Study On The Key Technology Of GNSS Multi-frequency Precise Point Positioning Based On Joint Enhancement Of Leo Constellation And Reference Station

With the modernization of GPS,the continuous improvement of Galileo,and the completion of China’s BDS-3,the new generation of space vehicles can all broadcast navigation signals at three or more frequencies,and GNSS has officially entered an era of multi-system co-existence.The multi-system and multi-frequency also provide new opportunities for the improvement of positioning performance.Based on the above background,this thesis focuses on the key technologies of rapid and precise positioning of multi-system multi-frequency precise point positioning(PPP),which mainly involves:(1)comprehensive compensation of real-time orbit and clock error based on broadcast ephemeris,(2)frequency bias extraction of multi-frequency PPP,(3)fast ambiguity resolution(AR)of multi-frequency PPP,and(4)low earth orbit(LEO)and precise atmosphere enhanced PPP for rapid precise positioning.Through the improvement of related models and algorithms,the positioning real-time,continuity and reliability have been improved.The main work and contributions are listed as follows:1.A comprehensive compensation method of real-time orbit and clock error based on broadcast ephemeris is proposed.Aiming at the problem that there is a certain degree of lag and uncertainty in receiving real-time orbit and clock corrections,based on the fixed solution of PPP,using ambiguities and satellite fractional cycle bias(FCB)of previous epoch and the short-term predicted tropospheric delay,combined corrections are retrieved with broadcast ephemeris,including orbit error,satellite clock and receiver-related error,so as to ensure the continuous and reliable positioning of PPP users even during outages of orbit and clock corrections.Experiments are carried out through networks with an average distance of about 150～350km.The results show that the average accuracy of the combined corrections are better than 2cm in case of different time lags of 5s～60s.Both the simulated kinematic and the real tracked unmanned aerial vehicle(UAV)dynamic experiments obtained centimeters-level positioning accuracy.2.The frequency bias of multi-system multi-frequency PPP is extracted,and their characteristics are analyzed.(1)The inter-frequency clock bias(IFCB)time series of GPS Block IIF and BDS-2satellites are extracted,and its characteristics are analyzed.The results show that the IFCB has nothing to do with the station distribution,and shows an obvious periodicity.Based on the periodicity,the 6th and 4th harmonic functions are used to fit and predict the IFCB of GPS and BDS-2,respectively,and both the fitting and prediction accuracy of are better than 2cm.The effect of IFCB on multi-frequency PPP is further analyzed.The results show that whether in static or kinematic mode,the positioning accuracy is improved to varying degrees after correcting IFCB,and the standard deviation of the post-fit residuals is reduced by 19.2%-75.0%.(2)The raw FCB is restored equivalently from the extra-wide-lane(EWL),wide-lane(WL)and narrow-lane(NL)FCB through a linear transformation method.Experimental results of globally distributed stations show that,the EWL/WL FCB for each system are relatively stable,and among them,the EWL FCB of Galileo E5b-E5 a and BDS-3 B1c-B1 I combination are very close to zero.There are certain fluctuations in NL and the raw frequency FCB.Among them,the raw frequency FCB of Galileo has achieved the best stability.The post-fit residuals of different combinations of FCB have no obvious systematic deviation,which means that the FCB has a high internal coincidence accuracy.(3)The specific components of the WL/NL FCB in ionosphere-free(IF)model and the raw frequency FCB in un-combined(UC)model are derived in detail.Further,the FCB equivalence between the two models is demonstrated,and experiments are carried out with specific examples.The results show that the WL FCB difference of the above two models can basically be maintained within 0.07 cycle.For GPS/Galileo/BDS-3/BDS-2 IGSO and BDS-2MEO,and the percentages of NL FCB difference within ±0.05 cycle are 99.93%,88.69%,99.38%,100.0% and 86.57%,respectively.Generally speaking,there is no obvious system deviation in the distribution of the NL FCB difference,and the two models have good consistency in the NL FCB.3.Established a stepwise AR model of fixing EWL/W/NL ambiguities sequentially for multi-frequency PPP high-dimensional ambiguityAccording to the difficulty of fixing the ambiguity,a stepwise AR model of fixing EWL/WL/NL ambiguities sequentially is established.Compared with the traditional method which directly fix the raw frequency ambiguity,the stepwise strategy can obviously shorten the time-to-first-fix(TTFF)of the NL ambiguity with constraints of EWL/WL AR,and the ratio value are also improved.In general,the stepwise strategy can shorten the TTFF of multi-system multi-frequency kinematic PPP to 2～3 min.In addition,taking advantage of the long wavelength and easy to be fixed characteristic of EWL/WL ambiguity,real-time single-epoch WL fixed solutions with decimeter-level accuracy can be realized.4.Study the technical theory and application of LEO and precise atmosphere enhanced PPP.Based on the multi-frequency stepwise AR strategy,in order to further shorten the initialization time of PPP,LEO is incorporated into the data processing of reference network,and a rapid and precise positioning method enhanced by both LEO and precise atmosphere is proposed.(1)LEO constellation enhanced PPP: Firstly,the feasibility of LEO enhanced GNSS positioning is preliminarily verified in theory based on simulated LEO/GNSS data,and the experimental results in case of six typical shielding scenarios show that both conventional multi-GNSS fusion and LEO-enhanced GNSS can improve the positioning performance of PPP in hash environments,and LEO can improved the positioning performance more obvious than pure GNSS fusion.Then,the combination of simulated LEO and real-tracked GNSS data is used for further verification.The results show that the initialization time of PPP can be shortened from a few minutes to about 1 minute through LEO enhancement.At the same time,the positioning accuracy and epoch fix rate are also improved to varying degrees.(2)Regional precise atmosphere enhanced PPP: Based on a unified multi-frequency uncombined PPP model,the high-precision atmosphere information of each base station is extracted throuth stepwise AR,and further interpolated according to users’ distribution.By adding pseudo-observation constraints,the corrclation between various estimated parameters is weakend,therby achieving rapid initialization.The real tracked data of 30 s and 1s sampling rates from differrent regions are used for verification.The results show that for the reference network with an average distance of 10～140km,the interpolation accuracy of the zenith tropospheric wet delay(ZWD)and slant ionosphere augmentation information is better than1 cm and 2cm,respectively.With the constraints of high-precision atmosphere,the positioning performance of the single-epoch WL ambiguity-fixed solution is significantly improved,from decimeter-level to a few centimeters,and at the same time,the NL ambiguity can also be instantaneously fixed.(3)LEO and precise atmosphere jointly enhanced PPP: A single-epoch PPP method with centimeter-level accuracy enhanced by both LEO and precise atmosphere is proposed.Compared with precise atmosphere only enhanced mode,the positioning accuracy and epoch fix rate of single-epoch PPP WL/NL ambiguity-fixed solution are improved to varying degrees.