Research On Multipath Error Correction Of BDS/GNSS Precise Point Positioning

The multipath effect reduces the accuracy of pseudorange and carrier phase observations,significantly affects the convergence of GNSS precise point positioning(PPP)in an urban high-obstruction environment.In recent years,multipath error correction method based on spatio-temporal repeatability has become a research hotspot in academia and industry because of its outstanding advantages in complexity and real-time.Compared with the traditional time-domain repeatability,the multipath spatial-domain repeatability method has the advantages of simple algorithm and easy implementation.As typical spatial repeatability correction methods,the multipath hemispherical map(MHM)and the MHM based on trend-surface analysis(T-MHM)have carried out a lot of research.At present,researchers mostly discuss the application of MHM and T-MHM in differential mode.Take the above into consideration,in this dissertation,we makes a systematic and in-depth study on the key technologies of PPP multipath correction.The main work and contributions of this thesis are as follows:1.Based on Allan variance theory,the characteristics of PPP residuals are discussed in detail,and the feasibility of PPP multipath error correction is analyzed.The results show that multipath error accounts for the main component in PPP residuals.Starting from the conclusion of PPP residual characteristics,the time-domain and spatial-domain repeatability of multipath errors of GPS,BDS and Galileo satellites are discussed respectively.Although the three systems all use MEO satellites,their orbital repetition times are different.2.Taking GPS satellite as an example,MHM and T-MHM methods are applied to PPP mode to verify the feasibility of PPP multipath error.We mainly analyze the unmodeled errors,evaluate the multipath correction effect of T-MHM,discuss the optimal modeling days,and test its sensitivity in the scale of the sky grid.The test results show that compared with MHM,T-MHM can effectively correct the highfrequency multipath error by trend surface fitting,and T-MHM has high robustness to PPP solution,which can effectively increase the grid size during modeling to improve the computational efficiency.3.Considering the particularity of BDS-3 orbital type,new signal,and traditional signal,we discuss the impact of two different multipath modeling and correcting methods on the PPP positioning results of new and traditional signals.The experiments showed that the positioning results of the IGSO and MEO satellites,modeled and corrected separately using the multipath,were slightly better than the results obtained when they were modeled and corrected together.And,the positioning accuracy and convergence time of B1C/B2 a before and after multipath error correction were slightly better than those of B1I/B3 I,which is related to the anti-multipath performance of BDS-3 new and traditional signals.4.We propose a multi-GNSS PPP multipath fusion modeling method.In multipath error modeling,the GNSS inter-system bias(ISB)and the optimal modeling days are comprehensively considered.When the multipath error is corrected,the system is not distinguished,and only the model value of the corresponding sky grid needs to be searched.Three compatible and interoperable multi-GNSS system combinations are selected,namely GPS+BDS-3,GPS+Galileo,and GPS+BDS-3+Galileo,the optimal modeling days of the corresponding combinations are given,and the improvement of positioning performance after correcting multipath by different combinations is evaluated,respectively.The test results show that compared with MHM,T-MHM method can improve the positioning accuracy of GPS+BDS-3 by 26.4% and shorten the convergence time by 39.4%.The positioning accuracy of GPS+Galileo can be improved by 33.2%,and the convergence time can be shortened by 29.4%.The positioning accuracy of GPS+BDS-3+Galileo can be improved by 24.6% and the convergence time can be shortened by 35.4%.