基于噪声估计的卡尔曼滤波用于GPS卫星钟差预报研究

Study on GPS satellite clock error prediction by Kalman filter based on estimating noise

利用重叠哈达玛方差确定卫星钟噪声随机模型,采用顾及钟差随机噪声模型的卡尔曼滤波进行钟差预报分析,并与最小二乘预报算法相比较,得出以下结论:卡尔曼滤波进行1 d以内的短期预报时,精度达到亚纳秒级,优于最小二乘预报算法,在长期预报或拟合数据量较少时,最小二乘预报精度优于卡尔曼滤波。

This article has studied on satellite clock prediction by using Kalman filter considering stochastic model of satellite clock noise,and stochastic model can be gotten by overlapping Hadamard Variance.Comparing with Least Squares(LS),we can have some conclusions: Kalman filter can get a precision better than 1 ns,and excel LS;but when plan time short than 1 day or predict time more than 10 days,LS clock error prediction is better than Kalman filter.