Autonomous Low-Thrust Station-Keeping for CubeSats subject to Maneuver Slot Constraints

This thesis investigates autonomous orbit control strategies for CubeSats, a class of miniaturized satellites, focusing on station-keeping maneuvers within Sun-synchronous low Earth orbit scenarios, particularly relevant for Earth observation missions. The main challenges posed by low-thrust contiguous-throttle electric propulsion are addressed, proposing control solutions able to deal with maneuver slot constraints and offering a more efficient fuel consumption compared to traditional chemical propulsion. The primary objective of this study is the design and implementation of orbit control algorithms, and their validation through mission simulations. Specifically, the thesis work comprehends the development of an orbit control simulator, implemented with Python programming language, tailored for autonomous space missions. The simulator models CubeSat systems dynamics, using Gauss's Variational Equations, and the environmental disturbances acting on them, such as gravity potential due to Earth's oblateness and atmospheric drag. The control feedback loop is initially closed with a Proportional Integrative Derivative (PID) controller, a computationally cheap approach that provides benchmark performance in terms of fuel consumption and control error. Its primary limitations are the absence of predictive features to address upcoming no-thrust time windows, imposed by maneuver slot constraints, and the possibility of performing only traditional reference tracking, while real orbit control missions accept a control error tolerance around the target orbit. To overcome them, a Model Predictive Control (MPC) strategy is implemented to successfully tackle maneuver slot constraints while further improving fuel efficiency, at the cost of a higher computational load. The MPC approach is further deepened by analyzing the sensitivity of the performance metrics, such as fuel consumption and control error, to variations of controller parameters, such as predictive model, slot constraints type, and prediction horizon duration. Both control approaches are tested in the very low Earth orbit range where atmospheric drag has a much greater impact, still achieving mission success and thus demonstrating the robustness of the designed controllers to external perturbations. Due to the limited computational capabilities that small satellite systems generally offer, the choice between the two proposed control strategies highly depends on the satellite's onboard hardware.