In-orbit Estimation of the Slow Varying Residual Magnetic Moment and Magnetic Moment Induced by the Solar Cells on CubeSat Satellites

External disturbance torques on Low Earth Orbit (LEO) satellites are considered as one of the most significant challenges to address for high-performance attitude control of small satellites. These disturbances represent a harsh environment for the successful operation of a LEO satellite mission. The magnetic disturbance torque is one of the dominant attitude disturbances torques, caused by the residual magnetic moment of the satellite when it interacts with the Earth's magnetic field. The main source of disturbance to generate a residual magnetic moment on the satellite are the solar panels. Therefore, the residual magnetic moment that originates in the solar panels should be estimated. This paper presents solar panel magnetic moment and slow varying magnetic moment estimation, based on an Extended Kalman Filter (EKF) by assuming the magnetic disturbance torque to represent the main model uncertainty. A 3-axis rate sensor (gyroscope) is used to obtain the filter innovations. The evaluation is performed on a 3U CubeSat satellite, and simulation results indicate that the EKF was able to estimate the satellite's body rate, slow varying magnetic moment and the induced magnetic torques of the solar panels. The results demonstrate a good convergence speed when the filter state is primed with zero initial conditions, i.e. no apriori knowledge of the magnetic disturbances.