灰色模型与自回归模型组合在钟差短期预报中的应用研究

Research on the Application of the Combination of Grey Model andAutoregressive Model in Short-term Clock Bias Prediction

研究卫星钟差预报的精度及稳定性对增强卫星导航系统的导航、授时等功能有重要作用。针对GM(1,1)模型在卫星钟差短期预报时存在的问题,引入AR模型对其预报结果进行修正,以提高钟差预报结果的精度和稳定性。实验结果表明:组合模型的预报精度较单一GM(1,1)模型有一定提高,其中铷钟的精度提高较为明显,铯钟的精度则提高较少;此外,组合模型的误差变化趋势与GM(1,1)模型相似,但稳定性相对更高。

It is important to study the accuracy and stability of satellite clock bias prediction for enhancing the navigation and timing functions of satellite navigation system.Aiming at the defects of GM(1,1)model in the short-term forecast of satellite clock bias,AR model is introduced to modify the forecast result to improve the accuracy and stability of the clock bias forecast results.The experimental results show that the prediction accuracy of the combined model is improved compared to the single GM(1,1)model,of which the accuracy improvement of the rubidium clock is more obvious,and the accuracy improvement of the cesium clock is less obvious.In addition,the error trend of the combined model is similar to that of the GM(1,1)model,but the stability is relatively higher.