SCI和EI收录∣中国化工学会会刊

Chinese Journal of Chemical Engineering ›› 2020, Vol. 28 ›› Issue (9): 2358-2367.DOI: 10.1016/j.cjche.2020.06.015

• Process Systems Engineering and Process Safety • Previous Articles     Next Articles

Deep learning technique for process fault detection and diagnosis in the presence of incomplete data

Cen Guo1,2, Wenkai Hu3, Fan Yang1, Dexian Huang1   

  1. 1 Department of Automation, Tsinghua University, Beijing 10084, China;
    2 Cornell University, NY 14850, United States of America;
    3 University of Alberta, Edmonton, AB T6G 1H9, Canada
  • Received:2020-01-22 Revised:2020-05-15 Online:2020-10-21 Published:2020-09-28
  • Contact: Fan Yang
  • Supported by:
    This work was supported by the National Natural Science Foundation of China (61433001) and Tsinghua University Initiative Scientific Research Program.

Deep learning technique for process fault detection and diagnosis in the presence of incomplete data

Cen Guo1,2, Wenkai Hu3, Fan Yang1, Dexian Huang1   

  1. 1 Department of Automation, Tsinghua University, Beijing 10084, China;
    2 Cornell University, NY 14850, United States of America;
    3 University of Alberta, Edmonton, AB T6G 1H9, Canada
  • 通讯作者: Fan Yang
  • 基金资助:
    This work was supported by the National Natural Science Foundation of China (61433001) and Tsinghua University Initiative Scientific Research Program.

Abstract: In modern industrial processes, timely detection and diagnosis of process abnormalities are critical for monitoring process operations. Various fault detection and diagnosis (FDD) methods have been proposed and implemented, the performance of which, however, could be drastically influenced by the common presence of incomplete or missing data in real industrial scenarios. This paper presents a new FDD approach based on an incomplete data imputation technique for process fault recognition. It employs the modified stacked autoencoder, a deep learning structure, in the phase of incomplete data treatment, and classifies data representations rather than the imputed complete data in the phase of fault identification. A benchmark process, the Tennessee Eastman process, is employed to illustrate the effectiveness and applicability of the proposed method.

Key words: Alarm configuration, Deep learning, Fault detection and diagnosis, Incomplete data, Stacked autoencoder

摘要: In modern industrial processes, timely detection and diagnosis of process abnormalities are critical for monitoring process operations. Various fault detection and diagnosis (FDD) methods have been proposed and implemented, the performance of which, however, could be drastically influenced by the common presence of incomplete or missing data in real industrial scenarios. This paper presents a new FDD approach based on an incomplete data imputation technique for process fault recognition. It employs the modified stacked autoencoder, a deep learning structure, in the phase of incomplete data treatment, and classifies data representations rather than the imputed complete data in the phase of fault identification. A benchmark process, the Tennessee Eastman process, is employed to illustrate the effectiveness and applicability of the proposed method.

关键词: Alarm configuration, Deep learning, Fault detection and diagnosis, Incomplete data, Stacked autoencoder