A robust autonomous star identification algorithm for ZY3 satellite

ZY-3 is China´s first civilian high-resolution stereo mapping satellite. It equipped multiple star trackers and gyros for attitude determination. The stellar images acquired by the autonomous Active Pixel Sensor (APS) based star tracker on ZY-3 satellite are downlinked to the ground for post-processing. The FOV (Field Of View) of the star tracker is 20^o×20^o, which is designed large to detect more stars in the FOV. However, the star identification becomes more complex. A robust autonomous star identification algorithm is proposed. The novel algorithm firstly chooses the brightest star in the stellar image as the base star, and links the base star to each neighbouring star for radiant pattern construction. Each “link” is compared with star database, and the star that has most times of meeting the matching condition is considered as the base navigation star. And the adjacent stars are constructed as a directed circularity pattern for unitary matching. The candidate matching group which has the longest chain will be selected as the ultimate matching result. The stellar images downlinked from ZY-3 are used to verify the effectiveness and reliability of this algorithm.