Research On The Ambiguity Resolution Algorithms Of GNSS Triple Frequency Precise Positioning

Integer ambiguity resolution is the key problem and technical difficulty in GNSS high precise positioning, and using three frequency GNSS observations can improve the ambiguity resolution efficiency. So the ambiguity resolution methods in GNSS precise positioning are researched in this dissertation, which concentrate on the combination theory and ambiguity resolution using three frequency observations based on the observed BDS three frequency data, in order to realize the fast ambiguity resolution, accordingly the precision and reliability of GNSS positioning and navigation are increased. The main research contents and conclusions are as follows:1. The combination theory of GNSS three frequency observations, the characteristics of combination observations and corresponding error feature are research in this dissertation, then the characteristics of extra-wide-lane combination, wide-lane combination, and narrow-lane combination using GPS and BDS observations, as well as the relationships in wavelength, lane-number, ionosphere-number and noise level are analyzed and discussed. On account of the main influence factors of short baseline and long baseline resolution precision, the optimal linear combinations of GPS and BDS, which is benefit to fix ambiguity, are selected respectively.2. In order to research the GNSS ambiguity resolution methods based on ambiguity domain applied in the effectiveness of ambiguity resolution using BDS and GPS three frequency observations, the algorithm including LAMBDA, Rounding and Bootstrapping are adopted in this thesis, the BDS zero-baseline and ultra-short baseline are processed in the same conditions, through which the C01 satellite and largest elevation satellite are selected as reference satellite respectively. The increments in ENU directions are compared with results obtained by TBC software. The results show that the good geometry distribution in space and the numbers of visible satellite not less than 8 of BDS satellites make the above two scheme can realize the ambiguity resolution in single epoch and effective baseline resolution, however, as the relative less number of visible satellites for GPS leading to the poor single epoch baseline resolution.3. The extra-wide-lane combination, which is easy to fix ambiguity, is used to replace the wide-lane combination in TCAR, then the resolution mode using EWL+EWL+NL combinations are formed. The TCAR method based on measure domain solving GNSS three frequency ambiguity is analyzed. The optimal combinations and the precision of float ambiguity resolution using GF_TCAR, GB_TCAR, GIF_TCAR and modified integrated TCAR are researched respectively, and the observed BDS data of zero baseline, short baseline, long baseline and ultra-long baseline are used to resolve ambiguity. The results indicate that the ambiguity resolution performance difference s of GF_TCAR and GB_TCAR in short baseline and ultra-long baseline are tiny. The modified integrated TCAR can improve the ambiguity resolution efficiency in a certain extent, whereby the improvement effect for narrow lane ambiguity resolution of short baseline is obvious. The GIF_TCAR makes combination noise magnified, but it is more suitable for long and ultra-long baseline ambiguity resolution.4. The BDS double frequency ambiguity resolution for raw frequency is calculated base on LAMBDA, and it is compared with the initial frequency ambiguity obtained from the above four TACR algorithms. Results show that the initial frequency ambiguity of TCAR algorithm relates to the combination ambiguity value and its coefficient, as well as the solution and computation amount are simple and small respectively. As the LAMBDA algorithm including decorrelation and search procedures, the calculation burden is too heavy. In terms of calculating results, zero baseline and short baseline results obtained by GIF_TCAR algorithm are relatively poor, and the other TCAR algorithms are almost the same compare with LAMBDA algorithm results.