探索深度学习对破裂流体回流和页岩气生产的预测性能
Shasha Sun1, Wenhua Bai2, Zhaoyuan Shao2
1Research Institute Petroleum Exploration and Development, Beijing, 10083, China. sunshasha501@163.com.
Scientific reports
|November 29, 2025
概括
一个新的深度学习模型,CNN-Transformer,通过整合卷积神经网络和变压器网络,准确地预测页岩气生产和流体回流. 该系统提高了页岩气运营预测的准确性,优于现有的模型.
科学领域:
- 石油工程是石油工程中的一个.
- 数据科学数据科学数据科学
- 人工智能的人工智能
背景情况:
- 页岩气产量预测受到复杂的液压压裂流体回流的挑战.
- 传统的数值模拟与双相流体流和杂的生产数据作斗争.
研究的目的:
- 开发一种深度学习系统,用于预测页岩气生产和流体回流.
- 提高页岩气产量估计的准确性和对产量变化的理解.
主要方法:
- 开发了一个CNN-Transformer深度学习模型,集成卷积神经网络 (CNN) 和变压器网络.
- 美国有线电视新闻网专注于观察数据中的局部相关性;变压器提取历史状态信息用于时间模式预测.
- 将CNN-变压器与CNN-LSTM和CNN-GRU-AM模型进行比较.
主要成果:
- 在预测回流量方面,CNN-Transformer的表现优于CNN-GRU-AM和CNN-LSTM (R2=0.644对比0.6068和0.5727).
- 在页岩气产量预测中,CNN-Transformer的分析误差显著降低 (R2=0.72与0.5705和0.4911对比).
结论:
- 开发的CNN-Transformer模型为预测页岩气生产和流体回流提供了一个强大的解决方案.
- 结果为平衡压裂液回流量量与生产预测准确度提供了基础.
相关概念视频
Major Losses in Pipes
1.9K
When a fluid flows through a pipe, it experiences energy losses due to frictional resistance along the pipe walls, known as major losses. These energy losses result in a pressure drop, which varies based on the flow conditions — whether laminar or turbulent — and the specific physical properties of the fluid and pipe.
Fluid flow can be classified as laminar or turbulent, primarily based on the Reynolds number. This dimensionless number reflects the relative influence of inertial to viscous...
Fluid flow can be classified as laminar or turbulent, primarily based on the Reynolds number. This dimensionless number reflects the relative influence of inertial to viscous...
1.9K
Rapidly Varying Flow
404
Rapidly varying flow (RVF) in open channels is characterized by abrupt changes in flow depth over a short distance, with the rate of depth change relative to distance often approaching unity. These flows are inherently complex due to their transient and multi-dimensional nature, making exact analysis difficult. However, approximate solutions using simplified models provide valuable insights into their behavior.Key Features of Rapidly Varying FlowRVF is commonly observed in scenarios involving...
404
Uniform Depth Channel Flow: Problem Solving
417
To calculate the flow rate for a trapezoidal channel, first, identify the bottom width, side slope, and flow depth of the channel. The cross-sectional area (A) corresponding to the depth of flow (y), channel bottom width (B), and side slope (θ) is determined by:Next, calculate the wetted perimeter, which includes the bottom width and the sloped side lengths in contact with the water. Using the values of the cross-sectional area and the wetted perimeter, determine the hydraulic radius by...
417
Hydraulic Jump
576
A hydraulic jump is a sudden rise in fluid depth in open channels, occurring when high-velocity (supercritical) flow transitions to low-velocity (subcritical) flow. This phenomenon requires an upstream Froude number greater than 1, as flows with Fr1<1 remain subcritical, making a hydraulic jump impossible due to the need for negative head loss, which violates thermodynamic principles.The characteristics of a hydraulic jump depend on the upstream Froude number and are classified as...
576


