但该方法对观测数据的质量要求较高,当观测出现异常时,传统的扩展Kalman方法容易导致滤波失真。为此,利用实测数据,通过虚拟掩模遮挡卫星模拟复杂的GNSS观测环境,研究了基于IGGIII的抗差EKF算法,确定了分位参数的合理经验值,并对其在GPS/GLONASS/BDS组合精密动态定位中的应用进行了分析。结果表明,在遮挡严重的复杂观测环境中,抗差EKF算法可有效地提高组合定位系统的模糊度的固定成功率和定位精度。The extended Kalman filter method is effective in multi-constellation integrated systems,but its application is restricted because the method demands high quality observations.Due to the fact that the classical EKF method will seriously degrade because of observation outliers,this paper studies the robust EKF method based on IGGIII model and determines the value ranges of fractile factors using actual measurement data from the integrated GPS/GLONASS/BDS system.Three masks are used to simulate the complex user observation environment,which are helpful when investigating the application of robust EKF method in precise relative positioning.Results show that the robust EKF improves the fix-rate for ambiguity and positioning accuracy.

关 键 词：多星座组合定位 抗差估计:扩展Kalman滤波 

分 类 号：P228.41[天文地球—大地测量学与测量工程]