Real-time kinematic software using robust Kalman filter and dual-frequency GPS signals for high precision positioning

摘要

The Global positioning System (GPS) started in the 1970’s with an ambitious project of U.S positioning service using satellites. It has now become one of the major technologies for positioning people and objects around the planet, with diverse application in mapping and localization. GPS has overcome all its expectation, providing signal continuously around the entire planet, providing positioning service more and more precise. Future constellation will now arise, like GALILEO for Europe or COMPASS for China, bringing more attractive, precise and powerful applications. The modernization programs of GPS and Russian GLONASS will brings even more capabilities for worldwide users.ududOne of the most interesting applications of GPS is the Real-Time Kinematic system. This technique emerged in the beginning of the 1990’s offers centimeter to millimeter precision to the GPS users, using carrier phase measurements and a reference GPS station. It uses differential corrections, and techniques of carrier phase ambiguity resolution ‘on the fly’. It has been successfully applied in geophysics and survey. Unfortunately, such RTK system is only precise in short range from the base station, that is to say less than about 20 km. When the distance from the base station increases, systematic errors are decorrelated. These errors reduce the ambiguity resolution success rate and decrease position precision and reliability. The purpose of this thesis is to overcome these limitations to bring full RTK precision and reliability for long baseline scenarios, up to 80 km.ududTo fulfill this purpose, a new concept of RTK system for real-time has been developed. This means the development of complete real-time GPS positioning software providing centimeter precision in a robust way for short and long baseline scenario. Different issues have been developed, such as real-time satellite management, robust Kalman filter implementation, and reliable ambiguity resolution technique. The long baseline problem has been developed and overcome using real-time atmospheric modeling and control of the geometric errors. This work presents the different new concepts used in the algorithm and the innovative technique for future system and developments using RTK positioning. ududTo demonstrate the reliability and the performance of the developed algorithm, data from Novatel and NRG-GNSS receiver have been intensively analyzed and processed. With static and dynamic short baseline real-time data, this new developed RTK software presents robust real-time centimeter to millimeter solution precision and fast and reliable ambiguity resolution. Results from the solution are analyzed and the parameters of the real-time solution are discussed. After validating these scenarios, long baseline dynamic data, coming from our industrial partner Gedex, have been processed in real-time mode. The solution uses the innovative concepts of ionospheric modeling in real-time, and the results present millimeter difference to the post-process Waypoint software. Impact of real-time management and ambiguity resolution technique are presented. The efficiency and precision of the solution opens the RTK solution to new purposes for research and development.