EMD-SVD静态短基线多径抑制方法  

Static short baseline multipath mitigation method based on EMD-SVD

作　　者：于凌宇 陈熙源 徐杨 YU Lingyu;CHEN Xiyuan;XU Yang(School of Instrument Science and Engineering,Southeast University/Key Laboratory of Micro-Inertial Instrument and Advanced Navigation Technology of Ministry of Education,Nanjing 210096,China)

机构地区：[1]东南大学仪器科学与工程学院/微惯性仪表与先进导航技术教育部重点实验室,南京210096

出　　处：《导航定位学报》2023年第3期138-146,共9页Journal of Navigation and Positioning

基　　金：国家自然科学基金项目(61873064);江苏现代农业产业关键技术创新项目(CX(21)2015);苏州市科技计划项目(SNG20200039)。

摘　　要：为了进一步抑制全球卫星导航系统(GNSS)高精度定位中的多径效应,针对在城市峡谷中短基线定位后的多径误差,结合经验模态分解(EMD)和奇异值分解(SVD)的优点,提出一种EMD-SVD多径抑制算法:对定位序列进行EMD分解,得到含有不同频率成分的固有模态函数(IMF);利用相关系数提取多径误差,相关系数最大的IMF分量为主要的多径误差序列;剔除多径误差后的序列往往含有变化很小的多径噪声分量,利用SVD分解对EMD预处理后的信号进一步降噪,剔除残存的多径噪声。实验结果表明,当多径噪声与信号的频率相近时,EMD-SVD方法的多径抑制效果明显优于EMD方法和SVD方法;经度、纬度、高度3个方向的定位精度能够分别提高50.32%、37.84%和56.95%。In order to further suppress the multipath effect in the high precision positioning of global navigation satellite system(GNSS),combining with the advantages of empirical mode decomposition(EMD)and singular value decomposition(SVD),the paper proposed an EMD-SVD multipath mitigation algorithm for the multipath errors after short baseline positioning in urban canyons:EMD decomposition was performed on the positioning sequence to obtain intrinsic mode function(IMF)components with different frequency components;and the correlation coefficient was used to extract the multipath error and the IMF component with the largest correlation coefficient was taken as the sequence of the main multipath error;after removing the multipath error,the sequence often contains the multipath noise component with little change,then SVD decomposition was used to further reduce the noise of the preprocessed signal by EMD and eliminate the residual multipath noise.Experimental result showed that when the frequency of multipath noise and signal would be similar,the multipath suppression effect of EMD-SVD method could be better than that of EMD method and SVD method;and the positioning accuracy of longitude,latitude and height could be improved by 50.32%,37.84%and 56.95%,respectively.

关 键 词：多径抑制 短基线静态定位 经验模态分解 奇异值分解 

分 类 号：P228[天文地球—大地测量学与测量工程]