基于粒子群寻优的高精度星敏感器在轨标定方法

星敏感器结构设计与安装过程会产生多种误差，主要误差源有星敏感器像平面主点误差、主距误差、倾斜误差与旋转误差，这些误差影响了星敏感器在轨标定的精度。本文根据星敏感器的误差模型，提出了一种高精度的星敏感器在轨标定方法。在已知含有误差的像平面的基础上，构造一个虚拟的像平面。当粒子群优化算法使含有误差的像点投影到虚拟像平面上的坐标与无误差时像点的坐标一致时，再利用Quest解算出三轴姿态角求得两个像平面之间的姿态矩阵，得出两像平面之间的关系。结果表明：星敏感器姿态确定精度较高且比较稳定。这种方法与传统标定方法的优势在于不依靠陀螺信息，原理简单，提高了数据的准确性。

A High Precision On-orbit Calibration Method for Star Sensors Based on Particle Swarm Optimization

The structural design and installation of the star sensor will produce a variety of errors. The main error sources include the plane principal point error, principal distance error, tilt error and rotation error of the star sensor, which affect the accuracy of the on-orbit calibration of the star sensor. Based on the error model of star sensor, a high precision calibration method for star sensor in orbit is proposed in this paper. When the particle swarm optimization algorithm makes the coordinates of the image point with errors projected onto the virtual image plane consistent with the coordinates of the image point without errors, the three-axis attitude angle is solved by using Quest to obtain the attitude matrix between the two image planes, and the relationship between the two image planes is obtained. The results show that the attitude determination accuracy of the star sensor is high and stable. The advantage of this method and the traditional calibration method is that it does not depend on the gyro information, the principle is simple, and the accuracy of the data is improved.