Stacked sparse autoencoders that preserve the local and global feature structures for fault detection

Although the model based on an autoencoder (AE) exhibits strong feature extraction capability without data labeling, such model is less likely to consider the structural distribution of the original data and the extracted feature is uninterpretable. In this study, a new stacked sparse AE (SSAE) based on the preservation of local and global feature structures is proposed for fault detection. Two additional loss terms are included in the loss function of SSAE to retain the local and global structures of the original data. The preservation of the local feature considers the nearest neighbor of data in space, while that of the global feature considers the variance information of data. The final feature is not only a deep representation of data, but it also retains structural information as much as possible. The proposed model demonstrates remarkable detection performance in case studies of a numerical process and the Tennessee Eastman process.

[1]  Tao Chen,et al.  Process fault diagnosis with model- and knowledge-based approaches: Advances and opportunities , 2020 .

[2]  Jay H. Lee,et al.  Machine learning: Overview of the recent progresses and implications for the process systems engineering field , 2017, Comput. Chem. Eng..

[3]  Shijin Wang,et al.  One-dimensional convolutional auto-encoder-based feature learning for fault diagnosis of multivariate processes , 2020 .

[4]  Herman Augusto Lepikson,et al.  Deep learning for diagnosis and classification of faults in industrial rotating machinery , 2020, Comput. Ind. Eng..

[5]  E. F. Vogel,et al.  A plant-wide industrial process control problem , 1993 .

[6]  Yue Li,et al.  Learning Representations With Local and Global Geometries Preserved for Machine Fault Diagnosis , 2020, IEEE Transactions on Industrial Electronics.

[7]  Yichao Liu,et al.  Data-Driven Fault Diagnosis Using Deep Canonical Variate Analysis and Fisher Discriminant Analysis , 2021, IEEE Transactions on Industrial Informatics.

[8]  Thomas J. McAvoy,et al.  Base control for the Tennessee Eastman problem , 1994 .

[9]  Orestes Llanes-Santiago,et al.  Data-driven monitoring of multimode continuous processes: A review , 2019, Chemometrics and Intelligent Laboratory Systems.

[10]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[11]  Yuan Li,et al.  A nonlinear method for monitoring industrial process , 2020 .

[12]  Wahiba Bounoua,et al.  Online monitoring scheme using principal component analysis through Kullback-Leibler divergence analysis technique for fault detection , 2019, Trans. Inst. Meas. Control.

[13]  Chenglin Wen,et al.  Deep learning fault diagnosis method based on global optimization GAN for unbalanced data , 2020, Knowl. Based Syst..

[14]  Yupu Yang,et al.  Automated feature learning for nonlinear process monitoring – An approach using stacked denoising autoencoder and k-nearest neighbor rule , 2018 .

[15]  Chung Bang Yun,et al.  Ultrasonic guided wave-based switch rail monitoring using independent component analysis , 2018, Measurement Science and Technology.

[16]  Zheng Liu,et al.  Intelligent Fault Diagnosis of Multichannel Motor–Rotor System Based on Multimanifold Deep Extreme Learning Machine , 2020, IEEE/ASME Transactions on Mechatronics.

[17]  Biao Huang,et al.  Survey on the theoretical research and engineering applications of multivariate statistics process monitoring algorithms: 2008-2017 , 2018, The Canadian Journal of Chemical Engineering.

[18]  R. A. Leibler,et al.  On Information and Sufficiency , 1951 .

[19]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[20]  Tao Liu,et al.  A planetary gear reducer backlash identification based on servo motor current signal and optimized fisher discriminant analysis. , 2020, ISA transactions.

[21]  Wende Tian,et al.  Fault monitoring based on mutual information feature engineering modeling in chemical process , 2019, Chinese Journal of Chemical Engineering.

[22]  Jing Wang,et al.  Quality-Relevant Fault Monitoring Based on Locally Linear Embedding Orthogonal Projection to Latent Structure , 2019 .

[23]  Huaiyu Zhu.  On Information and Sufficiency , 1997 .

[24]  Xuefeng Yan,et al.  Non-Gaussian chemical process monitoring with adaptively weighted independent component analysis and its applications , 2013 .

[25]  Zhao Xiaoqiang,et al.  Sparse representation preserving embedding based on extreme learning machine for process monitoring , 2020 .

[26]  Zhen Wang,et al.  Sparse stacked autoencoder network for complex system monitoring with industrial applications , 2020 .

[27]  Chaomin Luo,et al.  Low-Rank Joint Embedding and Its Application for Robust Process Monitoring , 2021, IEEE Transactions on Instrumentation and Measurement.

[28]  Xuefeng Yan,et al.  Double-Weighted Independent Component Analysis for Non-Gaussian Chemical Process Monitoring , 2013 .

[29]  Zhiqiang Ge,et al.  Review on data-driven modeling and monitoring for plant-wide industrial processes , 2017 .

[30]  Jun Liang,et al.  Online Process Monitoring Using Recursive Mutual Information-Based Variable Selection and Dissimilarity Analysis With No Prior Information , 2018, IEEE Access.

[31]  Na Zhang,et al.  A Process Monitoring Scheme for Uneven-Duration Batch Process Based on Sequential Moving Principal Component Analysis , 2020, IEEE Transactions on Control Systems Technology.

[32]  Xuefeng Yan,et al.  Adaptive parameter tuning stacked autoencoders for process monitoring , 2020, Soft Computing.

[33]  Xuefeng Yan,et al.  Layer-by-Layer Enhancement Strategy of Favorable Features of the Deep Belief Network for Industrial Process Monitoring , 2018 .

[34]  Zhi-huan Song,et al.  Global–Local Structure Analysis Model and Its Application for Fault Detection and Identification , 2011 .