《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (9): 3054-3064.DOI: 10.11772/j.issn.1001-9081.2025081033

• 前沿与综合应用 • 上一篇    

基于Shapelet-Transformer的油层甜点识别方法

刘明喆1, 熊小琴1(), 白世勋2, 赵仁保3, 王子薇4, 苟发强4, 徐亮4   

  1. 1.中国石油大学(北京)克拉玛依校区 工学院,新疆 克拉玛依 834000
    2.中国石油大学(北京)克拉玛依校区 石油学院,新疆 克拉玛依 834000
    3.中国石油大学(北京) 石油工程学院,北京 102249
    4.中国石油新疆油田公司 石西油田作业区,新疆 克拉玛依 834000
  • 收稿日期:2025-09-09 修回日期:2025-11-03 接受日期:2025-11-10 发布日期:2026-09-16 出版日期:2026-09-10
  • 通讯作者: 熊小琴
  • 作者简介:刘明喆(1999—),女,山西太原人,硕士研究生,主要研究方向:油层甜点智能识别
    熊小琴(1983—),女,重庆人,副教授,硕士,主要研究方向:油气储运、甜点产能匹配
    白世勋(1990—),男,山东东营人,副教授,博士,主要研究方向:非常规油气甜点识别与开发
    赵仁保(1971—),男,安徽和县人,教授,博士,主要研究方向:储层甜点预测
    王子薇(1999—),女,甘肃天水人,助理工程师,主要研究方向:油气田甜点定位
    苟发强(1995—),男,甘肃白银人,工程师,主要研究方向:油气田井位部署
    徐亮(1993—),男,四川彭州人,工程师,主要研究方向:油气甜点产能预测。
  • 基金资助:
    新疆天山创新团队项目(2022TSYCTD0002)

Oil reservoir sweet spot identification method based on Shapelet-Transformer

Mingzhe LIU1, Xiaoqin XIONG1(), Shixun BAI2, Renbao ZHAO3, Ziwei WANG4, Faqiang GOU4, Liang XU4   

  1. 1.School of Engineering,China University of Petroleum - Beijing at Karamay,Karamay Xinjiang 834000,China
    2.School of Petroleum,China University of Petroleum - Beijing at Karamay,Karamay Xinjiang 834000,China
    3.College of Petroleum Engineering,China University of Petroleum - Beijing,Beijing 102249,China
    4.Operation District of Shixi Oilfield,PetroChina Xinjiang Oilfield Company,Karamay Xinjiang 834000,China
  • Received:2025-09-09 Revised:2025-11-03 Accepted:2025-11-10 Online:2026-09-16 Published:2026-09-10
  • Contact: Xiaoqin XIONG
  • About author:LIU Mingzhe, born in 1999, M. S. candidate. Her research interests include intelligent identification of sweet spots in oil reservoirs.
    XIONG Xiaoqin, born in 1983, M. S., associate professor. Her research interests include oil and gas storage and transportation, sweet spot capacity matching.
    BAI Shixun, born in 1990, Ph. D., associate professor. His research interests include unconventional oil and gas sweet spot identification and development.
    ZHAO Renbao, born in 1971, Ph. D., professor. His research interests include oil reservoir sweet spot prediction.
    WANG Ziwei, born in 1999, assistant engineer. Her research interests include sweet spot positioning in oil and gas fields.
    GOU Faqiang, born in 1995, engineer. His research interests include well placement in oil and gas fields.
    XU Liang, born in 1993, engineer. His research interests include oil and gas sweet spot productivity prediction.
  • Supported by:
    Xinjiang Tianshan Innovative Research Team Project(2022TSYCTD0002)

摘要:

针对油藏地质数据表示复杂,现有方法难以充分提取油层甜点特征的问题,提出基于Shapelet-Transformer的油层甜点识别方法。该方法将测井数据重新表示为多变量深度序列,并构建融合局部与全局特征的网络模型,实现油层甜点的精准识别。首先,提出局部特征学习单元,通过Shapelet提取与筛选机制捕捉甜点特有的局部测井响应特征;其次,构建全局特征学习机制,挖掘测井序列的全局趋势与参数关联,提供全局地质特征;最后,设计融合分类模块,通过融合局部和全局特征输出油层甜点识别结果。实验结果表明,该方法在玛2井区的测井数据上的平均识别准确率达84.62%,较全卷积神经网络(FCNN)、BiLSTM-FCNN (Bidirectional Long Short-Term Memory and FCNN hybrid model)、图卷积时序网络(T-GCN)和多尺度时序分解模型(DisMS-TS)等基准模型至少提升了1.26个百分点,验证了该方法的有效性。

关键词: 测井响应序列, Shapelet, 局部特征学习, 全局特征学习, 油层甜点识别

Abstract:

To address the issues that oil reservoir geological data are complex to represent and the existing methods struggle to fully extract the features of oil reservoir sweet spots, an oil reservoir sweet spot identification method based on Shapelet-Transformer was proposed. In this method, well log data were re-represented as a multi-variable depth sequence and a network model integrating local and global features was constructed to achieve accurate identification of oil reservoir sweet spots. First, a local feature learning unit was proposed, and the unique local well log response features of sweet spots were captured through the Shapelet extraction and selection mechanism. Second, a global feature learning mechanism was constructed to mine the global trend and parameter correlation of well log sequences, thereby providing global geological features. Finally, a fusion and classification module was designed to output the identification results of oil reservoir sweet spots by fusing local and global features. Experimental results show that the average identification accuracy of this method on the well log data of Ma 2 well block is 84.62 %, which is at least 1.26 percentage points higher than that of Fully Convolutional Neural Network (FCNN), BiLSTM-FCNN (Bidirectional Long Short-Term Memory and FCNN hybrid model), Temporal Graph Convolutional Network (T-GCN), Disaggregated Multi-Scale Time Series decomposition model (DisMS-TS) and other benchmark models, which verifies the effectiveness of this method.

Key words: well log response sequence, Shapelet, local feature learning, global feature learning, oil reservoir sweet spot identification

中图分类号: