AAS 98-158

WAVELETS, SMOOTHERS, AND FILTERS - APPLICATION TO SATELLITE ATTITUDE DETERMINATION PROBLEMS

D. L. Mackison - University of Colorado

Abstract

Satellite attitude determination is dependent upon the choice of attitude sensors, geometrics, algorithms, dynamic and measurement models, and the appropriate use and combination of these resources. Even low noise systems including star sensors and inertial grade gyros have limits on their ultimate resolution. Approaches to reducing the effects of noise include the use of recursive Kalman filters, and the Extended Kalman Filter, and smoothing the data a priori or through the use of a forward-backward Kalman filter such as the Rauch-Tung-Striebel smoother. Here we apply smoothing techniques, wavelet denoising, and wavelet power spectra to satellite attitude data and demonstrate the comparative advantage and relative merits to problems in satellite attitude determination.

Back