Multi-fidelity prediction of spatiotemporal fluid flow

被引:9
|
作者
Mondal, Sudeepta [1 ]
Sarkar, Soumalya [1 ]
机构
[1] Raytheon Technol Res Ctr, 411 Silver Lane, E Hartford, CT 06118 USA
关键词
NEURAL-NETWORKS;
D O I
10.1063/5.0099197
中图分类号
O3 [力学];
学科分类号
08 ; 0801 ;
摘要
Data-driven prediction of spatiotemporal fields in fluid flow problems has received significant interest lately. However, the scarcity of data often plagues the accuracy of the prevalent approaches, particularly in situations where the cost associated with the generation of high-fidelity data is prohibitively expensive. Moreover, the data sources in many spatiotemporal fluid flow problems present a hierarchy of fidelities with respect to their computational cost/accuracy, such that the higher fidelity sources are more accurate and also more expensive than the low-fidelity sources. This article presents a novel multi-fidelity spatiotemporal modeling approach whereby the lower fidelity data source for a dynamical process is gainfully utilized in increasing the accuracy of predicting the higher fidelity fields, thereby reducing the overhead of high-fidelity simulations for data-driven prediction. The methodology is based on non-intrusive reduced order modeling using deep convolutional autoencoders, combined with a latent-space evolution framework based on multi-fidelity Gaussian processes. This framework results in probabilistic spatiotemporal predictions for unknown operating conditions of the dynamical system, which provides the end user with quantified levels of uncertainties associated with the data-driven predictions. The framework is validated on a prototypical fluid flow problem of interest: two-dimensional incompressible viscous flow around a cylinder.
引用
收藏
页数:18
相关论文
共 50 条
  • [1] Multi-fidelity prediction of fluid flow based on transfer learning using Fourier neural operator
    Lyu, Yanfang
    Zhao, Xiaoyu
    Gong, Zhiqiang
    Kang, Xiao
    Yao, Wen
    PHYSICS OF FLUIDS, 2023, 35 (07)
  • [2] Fast predictive multi-fidelity prediction with models of quantized fidelity levels
    Razi, Mani
    Kirby, Robert M.
    Narayan, Akil
    JOURNAL OF COMPUTATIONAL PHYSICS, 2019, 376 : 992 - 1008
  • [3] Multi-fidelity information fusion based on prediction of kriging
    Dong, Huachao
    Song, Baowei
    Wang, Peng
    Huang, Shuai
    STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION, 2015, 51 (06) : 1267 - 1280
  • [4] Multi-fidelity information fusion based on prediction of kriging
    Huachao Dong
    Baowei Song
    Peng Wang
    Shuai Huang
    Structural and Multidisciplinary Optimization, 2015, 51 : 1267 - 1280
  • [5] Multi-fidelity drag prediction for base bleed projectile
    Aziz, M. M.
    Ibrahim, A. Z.
    Ahmed, M. Y. M.
    Riad, A. M.
    19TH INTERNATIONAL CONFERENCE ON APPLIED MECHANICS AND MECHANICAL ENGINEERING (AMME-19), 2020, 973
  • [6] Multi-Fidelity Machine Learning Applied to Steady Fluid Flows
    Fuchi, Kazuko W.
    Wolf, Eric M.
    Makhija, David S.
    Schrock, Christopher R.
    Beran, Philip S.
    INTERNATIONAL JOURNAL OF COMPUTATIONAL FLUID DYNAMICS, 2022, 36 (07) : 618 - 640
  • [7] Multi-fidelity Gaussian process regression for prediction of random fields
    Parussini, L.
    Venturi, D.
    Perdikaris, P.
    Karniadakis, G. E.
    JOURNAL OF COMPUTATIONAL PHYSICS, 2017, 336 : 36 - 50
  • [8] Multi-fidelity prediction of molecular optical peaks with deep learning
    Greenman, Kevin P.
    Green, William H.
    Gomez-Bombarelli, Rafael
    CHEMICAL SCIENCE, 2022, 13 (04) : 1152 - 1162
  • [9] Multi-Fidelity Modeling in Rudder Efficiency Prediction of Large Deflection
    Han, Kunyu
    Cheng, Xide
    Nie, Peng
    Qi, Bo
    Ouyang, Xuyu
    Ship Building of China, 2023, 64 (06) : 250 - 260
  • [10] Multi-fidelity uncertainty quantification of particle deposition in turbulent flow
    Yao, Yuan
    Huan, Xun
    Capecelatro, Jesse
    JOURNAL OF AEROSOL SCIENCE, 2022, 166