From perturbative methods to machine learning techniques in space science

• Published in: Bollettino della Unione matematica italiana (2008) Vol. 18; no. 1; pp. 149 - 166
• Main Author: Celletti, Alessandra
• Format: Journal Article
• Language: English
• Published: Heidelberg Springer Nature B.V 01.03.2025
• Subjects: Celestial mechanics; Machine learning; Orbital mechanics; Perturbation theory; Space debris; Spacecraft breakup
• ISSN: 1972-6724; 2198-2759
• DOI: 10.1007/s40574-024-00422-x

Perturbation theory is a very useful tool to investigate the dynamics of models in space science. We start by presenting some results obtained implementing classical perturbation theory to investigate the motion of space debris, which are objects that populate the sky around the Earth after a satellite break-up event. When dealing with two or more break-up events, a clusterization of the fragments can be computed using machine learning techniques. We also present the celebrated KAM theory for symplectic and conformally symplectic systems. We recall several computer-assisted results in Celestial Mechanics in conservative and dissipative settings. Finally, we consider the spin-orbit problem and we show how machine learning methods can be conveniently used to classify regular and chaotic motions.