Laboratory calibration of star sensors using a global refining method

Laboratory calibration is critical to ensure the precise attitude determination of star sensors. Existing laboratory star sensor calibration methods exhibit disadvantages for large-field-of-view star sensors and large amounts of calibration data. Inspired by the least-squares method and Li's method, a global refining method is proposed to overcome the inherent disadvantages by simultaneously obtaining all of the star sensor's parameters. It first employs the maximum likelihood estimation method to optimize the initial estimation of the principal point and focal length. Next, a linear least-squares solution is used to initially estimate the star sensor distortion. Taking the installation error into account, we conduct a maximum likelihood estimation to estimate the installation angles from the estimated parameters of the first two steps. Finally, we determine a globally optimal solution to refine the star sensor parameters. Compared with the traditional method and Li's method under the same conditions, both the simulation and real data results demonstrate that the proposed method is more robust and can achieve high precision. In addition, the experimental results show that the calibration method can satisfy the precision requirements for large-field-of-view star sensors.