Publicado 4 números por año
ISSN Imprimir: 2689-3967
ISSN En Línea: 2689-3975
Indexed in
LIMITATIONS OF PHYSICS INFORMED MACHINE LEARNING FOR NONLINEAR TWO-PHASE TRANSPORT IN POROUS MEDIA
SINOPSIS
Deep learning techniques have recently been applied to a wide range of computational physics problems. In this paper, we focus on developing a physics-based approach that enables the neural network to learn the solution of a dynamic fluid-flow problem governed by a nonlinear partial differential equation (PDE). The main idea of physics informed machine learning (PIML) approaches is to encode the underlying physical law (i.e., the PDE) into the neural network as prior information. We investigate the applicability of the PIML approach to the forward problem of immiscible two-phase fluid transport in porous media, which is governed by a nonlinear first-order hyperbolic PDE subject to initial and boundary data. We employ the PIML strategy to solve this forward problem without any additional labeled data in the interior of the domain. Particularly, we are interested in non-convex flux functions in the PDE, where the solution involves shocks and mixed waves (shocks and rarefactions). We have found that such a PIML approach fails to provide reasonable approximations to the solution in the presence of shocks in the saturation field. We investigated several architectures and experimented with a large number of neural-network parameters, and the overall finding is that PIML strategies that employ the nonlinear hyperbolic conservation equation in the loss function are inadequate. However, we have found that employing a parabolic form of the conservation equation, whereby a small amount of diffusion is added, the neural network is consistently able to learn accurate approximation of the solutions containing shocks and mixed waves.
-
Aziz, K. and Settari, A., Petroleum Reservoir Simulation, London: Elsevier/8, Applied Science Publishers, 1979.
-
Brooks, R. and Corey, T., Hydraulic Properties of Porous Media, Hydrology Papers, Colorado State University, vol. 24,1964.
-
Crandall, M.G. and Lions, P.L., Viscosity Solutions of Hamilton-Jacobi Equations, Trans. Am. Math. Soc., vol. 277, no. 1,pp. 1-42, 1983.
-
Cybenko, G., Approximation by Superpositions of a Sigmoidal Function, Math. Control, Signals Syst., vol. 2, no. 4, pp. 303-314,1989.
-
Glorot, X. and Bengio, Y., Understanding the Difficulty of Training Deep Feedforward Neural Networks, in Proc. of the 13thInt. Conf. on Artificial Intelligence and Statistics, pp. 249-256,2010.
-
Gulshan, V., Peng, L., Coram, M., Stumpe, M.C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K., Madams, T., Cuadros, J., Kim, R., Raman, R., Nelson, P.C., Mega, J.L., and Webster, D.R., Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs, J. Am. Med. Assoc, vol. 316, no. 22, pp. 2402-2410,2016.
-
He, K., Zhang, X., Ren, S., and Sun, J., Deep Residual Learning for Image Recognition, in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, pp. 770-778,2016.
-
Hinton, G., Deng, L., Yu, D., Dahl, G.E., Mohamed, A.R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T.N., and Kingsbury, B., Deep Neural Networks for Acoustic Modeling in Speech Recognition, IEEE Signal Process. Mag., vol. 29, no. 6, pp. 82-97,2012.
-
Hornik, K., Stinchcombe, M., and White, H., Multilayer Feedforward Networks are Universal Approximators, Neural Networks, vol. 2, no. 5, pp. 359-366,1989.
-
Karpathy, A., Toderici, G., Shetty, S., Leung, T., Sukthankar, R., and Fei-Fei, L., Large-Scale Video Classification with Convolutional Neural Networks, Proc. of the 2014 IEEE Conf. on Computer Vision and Pattern Recognition, pp. 1725-1732,2014.
-
Krizhevsky, A., Sutskever, I., and Hinton, G.E., Imagenet Classification with Deep Convolutional Neural Networks, in Adv.. Neural Inf. Process. Syst, vol. 25, no. 2, pp. 1097-1105, 2012.
