一种基于自相关函数的GNSS时间序列噪声提取方法

Noise Extraction Method for GNSS Time Series Based on Autocorrelation Function

对GNSS时间序列的非线性研究一直是大地测量学和地球动力学研究的方向之一,从信号本身的特性去分析和处理非平稳信号的HHT方法在提取GNSS时间序列噪声方面缺乏一种有效的分界点确定方法来进行滤波。针对该问题,本文提出了一种基于自相关函数的GNSS时间序列噪声提取方法,先通过与CMSE方法的对比分析验证了该方法的有效性,随后利用该方法提取了stjo测站的噪声数据进行分析,得出最优噪声模型为WN+FN。

The nonlinear study of GNSS time series has been one of the directions for improving geodesy and geodynamic research.The hilbert-huang transform method was in a position to process non-stationary signals from the characteristics of the signal itself,but lack of a demarcation point determination method for filtering the noise of GNSS time series.Thus A noise extraction method of GNSS time series based on autocorrelation function was presented.The validity of the method was verified by comparing with the CMSE method.Then the noise data of stjo station was extracted by this method,and proved that the optimal noise model was WN+FN.