High accuracy attitude determination of a spacecraft with a fast-rotating Doppler radar reflector

The attitude dynamics of a spacecraft employing a large rotating appendage is complex, and control and determination of its attitude and pointing can be a real challenge considering the impact system uncertainties and sensor errors can have on the mission scientific products. The aim of this paper is to develop a multi-rigid-body model that describes the rotational motion of an Earth-observing scientific satellite equipped with a large conically-scanning antenna reflector. The antenna assembly is statically and dynamically balanced, and a Sliding Mode Controller is implemented for achieving the desired attitude and angular velocity. Then, the determination problem is addressed, in order to verify the system capabilities in guaranteeing the mission knowledge requirements through simulations. A full suite of sensor feeds measurements to an Extended Kalman Filter algorithm running a simplified model of the spacecraft dynamics, and the spacecraft attitude and antenna pointing are estimated. The spacecraft configuration and operational conditions are based on the WIVERN mission, currently under study at ESA, which aims at characterizing global winds via a Doppler radar fast-rotating antenna. In particular, the velocity component of the spacecraft along the antenna line-of-sight is taken into consideration, as errors in the determination of this quantity directly influence the quality of the observed wind speed. It is demonstrated that the determination system can achieve an Absolute Knowledge Error (AKE) under 100 μrad per axis in terms of attitude, and under 1 m/s in terms of line-of-sight (LOS) velocity error. Spectral analysis shows that the filter lowers high frequency contributions to the AKE, which is a desired outcome as lower frequency errors can be corrected in post-processing using data-driven calibration methods. Criticalities regarding antenna unbalance and star tracker bias uncertainties are discussed by means of worst-case scenarios simulations and Monte Carlo campaigns.