Attitude determination system based on vector observations for satellites experiencing sun-eclipse phases

Abstract

Due to mission requirements, fault detection and isolation protocols or budget restrictions, a satellite is required to use most reliable attitude determination hardware, such as magnetometers and sun sensors, in order to keep 3-axis attitude information available during its complete orbit. However, satellites experiencing sun-eclipse phases, sun sensors become no operational. In this paper, we propose an attitude determination system which provide 3-axis attitude information in both sun and eclipse phases, considering vector observations acquired from sun and magnetic measurements. To compensate the unavailability of sun sensors during eclipse phase, two variations of innovation processes merged into the Extended Kalman Filters are proposed. In order to keep the accuracy of attitude estimation process during eclipse mode, angular rates must be accurately estimated during sun phase. To solve this issue, rough angular rate information is calculated based on previous attitude information calculated by Gauss-Newton method, which fuse magnetic and sun sensor data. Numerical simulation results show the performance of the proposed attitude determination system, considering the use of vector measurement hardware with different precision degree.