arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.28909cs.LG
收藏

面向Norne油藏系统的序贯物理约束神经算子正演建模

Sequential Physics-Constrained Neural Operator Forward Modeling for the $\textit{Norne}$ Reservoir System

NVIDIA Corporation(NVIDIA公司)

Clement Etienam, Juntao Yang, Oleg Ovcharenko, Nick Luiken, Tsubasa Onishi, Nefeli Moridis, Issam Said

AI总结:

针对Norne油藏基准问题,提出基于傅里叶神经算子(FNO)及其物理信息变体(PINO)的序贯代理模型,通过理论分析(函数空间公式、协变量偏移量化、物理约束谱稳定性、K步TBPTT梯度分析)和实验验证,实现全3298天时间跨度上油、气、压力和水的高精度预测,并在单GPU上获得约10^4倍加速。

摘要:

我们开发了一个全面的数学和计算框架,用于使用神经算子对三相黑油油藏动态进行序贯代理建模,特别关注傅里叶神经算子(FNO)及其物理信息变体(PINO)。应用重点是Norne基准油藏,定义在非均匀的$46\times112\times22$网格($N=113,344$个单元)上,生产历史跨越$T=30$个时间步,覆盖3298天。我们的理论贡献围绕四个相互关联的问题组织:(1)在乘积Sobolev空间设置中的泛函分析公式,包括隐式时间步映射的适定性和尖锐的局部Lipschitz估计;(2)协变量偏移量化,证明Wasserstein-2距离增长为$W_2 \leq \varepsilon(L^n-1)/(L-1)$,对于$L>1$具有指数级总体风险差异;(3)物理约束谱稳定性,表明使用$\lambda_R \geq \lambda^*_R$的PINO训练将学习到的Jacobian谱半径减小到$\rho_F + C\lambda_R^{-1/2}$,产生时间一致展开误差$|\delta_n| \leq \varepsilon/(1-\rho)$;(4)$K$步TBPTT梯度分析,推导出几何偏差衰减$O(\rho^K)$、最优窗口$K^* = O(\log(T/\sigma^2))$以及Adam收敛$O(1/\sqrt{t}) + O(\rho^{K^*})$。实证验证确认了所有理论预测:自回归PINO代理在完整的3298天范围内保持$R^2>0.99$(油)、$R^2>0.90$(气)、$R^2\approx 0.80$(压力)以及单调改善的$R^2$(水),在八块NVIDIA B200 GPU上训练不到一小时。一个包含1000个成员的集合在单块B200 GPU上运行不到一分钟,相比OPM有限体积模拟器获得约$10^4$倍的挂钟加速。

英文摘要:

We develop a comprehensive mathematical and computational framework for sequential surrogate modeling of three-phase black-oil reservoir dynamics using neural operators, with particular emphasis on Fourier Neural Operators (FNO) and their physics-informed variant (PINO). The application focus is the Norne benchmark reservoir, defined on a heterogeneous $46\times112\times22$ grid ($N=113,344$ cells), with a production history spanning $T=30$ timesteps covering 3298 days. Our theoretical contributions are organized around four interlocking problems: (1) functional-analytic formulation in a product-Sobolev-space setting, including well-posedness of the implicit timestep map and sharp local Lipschitz estimates; (2) covariate shift quantification, proving that the Wasserstein-2 distance grows as $W_2 \leq \varepsilon(L^n-1)/(L-1)$, with exponential population-risk discrepancy for $L>1$; (3) physics-constrained spectral stability, showing PINO training with $λ_R \geq λ^*_R$ reduces the learned Jacobian spectral radius to $ρ_F + Cλ_R^{-1/2}$, yielding uniform-in-time rollout error $|δ_n| \leq \varepsilon/(1-ρ)$; and (4) $K$-step TBPTT gradient analysis, deriving geometric bias decay $O(ρ^K)$, optimal window $K^ = O(\log(T/σ^2))$, and Adam convergence $O(1/\sqrt{t}) + O(ρ^{K^*})$. Empirical validation confirms all theoretical predictions: autoregressive PINO surrogates sustain $R^2>0.99$ (oil), $R^2>0.90$ (gas), $R^2\approx 0.80$ (pressure), and monotonically improving $R^2$ (water) across the full 3298-day horizon, trained on eight NVIDIA B200 GPUs in under one hour. A 1000-member ensemble runs in under one minute on a single B200 GPU, giving a ${\sim}10^4\times$ wall-clock speedup over the OPM finite-volume simulator.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