动态灰色模型在卫星钟差预报中的应用及精度分析

Application and Accuracy Analysis of Dynamic Grey Model in Satellite Clock Error Prediction

在GPS实时精密单点定位中,卫星钟差的实时可靠预报是实现GPS实时高精度单点定位的关键之一。在探讨了二次多项式模型和灰色模型对钟差预报的局限性的基础上,从而将动态灰色模型用于钟差预报。动态灰色模型是不断地剔除早期信息并加入新信息,因此提高了预测区间的白度,使预报的精度明显提高。该文采用动态灰色模型建模,并与二次多项式模型和单一GM(1,1)模型进行对比,经实例分析得出动态灰色模型可以有效地提高卫星钟差预报的精度。

In the GPS real time precise single point positioning, the real-time and reliable prediction of satellite clock error is one of the key to realize the GPS real time high precise single point location. Based on discussion the limitations of clock error prediction by the two order polynomial model and grey model, the dynamic grey model is used to clock error prediction. The dynamic grey model can continuously re- moved the early information and add the new information, so the whiteness of the prediction interval is improved. And the prediction accuracy is improved obviously . This paper uses the dynamic grey model to build model, and compares it with two polynomial model and single GM （ 1,1 ） model. Through the practice analysing, it shows the dynamic grey model can improved the accuracy of satellite clock error prediction effectively by dynamic grey model.