短期接收机钟差预测方法及性能分析

Forecasting Methods and Performance Analysis of the Receiver Clock Error in Short-Term

接收机钟差建模与预测能缩小码相位不确定范围,可减少信号搜索范围,有效辅助信号的快速捕获。对实测的钟差数据分别利用二次多项式模型、AR模型和UC滑窗模型建模与预报,比较了三种方法得出的预报效果,结果表明:在建模的钟差观测点数为25,预报间隔为200s时,UC滑窗模型的预报误差的均值与二次项模型和AR模型相当,但误差的标准差比二次项模型和AR模型分别降低了18.9ns和0.1ns.

Modeling and forecasting of the receiver clock error can compress the uncer- tainty range of code phase, and reduce the searching range of the signal, effectively assist fast capture of signal. In this paper, the measured data of clock error were used to modeled and forecasted by quadratic polynomial model, AR model and UC sliding window model. Com- parison of the forecasting results of three methods showed that the prediction precision of UC sliding window is comparative with quadratic polynomial model and AR model, but the stand- ard deviation of error respectively decreases by 18.9 ns and 0.1 ns, when the length of the observed clock error in the model is 25 and the prediction interval is 200 s.