GPS卫星钟差短期预报研究

Short-term Forecast of Clock Error of GPS Satellite

针对传统灰色系统模型在卫星钟差预报时易受卫星钟差序列中的粗差、周期性误差等因素的影响,提出了一种奇异谱分析与灰色理论相结合的组合预报模型。该模型首先利用奇异谱分析对卫星钟差序列进行分析,提取卫星钟差的趋势性变化序列,并以此建立灰色系统模型,实现对卫星钟差的短期预报估计。通过与AR模型以及GM(1,1)模型预报结果的对比,验证了组合模型在预报精度和稳定性上的优越性;通过对4种类型GPS卫星的钟差序列进行预报,表明了组合模型的适用性更强。

Since the traditional grey system model is susceptible to factors such as gross errors and periodic errors when forecasting satellite clock error,a combined forecasting model of singular spectrum analysis and grey theory is proposed.It first analyzes the satellite clock error sequence through singular spectrum analysis,extracts the trend change sequence,and establishes a grey system model to realize the short-term forecast estimation of the satellite clock error.By comparing with the prediction results of AR model and GM(1,1)model,the forecasting accuracy and stability of the combined model is verified.After forecasting these 4 types of GPS satellite clock error series,the combined model is proved more applicable.