改进Vondrak滤波在削减BDS多路径误差中的应用  

Application of improved Vondrak filtering in reducing BDS multipath errors

作　　者：杨国刚[1] 蔡成林[1] 唐振辉[1] 孙凯[1] 

机构地区：[1]桂林电子科技大学广西精密导航技术与应用重点实验室,广西桂林541004

出　　处：《导航定位学报》2017年第4期78-85,共8页Journal of Navigation and Positioning

基　　金：国家自然科学基金项目(61263028);广西高校科学技术研究项目(KY2015ZD042)

摘　　要：针对传统Vondrak滤波在GNSS高精度定位数据处理中易受粗差值影响的问题,提出一种削减BDS多路径误差的改进型Vondrak滤波方法,即基于抗差的Vondrak滤波法:从函数模型方面论述结合最小二乘并通过引入校正函数在循环解算中逐步消减粗差影响的过程,并以滤波前、后序列RMS,序列间相关系数,误差改正前、后坐标序列RMS作为误差改正的效果评价准则,将RVF方法实际应用于削减BDS数据多路径误差的研究中,发现RVF模型可最大限度削弱BDS多路径观测资料中异常粗差的影响,提取出精确的多路径模型。实验结果表明,经RVF提取的多路径序列改正后,BDS定位结果坐标序列的精度能够提高30%~40%。Aiming at the problem that the traditional Vondrak filtering is subject to gross errors in the data processing of precise GNSS positioning,this paper proposed the robust Vondrak filtering(RVF)method based on least square:from the aspect of function model,the process of successively reducing the influence of gross errors in the cyclic solution by combining least squares and introducing the correction function was discussed in detail;and RVF was applied to lessen the BDS multipath errors,taking both correlation coefficients between sequences and RMS before and after filtering and correction as the evaluation criteria;then it was found that the RVF model could maximum weaken the gross errors and extract the accurate multipath sequence,and have a simpler and more convenient calculation process.Experimental result showed that the accuracy of coordinate sequence of BDS positioning could be improved 30 %~40 % after the correction of the multipath sequences extracted by RVF method.

关 键 词：BDS VONDRAK滤波 多路径 粗差 校正函数 

分 类 号：P228[天文地球—大地测量学与测量工程]