Methodology Of Multi-GNSS And Multi-frequcency Undifferenced And Uncombined Precise Point Positioning

Precise point positioning(PPP)is a kind of high-precision positioning technology based on pseudorange and carrier phase observations.It uses high-precision satellite orbit and clock error products and considers various error corrections accurately with a single receiver.The compatibility and development of multi-GNSS and multi-frequency signals is an important trend in the future development of GNSS with advantages of increasing observation redundancy information,improving system performance reliability,and improving navigation performance indicators.It is widely used in time service,atmospheric monitoring,earthquake prediction,meteorological prediction and other fields with important market application value.However,the abnormal processing of inter system and inter frequency bias and the loss of some band signals will lead to the degradation of the positioning performance of multi-GNSS multi frequency PPP.Therefore,how to construct undifferenced and uncombined model under the multi-GNSS mixed frequency mode is the key problem to be solved in the application and promotion of PPP technology.This paper makes full use of multi-GNSS and multi-frequency observation,and carries out theoretical research from multi-GNSS multi-frequency undifferenced and uncombined PPP,and finally build a PPP model for multi-GNSS mixed frequency mode,which makes the model universal and maximum degree to improve positioning performance.The main research contents of this paper are as follows:(1)This paper systematically expounds the basic theoretical knowledge of PPP.Based on the GNSS original observation,this paper introduces the characteristics and processing methods of various kinds of error sources,compares and analyzes the performance of dual frequency undifferenced and uncombined model and traditional ionospheric free combination model based on Kalman filter mode,and verifies that undifferenced and uncombined PPP has advantages in convergence time compared with traditional ionospheric free combination PPP through numerical example analysis The convergence time is increased by 20%.(2)Aiming at the problem that inter frequency bias(IFB)is introduced into the three frequency non difference non combination model due to the additional frequency,which leads to the slow convergence time,a Beidou tri-frequency undifferenced and uncombined PPP method considering the inter frequency bias is proposed.Based on the introduction of IFB parameters into the Beidou tri-frequency pseudo range observation equation,the time stability of IFB is analyzed,and the positioning performance of IFB under time constant,white noise and random walk random model is compared.The experimental results show that the strategy of estimating IFB as random walk process has more advantages in positioning accuracy and convergence time.In addition,compared with the Beidou dual frequency non difference non combination PPP,the introduction of tri frequency observation data can improve the PPP positioning performance.(3)Aiming at the problem that the abnormal processing of inter system and inter frequency deviation leads to the degradation of multi-mode multi frequency PPP positioning performance,the hardware delay deviation processing method for multi-mode multi frequency PPP combination model is studied.On the one hand,in the combination of multi-mode GNSS signals,the estimation model of inter system deviation ISB is constructed,and the time stability of ISB parameters is analyzed through experiments.The results show that the single day stability of ISB parameters is good,and one day's ISB parameters can be used as time constant estimation in PPP processing.On the other hand,according to the characteristics of frequency division multiple access(FDMA)of GLONASS signal,three traditional IFB processing methods,including ignoring pseudo range IFB,IFB linear model and IFB quadratic polynomial model,are constructed and compared.The experimental results show that the IFB processing method based on the quadratic polynomial function is the best.(4)Aiming at the problem that the geometric distribution of satellite becomes worse due to the lack of partial frequencies due to the fixed frequency of multi-mode multi frequency PPP,a multi system mixing non difference non combination PPP method is proposed in this paper.This method integrates the Beidou tri-frequency,GPS dual-frequency and GLONASS dual-frequency observation data,makes full use of the available observation information,and improves the redundancy of the observation model.The experimental results show that:compared with multi system dual frequency non difference non combination PPP,multi system mixing non difference non combination PPP improves the positioning accuracy by more than 5.4%,and has stronger robust performance.