Abstract:PreviousNext No AccessSEG 2020 Workshop: Broadband and Wide-azimuth Deepwater Seismic Technology, Beijing, China, 13–15 July 2020Inversion of reservoir parameters based on the rock physics model and neural ODEsAuthors: Jiawei LiuZhenwei GuoJianxin LiuWeitao SunJiawei LiuSchool of Geosciences and Info-Physics, Central South University, Changsha, ChinaSearch for more papers by this author, Zhenwei GuoSchool of Geosciences and Info-Physics, Central South University, Changsha, ChinaSearch for more papers by this author, Jianxin LiuSchool of Geosciences and Info-Physics, Central South University, Changsha, ChinaSearch for more papers by this author, and Weitao SunZhou Pei-Yuan Center for Applied Mathematics, Tsinghua University, Beijing, ChinaSearch for more papers by this authorhttps://doi.org/10.1190/bwds2020_26.1 SectionsAboutPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail Abstract Reservoir predictions using seismic wave propagation theories in rock physics typically involve complex numerical solutions of partial differential equations. For the deep-water hydrocarbon exploration, the lack of well logs data and rock samples on a massive scale bring difficulties to the numerical simulation methods. Furthermore, a precise identification of reservoir parameters from seismic data remains a big challenge. Thus, in this work, we propose a rapid prediction method based on the rock physics model and neural ODE nets. The prediction method uses a combination of multiple methodologies. Firstly, the wave propagation theories are applied to build the relationship between reservoir parameters and seismic data, including White model, Johnson model and et al. Then the velocity can be simulated as a function of reservoir parameters, such as porosity, via the neural network which based on the massive numerical simulations. We then consider the inversion of reservoir parameters with the functions in order to evaluate the application of wave propagation theories and neural ODEs in numerical calculations. Essentially, the velocity-porosity relationship results from neural network and traditional numerical methods are compared, and the computation time are successfully decreased. Keywords: reservoir characterization, seismic attributes, porosity, fractures, fluidPermalink: https://doi.org/10.1190/bwds2020_26.1FiguresReferencesRelatedDetails SEG 2020 Workshop: Broadband and Wide-azimuth Deepwater Seismic Technology, Beijing, China, 13–15 July 2020ISSN (online):2159-6832Copyright: 2020 Pages: 155 publication data© 2020 Published in electronic format with permission by the Society of Exploration GeophysicistsPublisher:Society of Exploration Geophysicists HistoryPublished Online: 09 Nov 2020 CITATION INFORMATION Jiawei Liu, Zhenwei Guo, Jianxin Liu, and Weitao Sun, (2020), "Inversion of reservoir parameters based on the rock physics model and neural ODEs," SEG Global Meeting Abstracts : 98-99. https://doi.org/10.1190/bwds2020_26.1 Plain-Language Summary Keywordsreservoir characterizationseismic attributesporosityfracturesfluidPDF DownloadLoading ...

Deep-learning-based Surrogate Model for Forward and Inverse Problems of Wave Propagation in Porous Media Saturated with Two Fluids

Rapid surrogate modeling of magnetotelluric in the frequency domain using physics-driven deep neural networks

Deep-neural-networks-based approaches for Biot–squirt model in rock physics

Effective Solution of Nonlinear Subsurface Flow Inverse Problems in Sparse Bases

Inversion of reservoir parameters based on the rock physics model and neural ODEs

Deep-Learning based Inverse Modeling Approaches: A Subsurface Flow Example

Seismic Inversion Based on Acoustic Wave Equations Using Physics-Informed Neural Network

Surrogate Modeling for Porous Flow Using Deep Neural Networks

An Unsupervised Deep Learning Approach for the Wave Equation Inverse Problem

Probabilistic physics-informed neural network for seismic petrophysical inversion#xD;

Solving multiphysics-based inverse problems with learned surrogates and constraints

Effective Porosity Seismic Inversion for Porous Media Saturated with an Ideal Fluid Using Simulated Annealing

Multiparameter Inversion of Reservoirs Based on Deep Learning

Biot's theory-based dynamic equation modeling using a machine learning auxiliary approach

High-resolution Seismic Impedance Inversion Integrating the Closed-Loop Convolutional Neural Network and Geostatistics: an Application to the Thin Interbedded Reservoir

Deep Graph Learning-Based Surrogate Model for Inverse Modeling of Fractured Reservoirs

Solving Inverse Problems in Steady State Navier-Stokes Equations using Deep Neural Networks

A Physics-Based Neural-Network Way To Perform Seismic Full Waveform Inversion

S-wave velocity inversion and prediction using a deep hybrid neural network

Rock-physics-driven Nonlinear Seismic Inversion for Petrophysical Parameters of Reservoir

Physics-informed neural network for seismic wave inversion in layered semi-infinite domain