Shallow Neural Networks for Fluid Flow Reconstruction with Limited Sensors

Abstract

In many applications, it is important to reconstruct a fluid flow field, or some other high-dimensional state, from limited measurements and limited data. In this work, we propose a shallow neural network-based learning methodology for such fluid flow reconstruction. Our approach 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. No prior knowledge is assumed to be available, and the estimation method is purely data-driven. We demonstrate the performance on three examples in fluid mechanics and oceanography, showing that this modern data-driven approach outperforms traditional modal approximation techniques which are commonly used for flow reconstruction. Not only does the proposed method show superior performance characteristics, it can also produce a comparable level of performance with traditional methods in the area, using significantly fewer sensors. Thus, the mathematical architecture is ideal for emerging global monitoring technologies where measurement data are often limited.

References in corpus (24)

• Adam: A Method for Stochastic Optimization
• Deep Convolutional Neural Network for Inverse Problems in Imaging
• Reconciling modern machine learning practice and the bias-variance trade-off
• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations
• Understanding deep learning requires rethinking generalization
• Super-resolution reconstruction of turbulent flows with machine learning
• A public turbulence database cluster and applications to study Lagrangian evolution of velocity increments in turbulence
• On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima
• Solving ill-posed inverse problems using iterative deep neural networks
• Data-Driven Sparse Sensor Placement for Reconstruction
• Estimating the intrinsic dimension of datasets by a minimal neighborhood information
• Data-Driven Forecasting of High-Dimensional Chaotic Systems with Long Short-Term Memory Networks
• Learning Activation Functions to Improve Deep Neural Networks
• A New Selection Operator for the Discrete Empirical Interpolation Method -- improved a priori error bound and extensions
• Robust flow field reconstruction from limited measurements via sparse representation
• Sparse Principal Component Analysis via Variable Projection
• Randomized Dynamic Mode Decomposition
• Randomized Matrix Decompositions using R
• Recovering missing CFD data for high-order discretizations using deep neural networks and dynamics learning
• Physics-informed Autoencoders for Lyapunov-stable Fluid Flow Prediction
• Theory of Deep Learning III: explaining the non-overfitting puzzle
• Forecasting Sequential Data using Consistent Koopman Autoencoders
• Prediction of wall-bounded turbulence from wall quantities using convolutional neural networks
• Memorization in Overparameterized Autoencoders

Cited by in corpus (47)

• Nonlinear mode decomposition with convolutional neural networks for fluid dynamics
• Assessment of supervised machine learning methods for fluid flows
• Super-Resolution Analysis via Machine Learning: A Survey for Fluid Flows
• Probabilistic neural networks for fluid flow surrogate modeling and data recovery
• The Neural Particle Method -- An Updated Lagrangian Physics Informed Neural Network for Computational Fluid Dynamics
• Convolutional neural networks for fluid flow analysis: toward effective metamodeling and low-dimensionalization
• From coarse wall measurements to turbulent velocity fields through deep learning
• Grasping Extreme Aerodynamics on a Low-Dimensional Manifold
• Data-driven recovery of hidden physics in reduced order modeling of fluid flows
• Experimental velocity data estimation for imperfect particle images using machine learning
• Generalization techniques of neural networks for fluid flow estimation
• Leveraging reduced-order models for state estimation using deep learning
• Sparse sensor reconstruction of vortex-impinged airfoil wake with machine learning
• A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence
• Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression
• Deep-learning assisted reduced order model for high-dimensional flow prediction from sparse data
• A Review of Mathematical and Computational Methods in Cancer Dynamics
• Data-driven transient lift attenuation for extreme vortex gust-airfoil interactions
• Forecasting Sequential Data using Consistent Koopman Autoencoders
• Reconstruction of three-dimensional turbulent flow structures using surface measurements for free-surface flows based on a convolutional neural network
• Semi Conditional Variational Auto-Encoder for Flow Reconstruction and Uncertainty Quantification from Limited Observations
• Velocity Reconstruction in Puffing Pool Fires with Physics-Informed Neural Networks
• Equation-free surrogate modeling of geophysical flows at the intersection of machine learning and data assimilation
• Informative and non-informative decomposition of turbulent flow fields
• Convolutional-network models to predict wall-bounded turbulence from wall quantities
• Deep learning fluid flow reconstruction around arbitrary two-dimensional objects from sparse sensors using conformal mappings
• Optimal sensor placement for reconstructing wind pressure field around buildings using compressed sensing
• Tensor-based flow reconstruction from optimally located sensor measurements
• Randomized Group-Greedy Method for Large-Scale Sensor Selection Problems
• Image features of a splashing drop on a solid surface extracted using a feedforward neural network
• Physics-Constrained Bayesian Neural Network for Fluid Flow Reconstruction with Sparse and Noisy Data
• Adaptive Activation Functions for Predictive Modeling with Sparse Experimental Data
• Reconstructing three-dimensional bluff body wake from sectional flow fields with convolutional neural networks
• Flow imaging as an alternative to pressure transducers through vision transformers and convolutional neural networks
• Inferring Displacement Fields from Sparse Measurements Using the Statistical Finite Element Method
• Reconstructing unsteady flows from sparse, noisy measurements with a physics-constrained convolutional neural network
• Wind Field Reconstruction with Adaptive Random Fourier Features
• Analyzing Koopman approaches to physics-informed machine learning for long-term sea-surface temperature forecasting
• Data assimilation empowered neural network parameterizations for subgrid processes in geophysical flows
• Robust State Estimation from Partial Out-Core Measurements with Shallow Recurrent Decoder for Nuclear Reactors
• ML enhanced measurement of the electrostatic charge distribution of powder conveyed through a duct
• Polynomial Ridge Flowfield Estimation
• Model order reduction with neural networks: Application to laminar and turbulent flows
• Convolutional causal learning for aerodynamic flows
• Reconstruction of three-dimensional turbulent flows from sparse and noisy planar measurements: A weight-sharing neural network approach
• An efficient offline sensor placement method for flow estimation
• Mapping surface height dynamics to subsurface flow physics in free-surface turbulent flow using a shallow recurrent decoder