Research On The Key Technologies And Methodologies Of Uncombined PPP-RTK

With the continuous development of the Global Navigation Satellite System(GNSS),multi-frequency GNSS observations have significantly improved the positioning performance of precise point positioning(PPP),which has been widely used in the field of low-orbit satellite determination,GNSS meteorology and geodynamics,and other fields.However,compared with real-time kinematic(RTK)technology,although the PPP ambiguity resolution(AR)has been well resolved,the long convergence time is still the technical bottleneck limiting the practical application of PPP,which makes PPP unable to meet the requirements of effectiveness,accuracy,and reliability of position services in the intelligent era.PPP-RTK integrates the advantages of PPP and RTK technologies while compensating for their respective technical shortcomings.With the increasing density of ground-based augmentation reference networks supporting multi-frequency GNSS observations,uncombined PPP-RTK is providing a new solution to support multi-frequency GNSS observations fusion and achieve fast and precise positioning due to its simple model and strong scalability.In this paper,to meet the urgent requirements for fast and precise positioning,the key technologies of uncombined PPP-RTK are studied systematically and thoroughly,focusing on the key issues of uncombined PPP ambiguity resolution,multi-frequency GNSS PPP-RTK,PPP-RTK considering prior ionospheric constraints,and singlesatellite STEC modeling and uncertainty determination,and the PPP-RTK data processing software platform based on the uncombined PPP model is developed.With a large number of real observations,the correctness and reliability of the algorithm and data processing software are effectively verified.The main work and contributions of this thesis are as follows:1.The uncombined PPP mathematic model,error sources,data pre-processing and quality control strategies,and parameter estimation are systematically described.The AR and validation methods are summarized in detail.To resolve the shortcomings of PPP AR utilizing the satellite with the highest elevation angle as the reference satellite in the same system(satellite’s elevation angle selection strategy,SE),a PPP partial ambiguity resolution(PAR)method combining data quality control and GramSchmidt(GS)is proposed.To validate the effectiveness of the proposed AR method,the performance of PPP AR and positioning can be evaluated.The experimental results show that the average epoch fixing rates of each day and each station are improved by 7.7% and 11.5% by the GS method,compared with the SE method in static PPP solution.The corresponding improvements of 7.9% and 7.8% can be achieved,respectively,in the pseudo-kinematic PPP solution.Similarly,in the static PPP solution,the average time to first fix(TTFF)of the GS method for each day and each station is improved by 22.3% and 25.4%,respectively,compared with the SE strategy.The 20.4% and 19.7% of improvement can be obtained from pseudokinematic PPP mode,respectively.In terms of the evaluation of the convergence time of multi-GNSS positioning performance,compared with the SE method,the GS method improved by 25.2%,13.2%,and 17.1%,respectively in terms of the horizontal convergence time of the BDS alone,GPS+BDS(GC)and GPS+Galileo+BDS(GEC).The vertical convergence time improved by 18.1%,18.4%,and 14.7%,respectively.In terms of positioning accuracy of multi-GNSS positioning performance,compared with the SE method,the GS method improved by 0.5%,22.5%,and 13.8% in the horizontal positioning accuracy of BDS alone,GC,and GEC,respectively,while the positioning accuracy of vertical direction can be improved by0.1%,3.8%,and 3.9%,respectively.2.The PPP-RTK model based on the multi-GNSS combined observations is established.The ambiguity resolution and positioning performance of both BDS+GPS PPP-RTK and Galileo+GPS PPP-RTK are assessed by utilizing the actual measurements.The experimental results demonstrate that compared with GPS PPPRTK,the convergence time of BDS+GPS PPP-RTK is improved by 50% on average.In terms of the accuracy evaluation of the Galileo+GPS PPP-RTK ionospheric correction,the proportion of GPS ionospheric residuals in the range of ± 0.30 TECU is95.9% and 85.2% derived from the small-and medium-reference network,respectively,while the proportion of Galileo is 82.8% and 81.7%,respectively.For the PPP-RTK convergence time,under the 90% confidence level,the horizontal convergence time of GPS alone is 2.0min and 2.5min for small-and medium-scale networks,respectively,and 2.0min and 3.0min of convergence time can be achieved for the vertical direction,respectively.In terms of Galileo+GPS,the 1.5min and2.0min of horizontal convergence times can be obtained for small-and medium-scale networks,while the 1.5min and 2.5min convergence time for the vertical direction,respectively,at a 95% confidence level.3.The GPS triple-frequency(TF)PPP-RTK model based on different AR strategies is established.The improvements of four PPP AR methods combined GramSchmidt(GS)and satellite’s elevation selection(SE)strategy with GPS dualfrequency(DF)and triple-frequency(TF)measurements are analyzed in detail,which includes the performance of ambiguity fixing,and PPP-RTK positioning.The experimental results represent that ambiguity fixing based on the GS strategy has better performance than that of the SE method in terms of the epoch fixing rates and TTFF.In general,triple-frequency GPS PPP-RTK based on the GS strategy shows the best positioning performance in terms of both positioning accuracy and convergence time.4.Due to the problem that the ionospheric delay includes receiver code biases(DCBs)in the uncombined PPP-RTK model,which affects the interpolation accuracy of slant ionospheric delays model,an uncombined PPP-RTK with the prior ionospheric constraints is proposed.Based on the uncombined PPP-RTK mathematical model utilizing multi-frequency GNSS measurement,an uncombined PPP-RTK model considering prior ionospheric constraints is developed by introducing the constraints condition of both external ionospheric products and accuracy.The prior ionospheric constraints not only enhance the model strength of the PPP-RTK server but also unifies the datum contained between the server and the user,avoiding causing systematic errors in user parameter estimation.Therefore,the accuracy of ionospheric delay extraction,and single-frequency(SF)and multifrequency(MF)PPP-RTK are evaluated to verify the advantages of PPP-RTK models with ionospheric a priori constraints.The experimental results show that in terms of evaluating the accuracy of ionospheric extraction,a significant improvement effect on the accuracy of ionospheric delay extraction can be obtained by the ionospheric prior constraints strategy.In the evaluation of SF and MF PPP-RTK,the accuracy of ionospheric correction based on dual-and triple-frequency observations can be improved by 65.4% and 66.4%,respectively.For the convergence time assessment of MF PPP-RTK,the PPP-RTK considering the ionospheric-constraints improves the convergence time of horizontal components by 84.0%,83.3%,87.9%,and 84.2%compared with PPP-RTK without ionospheric constraints,in the static DF,static TF,pseudo-kinematic DF,and pseudo-kinematic TF,respectively,while that of vertical components can be reduced by 80.0%,76.9%,76.4%,and 71.4%,respectively,at the95% confidence level.For the positioning accuracy assessment of MF PPP-RTK,compared with PPP-RTK without ionospheric constraints,the positioning accuracy of horizontal components of ionospheric-constrained PPP-RTK are improved by 31.8%,32.3%,45.1%,and 42.5%,in the static DF,static TF,pseudo-kinematic DF,and pseudo-kinematic TF,respectively,at the 95% confidence level.The positioning accuracy of vertical components of ionospheric-constrained PPP-RTK are improved by 33.1%,31.3%,48.4%,and 41.7% in the four solutions,respectively,compared with PPP-RTK without ionospheric constraints.5.For the problem that the size of the data broadcast is large and susceptibility to station spacing,a STEC single-satellite polynomial model and STEC uncertainty determination method are proposed.Using the measurements to evaluate the modeling accuracy,precision,and positioning performance of the STEC single-satellite polynomial model,and evaluate the PPP-RTK positioning performance considering the precision,the average of accuracy,and Gaussian approximation function as the ionospheric uncertainty.The experimental results show that,in terms of the accuracy evaluation of ionospheric TEC modeling,the mean precision of both the STEC modeling of order 5 and 9 based on the IPPs as the reference and the STEC modeling of order 5 and 9 based on the ground point as the reference are 9.55 cm,27.72 cm,1.44 cm,and 3.00 cm,respectively,the average of the corresponding modeling accuracy are 3.92 cm,2.85 cm,4.27 cm,and 2.91 cm.In terms of the assessment of ionospheric uncertainty,the RMS of the residual errors based on the precision,the average of modeling accuracy,and the Gaussian approximation of modeling accuracy are 11.35 cm,1.91 cm,and 0.62 cm,respectively.The precision as the ionospheric uncertainty of PPP-RTK,and the positioning accuracy in the east,north,and vertical directions are 13.00 cm,10.69 cm,and 19.04 cm,respectively.The average of the modeling accuracy as the ionospheric uncertainty of PPP-RTK,and the positioning accuracy of the corresponding direction are 4.82 cm,5.43 cm,and 12.22 cm,respectively.The Gaussian approximation of modeling accuracy as the ionospheric uncertainty of PPP-RTK,and the positioning accuracy of the corresponding direction are 0.25 cm,0.46 cm,and 4.16 cm,respectively.