摘要
在未知扰动噪声分布和强度先验信息前提下,提出了基于无迹卡尔曼滤波(UKF)状态估计的星敏感器和陀螺仪组合的高精度卫星姿态确定方法。设计了基于自回归(AR)模型的扰动噪声的建立方法,进而推导星敏感器与陀螺仪组合定姿的测量方程和过程方程,最终采用交互式多模型(IMM)定姿方法自适应调整模型集中不同噪声水平的概率以"高斯和"形式逼近未知的真实噪声,实现卫星高精度定姿。实验结果表明,IMM定姿精度优于单个模型定姿结果,并且具有较强的鲁棒性,能够为实际工程应用提供参考价值。
Without any prior information of disturbance noise＇s distribution and intensity, this paper proposed a state estimation method using star sensor and gyroscope to get high precision satellite attitude determination based on Unscented Kalman Filter （UKF）. This method first designed the disturbance noise based on AutoRegression （AR） model, and then deduced the measure and process equations of star sensor and gyroscope system. Finally, by adopting Interacting Multiple model （IMM）, this method could adaptively adjust the probabilities of different noise level, act the way as ＂Gaussian sum＂ to approximate the real noise, and achieve high precision satellite attitude determination. Simulation results verify that the IMM method has strong robustness and higher precision than single model, and provides reference for practical engineering application.
出处
《计算机应用》
CSCD
北大核心
2012年第A01期199-202,共4页
journal of Computer Applications
关键词
卫星定姿
扰动噪声
自回归模型
无迹卡尔曼滤波
satellite attitude determination
disturbance noise
AutoRegression （AR） model
Unscented Kalman Filter（UKF）
