Spaceplane model predictive control and guidance for aerogravity assisted maneuvers above Earth

Gravity assisted maneuvers resulting from close approaches to celestial bodies have been efficient in reducing the costs of interplanetary missions. If performed by a spaceplane passing through the planetary atmosphere, they will result in aerogravity assisted maneuvers. The main novelty of this work relies on the implementation of model predictive control for optimal guidance of an AGAM, looking to maximize the latitude, which increases the plane change at the end of the high-altitude atmospheric flight. Nevertheless, an open-loop solution can result in discrepancies when applied to a real scenario due to unpredictable perturbations and fluctuations of the atmosphere. Therefore, a closed-loop model predictive control is selected for this work. This type of controller satisfies the cost function, adhering to the nonlinear dynamics of the plant and the same constraints as the optimal control problem, thereby ensuring feasible trajectories across various scenarios. The dynamic optimization suite GEKKO was selected to solve the online optimal control problem in the finite horizon. Several configurations were tested, and the results highlight the ability of the model predictive controls to effectively minimize tracking error along the trajectory, while also demonstrating its robustness in ensuring a feasible path despite atmospheric perturbations and uncertainties in initial conditions. This controller proves to be an excellent candidate for the chosen problem while adhering to the constraints of the multi-input, multi-output system.