Research On Precise Orbit Determination Techniques Of LEO Satellite Based On GPS

Low Earth Orbit (LEO) satellites have a wide range of applications on areas such as atmospheric sounding, earth gravity field research, etc. Most of these scientific researches are based on a good knowledge of precise LEO absolute and relative trajectories. Global Positioning System (GPS), with its features of low cost, all weather and high precision, is the most widely used means of measurement for Precise Orbit Determination (POD) at present. Researches on precise absolute and relative orbit determination techniques for LEO using GPS are carried out in this thesis.Firstly, the basic GPS measurement models are given, the error sources of measurements are analyzed and ways of error reduction and elimination are pointed out. The theoretic priciples of the Extended Kalman Filter (EKF) used in the reduced-dynamic orbit determination method are elaborated. The dynamic model of LEO are also described in detail.Based on the above knowledge, using the Precise Point Positioning (PPP) method, the LEO POD simulation platform was implemented. The technical details are carefully analyzed, including the use of precise ephemeris products, numerical integration algorithm of dynamic model and explicit filter settings, also the overall algorithm flow. Against the problems that might potentially degrade the Orbit Determination (OD) performance, a new Receiver Autonomous Integrity Monitoring (RAIM) method and a new measurement noise modelling method based on study of historical observations are proposed. Simulation resulsts using real in-flight data show that these two methods improve OD precision by some extent.Relative positioning between formation-flying LEOs are also investigated. Relative positioning algorithms based on Single Differenced (SD) measurements, Double Differenced (DD) measurements with float ambiguity estimation and DD measurements with ambiguity fixing are implemented. Positioning performace of these3methods are compared using simulation results from GRACE mission flight data. It could be concluded that:Relative positioning using SD measurements are simpler and more robust, so it’s suitable for real-time applications; Algorithm using DD measurements with ambiguity fixing are more complex but outputs better positioning accuracy, thus is suitable for post-processing precise relative positioning applications.