Solving multiphysics-based inverse problems with learned surrogates and constraints

被引:0
|
作者
Ziyi Yin
Rafael Orozco
Mathias Louboutin
Felix J. Herrmann
机构
[1] Georgia Institute of Technology,
关键词
Fourier neural operators; Normalizing flows; Multiphysics; Deep learning; Learned surrogates; Learned constraints; Inverse problems;
D O I
暂无
中图分类号
学科分类号
摘要
Solving multiphysics-based inverse problems for geological carbon storage monitoring can be challenging when multimodal time-lapse data are expensive to collect and costly to simulate numerically. We overcome these challenges by combining computationally cheap learned surrogates with learned constraints. Not only does this combination lead to vastly improved inversions for the important fluid-flow property, permeability, it also provides a natural platform for inverting multimodal data including well measurements and active-source time-lapse seismic data. By adding a learned constraint, we arrive at a computationally feasible inversion approach that remains accurate. This is accomplished by including a trained deep neural network, known as a normalizing flow, which forces the model iterates to remain in-distribution, thereby safeguarding the accuracy of trained Fourier neural operators that act as surrogates for the computationally expensive multiphase flow simulations involving partial differential equation solves. By means of carefully selected experiments, centered around the problem of geological carbon storage, we demonstrate the efficacy of the proposed constrained optimization method on two different data modalities, namely time-lapse well and time-lapse seismic data. While permeability inversions from both these two modalities have their pluses and minuses, their joint inversion benefits from either, yielding valuable superior permeability inversions and CO2 plume predictions near, and far away, from the monitoring wells.
引用
收藏
相关论文
共 50 条
  • [1] Solving multiphysics-based inverse problems with learned surrogates and constraints
    Yin, Ziyi
    Orozco, Rafael
    Louboutin, Mathias
    Herrmann, Felix J.
    [J]. ADVANCED MODELING AND SIMULATION IN ENGINEERING SCIENCES, 2023, 10 (01)
  • [2] Discretization of Learned NETT Regularization for Solving Inverse Problems
    Antholzer, Stephan
    Haltmeier, Markus
    [J]. JOURNAL OF IMAGING, 2021, 7 (11)
  • [3] A multiphysics-based artificial neural networks model for atherosclerosis
    Soleimani, M.
    Dashtbozorg, B.
    Mirkhalaf, M.
    Mirkhalaf, S. M.
    [J]. HELIYON, 2023, 9 (07)
  • [4] COMSOL multiphysics-based modeling approach to solar cell development
    Beepat, Kevin Gurbani
    Sharma, Davinder Pal
    Pathak, Dinesh
    Mahajan, Aman
    [J]. INTERNATIONAL JOURNAL OF MODERN PHYSICS B, 2023, 37 (12):
  • [5] Solving Multiphysics, Multiparameter, Multimodal Inverse Problems: An Application to NMR Relaxation in Porous Media
    Li, Rupeng
    Shikhov, Igor
    Arns, Christoph H.
    [J]. PHYSICAL REVIEW APPLIED, 2021, 15 (05)
  • [6] Multiphysics-based Design Optimization of Medium Frequency Transformer with Experimental Validation
    Olowu, Temitayo O.
    Jafari, Hassan
    Sarwat, Arif
    [J]. 2021 THIRTY-SIXTH ANNUAL IEEE APPLIED POWER ELECTRONICS CONFERENCE AND EXPOSITION (APEC 2021), 2021, : 863 - 868
  • [7] Multiscale Multiphysics-Based Modeling and Analysis on the Tool Wear in Micro Drilling
    Niu, Zhichao
    Cheng, Kai
    [J]. JOURNAL OF MULTISCALE MODELLING, 2016, 7 (01)
  • [8] Iteratively solving linear inverse problems under general convex constraints
    Daubechies, Ingrid
    Teschke, Gerd
    Vese, Luminita
    [J]. INVERSE PROBLEMS AND IMAGING, 2007, 1 (01) : 29 - 46
  • [9] Solving SVM inverse problems based on clustering
    Wang, Xizhao
    Lu, Shuxia
    Zhu, Ruixian
    [J]. 2007 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN AND CYBERNETICS, VOLS 1-8, 2007, : 888 - 893
  • [10] Development of multiscale multiphysics-based modelling and simulations with the application to precision machining of aerofoil structures
    Shao, Yizhi
    Adetoro, Oluwamayokun Bamidele
    Cheng, Kai
    [J]. ENGINEERING COMPUTATIONS, 2021, 38 (03) : 1330 - 1349