基于一阶差分修正指数曲线法的GPS卫星钟差预报

GPS Satellite Clock Bias Prediction Based on First-order Difference MECM

针对二次多项式模型和修正指数曲线法模型在预报GPS卫星钟差时精度的不足,提出了基于一阶差分修正指数曲线法的GPS卫星钟差预报模型。该模型先对原始钟差数据序列相邻历元作一阶差分处理,以得到有效数字位数更少的一组钟差差值数据序列,以此来提高模型的预报性能;然后,基于一阶差分的钟差差值数据序列建立了GPS卫星钟差预报模型。试验结果与分析表明,该模型较二次多项式预报模型和修正指数曲线法预报模型的预报效果较好,中短期的预报精度均在1ns以内且只需采用较少的钟差数据进行建模即可达到高精度的预报,为GPS卫星高精度中短期钟差预报提供了一种新思路。

Aiming at the shortcomings of quadratic polynomial model(QPM)and modified exponential curve model(MECM)in GPS satellite clock bias(SCB)forecasting,a GPS SCB forecasting model based on first-order differential MECM was proposed.In this model,firstly,the algorithm made the firstvorder differential processing on the adjacent epochs of the original clock bias data sequence to obtain a set of data sequence of the clock bias sequence with a smaller number of effective digits,so as to improve the prediction performance of the model.Then,a GPS SCB forecasting model was established based on the first-order differential clock bias data sequence.The experimental results and analysis showed that the proposed algorithm was better than the forecasting accuracy of QPM and MECM,the accuracy of the short-term forecasts were less than 1 ns and only needed to use less clock data to achieve high-precision forecasting.