石油与天然气地质 ›› 2023, Vol. 44 ›› Issue (1): 226-237.doi: 10.11743/ogg20230119

• 方法技术 • 上一篇    下一篇

沉积过程模拟驱动下的深度学习地质建模方法

刘彦锋(), 段太忠(), 黄渊, 张文彪, 李蒙   

  1. 中国石化 石油勘探开发研究院,北京 102206
  • 收稿日期:2022-06-01 修回日期:2022-11-03 出版日期:2023-01-14 发布日期:2023-01-13
  • 通讯作者: 段太忠 E-mail:liuyf.syky@sinopec.com;duantz.syky@sinopec.com
  • 第一作者简介:刘彦锋(1986—),男,博士、副研究员,油气田开发地质。E-mail: liuyf.syky@sinopec.com
  • 基金项目:
    中国科学院A类战略性先导科技专项(XDA14010204);中国石化科技部项目(P21038-3)

Deep learning-based geological modeling driven by sedimentary process simulation

Yanfeng LIU(), Taizhong DUAN(), Yuan HUANG, Wenbiao ZHANG, Meng LI   

  1. Petroleum Exploration and Production Research Institute,SINOPEC,Beijing 102206,China
  • Received:2022-06-01 Revised:2022-11-03 Online:2023-01-14 Published:2023-01-13
  • Contact: Taizhong DUAN E-mail:liuyf.syky@sinopec.com;duantz.syky@sinopec.com

摘要:

油气藏勘探开发逐步向深层化、复杂化方向发展,观测数据不足、分辨率低等资料难题突显,传统的地质建模方法无法适应技术需求。以深度学习为代表的智能化地质建模方法可以充分整合多尺度、多维度的数据信息以及专家认识,是地质建模技术发展的重要方向。在综合分析地层沉积模拟和深度学习地质建模技术优缺点的基础上,探索形成了沉积过程模拟驱动的深度学习地质建模方法。首先,基于综合地质分析开展沉积正演模拟,分析参数不确定性,通过参数扰动形成大规模地质模型作为训练样本库;其次,利用条件化生成对抗网络学习样本库中蕴含的地质模式和规律,其中生成网络以井-震等条件数据作为输入、地质模型作为输出;最后,利用训练后生成网络在实际条件数据上的应用,得到目标区块的地质模型。通过在四川盆地普光气藏主力区块典型地质剖面的测试应用,该方法的可行性得到了验证,并分析了训练样本库大小对模拟结果的影响。沉积模拟和深度学习相结合,弥补了训练样本不足的缺陷,间接实现了知识驱动的深度学习地质建模,具有重要的推广意义。

关键词: 人工智能, 大数据, 知识驱动, 生成对抗网络, 沉积过程模拟, 深度学习地质建模, 普光气藏, 四川盆地

Abstract:

Problems including insufficient quantity and low resolution of data face exploration and development of deep and complex reservoir targets, and the traditional geological modeling methods have been inadequate in terms of technical needs. The intelligent geological modeling represented by deep learning capable of fully integrating multi-scale and multi-dimensional data as well as expert knowledge, is a key research and development direction of geological modeling technology. The study discusses the deep learning-based geological modeling driven by sedimentary process simulation following the comprehensive analysis of the advantages and disadvantages of stratigraphic forward modeling and deep learning-based geological modeling technology. First, forward modeling of sedimentation is carried out based on comprehensive geological analysis, parameter uncertainty is analyzed, and a large amount of geological models are established through parameter disturbance as a training dataset; Second, the geological patterns contained in the learning dataset are learned with the conditional Generative Adversarial Nets (cGAN), in which the Generative Adversarial Networks (GAN) takes the conditional data such as well and seismic data as the input, and the geological model as the output; Finally, the trained GAN is applied to the real conditional data to obtain the geological model of the target block. The feasibility of this method is verified through testing on the typical geological profiles of the main block of Puguang gas reservoir, and the impact of the training dataset scale on simulation results is analyzed. The combination of sedimentary simulation and deep learning could make up for the shortage of training data and indirectly realize knowledge-driven deep learning-based geological modeling. The method is therefore of great significance to popularization.

Key words: artificial intelligence (AI), big data, knowledge-driven, Generative Adversarial Networks (GAN), sedimentary process simulation, deep learning-based geological modeling, Puguang gas reservoir, Sichuan Basin

中图分类号: