GNSS Real-Time Precise Positioning Service System:Theory, Algorithm And Implementation

The Global Navigation Satellite System (GNSS) is a significant engineering system in the 20th century, which involves a large number of science and technology innovations. Since the advent of GPS in 1970s, the GNSS navigation and positioning theory and technique experience a rapid development. On the one hand, the GNSS system has been developing from the double-system of GPS and GLONASS to multi-system of GPS, GLONASS, Galileo and BDS (BeiDou Navigation Satellite System) etc., from dual-frequency signal to triple-frequency and multi-frequency signal, from the MEO-only (Median Earth Orbit) constellation to the GEO, IGSO and MEO-mixed constellation. On the other hand, the navigation and positioning algorithms and data processing theories have been evolving from post-processing to real-time processing, from differential processing to the single point processing, from low precision to high precision, from single system to multi-GNSS. It is an application and developing trend that using Multi-GNSS, multi-frequency signal and observation type to realize real-time and higher precision navigation and positioning application. This development not only completes the GNSS technology and theory, but also provides motivation for the modern development of GNSS technology theory and the new application requirements.GNSS real-time precision positioning technology is of important industrial and scientific application value, which can significantly improve the efficiency of social production, and is the inevitable development trend of satellite navigation and positioning technology. The GNSS real-time precision positioning users usually need to rely on external real time precision positioning service system that provide real-time positioning support products, such as ultra-fast precise orbit, real-time clock correction, satellite hardware delay and atmospheric delay correction. Thus, it is of great significance to study the GNSS real-time precision positioning service system of the related key technologies to promote the application of the GNSS real-time precision positioning technology.To implement and realize the GNSS real-time precise positioning service system, this dissertation focuses on some important issues in GNSS real-time precise positioning, such as data preprocessing and real-time quality control; GPS, GLONASS, Galileo and BDS quad-system real-time precise clock correction estimation; GPS/BDS satellite phase hardware delay estimation and the PPP ambiguity resolution, and modelling regional troposphere and ionosphere using the PPP-AR technique. These researches will complete and improve the GNSS data processing theory and algorithm, join the wide-area and local-area to generate the general real-time service products, form the wide-and local-area compatible data processing system for real-time precise positioning service, and provide a wide scales of real-time precision positioning service products, promote the further wide application of GNSS real-time precision positioning. Specifically, the work and contribution of this dissertation includes the following aspects:1. Proposed a real-time cycle-slip detection and repair algorithm which can make full use of the BDS, modernized GPS and QZSS triple-frequency observation to detect and repair cycle-slip in real-time. The single epoch detection and repair success rate is 99.14% for BDS GPS/QZSS 99.96% respectively. To satisfy the demand of real-time processing in the server end, a single station parallel distributed quality control strategy is proposed, by which the reference stations carry out PPP based on the real-time products from the server end and carry out quality control based on the PPP residual independently. This strategy can efficiently find problems observed quantity and gross error.2. Based on the traditional un-differenced real-time precision clock correction estimation method, the author developed a double-process parallel processing algorithm that the clock module runs a slow process and a fast process synchronously, and the former estimates all the satellite and receiver clock correction, the troposphere, and the ambiguity parameters, while the fast process fixes the ambiguity and troposphere parameters to the latest estimations of the slow process, with the satellite and receiver clock parameters estimated only. An experiment was carried out to estimate the quad-system real-time precise clock correction based on more than 60 MGEX reference stations. It is shown that the single epoch processing time of the slow process is<15 s, while that of the fast process time is<2 s satisfying the real-time demand. The dissertation has also studied how the relationship between the accuracy of clock correction and the parameterization strategy of inter-system bias, from which the optimal parameterization strategy of inter-system parameter is determined. To prevent the interruption of real-time service because of products discontinuity, a short-term clock prediction algorithm has been proposed by modelling the linear fitting residuals of the estimated clock. The test shows that the prediction model fitted from 3 h real-time clock products can contribute to predicted clock product as accurate as-0.3 ns,-0.4 ns and-0.5 ns for difference prediction time of 1 h,2 h and 3 h, respectively. This improve the stability of the system services dramatically.3. Studied the theory of satellite phase hardware delay (Fractional Cycle Bias, FCB) estimation and PPP ambiguity resolution algorithms. Analysis of different FCB estimation algorithms confirms their equivalence. The un-differenced single station was realized and improved in the mode of single station ambiguity fixing & FCB residuals network adjusting, which supports various input of ambiguity observations. The IGS MGEX global network and Chinese CMONOC network were used to estimate the GPS/BDS FCB products respectively. The wide-lane FCB of GPS and BDSIGSO, MEO satellites are rather stable, with the days variation<0.4 cycles, and daily mean variation<0.1 cycles, while the narrow-lane FCB products vary in 0.5 cycles, and daily mean changes largely between days. Most of the ambiguity residuals distribute between-0.25-0.25 cycles, with 95% confidence interval in ±0.1 cycles, which verify the consistency of the FCB products. Tests of PPP ambiguity resolution were carried out using these FCB products, i.e. the GPS-only, BDS-only and GPS/BDS combined PPP modes, of which the ambiguity fixing contributes to improvements of 1.65%-30.47%,7.43%-36.99% and 5.21%-43.53%, respectively.4. Proposed the regional zenith tropospheric delay and slant ionospheric delay modeling method by using the PPP-AR to retrieve the ZTD and STEC of each reference station which are used to interpolate the corrections of real-time users. Tested on different scales of network, the proposed algorithms exhibit ZTD interpolation accuracy of?1 cm RMS, STEC interpolation accuracy of-0.2 TECU STD for local CORS network, and-2 cm for ZTD and-0.5 TECU for STEC for wide-area network. A short baseline was used to compute the relative ZTD and double-differenced STEC by the PPP-AR method and traditional RTK method which are compared to each other. The ZTD difference of the two methods is<1 cm and the STEC difference is<0.1 TECU, which confirms the consistency between the PPP-AR method and double-difference method.5. Designed a compatible wide-area and local-area real-time precise positioning service system which estimates the real-time precise positioning products and customized to generate the corrections according to the positioning mode and precision demand of different users. The theory works of this dissertation has been implemented based on the PANDA software, forming a set of real-time GNSS precise data processing software.