基于UKF的航天器最小参数姿态矩阵估计方法

钊对航天器的惯性-星光姿态确定系统，选取基于矢量观测的最小参数姿态矩阵估计方法为定姿算法。但是由于系统模型是非线性的，针对系统模型的非线性，将UKF（Unscented Kalman Filter）与最小参数姿态矩阵估计方法结合，设计了一种针对MEMS陀螺和CMOS APS星敏感器的UKF姿念估计器。仿真结果表明：在敏感器精度较差的情况下，该UKF、姿态估计器能够满足大多数航天器对姿态确定精度的要求。此外，UKF和EKF（Extended Kalman Filter）的仿真对比结果表明：UKF滤波器的收敛速度高于EKF滤波器，状态估计的精度也优于EKF滤波器，且数值稳定性好。

Study of Spacecraft Minimal-Parameter Attitude Matrix Estimation Based on the Unscented Kalman Filter

This paper focused on the inertial-stellar attitude determination system consists of MEMS (Micro electromechanical Systems) gyros and CMOS APS star sensor, selected minimal-parameter attitude matrix estimation from vector observation as the method of attitude determination. But the system model of this method is nonlinear. In order to deal with the nonlinearity, a new UKF (Unscented Kalman Filter) attitude estimator based on UKF and method of minimal-parameter attitude matrix estimation is designed. The results of simulation show that the determination accuracy of the UKF estimator can meet the attitude determination requirements of the most spacecrafts. In addition, the UKF converges rapidly than the EKF, the state estimation precision and the stability of the UKF is better than the EKF.