Study On The Ambiguity Resolution Methods Based On Compass Triple-frequency Data

Compass satellite navigation system is steadily building with autonomy, openness, compatibility and progressive principle, in accordance with the overall strategy "first experiment, after the region, then the world", by the "three steps" to specific implementation plan. As the first satellite navigation system broadcasting triple-frequency data, multi-frequency system of Compass navigation satellite system is not only conducive to weaken the influence of tropospheric and ionospheric errors but also conducive to fix ambiguity of the triple-frequency data fast. Study on triple-frequency ambiguity resolution method of Compass system, have an important role in improving the navigation positioning precision and reliability, real-time.This dissertation lays emphasis on the study of the theory and method of triple-frequency ambiguity resolution of Compass data, and is mainly related to, multi-frequency combination observation, cycle-slips detection and repair, Ambiguity decorrelation algorithm,and multiple-carrier ambiguity resolution. The main achievements of this paper are listed as follows:1. According to the data characteristics of Beidou system, observation model of Compass Positioning System and the analysis of observation error and correction methods are given.2. The basic principle and the process of the calculation of Lambda algorithm have been researched; Aiming at studying the ambiguity decorrelation method, First of all, a detailed analysis of the reduced correlation principle of Gauss algorithm and the inverse integer Cholesky algorithm has been made, and ascending LDL algorithm is proposed based on improving the inverse integer Cholesky algorithm. Finally, according to the decorrelation degree of Gauss algorithm is not deep, An approach for ambiguity decorrelation based on precomputing and ordering diagonal elements is proposed by analyzing the disadvantages of the Gauss algorithm and the inverse integer Cholesky algorithm. The simulation data has been used to demonstrate the effect of ambiguity decorrelation, the result indicates that approach for ambiguity decorrelation based on precomputing and ordering diagonal elements is better than the other two methods.3. Triple-frequency combination observation has been studied,especially the method of searching the optimal triple-frequency combination observation of compass data with the mathematical knowledge of the integer linear transform. It is proved that the sum of three integer coefficients of extra-wide-lane or wide-lane combinations with long wavelength, low amplification factor for ionosphere and low noise should equal to zero, and the sum of three integer coefficients of narrow-lane combinations with small ionosphere effect and low noise should equal to one. There are two linearly independent the optimal combination observations at most,the sum of which equal to zero.So the new search is increased to look for the combination observations, the sum of which does not equal to zero. It is proved that This kind of combination observations with long wavelength,not too low amplification factor for ionosphere are suitable for calculating the short baseline ambiguity.4.The experiment has been done to verify the the performance of ambiguity resolution of the combination observations the sum of three integer coefficients of which equal to zero and combination observations the sum of three integer coefficients of which does not equal to zero using geometry-free pattern。The results show that the precision of float ambiguity solution calculated by optimal combination observations he sum of three integer coefficients of which equal to zero is too high to round directly to get the right solution.Due to the the double difference model that weaken the ionospheric delay error effectively, the wide-lane combination observations based on compass data he sum of three integer coefficients of which does not equal to zero can also be used to calculating the short baseline ambiguity.5. Matrix transform algorithm and geometry-free CIR algorithm have been researched. For the matrix transformation algorithm, pre-ordering ambiguity decorrelation algorithm has been used to obtain two different transformation matrixes for short baseline and long baseline. Compass triple-frequency data has been used to demonstrate matrix transform algorithm, the result indicates that matrix transform algorithm can achieve short baseline ambiguity resolution using single epoch data, but for long baseline data, it is difficult to calculate the ambiguity after many epochs of data smoothing. Primary error which affects on the CIR method for ambiguity resolution has been analyzed and study on the improvement effects on the ionospheric correction has been made. The result of Compass triple-frequency data indicates that that CIR algorithm can achieve short baseline ambiguity resolution using single epoch data, but it is difficult to achieve the ambiguity resolution due to the low precision of the ionospheric error correction.