Machine learning in viscoelastic fluids via energy-based kernel embedding

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorUniversity of Washington-
Autor(es): dc.contributorUniversidade Estadual Paulista (UNESP)-
Autor(es): dc.contributorCornell University-
Autor(es): dc.creatorOtto, Samuel E.-
Autor(es): dc.creatorOishi, Cassio M.-
Autor(es): dc.creatorAmaral, Fabio V.G.-
Autor(es): dc.creatorBrunton, Steven L.-
Autor(es): dc.creatorNathan Kutz, J.-
Data de aceite: dc.date.accessioned2025-08-21T21:53:20Z-
Data de disponibilização: dc.date.available2025-08-21T21:53:20Z-
Data de envio: dc.date.issued2025-04-29-
Data de envio: dc.date.issued2024-10-31-
Fonte completa do material: dc.identifierhttp://dx.doi.org/10.1016/j.jcp.2024.113371-
Fonte completa do material: dc.identifierhttps://hdl.handle.net/11449/296954-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/296954-
Descrição: dc.descriptionThe ability to measure differences in collected data is of fundamental importance for quantitative science and machine learning, motivating the establishment of metrics grounded in physical principles. In this study, we focus on the development of such metrics for viscoelastic fluid flows governed by a large class of linear and nonlinear stress models. To do this, we introduce energy-compatible families of kernel functions corresponding to a given viscoelastic stress model. Each kernel implicitly embeds flowfield snapshots into a Reproducing Kernel Hilbert Space (RKHS) in which distances and angles are computed and whose squared norm equals the total mechanical energy. Additionally, we present a solution to the preimage problem for these kernels, enabling accurate reconstruction of flowfields from their RKHS representations. Through numerical experiments on an unsteady viscoelastic lid-driven cavity flow, we demonstrate the utility of energy-compatible kernels for extracting energetically-dominant coherent structures in viscoelastic flows across a range of Reynolds and Weissenberg numbers. Specifically, the features extracted by Kernel Principal Component Analysis (KPCA) of flowfield snapshots using energy-compatible kernel functions yield reconstructions with superior accuracy in terms of mechanical energy compared to conventional methods such as ordinary Principal Component Analysis (PCA) with naïvely-defined state vectors or KPCA with ad-hoc choices of kernel functions. Our findings underscore the importance of principled choices of metrics in both scientific and machine learning investigations of complex fluid systems.-
Descrição: dc.descriptionFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)-
Descrição: dc.descriptionConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)-
Descrição: dc.descriptionArmy Research Office-
Descrição: dc.descriptionAI Institute in Dynamic Systems University of Washington-
Descrição: dc.descriptionDepartamento de Matemática e Computação Faculdade de Ciências e Tecnologia São Paulo State University-
Descrição: dc.descriptionDepartment of Mechanical Engineering University of Washington-
Descrição: dc.descriptionDepartment of Applied Mathematics University of Washington-
Descrição: dc.descriptionSibley School of Mechanical and Aerospace Engineering Cornell University-
Descrição: dc.descriptionDepartamento de Matemática e Computação Faculdade de Ciências e Tecnologia São Paulo State University-
Descrição: dc.descriptionFAPESP: 2013/07375-0-
Descrição: dc.descriptionFAPESP: 2021/07034-4-
Descrição: dc.descriptionFAPESP: 2021/13833-7-
Descrição: dc.descriptionFAPESP: 2023/06035-2-
Descrição: dc.descriptionCNPq: 305383/2019-1-
Descrição: dc.descriptionCNPq: 307228/2023-1-
Descrição: dc.descriptionArmy Research Office: W911NF-19-1-0045-
Idioma: dc.languageen-
Relação: dc.relationJournal of Computational Physics-
???dc.source???: dc.sourceScopus-
Palavras-chave: dc.subjectEnergy-based inner product-
Palavras-chave: dc.subjectKernel method-
Palavras-chave: dc.subjectMachine learning-
Palavras-chave: dc.subjectPrincipal component analysis-
Palavras-chave: dc.subjectReproducing kernel Hilbert space-
Palavras-chave: dc.subjectViscoelastic flow-
Título: dc.titleMachine learning in viscoelastic fluids via energy-based kernel embedding-
Tipo de arquivo: dc.typelivro digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.