Reduced Order Model of Diffusion Flames Based on Multi-scale Data from Detailed CFD: The Impact of preprocessing - Modelling and Optinal Decision for Uncertain Systems Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2022

Reduced Order Model of Diffusion Flames Based on Multi-scale Data from Detailed CFD: The Impact of preprocessing

Résumé

Machine learning techniques, such as reduced order models (ROM), have demonstrated low cost when creating models of complex systems while aiming at the same accuracy as high fidelity models, such as Computational Fluid Dynamics (CFD). Here, ROM are created using CFD simulations of non-premixed laminar flame detailed chemistry and transport. The data obtained for variable fuel velocity are reduced using the singular value decomposition (SVD) and then the genetic aggregation response surface (GARS) algorithm is applied to predict the properties fields for an arbitrary velocity. This work analyzes the effect of different data preprocessing approaches on the ROM, i.e., (1) the properties treated as a uncoupled or as a coupled system, (2) normalization of different properties, and (3) the logarithm of the chemical species. For all constructed ROM the energy content of the reduction process and the reconstructed fields of the flame properties are analyzed. This evidences the slower SVD modes convergence of the uncoupled ROM, and the faster one when the logarithm preprocessing is applied. Learning is shown to be achieved with a smaller number of modes for two of the coupled ROM and the ROM using the logarithm. The reconstruction of the mass fraction fields is characterized by regions of negative values, which underscores that the baseline ROM methodology does not preserve the properties monotonicity, positivity and delimitation. The logarithm preprocessing enables to overcome these problems and to accurately reproduce the original data.
Fichier principal
Vignette du fichier
ROM_diffusion_flame.pdf (1.29 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03738492 , version 1 (26-07-2022)
hal-03738492 , version 2 (02-12-2022)

Identifiants

  • HAL Id : hal-03738492 , version 1

Citer

Nicole Lopes Junqueira, Louise da Costa Ramos, Luis Fernando Figueira da Silva. Reduced Order Model of Diffusion Flames Based on Multi-scale Data from Detailed CFD: The Impact of preprocessing. 2022. ⟨hal-03738492v1⟩
99 Consultations
71 Téléchargements

Partager

Gmail Facebook X LinkedIn More