-
Lagaris, I.E., Likas, A., and Fotiadis, D.I., Artificial Neural Networks for Solving Ordinary and Partial Differential Equations, IEEE Trans. Neural Networks, vol. 9, no. 5, pp. 987-1000, 1998.
-
Lax, P.D., Hyperbolic Systems of Conservation Laws and the Mathematical Theory of Shock Waves, Philadelphia: Society for Industrial and Applied Mathematics, vol. 11, 1973.
-
Lax, P.D., Hyperbolic Partial Differential Equations, Providence, RI: American Mathematical Soc., vol. 14, 2006.
-
LeCun, Y. and Bengio, Y., Convolutional Networks for Images, Speech, and Time Series, in The Handbook of Brain Theory and Neural Networks, M.A. Arbib, Ed., Cambridge MA: The MIT Press, vol. 3361, no. 10, 1995.
-
LeCun, Y., Bengio, Y., and Hinton, G., Deep Learning, Nature, vol. 521, no. 7553, pp. 436-444, 2015.
-
Lee, H. and Kang, I.S., Neural Algorithm for Solving Differential Equations, J. Comput. Phys., vol. 91, no. 1, pp. 110-131, 1990.
-
Li, H., Xu, Z., Taylor, G., Studer, C., and Goldstein, T., Visualizing the Loss Landscape of Neural Nets, AdK Neural Inf. Process. Syst., pp. 6389-6399, 2018.
-
Lillicrap, T.P., Hunt, J.J., Pritzel, A., Heess, N., Erez, T., Tassa, Y., Silver, D., and Wierstra, D., Continuous Control with Deep Reinforcement Learning,, 2015. arXiv: 1509.02971.
-
Liu, Y., Gadepalli, K., Norouzi, M., Dahl, G.E., Kohlberger, T., Boyko, A., Venugopalan, S., Timofeev, A., Nelson, P.Q., and Corrado, G.S., Detecting Cancer Metastases on Gigapixel Pathology Images, 2017. arXiv: 1703.02442.
-
Meade Jr., A.J. and Fernandez, A.A., The Numerical Solution of Linear Ordinary Differential Equations by Feedforward Neural Networks, Math. Comput. Model, vol. 19, no. 12, pp. 1-25,1994.
-
Mnih, V., Badia, A.P., Mirza, M., Graves, A., Lillicrap, T., Harley, T., Silver, D., and Kavukcuoglu, K., Asynchronous Methods for Deep Reinforcement Learning, in Proc. of the 33rd Int. Conf. on Machine Learning, pp. 1928-1937,2016.
-
Nocedal, J. and Wright, S., Numerical Optimization, Berlin: Springer Science & Business Media, 2006.
-
Orr, F., Theory ofGas Injection Processes, Holte, Denmark: Tie-Line Publications, 2007.
-
Psichogios, D.C. and Ungar, L.H., A Hybrid Neural Network-First Principles Approach to Process Modeling, AIChE J., vol. 38, no. 10, pp. 1499-1511, 1992.
-
Raissi, M., Perdikaris, P., and Karniadakis, G.E., Physics Informed Deep Learning (Part I): Data-Driven Solutions of Nonlinear Partial Differential Equations, 2017. arXiv: 1711.10566.
-
Raissi, M., Perdikaris, P., and Karniadakis, G.E., Physics-Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Differential Equations, J. Comput. Phys, vol. 378, pp. 686-707,2019.
-
Stewart, R. and Ermon, S., Label-Free Supervision of Neural Networks with Physics and Domain Knowledge, in Proc. of the 31st AAAI Conference on Artificial Intelligence, pp. 2576-2582,2017.
-
Sutskever, I., Vinyals, O., andLe, Q.V., Sequence to Sequence Learning with Neural Networks, Adv. Neural Inf. Process. Syst., pp. 3104-3112,2014.
-
Zhu, Y., Zabaras, N., Koutsourelakis, P. S., and Perdikaris, P., Physics-Constrained Deep Learning for High-Dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data, J. Comput. Phys,., vol. 394, pp. 56-81,2019.
