GEO卫星电离层延迟时间序列抗差谱分析

ROBUST SPECTRAL ANALYSIS OF TIME SERIES OF GEO IONOSPHERIC CORRECTION

用北斗GEO卫星的观测数据,通过双频电离层改正算法获取卫星约一天内的电离层延迟数据。利用抗差谱分析方法,分离电离层延迟时间序列中的长期项、周期项和残差项,以控制异常误差对周期项的影响。计算表明,电离层时间序列谱分析拟合后的残差仍然受有色噪声的影响,因此,基于AR模型对残差项进行有色噪声拟合。结果表明,每天的电离层延迟不仅表现出趋势项、周期项,而且含有有色噪声。经过分段多项式、周期函数和有色噪声拟合后,电离层延迟剩余误差表现出明显的白噪声现象,证明了针对GEO卫星电离层建模的有效性。

Ionospheric delay was obtained with day＇s observation data from Compass GEO satellite,based on dual-frequency ionospheric correction model. The robust spectral analysis method was applied to analize ionosphere delay data for detaching the long term,period terms and random errors and controlling the effects of error on the terms. In order to control the influences of colored noises,the residuals are fitted by using an AR model. The result shows that the ionospheric correction is effected by not only trend term and period term changes,but also by color noises. After being fitted by piecewise polynomial,period terms and color noise,the ionospheric correction residual terms turn out to be white noises. The result proves the method is seffective.