检索结果-内蒙古大学图书馆

Invertible neural networks for uncertainty quantification in refraction tomography

Leading edge 2024年第6期43卷 358-366页

作者： Sun, Yen Williamson, paul TotalEnergies E&P Research and Technology HoustonTX United States TotalEnergies OneTech R&D Aquitaine France

Uncertainty quantification (UQ) should be an essential ingredient of geophysical inversion because it measures the confidence in the results and enables the assessment of the value of information in the data. However, UQ using established methods ranges from very expensive to prohibitively costly, and estimating noise levels and integrating prior information is challenging, so it is not yet widely undertaken. In this paper, we explore the capabilities of a machine learning-based UQ tool known as the invertible neural network (INN) and focus on its application to a 2D tomography problem within a complex foothills environment. We propose a novel approach to handle realistic problem dimensions that uses variational autoencoders to compress the velocity model and data. The INN relates the respective latent spaces, significantly reducing memory requirements. Our findings reveal that this INN-based workflow can perform tomographic inversion while integrating an implicit prior in the form of a set of velocity models with pertinent features. Furthermore, we can address both epistemic and aleatoric uncertainties by adopting a deep ensemble strategy. This integrated approach yields plausible estimates of relative confidence in the inverted velocities, showcasing the potential of INN as a tool for UQ in geophysical inversion. © 2024 Society of exploration Geophysicists. All rights reserved.

关键词： Tomography

来源：评论

学校读者我要写书评

暂无评论

Adding GpU acceleration to an industrial CpU-based simulator, development strategy and results

Adding GPU acceleration to an industrial CPU-based simulator...

引用

Spe Reservoir Simulation Conference 2021, RSC 2021

作者： Cao, Hui Zaydullin, Rustem Liao, Terrence Gohaud, Neil Obi, eguono Darche, Gilles Total E&P Research and Technology United States Total SE

ISBN: (纸本)9781613997475

Running multi-million cell simulation problems in minutes has been a dream for reservoir engineers for decades. Today, with the advancement of Graphic processing Unit (GpU), we have a real chance to make this dream a reality. Here we present our experience in the step-by-step transformation of a fully developed industrial CpU-based simulator into a fully functional GpU-based simulator. We also demonstrate significant accelerations achieved through the use of GpU technology. To achieve the best performance possible, we choose to use CUDA (NVIDIA GpU’s native language), and offload as much computations to GpU as possible. Our CUDA implementation covers all reservoir computes, which include property calculation, linearization, linear solver, etc. The well and Field Management still reside on CpU and need minor changes for their interaction with GpU-based reservoir. Importantly, there is no change to the nonlinear logic. The GpU and CpU parts are overlapped, fully utilizing the asynchronous nature of GpU operations. each reservoir computation can be run in three modes, CpU_only (existing one), GpU_only, CpU followed by GpU. The latter is only used for result checking and debugging. In early 2019, we prototyped two reservoir linearization operations (mass accumulation and mass flux) in CUDA;both showed very strong runtime speed-up of several hundred times, 1 p100-GpU (NVIDIA) vs 1 pOWeR8NVL CpU core rated at 2.8 GHz (IBM). encouraged by this success, we moved into linear solver development and managed to move the entire linear solver module into GpU. Again, strong speed-up of ~50 times was achieved (1 GpU vs 1 CpU). The focus for 2019 has been on standard Black-Oil cases. Our implementation was tested with multiple "million-cell range" models (Spe10 and other real field cases). In early 2020, we managed to put Spe10 fully on GpU, and finished the entire 2000 day time-stepping in ~35 sec with a single p100 card. After that our effort has switched to compositional AIM (Ada

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Data characterization and transfer learning for DL-driven velocity model building 1

Data characterization and transfer learning for DL-driven ve...

引用

1st International Meeting for Applied Geoscience and energy

作者： Sun, Yen Araya-polo, Mauricio Williamson, paul Total E&P Research and Technology United States

The success of hydrocarbon exploration and production imaging projects relies heavily on the accuracy of the velocity model. The standard methodology for model building is computationally expensive and requires extensive human interpretation and quality control. Several deep learning (DL) approaches have been proposed to reduce both computational and labor costs, but they are all confronted with the challenge of limited available pertinent training data. In this paper, we propose a hybrid workflow to tackle the generalization issue in DL-driven velocity model building frameworks with two major components: 1) Generative Adversarial Network (GAN) to enrich the training dataset and 2) Transfer Learning to improve the alignment of the predictions with the target output population. The results show that trained GANs can generate synthetic models which contain similar features to the targeted Gulf of Mexico (GOM) velocity models. Using the SASI evaluation metric we demonstrate that GAN-generated velocity models have similar data distribution for salt geometries as GOM models. To reconstruct velocity models, U-Net-like architectures are first trained with pseudo-randomly-generated velocity models and then, with transfer learning, re-trained with GAN-generated synthetic velocity models. Afterwards, the trained model is tested with synthetic data from velocity models representative of real GOM geological settings. The results demonstrate that the prediction accuracies for both salt geometries (IoU increased from 0.646 to 0.801) and background velocities (MAe decreased from 359.3 to 158.2 m/s) can be largely improved with the proposed hybrid workflow. © 2021 Society of exploration Geophysicists First International Meeting for Applied Geoscience & energy

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Seismic data interpolation via frequency-constrained 3D inception Unet 2

Seismic data interpolation via frequency-constrained 3D ince...

