A high accuracy star tracker using running sequential angular match technique

Various attitude sensors are available for determining a spacecrafts orientation in space. A star tracker provides the most accurate attitude solution amongst all the existing sensors. It operates in two modes: when no prior attitude is available, i.e. in the Lost In Space (LIS) mode and if the attitude is known, it turns itself into normal tracking mode of star position. The goal of our research effort is to investigate, analyze and solve the problems faced by the star tracker in LIS mode. Star pattern recognition is the most important part of a star tracker in LIS mode. In this paper, a novel pattern recognition technique based on sequential angular matching of a star cluster is developed, which achieves a high accuracy of 99.57% for recognizing stars. The results from the simulations show that this technique is also robust to the problems of mismatch between the star pattern database and image, missing stars, deviation in star positions and magnitude uncertainty as it maintains a high recognition accuracy compared to the conventional methods.