整合PINN与传统的井记录,以在超深源岩中进行少数拍摄TOC预测
Taohua He1,2,3,4, Jiayi He4, Jin Xu1,2
1State Key Laboratory of Shale Oil and Gas Enrichment Mechanisms and Effective Development, Beijing 100083, China.
ACS omega
|February 2, 2026
概括
这项研究引入了一个基于物理学的神经网络 (PINN),用于预测超深岩石中的总有机碳 (TOC). 皮恩集成地质规则与稀疏的数据,在充满挑战的环境中胜过传统方法.
科学领域:
- 石油地质科学 石油地质科学
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 精确的总有机碳 (TOC) 预测在超深源岩 (>8,000 m) 中至关重要,但由于数据稀缺和复杂的日志反应,具有挑战性.
- 传统方法在极端的埋葬环境中与非线性关系和有限的校准数据作斗争.
研究的目的:
- 用最小的校准数据开发一个强大的框架,用于在超深源岩中预测TOC.
- 将基本的石质物理约束集成到神经网络中,以提高预测的准确性和可靠性.
主要方法:
- 开发了一个物理信息神经网络 (PINN) 框架,将石质物理约束 (密度-TOC反向关系,马射线-TOC正相关性,岩石物理验证) 作为明确的损失条款.
- 该PINN模型应用于Tarim盆地的下堪布里亚Yuertus形成的有线日志数据 (马射线,电阻,密度,声学传输时间).
主要成果:
- 该PINN模型实现了高精度 (R2=0.9451,RMSE=0.8571),显著优于像ΔlogR,MRA,MLP,RF和SVM这样的传统方法.
- 与其他车型相比,PINN表现出卓越的性能,R2的改进率从11%到282%不等.
- 该模型产生了在地质上一致的TOC配置文件,避免了纯数据驱动方法中看到的不稳定性,并实现了RMSE和MAE低于单位.
结论:
- 基于物理学的学习提供了一个有效的范式,用于在数据稀缺的超深环境中进行石化物理特征.
- 在传统方法失败的情况下,PINN为形成评估提供了可靠的解决方案,解决了关键的行业需求.
- 这种方法在地质科学应用中弥合了数据驱动的灵活性和基于物理的可靠性之间的差距.
更多相关视频
07:47Use of Sacrificial Nanoparticles to Remove the Effects of Shot-noise in Contact Holes Fabricated by E-beam Lithography
Published on: February 12, 2017
7.6K
10:36Stress Distribution During Cold Compression of Rocks and Mineral Aggregates Using Synchrotron-based X-Ray Diffraction
Published on: May 20, 2018
10.1K
相关概念视频
01:19The Mantel-Cox Log-Rank Test
1.1K
The Mantel-Cox log-rank test is a widely used statistical method for comparing the survival distributions of two groups. It tests whether a statistically significant difference exists in survival times between the groups without assuming a specific distribution for the survival data, making it a non-parametric test. This flexibility makes the log-rank test particularly valuable in medical research and other fields where the timing of an event, such as death or disease recurrence, is of...
1.1K
01:30Sign Convention
3.5K
When analyzing a beam subjected to various loads, it is crucial to understand the internal forces and moments generated within the structure. These internal forces can be broadly classified into normal forces, shear forces, and bending moments. To determine these forces and moments, we use the method of sections and apply a specific sign convention based on their direction and the side of the section being analyzed.
The normal force acts perpendicular to the beam's cross-section and can...
The normal force acts perpendicular to the beam's cross-section and can...
3.5K
02:27Predicting Molecular Geometry
45.8K
VSEPR Theory for Determination of Electron Pair Geometries
45.8K
01:15Enolate Mechanism Conventions
2.9K
When a carbonyl compound is treated with a strong base, the α position gets deprotonated to give a resonance-stabilized intermediate called an enolate. Enolates are ambident nucleophiles because they possess two nucleophilic sites that can attack an electrophile owing to the delocalization of the negative charge between the α carbon and oxygen atoms. When the oxygen atom attacks an electrophile, it is called O-attack, whereas electrophilic attack via the α carbon is known as...
2.9K
01:03Prediction Intervals
3.4K
The interval estimate of any variable is known as the prediction interval. It helps decide if a point estimate is dependable.
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
3.4K
02:39The Integrated Rate Law: The Dependence of Concentration on Time
41.7K
While the differential rate law relates the rate and concentrations of reactants, a second form of rate law called the integrated rate law relates concentrations of reactants and time. Integrated rate laws can be used to determine the amount of reactant or product present after a period of time or to estimate the time required for a reaction to proceed to a certain extent. For example, an integrated rate law helps determine the length of time a radioactive material must be stored for its...
41.7K