2.765

2022影响因子

(CJCR)

  • 中文核心
  • EI
  • 中国科技核心
  • Scopus
  • CSCD
  • 英国科学文摘

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

一种新颖的深度因果图建模及其故障诊断方法

唐鹏 彭开香 董洁

唐鹏, 彭开香, 董洁. 一种新颖的深度因果图建模及其故障诊断方法. 自动化学报, 2022, 48(6): 1616−1624 doi: 10.16383/j.aas.c200996
引用本文: 唐鹏, 彭开香, 董洁. 一种新颖的深度因果图建模及其故障诊断方法. 自动化学报, 2022, 48(6): 1616−1624 doi: 10.16383/j.aas.c200996
Tang Peng, Peng Kai-Xiang, Dong Jie. A novel method for deep causality graph modeling and fault diagnosis. Acta Automatica Sinica, 2022, 48(6): 1616−1624 doi: 10.16383/j.aas.c200996
Citation: Tang Peng, Peng Kai-Xiang, Dong Jie. A novel method for deep causality graph modeling and fault diagnosis. Acta Automatica Sinica, 2022, 48(6): 1616−1624 doi: 10.16383/j.aas.c200996

一种新颖的深度因果图建模及其故障诊断方法

doi: 10.16383/j.aas.c200996
基金项目: 国家自然科学基金(U21A20483, 61873024, 61773053)资助
详细信息
    作者简介:

    唐鹏:北京科技大学自动化学院博士研究生. 2013年获得长沙理工大学电气与信息工程学院学士学位. 2016年获得北方工业大学电气与控制工程学院硕士学位. 主要研究方向为过程监测和故障诊断. E-mail: gnepgnat@163.com

    彭开香:北京科技大学自动化学院教授. 2007 年获得北京科技大学控制科学与工程博士学位. 主要研究方向为复杂工业过程的故障诊断与容错控制. 本文通信作者. E-mail: kaixiang@ustb.edu.cn

    董洁:北京科技大学自动化学院教授. 2007 年获得北京科技大学控制科学与工程博士学位. 主要研究方向为智能控制理论与应用, 过程监控与故障诊断和复杂系统建模与控制. E-mail: dongjie@ies.ustb.edu.cn

A Novel Method for Deep Causality Graph Modeling and Fault Diagnosis

Funds: Supported by National Natural Science Foundation of China (U21A20483, 61873024, 61773053)
More Information
    Author Bio:

    TANG Peng Ph.D. candidate at the School of Automation and Electrical Engineering, University of Science and Technology Beijing. He received his bachelor degree from Changsha University of Science and Technology in 2013. He received his master degree from North China University of Technology in 2016. His research interest covers process monitoring and fault diagnosis for process industries

    PENG Kai-Xiang Professor at the School of Automation and Electrical Engineering, University of Science and Technology Beijing. He received his Ph.D. degree in control science and engineering from University of Science and Technology Beijing in 2007. His research interest covers fault diagnosis and fault-tolerant control for complex industrial system. Corresponding author of this paper

    DONG Jie Professor at the School of Automation and Electrical Engineering, University of Science and Technology Beijing. She received her Ph.D. degree in control science and engineering from University of Science and Technology Beijing in 2007. Her research interest covers intelligent control theory and application, process monitoring and fault diagnosis, and complex system modeling and control

  • 摘要: 为了实现复杂工业过程故障检测和诊断一体化建模, 提出了一种新颖的深度因果图建模方法. 首先, 利用循环神经网络建立深度因果图模型, 将Group Lasso稀疏惩罚项引入到模型训练中, 自动地检测过程变量间的因果关系. 其次, 利用模型学习到的条件概率预测模型对每个变量建立监测指标, 并融合得到综合指标进行整体工业过程故障检测. 一旦检测到故障, 对故障样本构建变量贡献度指标, 隔离故障相关变量, 并通过深度因果图模型的局部因果有向图诊断故障根源, 辨识故障传播路径. 最后, 通过田纳西−伊斯曼过程进行仿真验证, 实验结果验证了所提方法的有效性.
  • 图  1  深度因果图的单节点预测模型网络结构

    Fig.  1  The network structure of single node prediction model for deep causality graph

    图  2  基于深度因果图模型的故障检测和诊断框架

    Fig.  2  The fault detection and diagnosis based on deep causality graph model

    图  3  TE过程工艺流程图

    Fig.  3  The flowchart of TE process

    图  4  变量连接数和预测误差的关系曲线

    Fig.  4  The relation curve between prediction error and the number of variable connections

