Physics-Informed Localized Learning for Advection-Diffusion-Reaction Systems (2305.03774v2)
Abstract: The global push to advance Carbon Capture and Sequestration initiatives and green energy solutions, such as geothermal, have thrust new demands upon the current state-of-the-art subsurface fluid simulators. The requirement to be able to simulate a large order of reservoir states simultaneously, in a short period of time, has opened the door of opportunity for the application of machine learning techniques for surrogate modelling. We propose a novel physics-informed and boundary condition-aware Localized Learning method which extends the Embed-to-Control (E2C) and Embed-to-Control and Observe (E2CO) models to learn local representations of global state variables in an Advection-Diffusion Reaction system. Trained on reservoir simulation data, we show that our model is able to predict future states of the system, for a given set of controls, to a great deal of accuracy with only a fraction of the available information. It hence reduces training times significantly compared to the original E2C and E2CO models, lending to its benefit in application to optimal control problems.
- doi: SPE-173252-MS. D031S011R007.
- Deep learning-based proxy models to simulate subsurface flow of three-dimensional reservoir systems. 2022(1):1–32, 2022. ISSN 2214-4609. doi: https://doi.org/10.3997/2214-4609.202244049.
- Model reduction and neural networks for parametric pdes, 2021.
- Linearized reduced-order models for subsurface flow simulation. J. Comput. Phys., 229(3):681–700, 2 2010. ISSN 0021-9991. doi: 10.1016/j.jcp.2009.10.004.
- Physics-aware deep-learning-based proxy reservoir simulation model equipped with state and well output prediction. Frontiers in Applied Mathematics and Statistics, 7, 2021. ISSN 2297-4687. doi: 10.3389/fams.2021.651178.
- Reduced-order modeling for compositional simulation using trajectory piecewise linearization. SPE Journal, 19, 02 2013. doi: 10.2118/163634-MS.
- Constraint reduction procedures for reduced-order subsurface flow models based on pod–tpwl. International Journal for Numerical Methods in Engineering, 103(1):1–30, 2015. doi: https://doi.org/10.1002/nme.4874.
- Enhanced linearized reduced-order models for subsurface flow simulation. J. Comput. Phys., 230(23):8313–8341, 9 2011. ISSN 0021-9991. doi: 10.1016/j.jcp.2011.06.007.
- Deep residual learning for image recognition, 2015.
- Reduced-order modeling of co2 storage operations. International Journal of Greenhouse Gas Control, 68:49–67, 2018. ISSN 1750-5836. doi: https://doi.org/10.1016/j.ijggc.2017.08.017.
- Deep-learning-based surrogate model for reservoir simulation with time-varying well controls. Journal of Petroleum Science and Engineering, 192:107273, 2020. ISSN 0920-4105. doi: https://doi.org/10.1016/j.petrol.2020.107273.
- Neural operator: Learning maps between function spaces, 2023.
- Multipole graph neural operator for parametric partial differential equations, 2020a.
- Neural operator: Graph kernel network for partial differential equations, 2020b.
- Fourier neural operator for parametric partial differential equations, 2021.
- Physics-informed neural operator for learning partial differential equations, 2022.
- Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators. Nature Machine Intelligence, 3(3):218–229, mar 2021. doi: 10.1038/s42256-021-00302-5.
- U-net: Convolutional networks for biomedical image segmentation, 2015.
- Reduced-order modeling for thermal recovery processes. Computational Geosciences, 18(3):401–415, 8 2014. ISSN 1573-1499. doi: 10.1007/s10596-013-9369-8.
- Embed to control: A locally linear latent dynamics model for control from raw images. In Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2, NIPS’15, pp. 2746–2754, Cambridge, MA, USA, 2015. MIT Press.
- Bayesian deep convolutional encoder–decoder networks for surrogate modeling and uncertainty quantification. Journal of Computational Physics, 366:415–447, 2018. ISSN 0021-9991. doi: https://doi.org/10.1016/j.jcp.2018.04.018.