Spacecraft attitude estimation using GPS: Methodology and results

This dissertation describes research in the area of spacecraft attitude determination using the Global Positioning System (GPS). Five algorithms for attitude determination that deal with various levels of prior knowledge and dynamic uncertainty have been developed, implemented, and studied. The purpose and description of each method, as well as test results and analysis are presented.; The attitude initialization algorithm is an iterative least squares batch process that, with very little a priori information, estimates the initial attitude of a spacecraft. Another method is the attitude point solution which invokes an iterative least squares solution at each measurement epoch. The attitude Kalman filter incorporates dynamic information into the least squares process, and as a result yields the most accurate attitude solutions. The on-orbit baseline estimator is a multi-step process that provides the accurate estimates of antenna locations needed for high performance attitude solutions. And finally, the combined filter is an algorithm that combines both attitude and baseline estimation into one process.; To verify the attitude and baseline estimation algorithms, a simulation modeling a near-Earth gravity gradient stabilized satellite using GPS for attitude determination was developed. The simulation generates differential phase measurements that are then used to test the algorithms.; Flight data from two satellites with GPS receivers onboard has also been used to demonstrate the algorithms. One is the RADCAL satellite which is a gravity gradient satellite used by the Air Force for radar calibration. The other is a 3-axis stabilized satellite used for atmospheric science called CRISTA-SPAS. A case study of the attitude determination results of each satellite is presented.