多核函数法对GNSS大空间域共模误差的识别

The Identification of GNSS Large Space Domain Common-Mode Error Based on Multi-Kernel Function Method

根据GNSS时序资料3分量E,N和U误差的构成和地壳微形变分析的需求,提出了基于多核函数大空间域共模误差识别的实用方法。并以川滇地区GNSS连续观测资料为例进行了实际计算与应用,获得了如下看法或结果:1)基于多核函数可有效识别大空间尺度GNSS时序3分量中的共模误差;2)进一步明确了GNSS连续站水平向时序观测值中共模误差是其误差的主要组成部分;3)通过对共模误差白噪声的剔除性处理可以更有效地分辨微形变或弱动态变化信息。

Based on the structure of three component errors that the N,E,U times series of GNSS and the demand of the crustal micro deformation analysis, an practical method of large space domain common-mode error identification was raised based multi-kernel function, and the actual calculation and application were carried out as an example by using the GNSS continuous observation data in Sichuan-yunnan area. The idea or results are obtained： 1） The common-mode error in GNSS three component time series of the large space domain can be effectively identified by multikernel function; 2） That the common-mode error in GNSS continuous time series observation in the horizontal direction is the main components of the errors was clarified further; 3） The micro deformation or weak dynamic change information can be distinguished through the process to eliminate the white noise part of common-mode error.