Research On Estimation,Modeling And Prediction Algorithms Of GPS Inter-frequency Clock Bias

In recent years,with the rapid development of Global Navigation Satellite System(GNSS),increasing numbers of satellites are transmitting three or even more than three frequencies,and the fusion of multi-frequency signals has become the development trend of GNSS.The abundant frequency resources can realize more combinations with excellent performance,which brings new opportunities and challenges to the rapid ambiguity resolution in precise point positioning(PPP),cycle slip detection and repair,carrier phase extraction,etc.The existence of IFCB on GPS satellites not only prevents traditional L1/L2 ionosphere-free(IF)combined precise clock products from the application of PPP involving L5 frequencies directly,but also has a significant impact on the performance of multi-frequency PPP.Therefore,the accurate estimation,modeling and predicting of IFCB are of great importance for the consistent use of traditional GPS precise clock products for multi-frequency PPP.Aiming at the demand of real-time application of PPP and the development trend of multifrequency,this article takes GPS satellite IFCB as the research object,a set of IFCB modeling and predicting strategy are established that can meet real-time multi-frequency PPP as the main research objective,an IFCB modeling and predicting method suitable for real-time PPP is proposed,which improves the short-term predicting accuracy and availability of IFCB.This article focuses on IFCB estimation,high-precision modeling and predicting methods.The main research contents and results of this paper are as follows:(1)A parameterization method for GPS multi-frequency hardware delay bias is established.Considering the correlation between parameters and maintaining compatibility with traditional precise clock products,A strategy for combining GPS multi-frequency hardware delay bias with satellite clock,receiver clock,slant ionospheric delay and undifferenced ambiguity parameters is formulated.According to the strategy,a multi-frequency undifferenced and uncombined PPP function model considering IFCB is established,and the multi-frequency PPP floating resolution is realized.(2)Combined with the observation quality control method and epoch-differenced method based on GFIF(Geometry-free and Ionospheric-free)combination,the efficient estimation of IFCB is realized.The effect of IFCB on multi-frequency PPP is investigated using the estimated IFCB.Compared with ignoring IFCB correction,the positioning accuracy of E,N,U three directions in static mode with IFCB correction is increased by 34.7%,18.3%,30.6%respectively,and the positioning accuracy in kinematic mode can be increased by 38.6%,43.6%,47.4%.At the same time,the correction of the IFCB can effectively eliminate the systematic error in the residual of the carrier phase.(3)The IFCB variation characteristics of GPS BLOCK IIF and BLOCK III satellite are revealed from the perspectives of time and frequency domains.Studies have shown that the daily IFCB variation of the BLOCK IIF satellite ranges from centimeters to decimeters,and the maximum is close to 20 cm,which has obvious time-varying characteristics.However,the daily IFCB variation of the BLOCK III satellite is usually less than 1 cm,and the IFCB time-varying characteristics of the Block IIF satellite have been significantly weakened.At the same time,the BLOCK IIF satellite IFCB has main period items of 12 hours and 6 hours,which are similar to the satellite orbit period or half of it,and there are also periodic items such as 8 hours and 4hours.Based on this,the single-day IFCB is modeled,and the model considering 6 period items is more consistent with the variation characteristics of the IFCB within single-day,and the single-day modeling accuracy reaches 6mm.(4)Aiming at the problems of initial deviation in traditional predicting method that fits the higher-order harmonic model directly and the difficulty of unifying the IFCB datum of different predicting arcs,an improved IFCB predicting method is proposed.This method takes into account the correction of the IFCB initial deviation,and significantly improves the short-term prediction accuracy by 66.7%,57.1%,43.5% and 37.0% for 3 hours,6 hours,12 hours and 24 compared with the traditional method.At the same time,the IFCB relative datum is introduced in the predicted arc,maintaining the uniformity of the IFCB relative datum in different predicted arcs.The effectiveness and usability of this method in continuous positioning is verified by PPP,which significantly improves the positioning performance of kinematic PPP at the junction of adjacent prediction arcs.Compared with the traditional method,the multiple reference stations have increased by 20.4%,19.1%,and 20% in the three directions of E,N,and U,respectively,which makes up for the shortcomings of traditional predicting methods in continuous or realtime PPP.