    图  5  故障4的故障检测结果

    Fig.  5  The fault detection result for Fault $4$

    图  6  故障4的VCI图

    Fig.  6  The plot of VCI for Fault $4$

    图  7  故障4的故障相关变量的因果关系

    Fig.  7  The causalities among fault-related variables for Fault $4$

    图  8  故障8的故障检测结果

    Fig.  8  The fault detection result for Fault $8$

    图  9  故障8的VCI图

    Fig.  9  The plot of VCI for Fault $8$

    图  10  故障8的故障相关变量的因果关系

    Fig.  10  The causalities among fault-related variables for Fault 8

    表  1  TE过程的因果矩阵

    Table  1  The causality matrix of TE process

    12345678910111213141516171819202122232425262728293031
    10000000000000000000000001000000
    20000000000000000000000100000000
    30000000000000000000000010000000
    40000000000000000000000000000000
    50000000000000000000000000000000
    60000000000000000000100000000000
    70000000000101001000011111100000
    80000000000000000000000000000000
    90000000000000000000001100000000
    100000001000001001000001000010000
    110000000100000000000101111110001
    120000000000000000000000000000000
    130000001000100001000111110110000
    140000000000000000000000000000000
    150000000000000000000000000000000
    160000001000101000000111111100000
    170000000000000000000000000000000
    180000000000100000000011000101000
    190000000000000000000000000000000
    200000000000100001010000000110000
    210000000000000000010101101110000
    220000001000101001010100000110001
    230000000000000000000000010000100
    240000000000000000000000100000100
    251000000000000000000000000100000
    260001000000000000000000001000100
    270000000010101001000111111100001
    280000000000000000010001000000100
    290000000000000000110100000000000
    300000000000000000000010000000000
    310000000000000000000000000000000
    下载: 导出CSV

    表  2  21个故障类型的FDRs (%)

    Table  2  The FDRs of 21 faults (%)

    Fault12345678
    FDR99.196.915.11004.110010092.4
    Fault910111213141516
    FDR11.98197.12993.699.46.642.9
    Fault1718192021
    FDR86.169.495.689.94.7
    下载: 导出CSV
  • [1] 赵春晖, 余万科, 高福荣. 非平稳间歇过程数据解析与状态监控—回顾与展望. 自动化学报, 2020, 46(10): 2072-2091

    Zhao Chun-Hui, Yu Wan-Ke, Gao Fu-Rong. Data analytics and condition monitoring methods for nonstationary batch processes — current status and future. Acta Automatica Sinica, 2020, 46(10): 2072-2091
    [2] Dong Y N, Qin S J. A novel dynamic PCA algorithm for dynamic data modeling and process monitoring. Journal of Process Control, 2018, 67: 1-11 doi: 10.1016/j.jprocont.2017.05.002
    [3] 周平, 刘记平, 梁梦圆, 张瑞垚. 基于KPLS鲁棒重构误差的高炉燃料比监测与异常识别. 自动化学报, 2021, 47(7): 1661-1671

    Zhou Ping, Liu Ji-Ping, Liang Meng-Yuan, Zhang Rui-Yao. KPLS Robust reconstruction error based monitoring and anomaly identification of fuel ratio in blast furnace ironmaking. Acta Automatica Sinica, 2021, 47(7): 1661-1671
    [4] Zhu J L, Ge Z Q, Song Z H. Non-Gaussian industrial process monitoring with probabilistic independent component analysis. IEEE Transactions on Automation Science and Engineering, 2017, 14(2): 1309-1319 doi: 10.1109/TASE.2016.2537373
    [5] Yang J, Dong J T, Shi H B, Tan S. Quality monitoring method based on enhanced canonical component analysis. ISA Transactions, 2020, 105: 221-229 doi: 10.1016/j.isatra.2020.06.008
    [6] Wang K, Chen J H, Song Z H. A sparse loading-based contribution method for multivariate control performance diagnosis. Journal of Process Control, 2020, 85: 199-213 doi: 10.1016/j.jprocont.2019.12.001
    [7] Alcala C F, Qin S J. Analysis and generalization of fault diagnosis methods for process monitoring. Journal of Process Control, 2011, 21(3): 322-330 doi: 10.1016/j.jprocont.2010.10.005
    [8] Van Den Kerkhof P, Vanlaer J, Gins G, Van Impe J F M. Analysis of smearing-out in contribution plot based fault isolation for statistical process control. Chemical Engineering Science, 2013, 104: 285-293 doi: 10.1016/j.ces.2013.08.007
    [9] 尹进田, 谢永芳, 陈志文, 彭涛, 杨超. 基于故障传播与因果关系的故障溯源方法及其在牵引传动控制系统中的应用. 自动化学报, 2020, 46(1): 47-57

