Comparative study of PCA approaches for fault detection: Application to a chemical reactor

Principal component analysis (PCA) has been successfully applied to several monitoring problems. However, classical PCA which is a fixed-model approach has some limitations: one of these is the inability to deal with parameter-varying process. Then an adaptation mechanism is recommended. This paper suggests a recursive PCA method for fault detection based on First Order Perturbation (RPCA-FOP). It also compares the effectiveness of the presented RPCA-FOP method and two other PCA techniques existing in literature such as the conventional PCA and the sliding window principal component analysis (SWPCA). The considered performances which are the average computation time, the missed detection rate and the false alarm rate are evaluated by simulation on a Continuous Stirred Tank Reactor (CSTR).

[1]  Mohamed-Faouzi Harkat.  Détection et localisation de défauts par analyse en composantes principales , 2003 .

[2]  T. Voegtlin,et al.  Recursive PCA and the structure of time series , 2004, 2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).

[3]  G. Irwin,et al.  Process monitoring approach using fast moving window PCA , 2005 .

[4]  Liangfei Xu,et al.  Proton exchange membrane fuel cell system diagnosis based on the signed directed graph method , 2011 .

[5]  L. Jun,et al.  Comparative study of PCA approaches in process monitoring and fault detection , 2004, 30th Annual Conference of IEEE Industrial Electronics Society, 2004. IECON 2004.

[6]  Babak Nadjar Araabi,et al.  Recursive spectral analysis of natural time series based on eigenvector matrix perturbation for online applications , 2011 .

[7]  Fuli Wang,et al.  Adaptive Monitoring Based on Independent Component Analysis for Multiphase Batch Processes with Limited Modeling Data , 2008 .

[8]  Didier Maquin,et al.  Fault Detection and Isolation with Robust Principal Component Analysis , 2008, 2008 16th Mediterranean Conference on Control and Automation.

[9]  Mohamed Benouaret,et al.  Sensor fault detection, isolation and reconstruction using nonlinear principal component analysis , 2007, Int. J. Autom. Comput..

[10]  Hector Budman,et al.  Fault detection, identification and diagnosis using CUSUM based PCA , 2011 .

[11]  Shu-Kai S. Fan,et al.  Multiple-input multiple-output double exponentially weighted moving average controller using partial least squares , 2010 .

[12]  Chun-Chin Hsu,et al.  Adaptive Kernel Principal Component Analysis (KPCA) for Monitoring Small Disturbances of Nonlinear Processes , 2010 .

[13]  Jyh-Cheng Jeng,et al.  Adaptive process monitoring using efficient recursive PCA and moving window PCA algorithms , 2010 .

[14]  Jianqiu Li,et al.  Proton exchange membrane fuel cell system diagnosis based on the multivariate statistical method , 2011 .

[15]  Hassani Messaoud,et al.  Online identification of nonlinear system using reduced kernel principal component analysis , 2010, Neural Computing and Applications.

[16]  D. Franke.  Application of extended Gershgorin theorems to certain distributed-parameter control problems , 1985, 1985 24th IEEE Conference on Decision and Control.

[17]  Chi Ma,et al.  Fault diagnosis of nonlinear processes using multiscale KPCA and multiscale KPLS , 2011 .

[18]  Ping Zhang,et al.  On the application of PCA technique to fault diagnosis , 2010 .

[19]  Gang Li,et al.  Total PLS Based Contribution Plots for Fault Diagnosis , 2009 .