The Research On Method Of GNSS Partial Ambiguity Resolution

With the gradual updating and construction of the four global navigation satellite systems,the number of satellites in orbit will reach more than 100 in the future,and the frequency of navigation signals will increase to three and even more,providing users with more observation information,which will greatly improve the precision,reliability and availability of satellite navigation and positioning services.Integer ambiguity resolution is the key problem in achieving high-precision positioning of GNSS.With the increase of the number of satellite observation equations,ambiguity resolution has a higher precision of real-value,which is beneficial to fix ambiguity accurately;but it also will inevitably lead to an increase in the ambiguity resolution dimension at the same time,which will increase the risk of fixing all ambiguities,thus maybe reduce the fixed success rate.Therefore,partial ambiguity resolution has been proposed to further improve the fixed success rate of ambiguity by fixing a suitable subset in the highdimensional ambiguity,which has important research significance.Up to now,although many scholars have carried out research on partial ambiguity resolution,there are still some problems that not solved well.First,the subset selection strategy is a key issue of partial ambiguity resolution,according to the selection strategy,partial ambiguity resolution can be divided into three levels(satellite level,frequency level and ambiguity level)and two categories(model-driven and data-driven).There are many methods for selecting ambiguity subset,however,it is difficult for users to choose and use due to most of the methods are independent of each other and have their own advantages and disadvantages,and they all lack a certain theoretical basis.In addition,since the purpose of fixing ambiguity is to improve the precision of baseline vector,which is directly related to ambiguity subset size,the larger the subset is,the higher baseline fixed solution precision is.However,the current subset selection methods mostly ignore the ambiguity subset size.Therefore,reasonably designing the ambiguity subset screening criteria to improve ambiguity fixed success rate as well as ensuring a high baseline solution precision,is another key problem that needs to be solved.Aiming at the above problems,this paper theoretically analyzes the existing methods of selecting ambiguity subset,explore its intrinsic essence and core idea,and then seek the key factors that affecting subset selection and ambiguity fixing.Combined with the influence of ambiguity subset on the precision of baseline fixed solution,a new partial ambiguity resolution method based on both model and data is proposed.Finally,Multi-frequency GNSS data from different scenarios are used to comprehensively evaluate the resolution performance of the proposed method.The main contents of this paper are as follows:(1)The analytical expressions of ambiguity parameters,baseline vector parameters and their variance-covariance matrix are deduced in detail.It is pointed out that the variance-covariance matrix of ambiguity contains both the precision information of observations and the satellite spatial configuration information,thus it can be used for selecting ambiguity subset.In addition,the theory of integer ambiguity resolution is systematically introduced,providing a theoretical basis for the study of partial ambiguity resolution method below.(2)The internal relations and differences of subset selection methods in satellite level,frequency level and ambiguity level are systematically analyzed.It is pointed out that all of the methods select ambiguity subset according to the precision order of observations/ambiguities.Then the theoretical consistency of these three levels is explained,and the subset selection can be solved directly from ambiguity level using ambiguity precision information.(3)The model-driven method and data-driven method in partial ambiguity resolution are compared and analyzed carefully.The model-driven method reflects the strength of observation model and the overall precision of ambiguity parameters.The data-driven method reflects the distinguishability among ambiguity candidate vectors.Combining model-driven method with data-driven method can sufficiently guarantee the precision and reliability of ambiguity fixed solution.(4)A baseline precision gain function is constructed to measure the degree of precision improvement of baseline fixed solution relative to baseline float solution.It can also reflect the relationship between ambiguity subset size and baseline precision,so that the size of ambiguity subset to be fixed can be determined according to the required baseline precision.It is proved that using the partial ambiguity fixed solution and the conditional updated remaining ambiguity float solution together to update baseline vector is equivalent to using only the partial ambiguity fixed solution to update baseline vector,so it is convenient to directly use partial fixed ambiguity to obtain baseline fixed solution.(5)Based on the above theory,this paper proposes a model and data dual driven partial ambiguity resolution method which uses conditional variance matrix order as the order of ambiguity precision,the model-driven Bootstrapping success rate as the first constraint to ensure a high ambiguity fixed success rate,the data-driven FFRT test as the second constraint to ensure the reliability of ambiguity fixed solution,the baseline precision gain as the third constraint to ensure an enough high baseline resolution precision,and the ambiguity dual-frequency consistency test as the last constraint to further ensure the correctness of the fixed ambiguity.The results show that the partial ambiguity resolution method proposed in this paper can strongly improve fixed success rate compared with fixing full ambiguity.Compared with the existing partial ambiguity resolution methods,this proposed method can improve correct fixed rate(the ratio of fixed success rate to fixed rate)to a certain degree,which is beneficial to ensure the reliability of ambiguity fixed solution and obtain a more accurate and stable baseline fixed solution.