Research On The Algorithms And Theory Of GNSS Multi-frequency Data Processing

The development and application of the GNSS multi-frequency technology have greatlyimproved the capability of the satellite navigation system, and have enlarged the ranges of GNSSnavigation and position. This dissertation lays emphasis on the study of the theory and method ofGNSS multi-frequency data processing, and is mainly related to ionospheric multi-frequencycorrection, multi-frequency combination observation, cycle-slips detection and repair, andmultiple-carrier ambiguity resolution. The main achievements of this paper are listed as follows:1. The influence of ionospheric second order error on the navigation and position has beenanalyzed. The model of ionospheric triple-frequency two order correction has been provided, andthe problem has also been analyzed that ionospheric triple-frequency two order correction canmagnify the errors greatly. The application of wavelet package de-noising in ionospheric twoorder correction has been studied, the experiment result shows that wavelet package de-noisingcan markedly weekend the errors, and that the precision after two order correction can be up tohigh-precision demand of position.2. The statistical analysis about a series of combination observations with the longerwavelength or less ionospheric errors has been made. The Compass, GPS and Galileocombination observations have been compared with each other. The methods of enumeration,integer programming and fuzzy cluster used to filtrate the better combination observation havebeen compared. The selection for multi-frequency combination observations based onself-adaptive clustering algorithm has been investigated profoundly, and the experiment resultshows that the method has less workload and great efficiency, and it can find the bestcombination.3. The theory of pseudo-range minus phase linear combination for cycle-slip detection andcorrection has been introduced, and the influence of estimate value on cycle-slip correction hasbeen analyzed. The disadvantage that the geometry-free carrier combination can not detect somespecial cycle-slip has been analyzed and the searching method has been studied primarily tocorrect the cycle-slip. Compass triple-frequency data have been used to demonstrate the effect ofmethod of pseudo-range minus phase linear combination detection and the method ofgeometry-free carrier combination for cycle-slip detection. The result indicates that the twomethods can be used in the dynamic, un-differential navigation and position, and can detect andcorrect continuous cycle-slips exactly.4. The reasons of the matrix changing algorithm can fast finish ambiguity resolution have been analyzed, and Compass triple-frequency data have been used to validate the algorithm forambiguity resolution in short and long baselines. The primary factor which affects on the CIRmethod for ambiguity resolution in long baseline has been analyzed. The algorithm ofgeometry-based CIR, the ionospheric correction, ionophere-free combination, geometry-free andionophere-free combination have been analyzed respectively to improve the original carrierambiguity resolution.