| Precise point positioning(PPP)was developed in the late 20th century,which determines the position of user in real-time or post processing modes by code and phase measurements.PPP has been widely used for scientific research and civilian applications.Currently,the satellite navigation world is facing dramatic changes with the rapid development of multi-constellation Global Navigation Satellite Systems(GNSS).The positioning accuracy and reliability can be significantly improved with GNSS combination.To shorten the convergence time and improve the positioning accuracy,PPP ambiguity resolution(AR)technique has been proposed and developed in recent years.Although remarkable progress has been achieved for GPS ambiguity resolution,GPS PPP AR still suffers from the problem of a long time to fix the first ambiguity.Multi-GNSS fusion can significantly shorten the PPP convergence time and improve the positioning accuracy.In this contribution,we developed the multi-GNSS(GPS/BDS/GLONASS/Galileo)UPD estimation model and a comprehensive analysis of the ambiguity resolved PPP with four-system observations was made.The main work and contributions of this thesis are as follows:The Multipath(MP)combination series of BDS GEO,IGSO and MEO satellites was calculated and analyzed.The satellite-induced code bias was identified to exist in BDS-2 code observations,which affects the precision and consistence of the WL ambiguity derived from MW combination.For IGSO and MEO satellites,the variation of this error reveals elevation-dependent and can be corrected with our proposed weighted least-squares method.Besides the satellite-induced code bias,the GEO observations are also affected by the multipath of the station seriously.To mitigate the effect of satellite-induced code bias and multipath of GEO observations,the sidereal filter is used to extract the systematic variation superimposed in code observation and apply it back as correction.The MP series of BDS-3 satellites was also investigated and results demonstrated that the satellite-induced code bias is negligible for the BDS-3 satellites.With observation data acquired from three tracking networks including Hong Kong,CMONOC,and MGEX,the spatial-temporal characteristic of estimated BDS UPDs are analyzed.The estimated WL and NL UPDs during the 10 days from DOY 024 to DOY 033,2016 are quite stable with a STD of less than 0.15 cycles.The STDs of WL UPDs are less than 0.06,0.03,0.07 cycles for Hong Kong,CMONOC,and MGEX networks,respectively.The average usage rates of all satellites for WL UPDs are 88.67%,94.97%and 78.33%for Hong Kong,CMONOC,and MGEX networks.We found that that WL UPDs estimated by CMONOC stations show the best temporal stability with a highest usage rate among three networks,which may benefit from its unified receiver type,denser distribution than MGEX and longer visible arc than Hong Kong network.For NL UPDs,the STDs range from 0.006 to 0.13 cycles for all BDS satellites and are on average 0.037,0.052 and 0.058 cycles for Hong Kong,CMONOC,and MGEX networks,respectively.Different from the situation of WL UPDs,the NL UPDs derived from Hong Kong network are more stable than two other networks.It is demonstrated that the temporal stability of NL UPDs together with the usage rate and residual distribution derived from a smaller network are generally better than those of global network because the orbit errors and atmospheric residual errors can be absorbed into the NL estimates in a small network.Based on high-quality UPD products,the BDS-only PPP ambiguity resolution with GEO/IGSO/MEO satellites is achieved.As for BDS-only static PPP AR,the average TTFF of Hong Kong and CMONOC stations are appropriately 40 min with the fixing percentage of 91.30%and 76.40%,respectively.The average TTFF of MGEX stations is 57.4 min,which is longer than that of CMONOC and Hong Kong stations,and the fixing percentage is 68.99%.After the ambiguity resolution,the positioning accuracy of Hong Kong daily PPP solutions is 0.72,0.54,and 3.21 in the east,north and vertical