Processing of Meteosat attitude measurement flight data and evaluation of the performance of an adaptive filtering technique

The attitude measurement data of the ESA geosynchronous spacecraft METEOSAT have been analyzed using an adaptive recursive filtering technique developed at NLR. The mathematical models of spacecraft dynamics and sensors have been kept very simple. Modelling errors are such that, without adaptive feed-back, strong divergence of the filtering process occurs. It is shown that the adaptive process stabilizes the filtering considerably. Complete error consistency, however, could not be attained in the present case. The background of this latter fact is analyzed.