A hybrid model for navigation satellite clock error prediction

Abstract

In order to improve navigation satellite clock error prediction accuracy, a hybrid model is proposed in this paper. According to the physics property of atomic clock, the model firstly fits the clock error series by polynomial model. Then it models for polynomial fitting residuals, using functional network. The functional network structure is defined by wavelet de-noising and phase space reconstruction. Finally the GPS satellites are taken for example and four separate predict tests are done, the simulation results show that the proposed method can fit and predict the clock error series effectively, whose predict accuracy is better than those of IGU-P and conventional methods. © Springer-Verlag Berlin Heidelberg 2013.