Research On Triple-frequency Ambiguities Resolution For GNSS Network RTK Reference Stations

With the development of national economic construction,GNSS(Global Navigation Satellite System)positioning technology has gradually expanded from the military field to the civilian market,and GNSS satellite positioning service has begun to move towards millions of households.At the same time,the technologies such as unmanned driving are developing vigorously,which put forward higher requirements for the accuracy and reliability of positioning service system.GNSS network RTK(Real-Time Kinematic)technology can provide centimeterlevel high-precision location service.It improves the limited coverage of the RTK which has single reference station.Network RTK includes three steps,determination of ambiguity between reference stations,generation of error correction and location of rover stations.The ambiguity between reference stations is indeed the premise and basis of reliable location services provided by network RTK.Up to now,the network RTK usually calculates the ambiguity between reference stations according to the double-frequency observations,and has limited means to deal with errors.Moreover,the distance between reference stations is limited.The use of the triple-frequency observations can effectively improve the accuracy and reliability of the ambiguity determination,and at the same time can enlarge the distance between reference stations.Up to now,GPS,BDS and Galileo these three GNSS systems are able to provide triple-frequency positioning services.Therefore,this paper mainly studies the determination of triple-frequency ambiguity between reference stations in GNSS network RTK.The contents and results of this paper are summarized as follows:1.The three processes of ambiguity determination between reference stations,error correction generation and rover station location in network RTK technology are introduced in detail.2.The TCAR(Three Carrier Ambiguity Resolution)method of triple-frequency ambiguity resolution is deduced,and the theoretical effect of each step is analyzed from the aspect of noise influence.The effect is compared with the WL(WideLane)AR(Ambiguity Resolution)process under the dual-frequency system,and noise has less influence on determination of triple-frequency ambiguity in theory.3.A GF(Geometry-Free)TCAR algorithm based on GF model is introduced in detail to solve the triple-frequency ambiguity between reference stations.This method selects the optimal triple-frequency observation,fix the ambiguity of two EWL(Extra-Wide-Lane)first,determine the ambiguity of WL through linear combination,and determine the original ambiguity after calculating the ionospheric delay.Experiments show that the method can reliably determine the ambiguity of EWL and WL combinations in three systems,but the original ambiguity can not be accurately fixed in BDS GEO(Geostationary Orbit)and Galileo IOV(In-Orbit Validation)satellites for the reason of amplified systematic noise.4.An IF(Ionospheric-Free)TCAR algorithm based on IF model is proposed,which makes full use of the known coordinates of reference stations.The ambiguity determination method of EWL is consistent with that of GF TCAR method which is determined by MW(Melbourne-Wübbena)combination.For WL AR,two IF combination ambiguities calculated by using the known coordinates of reference stations are used to calculate the WL ambiguity by linear combination.Experiments show that compared with the GF TCAR method in low latitudes,IF TCAR improves more than 30% both in accuracy and convergence speed.5.In the original ambiguity determination of IF TCAR method,Kalman filter is proposed to solve the ZTD(Zenith Tropospheric Delay)residual and IF combination ambiguities.The original ambiguity is determined by the filtered IF combination ambiguity and the process makes full use of prior information and absorbs unmodeled geometric errors,which is more rigorous in theory.Experiments show that the IF TCAR method is more reliable in the original ambiguity resolution in low latitudes.The IF TCAR performance of three systems is better than that of GF TCAR both in accuracy and convergence time.Especially in the BDS GEO and Galileo IOV satellites,GF TCAR method can not determine the original ambiguity,while IF TCAR can still reliably determine the ambiguity.6.A new method for solving the original ambiguity of combined system is proposed.Using the same ZTD residual of reference stations in different systems and Kalman filtering to determine the IF combination ambiguity of multi-system,the problem of limited number of satellites in single system is improved,and the reliability of ambiguity filtering results is increased.The experimental results show that the accuracy of the original ambiguity in combined system can be improved by about 10% compared with that in single system and convergence speed can be improved more than 30% in low latitudes.