基于模型误差确定卫星姿态的预测滤波算法  

MODEL-ERROR-BASED PREDICTIVE FILTER FOR SATELLITE ATTITUDE DETERMINATION

作　　者：林玉荣[1] 邓正隆[1] 

机构地区：[1]哈尔滨工业大学控制科学与工程系,哈尔滨150001

出　　处：《宇航学报》2001年第1期78-83,88,共7页Journal of Astronautics

基　　金：航天科技集团创新基金资助项目

摘　　要：本文给出了一种高精度、鲁棒性强的实时姿态估计算法 ,即预测滤波算法。该算法利用星敏感器所提供的观测矢量 ,对运动学方程中由陀螺漂移造成的角速度模型误差进行一步预测 ,从而准确地估计卫星的运动姿态和角速度。预测滤波算法的良好性能已通过仿真测试得到了验证。文中进一步提出的修正算法 ,使滤波器的预测性能得到了改进。A real-time predictive filter for attitude estimation, which is characteristic of both high accuracy and strong robustness, is presented in the paper The algorithm accurately estimates satellite attitude and attitude rate using measurement vectors from star sensors by predicting attitude rate errors in the kinematics equation one step ahead which is caused by drifting error and measurement noise of gyro Both high performance and effectiveness of the method are demonstrated through simulation study The modified algorithm presented in the paper improved performance of the predictive filter

关 键 词：预测滤液 姿态确定 观测矢量 四元数正交化 三轴稳定卫星 姿态控制 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计]