Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics

Abstract This work presents a machine learning based method for bi-fidelity modelling. The method, a Knowledge Based Neural Network (KBaNN), performs a local, additive correction to the outputs of a coarse computational model and can be used to emulate either experimental data or the output of a mor...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Nick Pepper, Audrey Gaymann, Sanjiv Sharma, Francesco Montomoli
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/3db603d25349419baf2ed4038d973d20
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:3db603d25349419baf2ed4038d973d20
record_format dspace
spelling oai:doaj.org-article:3db603d25349419baf2ed4038d973d202021-12-02T15:33:02ZLocal bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics10.1038/s41598-021-93280-y2045-2322https://doaj.org/article/3db603d25349419baf2ed4038d973d202021-07-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-93280-yhttps://doaj.org/toc/2045-2322Abstract This work presents a machine learning based method for bi-fidelity modelling. The method, a Knowledge Based Neural Network (KBaNN), performs a local, additive correction to the outputs of a coarse computational model and can be used to emulate either experimental data or the output of a more accurate, but expensive, computational model. An advantage of the method is that it can scale easily with the number of input and output features. This allows bi-fidelity modelling approaches to be applied to a wide variety of problems, for instance in the bi-fidelity modelling of fields. We demonstrate this aspect in this work through an application to Computational Fluid Dynamics, in which local corrections to a velocity field are performed by the KBaNN to account for mesh effects. KBaNNs were trained to make corrections to the free-stream velocity field and the boundary layer. They were trained on a limited data-set consisting of simple two-dimensional flows. The KBaNNs were then tested on a flow over a more complex geometry, a NACA 2412 airfoil. It was demonstrated that the KBaNNs were still able to provide a local correction to the velocity field which improved its accuracy. The ability of the KBaNNs to generalise to flows around new geometries that share similar physics is encouraging. Through knowledge based neural networks it may be possible to develop a system for bi-fidelity, computer based design which uses data from past simulations to inform its predictions.Nick PepperAudrey GaymannSanjiv SharmaFrancesco MontomoliNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-11 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Nick Pepper
Audrey Gaymann
Sanjiv Sharma
Francesco Montomoli
Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
description Abstract This work presents a machine learning based method for bi-fidelity modelling. The method, a Knowledge Based Neural Network (KBaNN), performs a local, additive correction to the outputs of a coarse computational model and can be used to emulate either experimental data or the output of a more accurate, but expensive, computational model. An advantage of the method is that it can scale easily with the number of input and output features. This allows bi-fidelity modelling approaches to be applied to a wide variety of problems, for instance in the bi-fidelity modelling of fields. We demonstrate this aspect in this work through an application to Computational Fluid Dynamics, in which local corrections to a velocity field are performed by the KBaNN to account for mesh effects. KBaNNs were trained to make corrections to the free-stream velocity field and the boundary layer. They were trained on a limited data-set consisting of simple two-dimensional flows. The KBaNNs were then tested on a flow over a more complex geometry, a NACA 2412 airfoil. It was demonstrated that the KBaNNs were still able to provide a local correction to the velocity field which improved its accuracy. The ability of the KBaNNs to generalise to flows around new geometries that share similar physics is encouraging. Through knowledge based neural networks it may be possible to develop a system for bi-fidelity, computer based design which uses data from past simulations to inform its predictions.
format article
author Nick Pepper
Audrey Gaymann
Sanjiv Sharma
Francesco Montomoli
author_facet Nick Pepper
Audrey Gaymann
Sanjiv Sharma
Francesco Montomoli
author_sort Nick Pepper
title Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
title_short Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
title_full Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
title_fullStr Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
title_full_unstemmed Local bi-fidelity field approximation with Knowledge Based Neural Networks for Computational Fluid Dynamics
title_sort local bi-fidelity field approximation with knowledge based neural networks for computational fluid dynamics
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/3db603d25349419baf2ed4038d973d20
work_keys_str_mv AT nickpepper localbifidelityfieldapproximationwithknowledgebasedneuralnetworksforcomputationalfluiddynamics
AT audreygaymann localbifidelityfieldapproximationwithknowledgebasedneuralnetworksforcomputationalfluiddynamics
AT sanjivsharma localbifidelityfieldapproximationwithknowledgebasedneuralnetworksforcomputationalfluiddynamics
AT francescomontomoli localbifidelityfieldapproximationwithknowledgebasedneuralnetworksforcomputationalfluiddynamics
_version_ 1718387137120829440