ORCID Number
0009-0008-7141-5189
Date of Award
Spring 5-4-2026
Embargo Period
5-4-2026
Access Type
Thesis - Open Access
Degree Name
Master of Science in Aeronautical Engineering
Department
Aerospace Engineering
Committee Chair
Riccardo Bevilacqua
Committee Chair Email
bevilacr@erau.edu
First Committee Member
David Canales Garcia
First Committee Member Email
canaled4@erau.edu
Second Committee Member
Morad Nazari
Second Committee Member Email
nazarim@erau.edu
College Dean
James W. Gregory
Abstract
Close-proximity operations in the vicinity of Near-Earth Asteroids (NEAs) are essential for scientific studies and possible future planetary-defense missions. Unlike motion around large celestial bodies, spacecraft dynamics near small, rotating asteroids are dominated by weak, highly irregular gravity fields. While full characterization of an asteroid’s shape enables high-fidelity gravitational modeling, such information is typically unavailable in realistic mission scenarios, and in-situ exploration is often required. As a result, the forces acting on the spacecraft cannot be modeled accurately in advance, leading to significant uncertainty in the equations of motion and challenging guidance and control strategies. To address these challenges, the local dynamical environment is modeled by representing the asteroid shape as a closed polyhedron and using a constant-density polyhedral gravity formulation to generate high-fidelity “truth” accelerations. The equations of motion are formulated in both inertial and uniformly rotating body-fixed frames, introducing the concept of effective potential. Equilibrium points are then computed from this effective potential, and their stability is assessed via Lyapunov’s direct method. To fully understand the dynamics in the vicinity of these bodies, zero-velocity curves are computed in addition to propagating the uncontrolled dynamics. This work then addresses the control challenge by combining online parameter estimation with adaptive control. An Integral Concurrent Learning (ICL) framework is used to estimate a reduced set of gravity expansion coefficients that represent the dominant perturbing effects of the asteroid’s irregular shape. These estimated coefficients are then fed into a Model Reference Adaptive Control (MRAC) law to track the desired trajectory. The reference model is defined by idealized two-body dynamics, while the spacecraft’s true dynamics are propagated using the high-fidelity polyhedral gravity model. By explicitly accounting for gravity-field irregularities through adaptive estimation and control, the proposed approach reduces dependence on detailed pre-mission asteroid characterization and ground-in-the-loop operations. The following coefficients, J2, C22, S22, C30, C31, S31, C32, and S32, are estimated for asteroid (101955) Bennu using a third-degree, second-order, spherical-harmonics-based regressor. Compared with published shape-based values, the results show that the ICL estimates converge toward the reference coefficients, with estimation errors ranging from 0.12% for J2 to 4.20% for S31. Overall, the proposed framework offers an additional path toward autonomous spacecraft control that can be applied to unvisited small irregular bodies where the gravity field is highly uncertain.
Scholarly Commons Citation
Diaz Rodrigo, Alvaro, "Adaptive Control Combined with Integral Concurrent Learning for Trajectory Tracking Near Asteroids" (2026). Doctoral Dissertations and Master's Theses. 982.
https://commons.erau.edu/edt/982
