The Theory And Application On Multi-frequency Data Processing Of GNSS 2

The most prominent characteristic of GNSS 2 is using the multi-frequency observations. The application of multi-frequency in GNSS 2 will give the traditional dual-frequency navigation and positioning model great changes. For example: the higher-order terms of ionospheric refraction error can be corrected through using multi-frequency observations; it shows more advantages than dual-frequency in such as wavelength, ionospheric refraction error and etc; the ambiguity can be quickly fixed by using the relationship between multi-frequency combinatorial wavelength and observation errors. In this dissertation these problems are deeper researched dealing with the multi-frequency observations data. The main studies and contributions are described as following:1 Compared different traditional models of atmospheric refraction error via examplesThe correction models of tropospheric refractive error are calculated and compared respectively such as Marini, Chao, CFA, Mit, Mtt, Hopfield and etc. The discrepancy of these models in different elevation angels and elevations are analyzed, the result shows that the discrepancy in different stations is related with the station's height, and this discrepancy in different models is decreasing with the elevation angel increased. For the ionospheric refractive error, the Klobachar model is adopted to make a single-frequency correction in different elevations, the result shows the difference of ionospheric delay correction between stations is mainly related with difference of elevation angles from stations to the satellites, in the other word is related with the geometric forms from satellites to stations.2 Higher-order terms correction of the ionospheric refraction errorThe main effect on the GPS signals of the ionosphere and the ionospheric refractive model are researched in this dissertation; then the dual-frequency correction model of ionospheric refractive error is concluded. Aiming at the third frequency added in GPS modernization, the triple-frequency correction model of ionospheric refractive error and the ionosphere-free combination model of triple-frequency is deduced. The second-order ionospheric effect can be corrected and the positing accuracy can be increased through those models. The results show that, when the electron density is too much(TEC=1.38e18) those models can decrease the ionospheric error to 5mm while dual-frequency ones only decrease to 5.7cm, and commonly these models can be to the level of inferior mm while dual-frequency correction is only 1.8cm. These all indicate the necessary of correction the higher-order terms of ionospheric error in high precision positioning.3 Researched the various kinds of combination patterns through the multi-frequency data, and analyzed the characteristics of those combinationsIn carrier observation equations, some observation errors are usually eliminated effectively via the linear combination due to the pertinence of each other, and in this way the success rate in resolution of the integer ambiguity can be improved in high precision positioning. Soon the basis of the definition of the multi-frequency phase combination, the error impact after combination is analyzed in detail. Aimed the error characteristics, the choosing standard of combination observation in the aspect of wavelength, ionospheric error and etc are presented, and a series of typical combination are listed. All work above establishes a theoretical basis to calculate ambiguity used by multi-frequency.4 The positioning precision of combination observation is researched theoretically and its concrete mathematic model is deducedSetting out from the angle of using a set of combination observation singly in navigation and positioning, the positioning precision of combination observation is analyzed theoretically. From the variance-covariance of multi-frequency carrier phase combination's observation equation, every observation error expression is deduced one by one via the propagation of errors, and the impact of positioning precision is analyzed with the changing of observation errors. When using only one carrier phase combination into high precision positioning, a method to get the optimal combination is brought forward, and validated via example.5 Research on cycle-slip and gross error detection by using multi-frequency dataAs the great random changes of the clock error, atmospheric refractive error and the interval factor of epoch rate, small cycle-slip can hardly be detected and repaired by using original phase observation sequence. So, how to repair the cycle-slip and gross error of original data quickly and accurately is a momentous problem. In this dissertation some characteristics of the multi-frequency are used to detect and repair the cycle-slip and gross error through the simulated data. The result shows that any cycle-slip can be repaired thoroughly at single epoch by using multi-frequency data in bigger interval situation; as to the gross error, which exceed 0.2 cycle will be detected, and partly can be repaired at single epoch.6 Research on the multi-frequency data simulationBy now we can't get the multi-frequency observation data, the numerical value simulation must be done before validate all kinds of multi-frequency ambiguity resolution methods. Since the ionospheric refraction error is related with the frequency, and the tropospheric refraction error, the clock error of the satellite and receiver are equality at the same epoch between different frequencies, so we can use those errors' characters to make data simulation of the original L5 observations in GPS modernization. Through the different types of observation error models the double difference L5 observations is simulated. Then we resolved the ambiguity by using simulated data via the method of matrix transformation, and compared the result with software TGO, it shows the feasibility of the simulation method researched.7 Research on the fast integer ambiguity resolution algorithm of different distance baseline by using multi-frequency combination observationsAimed at the deficiency in ambiguity resolution quickly by traditional GPS dual-frequency observations, the theoretical deduction to solute ambiguity in single epoch and validation viaexamples are made on the basis of the multi-frequency combination observation theory. This dissertation focuses on the research of TCAR/MCAR method. Aimed at the disadvantages such as the ideal hypothetical condition, and few combined observation used in TCAR/MCAR, on the case of other combinations are deeper studied on the basis of multi-frequency combination theory. And the influence to the success rate of ambiguity resolution is researched while adding all observation errors. Then on the above works we expand the method of TCAR/MCAR, and abundant examples and comparisons are carried out on this method. Besides, considering the ionospheric error as the main obstacle to the ambiguity resolution in long baseline, the attempt of ionospheric correction is drawn with multi-frequency observations. The result shows that, the integer ambiguity of each frequency can be solved in single epoch when the baseline is less than 10km, that means we can attain the aim to the real-time navigation and positioning with multi-frequency observations; as to the middling baseline, the ambiguity can be get directly by using the "cascade relationship" between the observation noise and a series of combination observations which have the long wavelength, weak ionospheric error and low noise. This method can obtain more higher success rate without the complicated ambiguity search; as to the long baseline, some weak ionospheric error, long wavelength observations' ambiguities can be get firstly by using the theory of TCAR/MCAR, then combined them with the non-ionospheric combinations to get the ambiguity of the single frequency, then attain the aim of navigation and positioning quickly.8 Software development of multi-frequency data processingOn the base of multi-frequency data processing theory and research of those algorithms, the software of multi-frequency data processing is developed. This software is formed with data simulation, data pretreatment and baseline processing, and its functions offer the technology flat for deeper research of multi-frequency.9 Analyzed an example of GPS network using multi-frequency observationsUsing the software developed, we analyzed the data processing result of a GPS network in the three aspects: the detection of gross error, processing of short baseline and long baseline. It shows that the work done in this dissertation attains a well effect.