利用奇异谱分析探测GPS坐标时间序列时变周期信号

Detecting Time-varying Periodic Signals of GPS Coordinate Time Series Through Singular Spectrum Analysis

GPS坐标时间序列呈现复杂的非线性特征,准确提取坐标时间序列中的季节性信号对理解区域形变具有重要意义。传统最小二乘拟合(Least Squares Fitting,LSF)无法顾及信号的年际变化特征,难以有效分离噪声和周期信号。介绍利用奇异谱分析(Singular Spectrum Analysis,SSA)方法提取GPS坐标时间序列中时变周期信号。结果表明,SSA可从含有噪声的GPS时间序列中提取各坐标分量的周期信号,相较于LSF,SSA在非线性信号的准确提取上具有更好的效果。

GPS coordinate time series have complex non-linear features,accurately extracting seasonal signals from which is of great significance for understanding regional deformation.Traditional least squares fitting(LSF)did not take into account the interannual variation characteristics of the signal,and is difficult to effectively separate noise from periodic signals.This paper uses singular spectrum analysis(SSA)method to extract time-varying periodic signals in GPS coordinate time series.The results show that SSA can extract the periodic signals of each coordinate component from the GPS time series containing noise,which outperforms LSF in extracting nonlinear signals.