引用

2nd International Meeting for Applied Geoscience and energy, IMAGe 2022

作者： Sun, Yen Williamson, paul TotalEnergies E&P Research & Technology United States

High-quality seismic imaging strongly relies on spatially well-sampled input data. However, due to budgetary or physical constraints during the acquisition process, seismic data are often irregularly and/or sparsely sampled in at least one dimension. We have investigated Deep-learning (DL) interpolation to assess its ability to transcend aliasing limits without the computational effort of going to a full 5D interpolation process. In particular we consider the problem of interpolating between streamers, in marine acquisitions. We started with a "standard", 3D convolutional neural network (CNN) architecture: while computationally intense, a 3D NN is required to optimize interpolation performance. This already outperforms sinc interpolation. We then enhanced the network training by adding a second label - the data transformed into the frequency-wavenumber domain;this requires the addition of a Fourier Transform layer to the architecture, which can be denoted a Frequency-Constrained NN (FC-NN). Tests show that FC-NN outperforms CNN on both structural similarity index (SSIM) and peak signal-to-noise ratio (pSNR) evaluation metrics. Adding inception blocks to FC-NN to capture features at multiple scales further improves the prediction accuracy. We demonstrate that the proposed architecture can provide high accuracy interpolated results for both synthetic and field data. For real data, we also test the hypothesis that an FC-NN trained to interpolate the streamers at a given spacing can also be used to further interpolate the streamers at a finer spacing. © 2022 Society of exploration Geophysicists and the American Association of petroleum Geologists.

关键词： Interpolation

来源：评论

学校读者我要写书评

暂无评论

Seismic data denoising by combining self-supervised and supervised learning 3

Seismic data denoising by combining self-supervised and supe...

引用

3rd International Meeting for Applied Geoscience and energy, IMAGe 2023

作者： Sun, Yen Williamson, paul TotalEnergies E&P Research & Technology United States TotalEnergies OneTech R&D France France

Noise removal is a standard part of seismic data processing;however, deterministic workflows may require specific adaptations and parameter tuning for each dataset. Conventional ground-roll noise attenuation methods involving domain transform or decomposition can be time-consuming. Recent studies have demonstrated that machine learning (including supervised, self-supervised and unsupervised approaches) can be an effective tool for seismic noise removal. However, in most applications, the proposed method targets one or other of coherent or incoherent noise. In this study, we propose a workflow for a field dataset containing both coherent (ground-roll) and incoherent noise components. The workflow is built by sequentially combining supervised and self-supervised methods. For the incoherent noise, we train a neural network directly with field data in a self-supervised approach;however, by careful design and selection of the training data, we can also target noise that is coherent over a short range For the coherent, ground-roll noise, we train a second neural network with supervised learning using synthetic data;we found that this can be directly applied to field data without any additional training which makes this approach very computationally efficient. The synthetic training data need to be thoughtfully augmented to include the key features of the target field test sets. We demonstrate that the proposed workflow can effectively and efficiently remove multiple types of noise from seismic data. © 2023 Society of exploration Geophysicists and the American Association of petroleum Geologists.

关键词： Seismic response

来源：评论

学校读者我要写书评

暂无评论

New Mechanistic Approach to Trapped Foam in population-Balance Model enabling Improved prediction of N2 and CO2 Foams Rheology in porous Media

New Mechanistic Approach to Trapped Foam in Population-Balan...

引用

2022 Spe Improved Oil Recovery Conference, IOR 2022

作者： Ma, Kun Mateen, Khalid Total Energies e and P Research and Technology United States

ISBN: (纸本)9781613998502

As reported in the literature, CO2 foam/emulsion flow in porous media is generally weaker than N2 foamunder the same experimental conditions. Additionally, the steady-state CO2-foam quality scan has relativelya more flattened shape in the low-quality foam regime, besides the foam apparent viscosity showing a moregradual transition from the high-quality to low-quality foam regime. These characteristics are not fullycaptured by the existing foam models and could be indicative of different underlying physics in CO2 foam. Our new model is based on the trapped bubble concept at the pore level. On trapping, the trapped foamlamellae undergo a series of movement and merging events driven by gas diffusion, and they end up at astationary position near the pore throat. In our approach, the yield stress of trapped foam, which is the totalpressure gradient resulting from the effective number of lamellae in the direction of flow, is taken equal tothe flowing foam pressure gradient estimated through the bubble population-balance approach. In this newmodel, the maximum flowing foam texture is updated based on the trapped foam texture established. Thereis, therefore, a consequential dependance of the flowing foam pressure gradient on the trapped foam texture. Automatic parameter regressions and sensitivity studies have been conducted with both an existingpopulation-balance model and our new model. The results show that both models can fit the N2-foam data,but our new model fits the CO2-foam steady-state data significantly better than the existing population-balance model in both high-and low-quality flow regimes. Our analysis shows that the limiting capillarypressure and the flowing foam texture are crucial for correctly fitting the transition between the high- andlow-quality regimes. Matching the gradual transition in the steady-state CO2-foam data helps better estimatethe foam generation/coalescence rates. An important novel feature of our model is that it distinguishes the flowin

