Single star identification and attitude determination in tracking mode

The autonomous star trackers are among the most attractive attitude sensors, as they are capable to provide three-axis attitude information with high accuracy. Autonomous star tracker's basic software algorithms include Star image centroiding, Star Pattern Recognition (SPR) and attitude determination. In the last few decades many SPR algorithms have been presented, almost all of them need at least two or three stars in star image, for successful SPR. Similarly almost all of the existing attitude determination algorithms need at least two identified stars in the star image. In the worst case if there is only one star in the image then SPR and hence the attitude determination becomes a problem. Just star magnitude alone can not be used to identify one star reliably due to poor magnitude measurement accuracy. However the scheme of automatic pattern recognition based on predictive centroiding in tracking mode, can be used to identify a single star. The position information of this identified single star can be used to form one or more empirical image stars from the reference catalog stars, for attitude determination. In this paper the problem of single star identification and attitude determination is addressed. Simulation results show that the proposed method to calculate the attitude has good accuracy, and it depends on the star tracker calibration and centroiding accuracies. This scheme is particularly useful in case where single star is identified and no other scheme can be used.