Machine-learning approach for mapping stable orbits around planets

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorUniversidade Estadual Paulista (UNESP)-
Autor(es): dc.contributorEberhard Karls Universität Tübingen-
Autor(es): dc.contributorSorbonne Université-
Autor(es): dc.creatorPinheiro, Tiago F. L. L.-
Autor(es): dc.creatorSfair, Rafael-
Autor(es): dc.creatorRamon, Giovana-
Data de aceite: dc.date.accessioned2025-08-21T19:35:52Z-
Data de disponibilização: dc.date.available2025-08-21T19:35:52Z-
Data de envio: dc.date.issued2025-04-29-
Data de envio: dc.date.issued2024-12-31-
Fonte completa do material: dc.identifierhttp://dx.doi.org/10.1051/0004-6361/202451831-
Fonte completa do material: dc.identifierhttps://hdl.handle.net/11449/297463-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/297463-
Descrição: dc.descriptionContext. Numerical N-body simulations are typically employed to map stability regions around exoplanets. This provides insights into the potential presence of satellites and ring systems. Aims. We used machine-learning (ML) techniques to generate predictive maps of stable regions surrounding a hypothetical planet. This approach can also be applied to planet-satellite systems, planetary ring systems, and other similar systems. Methods. From a set of 105 numerical simulations, each incorporating nine orbital features for the planet and test particle, we created a comprehensive dataset of three-body problem outcomes (star-planet-test particle). Simulations were classified as stable or unstable based on the stability criterion that a particle must remain stable over a time span of 104 orbital periods of the planet. Various ML algorithms were compared and fine-tuned through hyperparameter optimization to identify the most effective predictive model. All tree-based algorithms demonstrated a comparable accuracy performance. Results. The optimal model employs the extreme gradient boosting algorithm and achieved an accuracy of 98.48%, with 94% recall and precision for stable particles and 99% for unstable particles. Conclusions. ML algorithms significantly reduce the computational time in three-body simulations. They are approximately 105 times faster than traditional numerical simulations. Based on the saved training models, predictions of entire stability maps are made in less than a second, while an equivalent numerical simulation can take up to a few days. Our ML model results will be accessible through a forthcoming public web interface, which will facilitate a broader scientific application.-
Descrição: dc.descriptionCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)-
Descrição: dc.descriptionGrupo de Dinâmica Orbital e Planetologia São Paulo State University UNESP, Guaratinguetá-
Descrição: dc.descriptionEberhard Karls Universität Tübingen, Auf der Morgenstelle, 10-
Descrição: dc.descriptionLESIA Observatoire de Paris Université PSL CNRS Sorbonne Université, 5 place Jules Janssen-
Descrição: dc.descriptionGrupo de Dinâmica Orbital e Planetologia São Paulo State University UNESP, Guaratinguetá-
Descrição: dc.descriptionCAPES: 001-
Idioma: dc.languageen-
Relação: dc.relationAstronomy and Astrophysics-
???dc.source???: dc.sourceScopus-
Palavras-chave: dc.subjectMethods: numerical-
Palavras-chave: dc.subjectPlanets and satellites: dynamical evolution and stability-
Título: dc.titleMachine-learning approach for mapping stable orbits around planets-
Tipo de arquivo: dc.typelivro digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.