Study Of GNSS Precise Point Positioning And Its Quality Control Method

Precise Point Positioning(PPP)is a spatial positioning technology that began to emerge at the end of the 20th century.It integrates the technical points of standard single point positioning and relative positioning.Precision single-point positioning refers to the use of precision satellite orbits and precision satellite clock errors provided by the International GNSS Service Organization(IGS)or other analysis centers,and then comprehensively consider the precise correction of various error models,using pseudorange observations and carrier phase observations It is a spatial positioning technology that realizes precise absolute positioning of a single receiver.The rapid development and improvement of precision single-point positioning technology has made this technology no longer limited to users for spatial positioning,and is currently widely used in precision orbit determination,severe weather forecasting,marine surveying,agriculture,military and other fields.At present,the basic theoretical and practical problems of precision single-point positioning technology have been well solved,but most of the ambiguity solutions in PPP still use floating-point solutions,which are not fixed to integers,which is important to the accuracy and convergence speed of PPP Great influence.The solution to this problem mainly relies on the control of the imported data quality.Although the optimization of the algorithm can also improve the accuracy and reliability of PPP,it is difficult for ordinary users of PPP to understand the accuracy and reliability of the final positioning.Therefore,how to improve the positioning accuracy and reliability of PPP is a key issue that still needs to be resolved in the wider application of PPP technology.In response to this problem,the quality analysis and quality control of PPP data processing is still a research work that needs to be continued.The quality control of PPP aims to preprocess the input observation data,precision orbit,precision clock error and other data products,and then design a set of calculation algorithms to improve the PPP based on the use of observation models,random models and parameter estimation algorithms.Accuracy,reliability,and convergence.Therefore,the quality control of PPP can be considered from each stage of the PPP process.Therefore,in the PPP process,in order to ensure the accuracy and reliability of the PPP,the quality of the input data is critical.A good set of data is an important factor for obtaining high-precision and high-reliability PPP solution results.In addition,studies have found that when using Kalman filtering algorithm to estimate PPP parameters,it often encounters the effects of observation anomalies and dynamic model anomalies.For example,when the observed value at a certain moment contains gross errors,the accuracy of the PPP parameter estimation will be affected by it and decrease,and it may also lead to long-term failure to converge.The abnormality of the dynamic model is manifested in the slow convergence rate caused by the inaccurate state noise.Because in theory,the state parameter covariance will eventually slowly approach zero.When the state covariance is incorrect,the required convergence time will be longer.In order to improve the accuracy,reliability and convergence speed of PPP,this paper has done certain research work on the input data.In the parameter estimation stage,a method to solve the above problems is proposed.The main research work is as follows:(1)For the observation data input in the PPP solution,a variety of methods are used to analyze the observation data,mainly including the following items:1)Judging whether the number of satellites meets the necessary conditions through the visibility of the satellites,and the single system is least visible The number of satellites is 4.From the perspective of reliability,it is necessary to ensure that the number of instantaneous satellites is not less than 5.2)Determine the geometric strength of the satellite constellation by calculating the Dilution of Precision(DOP)value.Generally,the GDOP value is used as a quantitative standard;3)Through the statistical data integrity rate,select the data with higher integrity rate for experimental analysis;4)Analyze the multi-path effects and noise of the observations to provide a basis for the weighting of subsequent observations and the detection of data anomalies;5)Detect gross errors and cycle slips to provide relatively clean original data for subsequent parameter estimation,Avoid too many gross errors and cycle slips that lead to the collapse of the solution equation and the failure of subsequent quality control.For the precision ephemeris data input in the PPP calculation,this article takes the GPS system as an example,and selects the precision ephemeris from the six IGS analysis centers MIT,JPL,GFZ,ESA,GRG and WUM to compare with the final precision ephemeris products of IGS Analyze the consistency between the precision ephemeris of each analysis center and the comprehensive precision ephemeris of IGS.The experimental results show that in the precision rail products,the precision rails of the six analysis centers have good consistency with the IGS final precision rail products;in the precision clock error products,except for the precision clock error of the MIT Analysis Center and the IGS integrated precision clock Except for the large difference between the differences,the precision clock errors of the other five analysis centers are in good agreement with the final precision clock products of IGS.In addition,PPP positioning was performed on the precision ephemeris products of different analysis centers.Experiments show that using the precision orbit and precision clock offset of the same analysis center for PPP positioning,the positioning accuracy of different analysis centers is equivalent to that of IGS products.(2)Through studying the derivation process of the standard Kalman filter,it is found that when Kalman filter participates in the estimation of PPP parameters,the inaccuracy of dynamic noise will cause abnormal disturbance of the dynamic model.Because the inaccuracy of dynamic noise interferes with the state covariance prediction value,the gain matrix of the filter is not accurate enough.The gain matrix is one of the important factors that make the state parameter estimation reliable.If the gain matrix is unreliable,it will eventually cause bad interference to the accuracy of the parameter estimation and even the convergence time.Most of the constructions based on adaptive filtering start with innovation vectors.In this paper,an adaptive factor is constructed through the innovation residual,and the state covariance prediction value is corrected by the adaptive factor,making the state covariance prediction value and gain matrix more reliable,and finally reaches The purpose of improving parameter estimation accuracy and reducing convergence time.In addition,this paper studies the derivation process of robust adaptive Kalman filter,and compares it with standard Kalman filter.Finally,through self-programming,select the data of multiple IGS tracking stations for static/dynamic PPP test.The results show that in the static PPP calculation,whether it is the adaptive Kalman filter or the robust adaptive Kalman filter,it can improve the positioning accuracy and reduce the convergence time better than the standard Kalman filter;in the dynamic PPP calculation,the adaptive Kalman filter and robust adaptive Kalman filter are more stable in resisting noise than standard Kalman filter.Figure[34]Table[7]References[88]...