Toward the most accurate attitude determination system using star trackers

A recently introduced approach for evaluating attitude estimation error, namely, the Attitude Error Estimator Algorithm (AEEA), is utilized to quantify the attitude accuracy obtainable by the two novel DIGISTAR II and DIGISTAR III star trackers designs, which we introduce in this paper. These proposed sensors, by using one/two mirrors deflecting the sensor field-of-view to two/three orthogonal directions, provide the attitude estimation algorithm with the optimal condition data set of orthogonal observed directions. The AEEA, which can be used to evaluate the relative merits of competing sensor designs, is based on the sensor error model and on the spatial geometry of the observed vectors. Attitude estimation error statistics, such as expected error, standard deviation and variance, are shown to be dependent on the maximum value only. The AEEA shows by meaningful plots that, in comparison with an "equivalent" traditional star sensor, the proposed new star sensors achieve a substantial improvement in the obtainable attitude accuracy, which is also shown to be a weak function of the number of redundant observed stars and the sensor field-of-view size. Mirror misalignments are also analyzed. DIGISTAR III, which is about 10% more accurate than DIGISTAR II and presents the undoubted advantage of a higher probability value in having orthogonal directions, entails some drawbacks due to the additional mirror and volume/weight required.

Toward the most accurate attitude determination system using star trackers Academic Article