Modeling of an Orbital Simulator and Design of MPC-based Attitude Control System for VLEO maneuvers

In the recent years the space field has been increasingly open to investment from private companies, so getting a competitive edge in relation to other competitors is fundamental. For this thesis, a Very Low Earth Orbit (VLEO) mission is selected, considering its numerous advantages both in payload and in platform. For optical payloads the benefits are increased resolution, reduced aperture sizes and improved radiometric performance, while for radar/communication payloads are improved link budget, reduced antenna sizes and reduced transmission power and latency. VLEO is also convenient for the smaller radiation flux compared to higher orbits, for the reduced launch cost per unit mass and for being more environmentally sustainable, since increased drag is useful for the removal of orbiting debris. The increment of the above-mentioned disturbance, though, entails an increase effort to keep the satellite orbiting as intended. The purpose of the current thesis is to model a simulator in MATLAB\Simulink, to evaluate the possibility of exploiting VLEOs for space missions. In particular, this work is focused on designing a model of the attitude of a small satellite capable of operating in the aforesaid orbit environment. A near-circular sun-synchronous orbit is chosen with an altitude of 300 km. To better simulate VLEO environment, the following attitude disturbances are implemented: Gravity gradient and Magnetic Torque. The first disturbance is due to the different absolute values of the force of gravity experienced in each part of the spacecraft that can result in a torque. This torque tends to align the axis of minimum inertia of the vehicle with the radial axis originating from the center of gravity of the Earth. Magnetic torque disturbance is caused by the magnetic field generated from the planet that interacts with the residual magnetic dipole of the satellite. Quaternion Feedback Control (QFC) and Model Predictive Control (MPC) are implemented as attitude controllers with a comparison between the two. The result of both algorithms is the evaluation of the control torque needed to have the desired orientation in each phase of the simulation. With the QFC the output is obtained by calibrating the gains that are multiplied with the quaternion error and angular velocity. With the MPC the output is a set of expected control torques, which is the result of a convex optimization problem, evaluated from the current instant to a certain time in the future. To apply the desired torque to the spacecraft, actuators are needed, and for this project it is implemented in the simulator a cluster of Variable Speed Control Moment Gyros (VSCMGs). Those are advanced momentum exchange devices composed of a high-speed flywheel that can vary its speed, mounted within a rotating gimbal. This setup allows a VSCMG to generate torque in two orthogonal directions, providing higher degrees of control compared to similar devices. Results show that the target orientation can be maintained even including disturbances and that attitude errors are constrained into the given requirements.