基于时间序列的快速卫星钟差短期预报研究

Study on Short-term Prediction of Ultra-Rapid Satellite Clock Error Based on Time Series

卫星钟差预报在实时高精度的导航定位中起着重要的作用。为提高GPS实时精密单点定位的精度,需要有实时的、高精度的卫星钟差预报。利用时间序列系统理论与灰色模型对快速卫星钟差进行了短期预报。结果表明,基于时间序列预报模型的预测精度优于灰色模型,它对超快速星历的预报精度与IGS产品中的IGU超快速星历本身的精度相当。Rb钟的精度和稳定性优于Cs钟。

The prediction of satellite clock error plays high precise. To improve the precision of real-time an important role in navigation positioning with real-time precise point positioning, we have to obtain real-time ephemeris with high-precision. Time series system theory and GM model are used to predict short term satellite clock error. The final results show that the predicting accuracy of ARIMA model based on time series is better than that of GM model, the predicting accuracy of satellite clock error with ultra-rapid products is comparable with IGU products, and the precision and stability of Rb clock is better than that of Cs clock.