Spinor-Based Attitude Determination with Star Sensor Considering Depth

Abstract

A star sensor is a high-precision satellite attitude measurement device. Since its observation information has only two-dimensional direction vectors, when a star sensor is used for attitude determination the dimension of the observation information is less than the number of attitude angles determined, so mainstream algorithms usually only guarantee the accuracy of the pitch angle and the roll angle. In view of the lack of depth information in the observation's imaging geometric condition, this article proposes a spinor-based attitude determination model, which describes a straight line passing through two stars with the spinor and maps the depth information of the straight line with the pitch, to establish an imaging geometry model of the spinor coplanar condition. Experiments show that the yaw-angle attitude accuracy of the method is an order of magnitude better than that of mainstream algorithms, and the accuracy of the three attitude angles reaches the arc-second level.