Continuous conditional generative adversarial networks for data-driven solutions of poroelasticity with heterogeneous material properties. (October 2022)

• Record Type: Journal Article
• Authors: Kadeethum, T.; O'Malley, D.; Choi, Y.; Viswanathan, H.S.; Bouklas, N.; Yoon, H.
• Is Part Of: Computers & geosciences. Volume 167(2022)
• Publication Date: 2022-10
• Subjects: Non-intrusive -- Data-driven -- Reduced order modeling -- Generative adversarial networks -- Finite element -- Poroelasticity; Environmental policy -- Periodicals; 550.5
• DOI: 10.1016/j.cageo.2022.105212
• Languages: English
• ISSNs: 0098-3004

Abstract

Abstract: Machine learning-based data-driven modeling can allow computationally efficient time-dependent solutions of PDEs, such as those that describe subsurface multiphysical problems. In this work, our previous approach (Kadeethum et al., 2021d) of conditional generative adversarial networks (cGAN) developed for the solution of steady-state problems involving highly heterogeneous material properties is extended to time-dependent problems by adopting the concept of continuous cGAN (CcGAN). The CcGAN that can condition continuous variables is developed to incorporate the time domain through either element-wise addition or conditional batch normalization. Moreover, this framework can handle training data that contain different timestamps and then predict timestamps that do not exist in the training data. As a numerical example, the transient response of the coupled poroelastic process is studied in two different permeability fields: Zinn & Harvey transformation and a bimodal transformation. The proposed CcGAN uses heterogeneous permeability fields as input parameters while pressure and displacement fields over time are model output. Our results show that the model provides sufficient accuracy with computational speed-up. This robust framework will enable us to perform real-time reservoir management and robust uncertainty quantification in poroelastic problems. Highlights: Continuous conditional generative adversarial network (CcGAN) for poroelastic process CcGAN is generalized for heterogeneous fields and predictions at any timesteps Accurate results with up to 10K times speed-up compared to a finite element solver