摘要
Satellite attitude information is essential for pico-satellite applications requiring light-weight,low-power,and fast-computation characteristics.The objective of this study is to provide a magnetometer-only attitude estimation method for a low-altitude Earth orbit,bias momentum pico-satellite.Based on two assumptions,the spacecraft spherical symmetry and damping of body rates,a linear kinematics model of a bias momentum satellite's pitch axis is derived,and the linear estimation algorithm is developed.The algorithm combines the linear Kalman filter(KF) with the classic three-axis attitude determination method(TRIAD).KF is used to estimate satellite's pitch axis orientation,while TRIAD is used to obtain information concerning the satellite's three-axis attitude.Simulation tests confirmed that the algorithm is suited to the time-varying model errors resulting from both assumptions.The estimate result keeps tracking satellite attitude motion during all damping,stable,and free rotating control stages.Compared with nonlinear algorithms,such as extended Kalman filer(EKF) and square root unscented Kalman filer(SRUKF),the algorithm presented here has an almost equal performance in terms of convergence time and estimation accuracy,while the consumption of computing resources is much lower.
Satellite attitude information is essential for pico-satellite applications requiring light-weight, low-power, and fast-computation characteristics. The objective of this study is to provide a magnetometer-only attitude estimation method for a low-altitude Earth orbit, bias momentum pico-satellite. Based on two assumptions, the spacecraft spherical symmetry and damping of body rates, a linear kinematics model of a bias momentum satellite＇s pitch axis is derived, and the linear estimation algorithm is developed. The algorithm combines the linear Kalman filter （KF） with the classic three-axis attitude determination method （TRIAD）. KF is used to estimate satellite＇s pitch axis orientation, while TRIAD is used to obtain information concerning the satellite＇s three-axis attitude. Simulation tests confirmed that the algorithm is suited to the time-varying model errors resulting from both assumptions. The estimate result keeps tracking satellite attitude motion during all damping, stable, and free rotating control stages. Compared with nonlinear algorithms, such as extended Kalman filer （EKF） and square root unscented Kalman filer （SRUKF）, the algorithm presented here has an almost equal performance in terms of convergence time and estimation accuracy, while the consumption of computing resources is much lower.
基金
supported by the Program for New Century Excellent Talents in University (No. NCET-06-0514),China
the Postdoctoral Science Foundation of China (Nos. 20081458 and 20080431306)