-
Xu Rui, Zhang Dongxiao, Rong Miao, Wang Nanzhe, Weak form theory-guided neural network (TgNN-wf) for deep learning of subsurface single- and two-phase flow, Journal of Computational Physics, 436, 2021. Crossref
-
Sambo Chico , Feng Yin , Physics Inspired Machine Learning for Solving Fluid Flow in Porous Media: A Novel Computational Algorithm for Reservoir Simulation, Day 1 Tue, October 26, 2021, 2021. Crossref
-
Magzymov Daulet , Ratnakar Ram R., Dindoruk Birol , Johns Russell T., Evaluation of Machine Learning Methodologies Using Simple Physics Based Conceptual Models for Flow in Porous Media, Day 2 Wed, September 22, 2021, 2021. Crossref
-
Fraces Cedric G., Tchelepi Hamdi , Physics Informed Deep Learning for Flow and Transport in Porous Media, Day 1 Tue, October 26, 2021, 2021. Crossref
-
Mehmani Yashar, Anderson Timothy, Wang Yuhang, Aryana Saman A., Battiato Ilenia, Tchelepi Hamdi A., Kovscek Anthony R., Striving to translate shale physics across ten orders of magnitude: What have we learned?, Earth-Science Reviews, 223, 2021. Crossref
-
Almajid Muhammad Majid, Abu-Alsaud Moataz Omar, Prediction of Fluid Flow in Porous Media using Physics Informed Neural Networks, Day 2 Tue, November 10, 2020, 2020. Crossref
-
Colombo Daniele, Turkoglu Ersan, Li Weichang, Sandoval-Curiel Ernesto, Rovetta Diego, Physics-driven deep-learning inversion with application to transient electromagnetics, GEOPHYSICS, 86, 3, 2021. Crossref
-
Tang Meng , Liu Yimin , Durlofsky Louis J., History Matching Complex 3D Systems Using Deep-Learning-Based Surrogate Flow Modeling and CNN-PCA Geological Parameterization, Day 1 Tue, October 26, 2021, 2021. Crossref
-
Colombo Daniele, Turkoglu Ersan, Li Weichang, Rovetta Diego, A framework for coupled physics-deep learning inversion and multiparameter joint inversion, First International Meeting for Applied Geoscience & Energy Expanded Abstracts, 2021. Crossref
-
Patel Ravi G., Manickam Indu, Trask Nathaniel A., Wood Mitchell A., Lee Myoungkyu, Tomas Ignacio, Cyr Eric C., Thermodynamically consistent physics-informed neural networks for hyperbolic systems, Journal of Computational Physics, 449, 2022. Crossref
-
Wang Nanzhe, Chang Haibin, Zhang Dongxiao, Efficient Uncertainty Quantification and Data Assimilation via Theory-Guided Convolutional Neural Network, SPE Journal, 26, 06, 2021. Crossref
-
Ciesielski Peter N., Pecha M. Brennan, Thornburg Nicholas E., Crowley Meagan F., Gao Xi, Oyedeji Oluwafemi, Sitaraman Hariswaran, Brunhart-Lupo Nicholas, Bridging Scales in Bioenergy and Catalysis: A Review of Mesoscale Modeling Applications, Methods, and Future Directions, Energy & Fuels, 35, 18, 2021. Crossref
-
Colombo Daniele, Turkoglu Ersan, Li Weichang, Rovetta Diego, Coupled physics-deep learning inversion, Computers & Geosciences, 157, 2021. Crossref
-
Zhang Kai, Zuo Yuande, Zhao Hanjun, Ma Xiaopeng, Gu Jianwei, Wang Jian, Yang Yongfei, Yao Chuanjin, Yao Jun, Fourier Neural Operator for Solving Subsurface Oil/Water Two-Phase Flow Partial Differential Equation, SPE Journal, 27, 03, 2022. Crossref
-
Deng Lichi, Pan Yuewei, Application of physics-informed neural networks for self-similar and transient solutions of spontaneous imbibition, Journal of Petroleum Science and Engineering, 203, 2021. Crossref
-
