A Dynamic Post-Calibration Method Based on EKF for Dual-Head Star Tracker

The dual-head star tracker (DHST) is affected by the space environment during its orbit, and its optical parameters (i.e., principal point, focal length, and distortion) and the relative installation between the heads undergo continuous changes, which affect the attitude determination accuracy. The traditional static calibration method replaces the actual value with the estimated optimal solution of the parameters, which cannot express the real-time changes of the parameters. This article proposes a dynamic post-calibration method based on the extended Kalman filter (EKF). First, star points are selected using the convex hull area percentage, and the optical parameters are calculated using the least-squares (LS) method. Then, the relative installation between the optical heads is expressed in Euler angles, and the EKF method is employed for calibration. Finally, the star image downloaded from the GF-7 star tracker is utilized as the original data for the experiment. The experimental results demonstrate that this method can significantly improve the accuracy of attitude determination, with an accuracy that is superior to that of the LS method by 0.3 arcseconds.