宇航计测技术

• 论文 • 上一篇    下一篇

广义回归神经网络在卫星钟差短期预报中的应用

雷雨1,2,4；赵丹宁1,3,4   

  1. 1、中国科学院国家授时中心，西安 710600；
    2、中国科学院时间频率基准重点实验室，西安 710600；
    3、中国科学院精密导航定位与定时技术重点实验室，西安 710600；
    4、中国科学院大学，北京 100039
  • 出版日期:2013-08-15 发布日期:2013-08-15
  • 作者简介:雷雨（1983-），男，博士生，主要研究方向：GNSS时间传递。
  • 基金资助:
    国家自然科学基金青年项目（11103025）

Application of Generalized Regression Neural Network in Short-time Prediction for Satellite Clock Error

LEI Yu1,2,4；ZHAO Dan-ning1,3,4   

  1. 1、National Time Service Centre, Chinese Academy of Sciences, Xi′an 710600;
    2、Key Laboratory of Time and Frequency Primary Standards, National Time Service Center,Chinese Academy of Sciences, Xi′an 710600;
    3、Key Laboratory of Precision Navigation and Timing Technology, National Time Service Center,Chinese Academy of Sciences, Xi′an 710600;
    4、University of Chinese Academy of Sciences, Beijing 100039
  • Online:2013-08-15 Published:2013-08-15

PDF (PC)

780

摘要： 近年来，神经网络(Neural Network,简称NN)在非线性系统的预测方面取得了广泛的应用。考虑到卫星钟差包含了复杂的非线性因素，所以将一种新型神经网络-广义回归神经网络(Generalized Regression Neural Network,GRNN)应用于钟差预报中。采用“滑动窗”方式构建样本数据以提高数据利用率，为提高网络的泛化能力，利用K重交叉验证法(K-fold Cross-Validation)对网络进行训练学习，并根据最小均方根误差(Root Mean Square Error,RMSE)确定最优平滑因子。利用国际GNSS服务(International GNSS Service,IGS)公布的精密GPS卫星钟差数据进行预报实验，并与传统二次多项式模型对比分析。结果表明：GRNN模型在24h的预报跨度内的误差可达ns级，并较多项式模型有更好的稳定性；对于线性钟差，GRNN模型要逊于多项式模型，而对于非线性钟差，GRNN模型则明显优于多项式模型，初步验证了GRNN用于钟差预报的可行性、有效性以及实用性。

关键词: 广义回归神经网络, 二次多项式, 钟差预报, 交叉验证, 滑动窗

Abstract: Neural Network(NN) has been widely used in forecast of nonlinear systems over the past years. Generalized Regression Neural Network(GRNN), which is a new kind of NN, is proposed to predict satellite clock error because of its nonlinear characteristic. An slip window pattern is used organize sample data, which can raise utilization rate of data. In order to improve forecast ability of GRNN, the method called K-fold Cross-Validation is employed to train network. Furthermore, the optimal smoothing factor is determined in terms of Root Mean Square Error(RMSE). An experiment is carried out to verify GRNN effectiveness. Real satellite clock error data from International GNSS Service(IGS) is trained to construct the GRNN model, then this model is used to predict clock error. And also Clock error is forecasted using quadratic polynomial. Results show that GRNN model can reach ns-level prediction accuracy within 24 hour. Moreover, GNRR is more stable in comparison with quadratic polynomial. GRNN is worse than quadratic polynomial when predicting linear clock error, however, the former is obviously better than the latter when nonlinear clock error.

Key words: Generalized regression neural network, Quadratic polynomial, Clock error prediction, Cross validation, Slip window

京ICP备09048426号-4　 京公网安备 11010602060144号　版权所有 © 《宇航计测技术》编辑部