BDS/GPS Non-Difference Non-Combined Robust Adaptive PPP

Precise point positioning(PPP)is a high-accuracy positioning technology to obtain the absolute positioning by using a dual-frequency GNSS receiver's pseudorange and carrier phase observation measurements,which uses precision ephemeris and precision clock products provided by IGS and other organizations,based on error correction strategies to eliminate or reduce most of the errors,and using parameter estimation methods.Since precise point positioning technology was put forward,a large number of studies on PPP have been done,which are based on ionosphere-free combination model.Although ionosphere-free combination model can eliminate the influence of the first order of the ionosphere delay,it also causes the amplification of the observation error and the effect of the multipath effect,and the ionospheric delay in the atmosphere can not be obtained.The non-difference non-combination model based on the original GNSS observation value can make up the above shortcomings,but there are many parameters to be estimated and the problem of dealing with the error of frequency.When the observation condition of the receiver is not good,using a single system for positioning will result in the problems of insufficient visible satellites and large PDOP values.The combination of the two systems can improve the satellite geometry,improve positioning accuracy and reduce convergence time.With the vigorous development of China's BDS and the proposed GPS modernization in the United States,it has brought tremendous opportunities for development in the field of satellite navigation.BDS/GPS combination of precision point positioning is the direction of development in the future,how to apply multiple systems and multiple frequencies is an urgent problem to be solved in precision point positioning.Kalman filtering is a parameter estimation method,usually used in precision single-point positioning.Because the abnormal error of observations can not be completely eliminated in the data pre-processing,error model correction and other stages,and the commonly used dynamic model can not accurately describe the motion state of the carrier,the use of Kalman filter for precision point positioning will bring some problems that the results of the filter solution are unstable,the accuracy of the solution is not high,and the convergence time is long.Therefore,it is necessary to take a solution mode that takes the observed model error and the dynamic model error into account in performing parameter estimation.Aiming at the existing problems mentioned above,this paper adopts the non-difference non-combined model in precise point positioning,studies the observation equations of the non-difference non-combination model,and presents the corresponding stochastic model,and linearizes the model;The error sources in the non-difference non-combination process of original observations are classified and the corresponding correction models are given to eliminate or reduce the influence of errors on the positioning.The application of Kalman filtering in the non-difference non-combination model is derived in detail,and it is pointed out that the error of observation model and dynamic model can not be taken into account when the standard Kalman filtering algorithm is applied.Based on this,this paper deduces an adaptively robust Kalman filtering in non-difference non-combination PPP.Finally,the observation data of the MGEX tracking station are processed and compared with the standard Kalman filter to verify the effect of an adaptively robust Kalman filtering.Experimental results show that in non-difference non-combination precise point positioning,both the static PPP process and the dynamic PPP process,the use of an adaptively robust adaptive Kalman filtering,which positioning accuracy and convergence time are better than Kalman filter's filtering effect.