Enhanced ambiguity resolution and integrity monitoring methods for Precise Point Positioning

摘要

Centimetre-level Global Navigation Satellite System (GNSS) based positioning is increasingly relevant for a large number of applications. Currently, this level of GNSS positioning accuracy is most commonly achieved using the conventional Real Time Kinematic (cRTK) method. In order to achieve such high-accuracies with cRTK, the distance (baseline) between the user and reference station must typically be shorter than 50 km for dual-frequency GNSS receivers. To address the limitations of cRTK, the Precise Point Positioning (PPP) method, which does not require local reference networks, was developed. The principle of PPP is to model and correct error sources such as satellite orbit and clock errors using correction products and error modelling. PPP is not currently suitable for many applications, because of the long solution convergence time (from 20 to 60 min to achieve 10 cm accuracy), insufficient positioning accuracies and a lack of integrity monitoring. Current fixed ambiguity PPP methods are analysed and tested using the National Oceanic and Atmospheric Administration (NOAA) dataset in this thesis. Based on the analysis, the most reliable existing validation method has unacceptably large rate (12.7%) of incorrect ambiguity resolution. Therefore, this thesis develops an enhanced PPP method.The enhanced PPP method is based on using the enhanced ambiguity validation method (e.g. time-window based validation) and employing both GLONASS and GPS measurements to calculate a float position solution. In addition, integrity monitoring is improved in terms of failure exclusion and protection level calculation. When employing the enhanced PPP method, the rate of incorrect ambiguity resolution decreases to 5.3% and of correct ambiguity resolution increases to 82.2% when using the (NOAA) dataset. The average horizontal, vertical and 3D position errors at the initial ambiguity resolution epoch are reduced by 40.0%, 23.8% and 31.8%, respectively, compared to the most reliable existing PPP method.