Corpus ID: 221949087

A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder

  title={A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder},
  author={Youngkyu Kim and Y. Choi and David Widemann and T. Zohdi},
  • Youngkyu Kim, Y. Choi, +1 author T. Zohdi
  • Published 2020
  • Computer Science, Mathematics
  • ArXiv
  • Traditional linear subspace reduced order models (LS-ROMs) are able to accelerate physical simulations, in which the intrinsic solution space falls into a subspace with a small dimension, i.e., the solution space has a small Kolmogorov n-width. However, for physical phenomena not of this type, e.g., any advection-dominated flow phenomena, such as in traffic flow, atmospheric flows, and air flow over vehicles, a low-dimensional linear subspace poorly approximates the solution. To address cases… CONTINUE READING
    5 Citations
    Parameterized Neural Ordinary Differential Equations: Applications to Computational Physics Problems
    • PDF
    Component-wise reduced order model lattice-type structure design
    • PDF


    Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
    • 106
    • Highly Influential
    • PDF
    Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
    • 68
    • PDF
    Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data
    • 155
    • PDF
    fPINNs: Fractional Physics-Informed Neural Networks
    • 68
    • PDF
    Solving high-dimensional partial differential equations using deep learning
    • 355
    • PDF
    DeepXDE: A Deep Learning Library for Solving Differential Equations
    • 98
    • PDF
    SNS: A Solution-Based Nonlinear Subspace Method for Time-Dependent Model Order Reduction
    • 8
    PDE-Net: Learning PDEs from Data
    • 224
    • PDF