Research On Assessment Of Signal Quality And Methodelogy Of BDS/Galileo/GPS Precise Positioning

In recent years,each navigation satellite system has developed rapidly,and GPS and GLONASS are rapidly modernizing,while BDS and Galileo are accelerating the launch of global satellites.The availability and reliability of PNT could be significantly increased via the combination of multiple GNSS.Therefore,this thesis aims to evaluate the quality of new satellites and signals and study the algorithm of multi-GNSS precision positioning with special focus on estimation of inter-system biases and three-carrier ambiguity resolution.The main contribution of this thesis includes:1.The function model and statistical model in multi-GNSS multi-frequency precision positioning are deduced,and linear combinations of the original multi-frequency observations are investigated.Then all kinds of biases in GNSS model are classified,and the filtering model and parameter estimation method over different length of baseline are discussed.Similarities and differences between the three systems of BDS/Galileo/GPS are compared in terms of orbit design,signal modulation method,time system and coordinate system.Based on the original design and construction experience,both Galileo and BDS-3 have designed their own improved BOC(Binary Offset Carrier)signals and adopted some new technologies.2.A complete signal quality assessment method for BDS/Galileo/GPS new satellites and new signals is proposed,including carrier-to-noise ratio modeling,code multipath modeling,phase multipath modeling and double-difference residual algorithm.Galileo E5 has apparant advantages in terms of signal strength and multipath suppression performance,which may be related to its advanced modulation scheme.Being free of the systematic code errors of BDS-2,the RMS of BDS-3s MP errors are comparable with that of GPS and Galileo.Both the code and phase multipath errors show elevation-dependent and periodic characteristics.The accuracy of Galileo single point positioning is comparable to that of GPS,but better than that of BDS.Although there are no significant differences for the RTK results of the three systems,the double-differenced carrier phase and code residuals of E5 is the smallest among all the signals.3.Two methods for estimating inter-system biases(ISB)with and without constraints are proposed.Both the inter-system phase(ISPB)and code(ISCB)biases have relatively stable characteristics in a certain time domain.ISBs are related to the receiver type,firmware version and the selected overlapping frequency.Upgrade of receiver firmware version results in changes of ISB values.The effects of receiver firmware upgrade on ISPB and ISCB are also different.In the ISB estimation method without constraints,three time –varying models are adopted including the white noise model,random walk model and random constant model.The ISB value determined based on the random constant model is most consistent with the ISB value estimated with the constraint condition.4.The inter-satellite-type biases(ISTB)between BDS GEO and non-GEO satellites are investigated.It can be found that the ISTB have effects on the estimation of ISPB,which causes the difference of 0.5 cycles ISPB between GEO and IGSO satellites.Affecting by ISTB,both the single signal B2 and loosely combined B2/E5 b cannot obtain the correct integer ambiguity.After correction of ISTB,the single signal B2,loosely and tightly combined B2/E5 b all can obtain high precision positioning result.5.Aiming at the problem that the tightly combined positioning is susceptible to ISB,the method of using the determined ISB value to correct observations is studied.Each system selects their own reference satellite in the loosely combined positioning,while the common reference satellite is selected in the tightly combined positioning.After the calibration of ISBs,the positioning performances of tight combination are assessed with the single system and loose combination.The success rate of ambiguity resolution and accuracy of positioning for the tight combination are significantly improved in comparison with that for the loose combination over short baselines.For L5/E5 a,on which only few satellites can be observed,the maximum increase in success rates of ambiguity resolution can reach larger than 30%,i.e.,from smaller than 60% of loose combination to larger than 80% of tight combination.6.Using data of the globally distributed MGEX stations,a systematic and comprehensive evaluation of signal quality and positioning performance for BDS-3 global system is conducted with regard to carrier-to-noise ratio,code noise and multipath,ephemeris error,TGD and BDS-2/BDS-3 ISB.At the same frequency point such as B1 I and B3 I,the signal power of BDS-3 satellites is apparently higher than that of BDS-2 satellites.The new signal B2 a of BDS-3 is superior to other signals in regard to signal strength and multipath suppression performance.The ephemeris on almost every satellite of BDS-3 has large gross errors,and the TGD values also show abnormal changes.There is a consistent inter-system bias between BDS-2 and BDS-3 in different stations.Affected by the above factors,the BDS-2/BDS-3 combined single point positioning is not significantly improved in comparison to the BDS-2 single system.7.A modified three carrier ambiguity resolution(TCAR)method to improve the narrow-lane(NL)ambiguity resolution over medium-long baseline is presented,in which the estimated double-differencing(DD)ionospheric delay derived from the Kalman filter floating solution is adopted to modify the floating NL ambiguities.To further improve the NL ambiguity resolution,smoothing algorithm is applied to refine the estimated DD ionospheric delay.Modified by the smooth DD ionospheric delay,the NL ambiguity residuals are mostly within 0.5 cycles over medium-long baselines,showing a significant improvement in contrast to the classical TCAR method.8.In order to realize high-precision and fast positioning,a partial ambiguity resolution algorithm based on the modified TCAR method is proposed.Firstly,by introducing the "fix and hold" mode into the three-carrier ambiguity resolution,the satellite elevation is used to determine the ambiguity subset.Considering the fact that even one cycle in EWL(0,-1,1)or(1,4,-5)may lead to very large cycles in NL(4,0,-3)due to the small wavelength of NL,the EWL combinations are restricted to guarantee the success rate of NL ambiguity resolution.Results show that,the positioning performance of the modified TCAR method is comparable to that of the LAMBDA method,and high-precision positioning can be achieved in a very short time.