Structured Joint Sparse Principal Component Analysis for Fault Detection and Isolation

Principal component analysis (PCA) has been widely applied in process monitoring of modern industrial systems. PCA performs fault detection by mapping the process data into a low dimensional subspace and tracking the process behavior using T2 and SPE statistics, whilst in fault isolation, it heavily relies on contribution plot or reconstruction based approaches. However, conventional methods based on contribution plot and reconstruction suffer from insufficient fault isolation capabilities. In order to improve the fault isolation performance, this article proposes a novel fault detection and isolation approach based on the Structured Joint Sparse PCA (SJSPCA). The objective function of SJSPCA involves two regularization terms: l2,1 norm and the graph Laplacian. By imposing the l2,1 norm term, SJSPCA is able to achieve row-wise sparsity, introducing the graph Laplacian regularization term can incorporate structured variable correlation information. In fault detection, conventional T2 and SPE statistics are constructed to detect abnormal situations. Once a fault is detected, a two stage fault isolation strategy is considered and a score index is calculated for each variable. The row-sparsity property of l2,1 norm ensures that the score indices associated to normal variables approach zero and the graph Laplacian constraint helps isolation of correlated faulty variables. The validity of SJSPCA in fault detection and isolation is illustrated by a process fault observed in an industrial blast furnace iron-making process.

[1]  S. Joe Qin,et al.  Reconstruction-based Contribution for Process Monitoring , 2008 .

[2]  Tiago J. Rato,et al.  Markovian and Non-Markovian sensitivity enhancing transformations for process monitoring , 2017 .

[3]  Steven X. Ding,et al.  A Review on Basic Data-Driven Approaches for Industrial Process Monitoring , 2014, IEEE Transactions on Industrial Electronics.

[4]  Flavio Sanson Fogliatto,et al.  Variable selection methods in multivariate statistical process control: A systematic literature review , 2018, Comput. Ind. Eng..

[5]  Feiping Nie,et al.  Efficient and Robust Feature Selection via Joint ℓ2, 1-Norms Minimization , 2010, NIPS.

[6]  I. Jolliffe,et al.  A Modified Principal Component Technique Based on the LASSO , 2003 .

[7]  Lei Huang,et al.  Bayesian Networks in Fault Diagnosis , 2017, IEEE Transactions on Industrial Informatics.

[8]  Chenglin Wen,et al.  Fault Isolation Based on k-Nearest Neighbor Rule for Industrial Processes , 2016, IEEE Transactions on Industrial Electronics.

[9]  Zhiqiang Ge,et al.  Distributed Parallel PCA for Modeling and Monitoring of Large-Scale Plant-Wide Processes With Big Data , 2017, IEEE Transactions on Industrial Informatics.

[10]  Marc Teboulle,et al.  A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems , 2009, SIAM J. Imaging Sci..

[11]  Hai Lin,et al.  Adaptive sparse principal component analysis for enhanced process monitoring and fault isolation , 2015 .

[12]  Michael E. Tipping,et al.  Probabilistic Principal Component Analysis , 1999 .

[13]  H. Zou,et al.  Regularization and variable selection via the elastic net , 2005 .

[14]  C. Vogel.  Computational Methods for Inverse Problems , 1987 .

[15]  Lei Xie,et al.  Shrinking principal component analysis for enhanced process monitoring and fault isolation , 2013 .

[16]  Bernhard Schölkopf,et al.  Kernel Principal Component Analysis , 1997, ICANN.

[17]  Xuefeng Yan,et al.  Multivariate Statistical Monitoring of Key Operation Units of Batch Processes Based on Time-Slice CCA , 2019, IEEE Transactions on Control Systems Technology.

[18]  Stephen P. Boyd,et al.  Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers , 2011, Found. Trends Mach. Learn..

[19]  Murat Kulahci,et al.  Real-time fault detection and diagnosis using sparse principal component analysis , 2017, Journal of Process Control.

[20]  Yang Liu,et al.  Compressive sparse principal component analysis for process supervisory monitoring and fault detection , 2017 .

[21]  Yuan Yao,et al.  Robust Multivariate Statistical Process Monitoring via Stable Principal Component Pursuit , 2016 .

[22]  Yuan Yao,et al.  Variable selection method for fault isolation using least absolute shrinkage and selection operator (LASSO) , 2015 .

[23]  Xilin Shen,et al.  Perturbation of the Eigenvectors of the Graph Laplacian: Application to Image Denoising , 2012, ArXiv.

[24]  Jieping Ye,et al.  Multi-Task Feature Learning Via Efficient l2, 1-Norm Minimization , 2009, UAI.

[25]  Zhiqiang Ge,et al.  Dynamic Probabilistic Latent Variable Model for Process Data Modeling and Regression Application , 2019, IEEE Transactions on Control Systems Technology.

[26]  Biao Huang,et al.  Process monitoring using a generalized probabilistic linear latent variable model , 2018, Autom..

[27]  Babatunde A. Ogunnaike,et al.  Hierarchical monitoring of industrial processes for fault detection, fault grade evaluation, and fault diagnosis , 2017 .

[28]  Ying Zheng,et al.  Normalized Relative RBC-Based Minimum Risk Bayesian Decision Approach for Fault Diagnosis of Industrial Process , 2016, IEEE Transactions on Industrial Electronics.

[29]  R. Tibshirani,et al.  Sparse Principal Component Analysis , 2006 .

[30]  Jianhua Z. Huang,et al.  Sparse principal component analysis via regularized low rank matrix approximation , 2008 .

[31]  Janos Gertler,et al.  Design of optimal structured residuals from partial principal component models for fault diagnosis in linear systems , 2005 .

[32]  Guang Wang,et al.  A Kernel Least Squares Based Approach for Nonlinear Quality-Related Fault Detection , 2017, IEEE Transactions on Industrial Electronics.

[33]  Qibing Jin,et al.  Fault isolation based on residual evaluation and contribution analysis , 2017, J. Frankl. Inst..

[34]  Nir Friedman,et al.  Probabilistic Graphical Models - Principles and Techniques , 2009 .

[35]  Junghui Chen,et al.  On-line batch process monitoring using dynamic PCA and dynamic PLS models , 2002 .

[36]  Béla Bollobás,et al.  Modern Graph Theory , 2002, Graduate Texts in Mathematics.

[37]  Donghua Zhou,et al.  Isolating incipient sensor fault based on recursive transformed component statistical analysis , 2018 .

[38]  Kaixiang Peng,et al.  Joint Data-Driven Fault Diagnosis Integrating Causality Graph With Statistical Process Monitoring for Complex Industrial Processes , 2017, IEEE Access.

[39]  Dexian Huang,et al.  Canonical variate analysis-based contributions for fault identification , 2015 .

[40]  Hongliang Fei,et al.  Anomaly localization for network data streams with graph joint sparse PCA , 2011, KDD.

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

[42]  Lei Xie,et al.  Nonparametric Density Estimation of Hierarchical Probabilistic Graph Models for Assumption-Free Monitoring , 2017 .

[43]  Steven X. Ding,et al.  Canonical correlation analysis-based fault detection methods with application to alumina evaporation process , 2016 .

[44]  Guang Wang,et al.  A Kernel Direct Decomposition-Based Monitoring Approach for Nonlinear Quality-Related Fault Detection , 2017, IEEE Transactions on Industrial Informatics.

[45]  Pascal Frossard,et al.  Learning Laplacian Matrix in Smooth Graph Signal Representations , 2014, IEEE Transactions on Signal Processing.

[46]  Zhenyu He,et al.  Joint sparse principal component analysis , 2017, Pattern Recognit..