Research On Prediction And Characteristics Of BDS/Galileo Onboard Satellite Atomic Clock

In the navigation and positioning system, precise positioning is essentially precise time measuring, the quality of onboard atomic clocks decide the performance of the navigation, positioning and timing(PNT) services directly. So, it is necessary to analyze the characteristics of onboard atomic clocks, to provide reliable guarantee for PNT services. Based on this, this paper using the BDS/Galileo precision clock error time series products to research the mid-long term characteristics of BDS/Galileo on-board satellite clocks, and study on the short predicting model of BDS, summarized as follow:(1) For BDS/Galileo onboard atomic clocks, in order to fit the satellite clock error model, the paper using the quadratic polynomial model, calculating BDS/Galileo on-board satellite clocks’ phase, frequency, frequency drift and residual, choosing Hadamard total variance to calculate the stability of BDS and Galileo onboard atomic clocks, analyzing the on-board satellite clocks’ characteristics, revealing the variation law of the BDS and Galileo satellite clocks’ phase, frequency, drift and residual; finding the performance of BDS MEO satellite clocks are better than the GEO and IGSO satellites; for Galileo, the paper finding out the performance of No.20 satellite clock is worse than the others obviously, and finding out that there is obvious cyclical change in the Galileo satellites clocks when adopting the quadratic polynomial model to fit Galileo satellites clock error model.(2) The results of BDS on orbit satellite clocks’ frequency stability shows that, the ten thousand-second stability of the BDS satellite clocks on orbit remain at around 10-13, it is consistent with the design index of the BDS atomic clocks; and the thousand-second frequency stability of Galileo satellite clocks shows that, during operation, the frequency stability of No.11,NO.12,NO.19 satellite clocks appear a number of abnormal jump, initial stage basically remain at around 10-14 order of magnitude; mid-term jumping to 10-13, and then restoring to 10-14 at last. The frequency stability variation law is in consistent with frequency drift, so speculating that there may be a number of switching actions between main and preparing clocks to Galileo satellite clocks.(3) For short predicting model of BDS clock errors, by comparing the quadratic polynomial model to the grey model, this paper finding out when predicting 6 hours, the accuracy degree of the quadratic polynomial model is in the same order of magnitude with the grey model basically. And the results of predicting 12 and 24 hours showing that, when increasing the initial data amount, the accuracy degree of the quadratic polynomial model increases gradually, the grey model decreases gradually. When the initial data is more than 12 hours, the predicting accuracy of the quadratic polynomial model is better than the grey model. Totally, for the stable characteristics BDS clock errors, adopting the quadratic polynomial model doing short predicting is better than the grey model.