Extreme learning machine for reduced order modeling of turbulent geophysical flows.

@article{San2018ExtremeLM,
  title={Extreme learning machine for reduced order modeling of turbulent geophysical flows.},
  author={Omer San and Romit Maulik},
  journal={Physical review. E},
  year={2018},
  volume={97 4-1},
  pages={
          042322
        }
}
We investigate the application of artificial neural networks to stabilize proper orthogonal decomposition-based reduced order models for quasistationary geophysical turbulent flows. An extreme learning machine concept is introduced for computing an eddy-viscosity closure dynamically to incorporate the effects of the truncated modes. We consider a four-gyre wind-driven ocean circulation problem as our prototype setting to assess the performance of the proposed data-driven approach. Our framework… 
View on PubMed
link.aps.org
65 Citations
Highly Influential Citations
2
Background Citations
12
Methods Citations
22

65 Citations

Citation Type

Citation Type

A Hybrid Approach for Model Order Reduction of Barotropic Quasi-Geostrophic Turbulence
TLDR
A robust reduced-order modeling approach for near real-time prediction of mesoscale flows that combines physics-based projection methods with neural network closures to account for truncated modes and introduces a weighting parameter between the Galerkin projection and extreme learning machine models.
An artificial neural network framework for reduced order modeling of transient flows
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.
A non-intrusive reduced order modeling framework for quasi-geostrophic turbulence
TLDR
Findings suggest that the proposed fully nonintrusive ROM framework is robust and capable of predicting chaotic nonlinear fluid flows in an extremely efficient way compared to the conventional projection-based ROM framework.
Projection-Based Reduced Order Model and Machine Learning Closure for Transient Simulations of High-Re Flows
The paper presents a Projection-Based Reduced-Order Model for simulations of high Reynolds turbulent flows. The PBROM are enhanced by incorporating various models of turbulent viscosity and residual…
Model Order Reduction with Neural Networks: Application to Laminar and Turbulent Flows
TLDR
The capability of neural network-based model order reduction, i.e., autoencoder (AE), for fluid flows, which comprises of convolutional neural networks and multi-layer perceptrons is investigated, finds that the AE models are sensitive to the choice of the aforementioned parameters depending on the target flows.
Wavelet-Powered Neural Networks for Turbulence
TLDR
This paper takes advantage of the balance between strong mathematical foundations and the physical interpretability of wavelet theory to build a spatio-temporally reduced dynamical map which fuses wavelet based spatial decomposition with spatIO-temporal modeling based on Convolutional Long Short Term Memory (C-LSTM) architecture.
A dynamic closure modeling framework for model order reduction of geophysical flows
TLDR
A dynamic closure modeling approach has been derived to stabilize the projection-based reduced order models in the long-term evolution of forced-dissipative dynamical systems, considering the analogy between large eddy simulation (LES) and truncated modal projection.
Assessment of supervised machine learning methods for fluid flows
TLDR
Four machine learning architectures are examined in terms of their characteristics, accuracy, computational cost, and robustness for canonical flow problems, and the use of convolutional neural network is considered in the context of super-resolution analysis of two- dimensional cylinder wake, two-dimensional decaying isotropic turbulence, and three-dimensional turbulent channel flow.
Probabilistic neural networks for fluid flow model-order reduction and data recovery
TLDR
The present results indicate that the probabilistic neural network not only produces a machine-learning-based fluid flow model but also systematically quantifies the uncertainty therein to assist with model interpretability.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 44 REFERENCES
Neural network closures for nonlinear model order reduction
TLDR
This study puts forth a robust machine learning framework for projection-based reduced-order modeling of such nonlinear and nonstationary systems as the viscous Burgers equation and shows that the proposed model yields significant improvements in accuracy over the standard Galerkin projection methodology with a negligibly small computational overhead.
Dispersive–Dissipative Eddy Parameterization in a Barotropic Model
Abstract Recently a new class of coarse-grained equations, known as α models, have been proposed for the mean motion of an ideal incompressible fluid. The use of one such model to represent the…
Proper orthogonal decomposition closure models for turbulent flows: A numerical comparison
Approximate deconvolution large eddy simulation of a barotropic ocean circulation model
A stabilized proper orthogonal decomposition reduced-order model for large scale quasigeostrophic ocean circulation
TLDR
A stabilized proper orthogonal decomposition (POD) reduced-order model (ROM) framework is developed for the barotropic vorticity equation and shows that they could represent a computationally efficient tool for large scale oceanic simulations over long time intervals.
A reduced‐order approach to four‐dimensional variational data assimilation using proper orthogonal decomposition
Four‐dimensional variational data assimilation (4DVAR) is a powerful tool for data assimilation in meteorology and oceanography. However, a major hurdle in use of 4DVAR for realistic general…
An intrinsic stabilization scheme for proper orthogonal decomposition based low-dimensional models
TLDR
This work follows the methodology of direct numerical simulation followed by proper orthogonal decomposition of temporally sampled DNS data to derive temporal and spatial eigenfunctions and demonstrates that although these low-dimensional models reprod...
Proper orthogonal decomposition closure models for fluid flows: Burgers equation
This paper puts forth several closure models for the proper orthogonal decomposition (POD) reduced order modeling of fluid flows. These new closure models, together with other standard closure…
Estimation of Reynolds number for flows around cylinders with lattice Boltzmann methods and artificial neural networks.
TLDR
The present work investigates the application of artificial neural networks (ANNs) to estimate the Reynolds (Re) number for flows around a cylinder with good accuracy with errors of less than 4% in all the studied cases.
An Efficient Description of the Dynamics of Barotropic Flow
Abstract A potentially efficient description of the atmospheric circulation is investigated in the context of a barotropic spectral model, truncated to T21. The model circulation evolves around a…
...
1
2
3
4
5
...