directions,with improvements of 58.6%,50%,and 41.8%.The positioning accuracy of CMONOC is improved from(2.24,2.31,5.64)cm to(1.18,0.79,3.30)cm by ambiguity resolution in the east,north and up directions,and from(2.71,1.80,6.00)cm to(1.58,1.15,4.33)cm for MGEX.For kinematic PPP,the TTFF of Hong Kong,CMONOC and MGEX are 49.0,48.3 and 73.4 min,and the fixing percentage are 84.44%,63.50%and 60.17%,respectively.The positioning accuracy of Hong Kong fixed solutions is(1.22,0.95,4.37)cm,with the improvements of 80.10%,59.57%and 63.64%in the east,north and up directions compared to the float solutions.The improvement of CMONOC is 43.85%,66.67%and 56.96%in east,north and up directions,and 40.41%,34.33%and 37.17%for MGEX.The multi-GNSS UPD estimation model was developed and the multi-GASS PPP ambiguity resolution was achieved.A dataset of 30 days from DOY 001 to 030 of 2017 with a tracking network consisting of about 148 MGEX/IGS stations was used for GPS UPD estimation.The mean STD of the 30-day WL UPDs is 0.023 cycles while the mean STD of NL UPDs is 0.04 cycles.A global tracking network containing 67 MGEX stations was used to estimate BDS UPDs.The impact study of satellites-induced biases for BDS resulted in a significant improvement of WL UPDs,particularly for IGSO and MEO satellites(>75%).Minor improvement was observed for NL UPDs(<25%).Besides,the WL UPDs estimated from the CMONOC and Hong Kong CORS network were also improved after the code bias correction,especially for MEO satellites(>70%).No obvious improvement was found for NL UPDs of regional network.With a network of homogeneous receivers,the GLONASS UPDs were estimated with three commonly used receivers(TRIMBLE NETR9,JAVAD TRE G3TH DELTA and LEICA),respectively.The stable WL UPDs with the mean STD less than 0.06 cycles and NL UPDs with the mean STD less than 0.11 cycles can be estimated with all three types of receivers.Results demonstrate that the UPD values of inhomogeneous receivers are different each other,however,different firmware versions seem to have no effect of the UPDs for the same type of receiver for TRIMBLE NETR9 or LEICA receivers.Global and European networks were applied for the estimation of Galileo UPDs.Both two networks can estimate stable WL UPDs with the mean STD less than 0.02 cycles and NL UPDs with the mean STD less than 0.11 cycles.The GREC UPD results were then applied to multi-GNSS PPP AR and the performance of the GCRE combined PPP AR was also investigated and compared with single-system(G)and dual-system(GR,GC,GE).Compared to the traditional float solutions,PPP AR can shorten the convergence time and improve the positioning accuracy.The combined GCRE PPP AR results present the fastest convergence and the highest accuracy for all three coordinate components compared to single-and dual-system PPP AR solutions.When the cut-off elevation angle is increased to 30°,the convergence time for GPS-only and GE PPP float solutions increase to 50.68 and 47.18 min while that for GR,GC and GCRE is still less than 30 min.Average TTFF of static PPP AR solution with 7° cut-off elevation angle for GCRE is 9.21 min,which is much shorter than that for GPS(18.07 min),GR(12.10 min),GE(15.36 min)and GC(13.21 min).When the cut-off elevation angle is increased to 30° from 7°,the TTFF of GPS and GE PPP AR solutions is increased significantly(>35 min)while the results of GCRE is still very stable with the TTFF of 13.24 min.With the same session length,the PPP AR results achieve obviously higher accuracy than float solutions.At the 2h-observation session,GCRE PPP AR solutions improve the positioning accuracy of the float solutions with the improvements of 69.6%,15.2%and 10.6%for the east,north and vertical direction.Compared to the single-and dual-system results,GCRE PPP AR can achieve the best accuracy with observations of 10 min.The positioning accuracy of GCRE fixed solution within 10 min is(1.84,1.11,1.53)cm while the GPS-only result is(2.25,1.29,9.73)cm for the east,north and vertical components. |
PDF Full Text Request