Single-epoch Ambiguity Resolution Methods For Multi-frequency GNSS

Global Navigation Satellite System(GNSS)is the main means to quickly obtain accurate location information in the fields of intelligent transportation and autonomous driving in urban environment,and GNSS relative positioning is one of the main methods to achieve real-time,even single-epoch precise positioning in urban environment.The key lies in the fast resolution of the integer ambiguity,even the single epoch ambiguity fixing.At present,all GNSS systems can broadcast three-frequency signals,and there has been extensive research and application of three-frequency GNSS ambiguity solution method.With China’s Beidou-3 Navigation Satellite System(BDS-3)publicly provides observation data of five frequency point,it is of great practical significance to study how to realize multifrequency(four-frequency,five-frequency)GNSS single-epoch ambiguity resolution.Under this background,this thesis mainly focuses on the selection method of optimal linear combination of multi-frequency GNSS and the methods of single-epoch ambiguity resolution for multi-frequency GNSS,and obtains some innovative results.The main research work and results obtained in this paper are as follows:1.Aiming at the linear combination problem required for the multi-frequency GNSS single-epoch ambiguity resolution,the basic relationship and linear combination characteristics of the multi-frequency GNSS linear combinations of wavelength,frequency and ambiguity are deduced and analyzed.It is found that the number of high-quality linear combinations of GNSS increases sharply with the increase of the number of GNSS frequencies.2.Aiming at the problem of "either or the other" caused by the rapid increase in the number of multi-frequency GNSS linear combinations,the existing linear combination selection methods are inefficient and subject to subjective judgment.A new method of optimal linear combination selection method was proposed based on fuzzy clustering analysis.The method uses the characteristic parameters of combinations as the indicators to characterize the linear combinations,and realizes the objective ranking of the linear combinations of multi-frequency GNSS under different conditions through data standardization,calibration,clustering and other steps,and then determines the optimal linear combination of multi-frequency GNSS.The single-epoch ambiguity resolution experiment was carried out using the GF＿TCAR method,and the applicability and effectiveness of the fuzzy clustering analysis method in the selection of multi-frequency GNSS linear combinations were verified.3.For the traditional GF＿TCAR method to fix the extra-wide lane,wide lane and narrow lane step by step,it is difficult to fix the narrow lane ambiguity caused by the low accuracy of the wide lane ambiguity resolution.A multi-frequency GNSS single-epoch Geometry-free model ambiguity resolution(GF＿MCAR)method using the "(k-1)EWL+NL" strategy was proposed.And the Geometry-free model was updated based on the first extra-wide lane combination whose ambiguity has been fixed instead of the pseudo-range combination,and the other extra-wide lane and narrow lane ambiguities were resolved.The single-epoch ambiguity resolution analysis of the GF＿MCAR method was carried out using the short baseline data measured by BDS.The results show that with the increase of the number of frequencies,the performance of the multi-frequency GNSS extra-wide lane combination is better,which is more conducive to the realization of the single-epoch extra-wide lane ambiguity reliably fixed.4.In order to improve the efficiency and reliability of traditional Geometry-based model ambiguity resolution,a "two-step" multi-frequency GNSS single-epoch ambiguity resolution Geometry-based integrated MCAR method was proposed.The method first fixes k-1 extrawide lane ambiguities,and then updates the integrated MCAR model with the extra-wide lane ambiguities whose ambiguities have been fixed to solve the narrow-lane ambiguity.Using the measured BDS multi-frequency data to analyze the single-epoch ambiguity resolution,the results show that,whether it is the three-frequency ambiguity resolution,the four-frequency ambiguity resolution or the five-frequency ambiguity resolution,the solution results of the integrated MCAR method were better than those of the Geometry-free model,especially the ambiguity resolution performance of narrow lane was significantly improved.5.Aiming at the influence of the ionospheric delay error that is not effectively corrected in the Geometry-free model and the Geometry-based model,based on the GNSS doubledifference pseudorange/phase geometry-free and ionosphere-free equation,a method suitable for multi-frequency GNSS ambiguity resolution based on Geometry-free and Ionosphere-free model was derived.According to the two conditions of Geometry-free and Ionosphere-free,a new Geo-Iono-free model combination was constructed using two extra-wide-lane combinations with fixed ambiguities and one narrow-lane combined observation to resolve narrow-lane ambiguity.The BDS short baseline and BDS-2 long baseline are used to analyze the single-epoch ambiguity resolution experiments,respectively.The results show that with the increase of the frequency number,the observation noise of the Geo-Iono-free model narrow-lane combination increases.It is difficult to use simple rounding to achieve narrowlane combination ambiguity single-epoch fixing.The theoretical research and experimental analysis of this thesis show that with the increase of GNSS frequency,the number of high-quality GNSS linear combinations increases sharply,and the fuzzy clustering analysis method proposed in this thesis can effectively select the optimal linear combination of different GNSS.The three MCAR models proposed in this thesis can all achieve reliable fixation of the multi-frequency GNSS single-epoch extra-wide lane ambiguity,but the Geo-Iono-free model narrow-lane combined observation noise in the GIF＿MCAR method was seriously amplified,which makes the narrow-lane combination ambiguity difficult to fix by single-epoch.The GB＿MCAR method has a significant improvement in the single epoch narrow lane ambiguity solution performance compared with the GF＿MCAR method,and can basically realize the narrow lane ambiguity single-epoch fixation.