关键词： Carbon dioxide

来源：评论

学校读者我要写书评

暂无评论

Multi-mode wave-equation based Scholte wave inversion 1

Multi-mode wave-equation based Scholte wave inversion

引用

1st International Meeting for Applied Geoscience and energy

作者： Gao, Fuchun Appe, Mathieu Ren, Li Williamson, paul Total E&P Research and Technology University of Texas Dallas United States

Scholte waves are interface waves propagating along fluid-solid boundaries, such as the sea bottom. They are often recorded by ocean bottom nodes (OBN)/cables (OBC) and regarded as noise in conventional data processing. In this study we present a workflow to estimate a shallow shear wave velocity model below the sea bottom by inverting Scholte waves, and demonstrate it on an OBN dataset from North Sea. We use a spectral element (SeM) seismic propagator to simulate the propagation of the Scholte waves on the irregular bathymetry, and to calculate the model updates. Technically the workflow consists of two components: dispersion curve picking using a machine learning (ML)-based method and dispersion curve inversion. The method is shown to be robust and it is believed to be a good candidate for an early stage of an elastic model building workflow when Scholte waves are observed. © 2021 Society of exploration Geophysicists First International Meeting for Applied Geoscience & energy

关键词： Shear waves

来源：评论

学校读者我要写书评

暂无评论

On application issues of automatic dispersion curves picking by machine learning 1

On application issues of automatic dispersion curves picking...

引用

1st International Meeting for Applied Geoscience and energy

作者： Ren, Li Gao, Fuchun Williamson, paul McMechan, George A. The University of Texas Dallas United States Total E&P Research and Technology

Multi-mode dispersion curve picking is required by many approaches for shallow-subsurface characterization using surface waves, but is typically a labor-intensive process. previously we automated the picking procedure using convolutional neural network (CNN) based machine learning (ML) method, which made industry-scale application possible in terms of efficiency. This study improves the picking algorithm in three ways: (1) it is more stable by using a new loss function which combines the conventional wavenumber differences and the wavenumber slope differences;(2) it is more efficient because we pretrain with synthetic data and then use transfer learning;(3) it is more complete because we estimate uncertainties in the picking using random dropout. These improvements make the picking algorithm more mature technically. © 2021 Society of exploration Geophysicists First International Meeting for Applied Geoscience & energy

关键词： Surface waves

来源：评论

学校读者我要写书评

暂无评论

Simultaneous deblending, deconvolution and Doppler-shift correction of marine vibrator data 84

Simultaneous deblending, deconvolution and Doppler-shift cor...

引用

84th eAGe Annual Conference and exhibition

作者： Guitton, A. Ildarov, M. Zamboni, e. TotalEnergies E&P Research and Technology France École et Observatoire des Sciences de la Terre France TotalEnergies OneTech France

ISBN: (纸本)9781713884156

Blending of multiple marine vibrator (MVib) sources appears as a viable option to increase productivity and mitigate the challenge of having a long emitting signal. This strategy generates both blending noise and Doppler shift in the recorded data that need to be handled in processing. We deblend, correct for the Doppler effect and deconvolve MVib data simultaneously in an inversion framework that promotes sparsity of the signal in a given domain (Radon in this work) and robustness to non-Gaussian noise. These optimization goals are fulfilled thanks to a hyperbolic L1-L2 functional. After successfully deblending synthetic gathers, we illustrate our method by blending/deblending shot gathers from a recent 2D pilot seismic survey recorded with moving marine vibrators. © (2023) by the european Association of Geoscientists & engineers (eAGe). All rights reserved.

关键词： Blending

来源：评论

学校读者我要写书评

暂无评论

Geometric-mean reverse-time migration for elastic media 1

Geometric-mean reverse-time migration for elastic media

引用

1st International Meeting for Applied Geoscience and energy

作者： Bai, Tong Lyu, Bin Gao, Fuchun Williamson, paul Nakata, Nori Massachusetts Institute of Technology United States Total E&P United States

It has been shown that Geometric-mean Reverse-Time Migration (GmRTM) can produce source images with improved resolution than conventional time-reversal imaging in acoustic media. Here, we extend this cross-correlation-based imaging method to elastic media. Back-propagation and Helmholtz decomposition are conducted individually for each receiver wavefield, after which we apply zero-lag cross-correlations over the decoupled p- or/and S- wavefields to obtain the corresponding p-, S- and pS-source images, respectively. We validate the proposed method and demonstrate its advantages using a synthetic example with the elastic Marmousi model and an Oklahoma field data example. © 2021 Society of exploration Geophysicists First International Meeting for Applied Geoscience & energy

关键词： Backpropagation

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：