跳到正文
arXiv cs.AI· Marcio A. Sampaio, Paulo H. Ranazzi, Martin J. Blunt·· 4 小时前AI 评分35

基于 StyleGAN 中间隐空间的油藏属性参数化集合数据同化方法研究

Parameterization method of reservoir properties for ensemble-based data assimilation using intermediate latent space of StyleGAN

AI 导读

一项研究将 StyleGAN2 的隐空间与中间 w 空间分别用于集合数据同化的油藏属性参数化,并与 VAE-GAN、Latent Diffusion 比较。在两类(三种类别)和连续型两个 2D 算例中,StyleGAN2 生成的样本地质真实性更高、历史拟合效果优异。研究发现 ESMDA 线性更新下 w 空间比高度纠缠的 z 空间更线性、更解耦,同化结果更好,并通过地质统计与历史拟合指标验证。

正文 · 原文

View PDF

Abstract:Ensemble smoothers are the most successful and efficient techniques currently available for history matching. However, because these methods rely on Gaussian assumptions, their performance is severely degraded when the prior geology is described in terms of complex facies distributions (non-Gaussian). In this way, for these methods, we need to apply efficient parameterization techniques. Currently, the most efficient methods for performing parameterization are deep learning models. However, given the variety of existing deep learning models, studies have not identified which is most suitable for use with ensemble-based methods, although some important models had already been evaluated. Based on a recent literature review, the most promising models selected were VAE-GAN, Latent Diffusion, and StyleGAN models. As a novel aspect of this work, data assimilation with the second generation of StyleGAN (StyleGAN2) model was performed using the latent z-space and intermediate w-space, separately. They were applied in two 2D case studies: one categorical (three facies) and the other continuous. The results demonstrated that all three models are highly efficient, with the StyleGAN2 model standing out for generating samples with geological realism and achieving excellent data matching in the cases studied. Our findings show that performing data assimilation with StyleGAN2 using the intermediate space (w-space) yielded better results than the traditional application in the latent space (z-space). This is due to the fact that ESMDA uses linear updates and the w-space is much more linear and disentangled than the highly entangled z-space, thereby ensuring that the updated vectors remain close to realistic geological patterns. These results were validated using main geostatistical and history matching metrics.
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
Cite as: arXiv:2609.39626 [cs.LG]
  (or arXiv:2609.39626v2 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2609.39626

arXiv-issued DOI via DataCite

Submission history

From: Marcio Augusto Sampaio M. A. Sampaio [view email]
[v1] Wed, 30 Sep 2026 12:35:56 UTC (9,480 KB)
[v2] Fri, 13 Nov 2026 18:17:13 UTC (4,341 KB)

来源:arXiv cs.AI · arxiv.org