Robust Estimation Algorithm For BDS/GNSS Continuous Relative Positioning

In December 2018,the third-generation Beidou satellite navigation system independently developed and operated by China was basically completed.It is expected that Beidou 3 will be completed in May 2020,providing global users with all-weather,all-day,high-precision Positioning,Navigation and Time(PNT)service.In addition,with the development of technologies such as self-driving cars,drones,unmanned boats,and smart cities,the demand for high-precision and high-reliability positioning has become stronger.Relative positioning,as one of the most mature satellite positioning technologies currently applied,can provide users with m-level to cm-level positioning services.However,when the environmental occlusion causes satellite lock loss and pseudorange multipath or low-cost receiver observations contain large errors,it will destroy the reliability and continuity of the relative positioning results.In practical applications,in addition to accuracy,the reliability and continuity of positioning results are also the focus of users' attention.In view of these problem,this paper proposes a “robust estimation algorithm for Beidou / GNSS continuous relative positioning”,which studies from multiple aspects such as pseudorange error,phase error,and continuous positioning after observation interruption to improve the reliability and continuousness of relative positioning results.The main work and conclusions of the paper are as follows: 1)The development of Beidou / GNSS relative positioning technology and the current problems are summarized systematically.The basic theories and methods of GNSS relative positioning parameter estimation are derived and introduced in detail from the functional model,stochastic model,pre-processing and posterior quality control,and parameter estimation.2)A "relative positioning quality control algorithm considering both pseudorange and phase errors" is proposed.In the posterior residual check stage,this method first checks the pseudorange residuals,and deletes the code observations and initializes the ambiguity for the pseudorange and phase errors,respectively,to ensure the reliability of the positioning results.In addition,by using relative and absolute residual check quantities,the accuracy of detecting the observations containing errors is improved.Two sets of static and one set of ship-borne dynamic results show that the method proposed in this paper improves the positioning results very little when the observation environment is good(the pseudorange error is small);when the observation environment is poor or highly dynamic(pseudorange error is larger),which greatly improves the convergence and precision of positioning results.3)A "GNSS continuous relative positioning algorithm considering velocity and acceleration constraints" is proposed.When an observation interruption occurs,the method uses the historical epoch information before interruption to construct a Kalman filter prediction equation constrained by velocity and acceleration,and combines the observation information of the current epoch to perform filter solution.In addition,based on the error propagation law,the variance-covariance information of the current epoch parameters is constructed to make full use of historical information to ensure the continuousness of the positioning results.The results of short baseline,medium-long baseline,and long baseline experiments show that the conventional method will lead to the re-convergence of the positioning results under the interruption of different durations,and the method proposed in this paper can guarantee the stability of the positioning results and the double-difference ambiguity sequence.In summary,the robust estimation algorithm of Beidou / GNSS continuous relative positioning proposed in this paper can improve the reliability and continuity of Beidou / GNSS positioning results,and it has certain significance for the application of Beidou in the global scope and multiple fields.