    Yin Jin-Tian, Xie Yong-Fang, Chen Zhi-Wen, Peng Tao, Yang Chao. Fault tracing method based on fault propagation and causality with its application to the traction drive control system. Acta Automatica Sinica, 2020, 46(1): 47-57
    [10] Ma L, Dong J, Peng K X. Root cause diagnosis of quality-related faults in industrial multimode processes using robust Gaussian mixture model and transfer entropy. Neurocomputing, 2018, 285: 60-73 doi: 10.1016/j.neucom.2018.01.028
    [11] Li Z C, Tian L, Jiang Q C, Yan X F. Distributed-ensemble stacked autoencoder model for non-linear process monitoring. Information Sciences, 2021, 542: 302-316 doi: 10.1016/j.ins.2020.06.062
    [12] Tang P, Peng K X, Zhang K, Chen Z W, Yang X, Li L L. A deep belief network-based fault detection method for nonlinear processes. IFAC-PapersOnLine, 2018, 51(24): 9-14 doi: 10.1016/j.ifacol.2018.09.522
    [13] Zhang Z H, Jiang T, Zhan C J, Yang Y P. Gaussian feature learning based on variational autoencoder for improving nonlinear process monitoring. Journal of Process Control, 2019, 75: 136-155 doi: 10.1016/j.jprocont.2019.01.008
    [14] Chen X L, Wang J, Zhou J L. Probability density estimation and Bayesian causal analysis based fault detection and root identification. Industrial & Engineering Chemistry Research, 2018, 57(43): 14656-14664
    [15] Chen X L, Wang J, Zhou J L. Process monitoring based on multivariate causality analysis and probability inference. IEEE Access, 2018, 6: 6360-6369 doi: 10.1109/ACCESS.2018.2795535
    [16] Zerrouki H, Estrada-Lugo H D, Smadi H, Patelli E. Applications of Bayesian networks in chemical and process industries: A review. In: Proceedings of the 29th European Safety and Reliability Conference. Hannover, Germany: ESREL 2019, 2020. 3122−3129
    [17] Mehranbod N, Soroush M, Piovoso M, Ogunnaike B A. Probabilistic model for sensor fault detection and identification. AIChE Journal, 2003, 49(7): 1787-1802 doi: 10.1002/aic.690490716
    [18] Mehranbod N, Soroush M, Panjapornpon C. A method of sensor fault detection and identification. Journal of Process Control, 2005, 15(3): 321-339 doi: 10.1016/j.jprocont.2004.06.009
    [19] Azhdari M, Mehranbod N. Application of Bayesian belief networks to fault detection and diagnosis of industrial processes. In: Proceedings of the 2010 International Conference on Chemistry and Chemical Engineering. Kyoto, Japan: IEEE, 2010. 92−96
    [20] Gonzalez R, Huang B, Lau E. Process monitoring using kernel density estimation and Bayesian networking with an industrial case study. ISA Transactions, 2015, 58: 330-347 doi: 10.1016/j.isatra.2015.04.001
    [21] Chen G J, Ge Z Q. Hierarchical Bayesian network modeling framework for large-scale process monitoring and decision making. IEEE Transactions on Control Systems Technology, 2020, 28(2): 671-679 doi: 10.1109/TCST.2018.2882562
    [22] Chen G J, Ge Z Q. Robust Bayesian networks for low-quality data modeling and process monitoring applications. Control Engineering Practice, 2020, 97: Article No. 104344 doi: 10.1016/j.conengprac.2020.104344
    [23] Yu J, Rashid M M. A novel dynamic Bayesian network-based networked process monitoring approach for fault detection, propagation identification, and root cause diagnosis. AIChE Journal, 2013, 59(7): 2348-2365 doi: 10.1002/aic.14013
    [24] Zhang Z D, Dong F L. Fault detection and diagnosis for missing data systems with a three time-slice dynamic Bayesian network approach. Chemometrics and Intelligent Laboratory Systems, 2014, 138: 30-40 doi: 10.1016/j.chemolab.2014.07.009
    [25] Chung J, Gulcehre C, Cho K, Bengio Y. Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv: 1412.3555, 2014
    [26] Yuan M, Lin Y. Model selection and estimation in regression with grouped variables. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 2006, 68(1): 49-67 doi: 10.1111/j.1467-9868.2005.00532.x
  • 加载中
图(10) / 表(2)
计量
  • 文章访问数:  1694
  • HTML全文浏览量:  798
  • PDF下载量:  487
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-11-30
  • 录用日期:  2021-03-19
  • 网络出版日期:  2021-05-12
  • 刊出日期:  2022-06-02

目录

    /

    返回文章
    返回