半参数模型在GPS季节性信号分离中的应用

The Application of Semi-Parametric Model in the Separation of GPS Seasonal Signals

根据噪声序列的Hurst指数特性,提出一种新的确定半参数模型平滑因子的方法——基于Hurst指数的二分搜索法,并将该方法应用于分离时变GPS季节性信号。通过对模拟数据及实测GPS坐标序列的分析,验证了该方法是一种有效的确定平滑因子的方法。将计算得到的平滑因子代入半参数模型,能够将GPS坐标序列中的季节性信号充分地分离出来。

Based on the Hurst index characteristics of noise series, a new method determining the smoothing parameter for semi-parametric models is proposed in this paper, a binary search method based on the Hurst index. We apply it to analyze simulated and real GPS coordinate series, and the numerical results show the proposed method can effeetively determine the smoothing parameter, and the seasonal signal of GPS coordi- nates time series can be completely separated by using the determined semi-parametric model.