基于模型误差补偿的星敏感器/陀螺卫星定姿方法及其仿真

Novel Satellite Attitude Determination Algorithm and Simulation Based on Model-error Compensation with Star-sensor/Gyro

针对经典扩展卡尔曼滤波(EKF)依赖于未知参数动态模型的准确程度这一问题,提出了一种基于当前星敏感器观测信息的模型误差补偿算法。该算法利用状态估计信息及当前矢量观测值,以预报误差最小为准则,对陀螺输出模型中的陀螺漂移进行补偿,进而利用修正后的陀螺测量模型预报卫星姿态,从而显著地提高了系统的定姿性能。仿真实验表明,在无陀螺漂移先验信息的情况下,该算法能够有效地补偿陀螺漂移引起的模型误差,定姿效果优于EKF。

Considering that the effect of extended Kalman filter （EKF） technique depends on the accuracy of the dynamic model with unknown parameters, a novel model-error compensation method was proposed. The approach adopts the ＇predicted error minimization＇ principle and employs both previously estimated information and current observing vectors to compensate the gyros＇ excursion. Based on the modified model, satellite attitude could be predicted accurately. Simulated results show that, even without prior knowledge of gyros＇ excursion, the proposed method can compensate the model＇s error induced by gyros＇ excursion effectively and outperforms traditional EKF.