Towards Physics-informed Deep Learning for Turbulent Flow Prediction

  title={Towards Physics-informed Deep Learning for Turbulent Flow Prediction},
  author={Rui Wang and K. Kashinath and M. Mustafa and Adrian Albert and R. Yu},
  journal={Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining},
  • Rui Wang, K. Kashinath, +2 authors R. Yu
  • Published 2020
  • Computer Science, Physics, Mathematics
  • Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining
  • While deep learning has shown tremendous success in a wide range of domains, it remains a grand challenge to incorporate physical principles in a systematic manner to the design, training, and inference of such models. In this paper, we aim to predict turbulent flow by learning its highly nonlinear dynamics from spatiotemporal velocity fields of large-scale fluid flow simulations of relevance to turbulence modeling and climate modeling. We adopt a hybrid approach by marrying two well… CONTINUE READING
    21 Citations
    Deep Orthogonal Decompositions for Convective Nowcasting
    • PDF
    Embedding Hard Physical Constraints in Neural Network Coarse-Graining of 3D Turbulence
    • 14
    • PDF
    Incorporating Symmetry into Deep Dynamics Models for Improved Generalization
    • 9
    • PDF
    MeshfreeFlowNet: A Physics-Constrained Deep Continuous Space-Time Super-Resolution Framework
    • 8
    • PDF
    Frequency-compensated PINNs for Fluid-dynamic Design Problems
    • PDF
    Data‐Driven Equation Discovery of Ocean Mesoscale Closures
    • 4
    • PDF
    Bridging Physics-based and Data-driven modeling for Learning Dynamical Systems
    • PDF


    Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge
    • 88
    • Highly Influential
    • PDF
    Generative Adversarial Nets
    • 20,163
    • Highly Influential
    • PDF
    Deep Residual Learning for Image Recognition
    • 57,702
    • Highly Influential
    • PDF