Corpus ID: 227151170

Generalization techniques of neural networks for fluid flow estimation

@article{Morimoto2020GeneralizationTO,
  title={Generalization techniques of neural networks for fluid flow estimation},
  author={Masaki Morimoto and Kai Fukami and Kai Zhang and K. Fukagata},
  journal={arXiv: Fluid Dynamics},
  year={2020}
}
We demonstrate several techniques to encourage practical uses of neural networks for fluid flow estimation. In the present paper, three perspectives which are remaining challenges for applications of machine learning to fluid dynamics are considered: 1. interpretability of machine-learned results, 2. bulking out of training data, and 3. generalizability of neural networks. For the interpretability, we first demonstrate two methods to observe the internal procedure of neural networks, i.e… Expand
Convolutional neural networks for fluid flow analysis: toward effective metamodeling and low-dimensionalization
We focus on a convolutional neural network (CNN), which has recently been utilized for fluid flow analyses, from the perspective on the influence of various operations inside it by considering someExpand
Supervised convolutional network for three-dimensional fluid data reconstruction from sectional flow fields with adaptive super-resolution assistance
The recent development of high-performance computing enables us to generate spatiotemporal high-resolution data of nonlinear dynamical systems and to analyze them for deeper understanding of theirExpand
Inserting machine-learned virtual wall velocity for large-eddy simulation of turbulent channel flows
We propose a supervised-machine-learning-based wall model for coarse-grid wall-resolved largeeddy simulation (LES). Our consideration is made on LES of turbulent channel flows with a first grid pointExpand
Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning
TLDR
The use of Voronoi tessellation is considered to obtain a structured-grid representation from sensor locations enabling the computationally tractable use of convolutional neural networks for real-time global field estimation. Expand
Comparison of linear regressions and neural networks for fluid flow problems assisted with error-curve analysis
Neural networks (NNs) and linear stochastic estimation (LSE) have widely been utilized as powerful tools for fluid-flow regressions. We investigate fundamental differences between them consideringExpand

References

SHOWING 1-10 OF 77 REFERENCES
Shallow Learning for Fluid Flow Reconstruction with Limited Sensors and Limited Data
TLDR
This work proposes a shallow neural network-based learning methodology for fluid flow reconstruction that learns an end-to-end mapping between the sensor measurements and the high-dimensional fluid flow field, without any heavy preprocessing on the raw data. Expand
Predictions of turbulent shear flows using deep neural networks.
In the present work we assess the capabilities of neural networks to predict temporally evolving turbulent flows. In particular, we use the nine-equation shear flow model by Moehlis et al. [New J.Expand
Prediction of turbulent heat transfer using convolutional neural networks
With the recent rapid development of artificial intelligence (AI) and wide applications in many areas, some fundamental questions in turbulence research can be addressed, such as: ‘Can turbulence beExpand
Construction of reduced-order models for fluid flows using deep feedforward neural networks
TLDR
The DNN approach is able to learn transient features of the flow and presents more accurate and stable long-term predictions compared to sparse regression and the robustness of the current ROMs constructed via DNNs is demonstrated through a comparison with sparse regression. Expand
Machine Learning for Fluid Mechanics
TLDR
Fundamental machine learning methodologies are outlined and their uses for understanding, modeling, optimizing, and controlling fluid flows are discussed and the strengths and limitations are addressed from the perspective of scientific inquiry. Expand
Deep learning in fluid dynamics
  • J. Kutz
  • Computer Science
  • Journal of Fluid Mechanics
  • 2017
TLDR
Although neural networks have been applied previously to complex fluid flows, the article featured here is the first to apply a true DNN architecture, specifically to Reynolds averaged Navier Stokes turbulence models, suggesting that DNNs may play a critically enabling role in the future of modelling complex flows. Expand
Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations
TLDR
Hidden fluid mechanics (HFM), a physics-informed deep-learning framework capable of encoding the Navier-Stokes equations into the neural networks while being agnostic to the geometry or the initial and boundary conditions, is developed. Expand
A review on Deep Reinforcement Learning for Fluid Mechanics
TLDR
Understanding of DRL capabilities along with state-of-the-art applications in fluid dynamics to researchers wishing to address new problems with these methods is provided. Expand
Special issue on machine learning and data-driven methods in fluid dynamics
TLDR
This special issue is excited to introduce this special issue highlighting a number of promising avenues of ongoing research to integrate machine learning and data-driven techniques in the field of fluid dynamics, and attempting to provide a broader perspective, outlining recent successes, opportunities, and open challenges. Expand
Model identification of reduced order fluid dynamics systems using deep learning
Summary This paper presents a novel model reduction method: deep learning reduced order model, which is based on proper orthogonal decomposition and deep learning methods. The deep learning approachExpand
...
1
2
3
4
5
...