Yan Bicheng, Chen Bailian, Robert Harp Dylan, Jia Wei, Pawar Rajesh J., A robust deep learning workflow to predict multiphase flow behavior during geological CO2 sequestration injection and Post-Injection periods, Journal of Hydrology, 607, 2022. Crossref
-
Martinez-Gonzalez Daniel A., ROAM-ML: A reduced order aerodynamic module augmented with neural network digital surrogates, AIAA SCITECH 2022 Forum, 2022. Crossref
-
Cornelio Jodel, Mohd Razak Syamil, Cho Young, Liu Hui-Hai, Vaidya Ravimadhav, Jafarpour Behnam , Residual Learning to Integrate Neural Network and Physics-Based Models for Improved Production Prediction in Unconventional Reservoirs, SPE Journal, 2022. Crossref
-
Kyas Svetlana, Volpatto Diego, Saar Martin O., Leal Allan M. M., Accelerated reactive transport simulations in heterogeneous porous media using Reaktoro and Firedrake, Computational Geosciences, 26, 2, 2022. Crossref
-
Ekechukwu Gerald Kelechi, de Loubens Romain, Araya-Polo Mauricio, Long short-term memory-driven forecast of CO2 injection in porous media, Physics of Fluids, 34, 5, 2022. Crossref
-
Vadyala Shashank Reddy, Betgeri Sai Nethra, Matthews John C., Matthews Elizabeth, A review of physics-based machine learning in civil engineering, Results in Engineering, 13, 2022. Crossref
-
Rodriguez-Torrado Ruben, Ruiz Pablo, Cueto-Felgueroso Luis, Green Michael Cerny, Friesen Tyler, Matringe Sebastien, Togelius Julian, Physics-informed attention-based neural network for hyperbolic partial differential equations: application to the Buckley–Leverett problem, Scientific Reports, 12, 1, 2022. Crossref
-
Almajid Muhammad M., Abu-Al-Saud Moataz O., Prediction of porous media fluid flow using physics informed neural networks, Journal of Petroleum Science and Engineering, 208, 2022. Crossref
-
Maucec Marko, Jalali Ridwan, GeoDIN - Geoscience-Based Deep Interaction Networks for Predicting Flow Dynamics in Reservoir Simulation Models, SPE Journal, 27, 03, 2022. Crossref
-
Trask Nathaniel, Huang Andy, Hu Xiaozhe, Enforcing exact physics in scientific machine learning: A data-driven exterior calculus on graphs, Journal of Computational Physics, 456, 2022. Crossref
-
Huang YuHao, Xu JiaRong, Fang ShaoMei, Zhu ZuPeng, Jiang LinFeng, Liang XiaoXin, Parallel Physics-Informed Neural Networks with Bidirectional Balance, 2022 the 6th International Conference on Innovation in Artificial Intelligence (ICIAI), 2022. Crossref
-
Haghighat Ehsan, Amini Danial, Juanes Ruben, Physics-informed neural network simulation of multiphase poroelasticity using stress-split sequential training, Computer Methods in Applied Mechanics and Engineering, 397, 2022. Crossref
-
Li Jian, Zhang Dongxiao, Wang Nanzhe, Chang Haibin, Deep Learning of Two-Phase Flow in Porous Media via Theory-Guided Neural Networks, SPE Journal, 27, 02, 2022. Crossref
-
Andersen Pål Østebø, Nygård Jan Inge, Kengessova Aizhan, Prediction of Oil Recovery Factor in Stratified Reservoirs after Immiscible Water-Alternating Gas Injection Based on PSO-, GSA-, GWO-, and GA-LSSVM, Energies, 15, 2, 2022. Crossref
-
Liu Di, Zou Changchun, Song Qianggong, Wan Zhonghong, Zhao Haizhen, A hybrid physics and machine learning approach for velocity prediction, The Leading Edge, 41, 6, 2022. Crossref
-
Liu Wei, Wang He, Xi Zhenzhu, Zhang Rongqing, Huang Xiaodi, Physics-Driven Deep Learning Inversion with Application to Magnetotelluric, Remote Sensing, 14, 13, 2022. Crossref
