Attitude control of a small spacecraft for earth observation via tube-based robust model predictive control

Precise attitude tracking control of small satellites is a demanding task due to the limited hardware resources of sensors, actuators, and processors equipped on board. Moreover, the effects of model uncertainty and persistent disturbance, such as gravity-gradient, magnetic, aerodynamic, and solar radiation pressure torques, on the satellite dynamics, make the attitude control more challenging. Consequently, under actuator limitations and stringent pointing requirements, robust controllers represent a more reliable solution to comply the mission constraints. In this paper, a tube-based robust model predictive control algorithm is exploited to address robust attitude control of CubeSat, within a Earth-observation scenario, and its performance are evaluated. The proposed algorithm successfully achieves the three-axis stabilization required by the payload, i.e., a small digital camera, using reaction wheels and applying constraints on the actuator torque and speed saturation as well as on pointing accuracy. To properly estimate the effectiveness of the proposed approach, the comparison with a nonlinear model predictive control is presented with the consideration of different sample times. Then, the robustness of these two control scheme is compared running 100 simulations, each one with random initial condition, and the results demonstrate the effectiveness of the proposed robust algorithm.