Deep learning in fluid dynamics

By J. Nathan Kutz. Journal of Fluid Mechanics, 2017

811 citations recorded by the source corpus.

Subfields: Computational Mechanics, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/1071cf2e59793862cb5cedb21ae99f1356aff26c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Statistical and Nonlinear Physics

Citations in the corpus

Integrating Scientific Knowledge with Machine Learning for Engineering and Environmental Systems (2020, background): “…which have already found tremendous success in commercial applications, are beginning to play an important role in advancing scientific discovery in environmental and engineering domains traditionally dominated by mechanistic (e.g. first principle) models [30, 124, 128, 141, 142, 157, 232, 283].”
Subgrid modelling for two-dimensional turbulence using neural networks (2018, background): “An excellent review of the strengths and opportunities of using artiﬁcial neural networks for ﬂuid dynamics applications may also be found in Kutz (2017).”
Machine learning methods for turbulence modeling in subsonic flows around airfoils (2018, background): “It is perhaps time for the turbulence modeling community to adopt challenge datasets [26].”
Machine-learning-based spatio-temporal super resolution reconstruction of turbulent flows (2020, background): “Applications of machine learning for turbulence modeling have been particularly active in ﬂuid dynamics (Duraisamy et al. 2019; Kutz 2017).”
Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification (2018, background): “Deep learning has been explored as a competitive methodology across fields such as fluid mechanics [21], hydrology [22,23], bioinformatics [24], high energy physics [25] and other.”
Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers (2020, background): “Deep learning methods have received significant amounts of attention in this area [31].”
Enhancing computational fluid dynamics with machine learning (2021, methodology): “As argued by Kutz [44], machine learning can produce more sophisticated models for the Reynolds stresses by using adequate data, in particular if the examples used for training represent a sufficiently rich set of flow configurations.”
Unsupervised deep learning for super-resolution reconstruction of turbulence (2020, methodology): “Ling et al. (2016) proposed a tensor-based NN by embedding the Galilean invariance of a Reynolds-averaged Navier–Stokes (RANS) model, showing a greater performance improvement than linear and nonlinear eddy viscosity models.”
Integrating Physics-Based Modeling with Machine Learning: A Survey (2020, background): “Machine learning (ML) models, which have already found tremendous success in commercial applications, are beginning to play an important role in advancing scientific discovery in domains traditionally dominated by mechanistic (e.g. first principle) models [22, 112, 117, 126, 127, 138, 214, 261].”
Nonlinear mode decomposition with convolutional neural networks for fluid dynamics (2019, background): “In recent years, machine learning has been widely applied in the field of fluid dynamics, and is highly expected for its strong ability to account for nonlinearity (Brunton & Noack 2015; Kutz 2017; Taira et al. 2019; Brunton et al. 2019).”
Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control (2018, background): “It has been speculated that Deep Neural Networks will bring advances also to fluid mechanics (Kutz, 2017), but until this day those have been limited to a few applications, such as the definition of reduced order models (Wang et al., 2018), the effective control of swimmers (Verma et al., 2018), or…”
Informed Machine Learning – A Taxonomy and Survey of Integrating Prior Knowledge into Learning Systems (2019, background): “In addition to these classical domains, machine learning and in particular deep learning are increasingly important and successful in engineering and the sciences [5], [6], [7].”
Constrained sparse Galerkin regression (2016, methodology): “…have long been used for flow modelling (Milano & Koumoutsakos 2002; Zhang & Duraisamy 2015) and control (Lee et al. 1997), and recently deep neural networks (Krizhevsky, Sutskever & Hinton 2012) have been used for Reynolds averaged turbulence modelling (Ling, Kurzawski & Templeton 2016; Kutz 2017).”
The Role of Machine Learning in the Understanding and Design of Materials (2020, background): “165 Adapting these for process modeling will not only drop the computational costs but also allow the addition of more levels of complexity in modeling by including nonideal e ﬀ ects that are often ignored in such modeling.”