Abstract：Through testing and studying on the technique of multi-kernel function, which is a new numericalapproximation technique proposed in recent years, it is found that this method has the characteristics as follows:1) It is easy to accomplish the extraction and separation of data of various frequency band during the filtering of GPS time series.2)It is simple and accurate to complete the interpolation when the continuous lack of data is not very badly.3)This method owns analyticity, and the accurate evaluation of parameters and filtering values is rigorous; 4)The computing result is stable, and it is not easy to produce illconditioned matrix.5)The computing is simple and costs short time. Through some examples of data processing of GPS time series, the suitable multikernel function is recommended, and the problems needing attention are pointed out during concrete applications.
Yang Bo,Zhang Fengshuang,and Han Yueping. METHOD OF MULTI-KERNEL FUNCTION AND ITS APPLICATION IN GPS TIME SERIES DATA PROCESSING[J]. jgg, 2010, 30(4): 137-141.
