GPS-based Kinematic Orbit Determination For Low Earth Satellites

The development of researches and application techniques such as gravity filed recovery and atmosphere sounding by on board GPS receivers pose very high requirements on precise orbit determination (POD) of Low-Earth-Orbiters (LEOs). Unfortunately, the technique of GPS-based Orbit Determination (OD) (also called kinematic OD) using GPS observations especially carrier-phase measurements in our country is relatively weak. So the theory, methods and some key issues of kinematic orbit determination are discussed in this dissertation:1 A new data pre-processing method of kinematic OD is proposedIn general, all cycle slip detection procedures are dependent (implicitly or explicitly) on analyses of Least square (LS) pattern of the data from epoch to epoch. But unfortunately, the LS residuals cannot correctly reflect the location and sizes of gross errors because of LS sharing out the gross errors to each observation equally. The method called "QUasi-Accurate Detection of gross errors (QUAD)" is extended so as to deal with the correlation observations. When using QUAD method for detection and correction gross errors and cycle slips in GPS observations, the key is how to select the correct Quasi-Accurate-Observations (QAOs). To do so, we divide the " selection procedures of QAOs into two different steps: preliminary selection and fine selection. In the preliminary selection step, we do not select all the observations but part of observations that meet certain criterion as QAOs. In fine selection step, we select QAOs according to the distribution characters of real errors ("hive off' phenomenon). So the success rate and reliability are relatively high. The results obtained from examples indicate that: after well elimination of systematic errors, the new method can locate the gross errors and cycle slips accurately and can give comparatively accurate estimators of them. This is rather significant to kinmatic OD with fewer observations2 A comprehensive Kalman filtering technique is presentedIn order to cope with the limitations when using traditional Kalman filter for kinematic OD. a comprehensive Kalman filter is presented in this thesis. This filter is a combination of adaptive UD decomposition Kalman filter with QUAD method. It use QUAD method to detect and correct the gross errors in observations, use UD decomposition technique to improve computation precision and overcome the instability of filter caused by instability of values, when divergence of Kalman filter had been detected, an adaptive filter is employed to adjust the prediction error covariance matrix. The results indicate that this comprehensive Kalman filtering technique has the features of good stability of computation values, rather strong adaptability and well elimination of the effect caused by gross errors.3 The procedure of GPS data simulation is systematically studied and correspondingsimulation software is compliedThe GPS simulation data can be used to compare the different schemes of OD designs and the different algorithms in one design. Although the onboard GPS simulation data are not real measurements, if the characteristics of simulation data are very similar to the real data, it would be very suitable for study on some specific methods such as for cycle slip detection and ambiguity resolution. The significance, principle and flow of GPS data simulation are discussed in details. Then some previous good error models and some latest outcomes presented in the world are made full use of in order to make simulated GPS data be very close to the real observations. The results obtained from the simulation example suggest a good effect and significance of using GPS simulation data for demonstrating the feasibility and validity of an OD design or an algorithm of certain problem.4 The ill-conditioned problem existing in relative kinematic OD using GPS carrier-phase data is discussed and then some preliminary efficient approachesfor solving this problem are given outThere are several reasons that cause the ill-condition pro...