Eventos Anais de eventos
ENCIT 2018
Brazilian Congress of Thermal Sciences and Engineering
Application of Deep Learning and Proper Orthogonal Decomposition for Reduced Order Models of Unsteady Flows
Submission Author:
Hugo Felippe da Silva Lui , SP
Co-Authors:
Hugo Felippe da Silva Lui, William Wolf
Presenter: Hugo Felippe da Silva Lui
doi://10.26678/ABCM.ENCIT2018.CIT18-0814
Abstract
In this work, we present a numerical methodology which combines flow modal decomposition via proper orthogonal decomposition and sparse regression using deep neural networks. The framework is implemented in the context of the sparse identification of non-linear dynamics algorithm recently proposed in the literature. The framework is applied for the construction of reduced order models of unsteady compressible flows. The results demonstrate that the technique provides accurate and stable reconstructions of the full order model beyond the training window of the deep neural network. In this paper, we describe the numerical techniques employed and show the results obtained by the current methodology.
Keywords
reduced order models, Deep learning, Proper Orthogonal Decomposition, Computational Fluid Dynamic (CFD)
