Star pattern recognition and mirror assembly misalignment for DIGISTAR II and III multiple FOVs star sensors

Two crucial aspects of the data processing of the recently proposed DIGISTAR II and DIGISTAR III multiple field-of-view (FOV) star trackers, which use one/two mirrors deflecting the sensor FOV to two/three orthogonal directions, respectively, are here analyzed. These aspects are the star identification process and the estimation of the mirror misalignment. In particular it is shown how to adapt the existing Search-Less Algorithm (SLA), which does not require searching phases, to these sensors and, in order to keep high the predicted gain in the attitude accuracy, two different methods to compute the mirror misalignment are provided. The first method, which evaluates the misalignment by a least-square approach, is used if the misalignment can be considered small enough, that is, such that the star-identification process can still be accomplished. The second method, which is used for any value of the mirror misalignment, implies, however, that the star identification process can be performed, separately, for each sub-FOV. Results and tests for the SLA adaptation as well as for the two methods to evaluate the mirror misalignment, are given.

Overview