164 / 2025-06-09 00:52:20
A Series of Structural Head Loop Bottleneck Attention in Fault Diagnosis for Pumping Unit
fault diagnosis,indicator diagram,deep learning,attention mechanism
全文待审
Yi Yang / Beihang University (Beijing University of Aeronautics and Astronautics);Peng Cheng Laboratory;Hangzhou Innovation Institute, Beihang University;Zhejiang Hehu Technology Co., Ltd.
Yanjing Li / Beihang University (Beijing University of Aeronautics and Astronautics)
Hanyu Cen / Beihang University (Beijing University of Aeronautics and Astronautics)
Minghao Wang / Beihang University (Beijing University of Aeronautics and Astronautics)
Tengtuo Chen / COMAC Beijing Aircraft Technology Research Institute, Beijing 102211, China
Oil extraction is a crucial component of the petroleum industry, and oil extractors play a key role in this process. However, the operation and maintenance of oil extractors face numerous challenges such as the dispersed distribution of equipment across vast oil fields and the frequent occurrence of various mechanical faults. These issues not only reduce operational efficiency, but also increase maintenance costs and downtime. In addition, with the increasing complexity and diversity of faults, traditional manual identification methods are becoming inadequate to meet the growing demands of the industry. Manual approaches are often time-consuming, error-prone, and require experienced personnel, which limits their scalability and effectiveness in large-scale operations. To address these challenges, this paper proposes a novel method for fault diagnosis of oil extractors. Firstly, the paper collects a multi-class dataset, which contains more than 3k images and 8 different types of faults. Secondly, a fault diagnosis model is introduced, which incorporates a series of attention modules to enhance feature extraction and improve the accuracy of the fault diagnosis. Finally, comparative experiments demonstrate that the proposed method significantly improves classification accuracy. This method shows potential for practical applications, offering a reliable solution for fault diagnosis in the oil extraction industry.
重要日期
  • 会议日期

    08月01日

    2025

    08月04日

    2025

  • 06月15日 2025

    初稿截稿日期

主办单位
中国机械工程学会设备智能运维分会
承办单位
新疆大学
移动端
在手机上打开
小程序
打开微信小程序
客服
扫码或点此咨询