Online fault monitoring and diagnosis using recursive sparse principal component analysis
(2.School of Automation, Central South University, Changsha, Hunan Province, China 410083)
(3.School of Physics and Electronics, Hunan Normal University, Changsha, Hunan Province, China 410081)
【Abstract】A fault monitoring and diagnosis method for industrial processes based on recursive sparse principal component analysis (RSPCA) is proposed, which can be used for adaptive fault monitoring and diagnosis of time-varying industrial processes. By introducing the elastic regression net, the principal component analysis is changed to the convex optimization of Lasso–Ridge regression. Successively, the covariance matrix is decomposed recursively by rank-1 matrix correction, which leads to the recursive updating of the sparse load matrix and the process control limit of monitoring statistics to realize the long-time self-adaptive fault monitoring of continuous industrial processes. With regard to the detected faults, the fault diagnosis is realized by using the contribution plot method. Experiments in the Tennessee-Eastman (TE) process show that the proposed method effectively reduces false positive rate and false negative rate with time complexity lower than those of traditional fault monitoring methods, which can ensure the sensitivity and real-time performance of the fault monitoring.
【Keywords】 recursive sparse principal component analysis; fault monitoring of industrial process; elastic regression network; Tennessee-Eastman process;
 Zhang K P, Jiang B, Chen F Y, et al. Composite fault diagnosis of high speed train based on time-varying model identification [J]. Control and Decision, 2019, 34 (2): 274–278 (in Chinese).
 Niu Y, Wang S, Lin Z, et al. Fault detection of industrial processes based on multivariate statistical process monitoring [J]. Journal of Chinese Society of Power Engineering, 2017, 37 (10): 829–836.
 Bing S, Shi H. Fault detection and classification using quality-supervised double-Layer method [J]. IEEE Transactions on Industrial Electronics, 2018, 65 (10): 8163–8172.
 Song B, Zhou X, Tan S, et al. Process monitoring via key principal components and local information based weights [J]. IEEE Access, 2019, 99 (7): 15357–15366.
 Zou H, Hastie T, Tibshirani R. Sparse principal component analysis [J]. Journal of Computational and Graphical Statistics, 2006, 15 (2): 265–286.
 Alameer A, Ghazaei G, Degenaar P, et al. Object recognition with an elastic net-regularized hierarchical MAX model of the visual cortex [J]. IEEE Signal Processing Letters, 2016, 23 (8): 1062–1066.
 Yu H, Khan F, Garaniya V. A sparse PCA for nonlinear fault diagnosis and robust feature discovery of industrial processes [J]. Aiche Journal, 2016, 62 (5): 1494–1513.
 Sapena-Bano A, Burriel-Valencia J, Pineda-Sanchez M,et al. The harmonic order tracking analysis method for the fault diagnosis in induction motors under time-varying conditions [J]. IEEE Transactions on Energy Conversion, 2017, 32 (1): 244–256.
 Du Y, Budman H, Duever T A, et al. Fault detection and classification for nonlinear chemical processes using lasso and gaussian process [J]. Industrial & Engineering Chemistry Research, 2018, 57 (27): 8962–8977.
 Yan R, Xiao H, Wang Z, et al. Detection, isolation and diagnosability analysis of intermittent faults in stochastic systems [J]. International Journal of Control, 2017, 91 (8): 1–24.
 Zhang L, Yong L, Chang Q. Intermittent connection fault diagnosis for CAN using data link layer information [J]. IEEE Transactions on Industrial Electronics, 2017, 64 (3): 2286–2295.
 Hastie T, Mazumder R, Zadeh R, et al. Matrix completion and low-rank SVD via fast alternating least squares [J]. Journal of Machine Learning Research, 2015, 16 (1): 3367–3402.
 Yue L, Ge Z. Weighted random forests for fault classification in industrial processes with hierarchical clustering model selection [J]. Journal of Process Control, 2018, 64 (4): 62–70.
 Jiang Y, He Y, Zhang H. Variable selection with prior information for generalized linear models via the prior LASSO method [J]. Journal of the American Statistical Association, 2016, 111 (5): 355–376.
 Bien J, Taylor J, Tibshirani R. A lasso for hierarchical interactions [J]. Annals of Statistics, 2018, 41 (3): 1111–1141.
 Vanhatalo E, Kulahci M. The effect of autocorrelation on the hotelling T2 control chart [J]. Quality & Reliability Engineering International, 2016, 31 (8): 1779–1796.
 Vouillamoz N, Mosar J, Deichmann N. Multi-scale imaging of a slow active fault zone: Contribution for improved seismic hazard assessment in the Swiss Alpine foreland [J]. Swiss Journal of Geosciences, 2017, 110 (2): 547–563.
 Gajjar S, Kulahci M, Palazoglu A. Real-time fault detection and diagnosis using sparse principal component analysis [J]. Journal of Process Control, 2017, 67: 112–128.
 Yi J, Di H, He H, et al. A novel framework for fault diagnosis using kernel partial least squares based on an optimal preference matrix [J]. IEEE Transactions on Industrial Electronics, 2017, 64 (5): 4315–4324.
 Amin M T, Imtiaz S, Khan F. Process system fault detection and diagnosis using a hybrid technique [J]. Chemical Engineering Science, 2018, 189: 191–211.