有色噪声观测量的逐次静态滤波方法

155-15Progressive static filtering for the measurements with colored noises

根据广义最小二乘原理推导出有色噪声情况下的逐次静态滤波的关系式,即通过改化各观测方程,将有色噪声观测值改化为白噪声的虚拟观测值,然后按白噪声逐次滤波公式进行滤波计算。同时,推导出有色噪声观测量逐次滤波后精度评定的理论公式。有色噪声的逐次静态滤波理论有效地解决了GPS载波相位观测值经过测站间、卫星间、历元间的三次差分计算后,观测方程的相关性造成的协方差矩阵容量过大问题,利用这一理论,消除了各历元间隔观测方程的相关性,节省了数据处理时间和计算机内存。

On the basis of least squares principle the regular formulas of progressive static filtering with colored noises are derived, i.e. by the observation equations reducing, the measurements with colored noise can bereduced as the suppositional ones with white noise, and then are calculated in filtering by means of the formula for the one with white noise. Meanwhile, the theoretical formula of precision evaluation after filtering is also derived in the paper. The theory can effectively solve the problem that the capacity of covariance matrix is too much, which results from the correlativity of observation equations after three times of differential computing among stations, satellites and calendars for the measurements of GPS carrier phases so that the time and memory for data processing can be saved.