EMD用于GPS时间序列降噪分析

发布时间：2018-02-04 08:21

  本文关键词： GPS时间序列 经验模态分解 降噪 本征模态分量　出处：《大地测量与地球动力学》2017年12期 　论文类型：期刊论文

【摘要】：针对GPS位置时间序列噪声成分复杂、建模困难、噪声与信号难以有效分离等特点,用自适应信号分析方法经验模态分解(EMD)对中国区域9个IGS站时间序列的N、E、U方向作降噪处理。首先对各方向分量作EMD分解,获取各方向噪声强度依次降低的一系列本征模态分量及其趋势项;其次利用相关系数辨别出噪声与信号的分界本征模态分量,并按重构方法重构各本征模态分量;最后用相关系数、信噪比、能量百分比等参数评价EMD方法降噪的有效性。结果表明,EMD方法能合理地分离序列中的信号与噪声,有效削弱噪声对GPS位置时间序列的影响,进一步提高GPS位置时间序列的精度。
[Abstract]:In view of the complex noise components of GPS position time series, the difficulty of modeling, and the difficulty of effectively separating the noise from the signal, and so on. In this paper, the empirical mode decomposition (EMD) method of adaptive signal analysis is used to reduce the noise of the time series of nine IGS stations in China. Firstly, the EMD decomposition of the components in each direction is performed. A series of eigenmode components and their trend terms are obtained which the noise intensity in each direction decreases in turn. Secondly, the boundary eigenmode components of noise and signal are distinguished by correlation coefficient, and each eigenmode component is reconstructed according to the reconstruction method. Finally, the correlation coefficient, signal-to-noise ratio, energy percentage and other parameters are used to evaluate the effectiveness of the EMD method. The results show that the EMD method can reasonably separate the signal from the noise in the sequence. The influence of noise on GPS position time series is weakened effectively, and the accuracy of GPS position time series is further improved.
【作者单位】： 长安大学地质工程与测绘学院;
【基金】：国家自然科学基金(41104019,41674001) 中央高校基本科研业务费专项基金(310826172202)~~
【分类号】：P228.4
【正文快照】： 如何有效分离GPS时间序列中包含的各类噪声已成为目前研究的热点。王敏等[1]、张飞鹏等[2]、姜卫平等[3]、袁林果等[4]研究了我国境内IGS站及区域CORS站的GPS时间序列的运动特性,结果表明GPS时间序列在水平方向主要以线性运动为主,而高程方向则表现出复杂的周期振荡与非线性运，

本文编号：1489799