Fault detection and diagnosis based on sparse representation classification (SRC)

Fault detection and diagnosis (FDD) play an important role in process monitoring applications and remain challenging open problems. Some of existing methods treating fault detection and diagnosis separately are cumbersome and their effects are non-ideal. Some of them may trend to fail when multiple kinds of faults occur owing to the limitation of the typical classification strategy. In this paper, we propose a novel FDD method based on sparse representation classification (SRC), where main contributions are devoted in terms of model training and classification strategy. The motivation behind the SRC is that the reconstruction residuals are very effective to multi-class classification when a faults dictionary is well constructed based on the training samples. Extensive experiments performed on the Tennessee Eastman Process (TEP) demonstrate the effectiveness of the proposed method.

[1]  Jianjun Shi,et al.  Causation-Based T2 Decomposition for Multivariate Process Monitoring and Diagnosis , 2008 .

[2]  George W. Irwin,et al.  PLS modelling and fault detection on the Tennessee Eastman benchmark , 2000, Proceedings of the 1999 American Control Conference (Cat. No. 99CH36251).

[3]  Leo H. Chiang,et al.  Process monitoring using causal map and multivariate statistics: fault detection and identification , 2003 .

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

[5]  Jason Weston,et al.  Gene Selection for Cancer Classification using Support Vector Machines , 2002, Machine Learning.

[6]  Arthur K. Kordon,et al.  Fault diagnosis based on Fisher discriminant analysis and support vector machines , 2004, Comput. Chem. Eng..

[7]  George W. Irwin,et al.  Improved principal component monitoring of large-scale processes , 2004 .

[8]  René Vidal,et al.  Robust classification using structured sparse representation , 2011, CVPR 2011.

[9]  D. Donoho.  For most large underdetermined systems of equations, the minimal 𝓁1‐norm near‐solution approximates the sparsest near‐solution , 2006 .

[10]  Michael Elad,et al.  Double Sparsity: Learning Sparse Dictionaries for Sparse Signal Approximation , 2010, IEEE Transactions on Signal Processing.

[11]  Michael Elad,et al.  Stable recovery of sparse overcomplete representations in the presence of noise , 2006, IEEE Transactions on Information Theory.

[12]  Winfrid G. Schneeweiss,et al.  Fault tree analysis in case of multiple faults, especially covered and uncovered ones , 1998 .

[13]  Shi Zhang,et al.  An integrated method of independent component analysis and support vector machines for industry distillation process monitoring , 2010 .

[14]  G. W. Irwin,et al.  PLS modelling and fault detection on the Tennessee Eastman benchmark , 2000, Int. J. Syst. Sci..

[15]  Moisès Graells,et al.  A semi-supervised approach to fault diagnosis for chemical processes , 2010, Comput. Chem. Eng..

[16]  Gang Li,et al.  Reconstruction based fault prognosis for continuous processes , 2010 .

[17]  Zhi-Bo Zhu,et al.  Fault diagnosis based on imbalance modified kernel Fisher discriminant analysis , 2010 .

[18]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[19]  Allen Y. Yang,et al.  Robust Face Recognition via Sparse Representation , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  D. Donoho.  For most large underdetermined systems of linear equations the minimal 𝓁1‐norm solution is also the sparsest solution , 2006 .

[21]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part II: Qualitative models and search strategies , 2003, Comput. Chem. Eng..

[22]  Jingzhu Shen,et al.  A hybrid ANN-ES system for dynamic fault diagnosis of hydrocracking process , 1997 .

[23]  Jing Li,et al.  Causation-based T 2 decomposition for multivariate process monitoring and diagnosis , 2006 .

[24]  Richard D. Braatz,et al.  Fault Detection and Diagnosis in Industrial Systems , 2001 .

[25]  Sirish L. Shah,et al.  Fault detection and diagnosis in process data using one-class support vector machines , 2009 .

[26]  Raghunathan Rengaswamy,et al.  A review of process fault detection and diagnosis: Part I: Quantitative model-based methods , 2003, Comput. Chem. Eng..