-
Jagtap Ameya D., Mao Zhiping, Adams Nikolaus, Karniadakis George Em, Physics-informed neural networks for inverse problems in supersonic flows, Journal of Computational Physics, 466, 2022. Crossref
-
Abbasi Amirhassan, Nataraj C., Physics-Informed Machine Learning for Uncertainty Reduction in Time Response Reconstruction of a Dynamic System, IEEE Internet Computing, 26, 4, 2022. Crossref
-
Chiu Pao-Hsiung, Wong Jian Cheng, Ooi Chinchun, Dao My Ha, Ong Yew-Soon, CAN-PINN: A fast physics-informed neural network based on coupled-automatic–numerical differentiation method, Computer Methods in Applied Mechanics and Engineering, 395, 2022. Crossref
-
Bandai Toshiyuki, Ghezzehei Teamrat A., Forward and inverse modeling of water flow in unsaturated soils with discontinuous hydraulic conductivities using physics-informed neural networks with domain decomposition, Hydrology and Earth System Sciences, 26, 16, 2022. Crossref
-
Mondal Sudeepta, Sarkar Soumalya, Multi-fidelity prediction of spatiotemporal fluid flow, Physics of Fluids, 34, 8, 2022. Crossref
-
Sajjadinia Seyed Shayan, Carpentieri Bruno, Shriram Duraisamy, Holzapfel Gerhard A., Multi-fidelity surrogate modeling through hybrid machine learning for biomechanical and finite element analysis of soft tissues, Computers in Biology and Medicine, 148, 2022. Crossref
-
Brandao Eduardo, Colombier Jean-Philippe, Duffner Stefan, Emonet Rémi, Garrelie Florence, Habrard Amaury, Jacquenet François, Nakhoul Anthony, Sebban Marc, Learning PDE to Model Self-Organization of Matter, Entropy, 24, 8, 2022. Crossref
-
Amini Danial, Haghighat Ehsan, Juanes Ruben, Physics-Informed Neural Network Solution of Thermo–Hydro–Mechanical Processes in Porous Media, Journal of Engineering Mechanics, 148, 11, 2022. Crossref
-
Zhang Wenjuan, Al Kobaisi Mohammed, On the Monotonicity and Positivity of Physics-Informed Neural Networks for Highly Anisotropic Diffusion Equations, Energies, 15, 18, 2022. Crossref
-
Diab Waleed, Chaabi Omar, Zhang Wenjuan, Arif Muhammad, Alkobaisi Shayma, Al Kobaisi Mohammed, Data-Free and Data-Efficient Physics-Informed Neural Network Approaches to Solve the Buckley–Leverett Problem, Energies, 15, 21, 2022. Crossref
-
Li Qing, Evje Steinar, Learning the nonlinear flux function of a hidden scalar conservation law from data, Networks and Heterogeneous Media, 18, 1, 2022. Crossref
-
Tori Floriano, Ginis Vincent, Phase space approach to solving higher order differential equations with artificial neural networks, Physical Review Research, 4, 4, 2022. Crossref
-
Wang TongSheng, Wang ZhiHeng, Huang Zhu, Xi Guang, Multi-domain physics-informed neural network for solving heat conduction and conjugate natural convection with discontinuity of temperature gradient on interface, Science China Technological Sciences, 65, 10, 2022. Crossref
-
Gong Bing, Langguth Michael, Ji Yan, Mozaffari Amirpasha, Stadtler Scarlet, Mache Karim, Schultz Martin G., Temperature forecasting by deep learning methods, Geoscientific Model Development, 15, 23, 2022. Crossref
-
Yan Bicheng, Harp Dylan Robert, Chen Bailian, Pawar Rajesh J., Improving deep learning performance for predicting large-scale geological $${{CO}_{2}}$$ sequestration modeling through feature coarsening, Scientific Reports, 12, 1, 2022. Crossref