Gaglio, Emanuela (2025) Aerodynamic Modeling and AI-Based Optimal Guidance and Control for Low Earth Orbit Missions. [Tesi di dottorato]

Documento PDF Gaglio_emanuela_37.pdf Download (12MB)

Abstract

The Low Earth Orbit (LEO) environment is currently experiencing a significant evolution, largely driven by the surge of small satellites, such as CubeSats, and the deployment of large-scale constellations. While LEO provides strategic advantages such as proximity to Earth, reduced communication latency, and lower launch costs, these benefits come at the expense of increasingly complex operational challenges. Among the most significant issues are the non-negligible influence of aerodynamic actions, which necessitate continuous orbit maintenance and maneuvers, as well as the growing threat posed by space debris and potential collisions. These factors demand a new generation of guidance and control (G&C) strategies capable of ensuring the success of the mission and, at the same time, the long-term sustainability. The situation is particularly critical for small satellite platforms, including CubeSats, often operating with stringent constraints on mass, volume, and safety, making traditional propulsion systems impractical or unfeasible. As a result, there is an growing need to explore alternative, propellantless methods of control that are not only efficient but also compatible with the limited resources available on board these platforms. At the same time, the increasing number of satellites in the LEO environment requires these methods to be highly precise, adaptive, and capable of real-time execution to support autonomous maneuvering, de-orbiting, and collision avoidance. This dissertation fits in this framework by proposing an integrated G&C framework actively exploiting aerodynamic forces as the primary mechanism for orbital maneuvering and control. The research is developed through two complementary and interdependent steps. The first involves a high-fidelity aerodynamic characterization of re-entry systems, including both inflatable and deployable systems with flaps embedded at the end of the heatshield, which are designed to exploit atmospheric forces for orbit modification. This is achieved through an extensive campaign of numerical simulations combining Computational Fluid Dynamics (CFD) and Direct Simulation Monte Carlo (DSMC) techniques. By capturing the complex flow interactions across a wide range of altitudes and flow regimes, from the rarefied environment to transitional and continuum, these models provide a robust foundation for an accurate prediction of the forces. The deployable concept, featuring control flaps at the end of the heatshield, was further analyzed to assess its capability for attitude control through multiple flap deflection combinations. The second component of this work introduces a novel hybrid G&C architecture that merges optimal control theory with Artificial Intelligence (AI) techniques to enable autonomous, onboard decision-making. This approach is designed to preserve optimality while exploiting AI-driven methods to significantly reduce computational cost and response time, as well as to robustly manage unpredictable perturbations. This aspect is particularly critical in LEO, where aerodynamic forces are difficult to predict with accuracy and external disturbances can be substantial. As a result, the system is capable of solving complex control problems in real-time, a critical requirement for autonomous space operations. The proposed methodology has been rigorously validated through three representative mission scenarios. The first focuses on the development of a time-optimal drag-based de-orbiting algorithm, capable of precisely targeting locations at the re-entry interface. The second investigates a controlled atmospheric re-entry and landing strategy based on AI for adaptive feedback in response to variable aerodynamic conditions. The third explores a collision avoidance framework based on aerodynamic maneuvering, designed to maximize the miss distance from incoming debris while minimizing orbital perturbations and preserving mission objectives. As a whole, the proposed work delivers a comprehensive and forward-looking contribution to the field of aerodynamics and G&C in LEO. The dissertation not only advances the state-of-art in propellantless orbital maneuvering but also lays the groundwork for more sustainable, autonomous, and resilient operations in a progressively more congested and contested orbital environment.

Downloads

Actions (login required)

Modifica documento