Three-axis rotation angle solving method of star sensor under multi-vector information

In order to improve the real-time performance and solution accuracy of star sensor, based on multi-vector information, a quaternion representation method for solving three-axis rotation angle of star sensor was proposed, and the method was derived in detail theoretically. Based on the three-dimensional coordinates of a single starlight vector in celestial coordinate system and star sensor coordinate system, the direction cosine matrix transformation form was transformed into quaternion transformation form. The quadratic quaternion transformation form was reduced in order to facilitate the subsequent solution. Considering the weight of different starlight vectors, the information of all starlight vectors was combined to solve the three-axis rotation angle of the star sensor. A specific solution was given, aiming at the iteration uncertainty and the direction singularity of quaternion vector in the process of solving three-axis rotation angle, The performance of the proposed algorithm was compared with that of the traditional algorithm. Simulation results show that the proposed algorithm has faster solving speed and higher accuracy in solving three-axis rotation angle than the traditional algorithm.

This research algorithm derives the quaternion transformation form based on the three-dimensional coordinates of a single star light vector in the celestial coordinate system and the star sensitive coordinate system. It fully utilizes the properties of the pseudo inverse matrix to reduce the degree of the quadratic quaternion matrix equation, and transforms the solution of the optimal rotation quaternion into the problem of finding the maximum eigenvalue of the solution matrix and its corresponding eigenvectors This study theoretically proves the range of eigenvalues and solves the prerequisite for iterative problems in the process of solving the three-axis rotation angle; A detailed analysis was conducted on the direction problem of quaternion vectors, eliminating the problem of solving the three-axis rotation angle caused by directional singularity The performance simulation and experimental comparison analysis of the algorithm in this study and traditional algorithms show that the algorithm, SVD algorithm, and QUEST algorithm have the highest accuracy in solving the three-axis rotation angle. The algorithm in this study has high computational efficiency and good overall performance.