附有周期项的预报模型及其在GPS卫星钟差预报中的应用研究

为了有效进行GPS卫星钟差预报和更好地反映卫星钟差特性,除了考虑卫星原子钟频移、频漂和频漂率等物理性质外,还应考虑到卫星钟差的周期性变化特点.在二次多项式模型基础上,增加了周期项因素,构造了新的预报模型.选取部分GPS卫星铯钟(Cs.clock)和铷钟(Rb.clock)钟差资料,根据钟差变化趋势分3种情况,按不同时间长度进行钟差预报分析,并与二次多项式模型的预报结果比较分析,大量数据分析表明:附有周期项的二次多项式模型预报精度优于二次多项式模型,铷钟预报精度略优于铯钟.

Prediction Model with Periodic Item and Its Application to the Prediction of GPS Satellite Clock Bias

In real-time GPS precise point positioning(PPP),real-time and reliable satellite clock bias(SCB)prediction is a key to implement real-time GPS PPP.It is difficult to hold the nuisance and inenarrable performance of space-borne GPS satellite atomic clock because of its high-frequency,sensitivity and impressionable.To predict SCB availablv and show its property clearer,we should consider the periodic variety of SCB besides the perspicuous characters of GPS satellite clocks such as the frequency deviation,frequency drift and kequency drift rate and so on.A new prediction model taking account of the Deriodic item based on the quadratic polynomial model is put forward in this paper.Taking GPS satellite clock data from cesium clock and rubidium clock for example,we analyze the clock bias prediction in different time scales and different cases,compared with the predictiOn of quadratic polynomial model.It shows that the prediction accuracy of quadratic polynomial model with periodic item is prior to that of quadratic polynomial model,and the prediction accuracy of rubidium clock data is a bit prior to that of cesium clock.