Research On GNSS Precision Satellite Clock Error And Clock Combination

GNSS Precise Point Positioning(PPP)technology only uses a single receiver to obtain high-precision station coordinates,receiver clock offset and atmospheric delay in the global range.Therefore,it is widely used in precision positioning,navigation,timing,atmospheric monitoring and other fields.Satellite clock error is an important factor restricting the precision point positioning accuracy,but the current satellite clock error products still have the following problems:the traditional satellite clock error products due to the use of one-day solution strategy,the clock phase ambiguity at the boundary does not overlap lead to jump,the clock sequence obtained by using this clock error product for time transfer will also h ave obvious sky jump phenomenon;although the prediction satellite clock error products can be obtained in real time,the accuracy cannot meet the needs of high precision satellite single point positioning.The clock error products released by each analysis center also have the problems of gross error and missing clock error of some epoch satellites.Based on the above problems,the main research contents and results of this paper are as follows:(1)Aiming at the sky boundary jump phenomenon in continuous time transfer,a continuous clock error product based on multi-day observation data estimation is adopted.Using this continuous clock error product for continuous time transfer,the maximum jump at the sky boundary is only 0.034 ns,which effectively weakens the sky jump of time transfer.(2)Aiming at the time-consuming problem of near-real clock error estimation,a global tracking station network selection strategy based on K-mean++clustering algorithm is proposed.The station selection strategy takes into account not only the global uniform distribution of the tracking station but also the quality of the observation data of the tracking station.Only 50 tracking stations can be used to calculate the post-satellite clock product with a precision of 0.016 ns.Aiming at the problem of low prediction accuracy of IGS Ultra-rapid products(IGU)and long release time of measured clock error,a near-real clock error product based on state space representation(SSR)correction is adopted.The delay is only 1 h,and the clock error accuracy is better than 0.09 ns.The PPP experiment was carried out using the clock error product and compared with the IGS circumferential solution coordinates.The RMS value of GPS phase residual is better than 1 cm,and the RMS value of GALILEO phase residual is about 1 cm.(3)In view of the inevitable problems of gross error and missing of some epoch satellite clock errors in clock error products,the time benchmark of clock error products in each analysis center is unified and the gross error in clock error products is eliminated.The multi-angle cap weighting method is used to weighted average the clock error products in each analysis center to generate a comprehensive clock error product with higher reliability.The STD of GPS integrated clock error and IGS final products(IGS)secondary error is better than 40ps,and the STD of GALILEO integrated clock error and integrated clock error products are better than 100ps.(4)The open-source software GPSTools(GT)developed by Professor Tomoji Takasu of Tokyo Ocean University can only process the data of GPS satellite system.In this paper,the secondary development of GT software is carried out,including the calculation of GALILEO satellite system,the comparison of clock error quadratic difference,and the selection of tracking station network,and some data processing procedures are optimized.