基于小波与ARMA模型的卫星钟差预报方法

A GPS PRECISE CLOCK ERRORS PREDICTION MODEL BASED ON WAVELET AND ARMA

针对导航卫星钟差预报中通常采用的二次项拟合模型(短期预报)与灰色模型(长期预报),只考虑钟差趋势项而忽略其随机项的问题,提出了一种结合小波与ARMA模型的钟差预报改进算法。在算例中采用IGS提供的精密钟差进行预报,结果表明,改进后的算法进一步提高了钟差预报的精度。

The quadratic polynomial model and grey model used in short and long period clock errors predictions are analyzed,considering that these models only think about trend item but not random item.An improved method is given,which combines wavelet and ARMA.The computation results show that through an example by using the precise clock errors from IGS（International GPS Service） the improved model has higher accuracy.