Research On Theory And Methodoloy Of Kinematic Postioning Based On Continuous Operatioal Reference Stations

Technology based on continuously operating reference stations is also known as network real-time kinematic positioning technology or multi-reference stations kinematic positioning technology, which is a new high-precision kinematic positioning technologies based on other technologies such as conventional kinematic positioning, communications, networking and computers in recent years. Network real-time kinematic technology is one of the trends of the satellite navigation and positioning technology development. This thesis systematically studies the kinematic positioning and programs the algorithm based on continuously operating reference stations. Extensive example results verify that the theories and methods involved in this thesis are correct. the main work and innovation are summarized as follows:1,The observation model, the error model and the stochastic model for satellite positioning are studied and summarized. The observation equation and linear equation of raw observation data and their combinations were given and the characteristics of some commonly used combinations are analyzed. The definition, characteristics, affect value and correction of systematic errors and accidental errors in satellite relative positioning are analyzed. The systematic errors include satellite ephemeris error, ionospheric delay, tropospheric delay; accidental errors include multi-path effect, observation noise. The observation stochastic model with simple form using the single-difference matrix is derived.2,The mathematical model of dynamic positioning based on continuously operating base station is studied. The dynamic positioning calculation is divided into three main processes: baseline solution between base stations, constructing observation data of the virtual reference station and positioning calculation of moving station users. The ambiguity resolution of medium-to long-range baseline and the interpolation algorithm of error correction data are two vital factors of dynamic positioning. On this basis, the common mathematical model of dynamic positioning has been improved: First, the synthesis error processing has been improved into the single-difference Ionosphere delay between stations related to satellites and the troposphere delay related to base station. It is beneficial to separate and refine different physical properties and construct successive observation data of virtual reference station; second, double-difference ambiguity parameters has been changed into single-difference ambiguity between two stations. It is beneficial that positioning calculation of moving station users become smoother when switching reference satellite so that the positioning result become more continuous. However, this improved mathematical model will increase the number of parameters to be estimated. 3,This thesis analyses the cycle-slip detection and repair of GNSS observation data. Analyzed several practical methods the combination method using ionosphere residual error and TurboEdit is designed for cycle-slip detection and repair of kinematic positioning. Accuracy, applicability, and dependability of this method are analyzed. The example with reference station data shows that this method can detect and repair cycle-slip more than 0.5 cycle.4,The analysis method of the quality of reference stations observation data is studied. The observation noise and multi-path effect of GNSS observation belong to accidental errors, the influence of which on positioning can not be neglected but hard to be separated. To solve this problem, a experimental method of receiver noise analysis is designed, and the mathematical model of receiver noise estimation is derived, the example results show: The magnitude of the receiver noise is related to the elevation angle. Pseudo-range observation noise and carrier phase observation noise have a relationship of certain percentage. An analysis method of carrier-phase measurement multi-path effects is proposed. The experimental method is designed, and multi-path solution model is deduced. Actual observation data is used to analyze the difference and relation of multi-path effects among different types of measurement or measurement at different frequency points, the relation between multi-path effects and satellite elevation angle is analyzed.5,This thesis uses the Kalman filter as a kinematic positioning parameter estimation method, and derives the mathematical model of Kalman filter used for kinematic positioning data processing, furthermore, gives the solution though the observation model is non-linear and the estimated parameters change frequently. Next, the forward and the bi-directional filter are applied in kinematic positioning based on continuously operating reference stations, the former one can be used for real-time positioning and the latter can be used for high-precision applications. The performance of the forward filter and the bi-directional filter are compared by the results of calculative experiments. The parameter estimation method taking the known coordinates of base station into account is proposed which fix the base station coordinates when parameters are estimated. The results show that this method can improve the success rat of the ambiguity solution and the proportion of the fixed solution in the relative positioning data processing of the medium and long distance, and this method is correct and valid. Besides, the adaptability of this algorithm is analyzed in this thesis.6,In this paper ambiguity solution is divided into two main steps: a real ambiguity estimation and integer ambiguity search. The common integer ambiguity search method are described. and the characteristics of various methods are analyzed. Taking into account that the LAMBDA method is more mathematically rigorous, takes full advantage of the variance-covariance matrix, provides information to provide the optimal Ambiguity unbiased estimate,and optimize the search space and can even improve the search efficiency, this paper adopts the LAMBDA method in ambiguity search and RATIO value as a confirmation.7,Building measurement of virtual reference station consist of two main steps. One is estimation of ionospheric and tropospheric delays,the other is error interpolation (of virtual reference station) . Deterministic model as well as stochastic model of ionospheric and tropospheric delays are analyzed, and then a procedure considering ? 2testing residuals of observables for ambiguity fixation is proposed. Experimental result indicated that when changes of tracked satellite number cause conditions of ambiguity fixing change,this procedure can solve the uncontinuous estimation of ionospheric and tropospheric delays.Experimental result indicated that the maximum difference between interpolation of ionospheric delay and parameter estimates is 2 cm. For tropospheric delays, this difference is 0.8. In experiment, the ionospheric and tropospheric delay estimation algorithms as well as interpolation algorithm are verified to be correct.8,The mathematic model is derived. The differences between base stations relative positioning and rover user kinematic positioning are compared and analyzed in application requirement, observational condition, observation model and kinematic model. The solution program of rover user positioning is made. And then the mathematic model of rover user kinematic positioning is validated through the example of kinematic positioning with regard to different vehicle and different distances.9,Considering the characteristic of observational data of different carriers, numbers of data proceeding strategies are analyzed in kinematic positioning. The correlation between data sampling rate and vehicle dynamic performance, and reducing data interval is proposed in data proceeding of moderate and low user kinematic positioning. The positioning results of single and dual frequency data are compared, and to moderate and low kinematic user in the good observational condition, the accuracy of centimeter is reached if dual-frequency receiver is replaced by single-frequency receiver. The kinematic positioning results are researched when the height difference between base station and rover station are too large, and the tropospheric delay must be estimated when the distance is too far or height difference is to large. The effects to positioning results of different resolution mode are researched and analyzed. The number of unknown quantity should be reduced in kinematic data proceeding, that is beneficial to advance success rate of solution and fixed solution.