Enhanced dynamic data-driven fault detection approach: Application to a two-tank heater system

Principal components analysis (PCA) has been intensively studied and used in monitoring industrial systems. However, data generated from chemical processes are usually correlated in time due to process dynamics, which makes the fault detection based on PCA approach a challenging task. Accounting for the dynamic nature of data can also reflect the performance of the designed fault detection approaches. In PCA-based methods, this dynamic characteristic of the data can be accounted for by using dynamic PCA (DPCA), in which lagged variables are used in the PCA model to capture the time evolution of the process. This paper presents a new approach that combines the DPCA to account for autocorrelation in data and generalized likelihood ratio (GLR) test to detect faults. A DPCA model is applied to perform dimension reduction while appropriately considering the temporal relationships in the data. Specifically, the proposed approach uses the DPCA to generate residuals, and then apply GLR test to reveal any abnormality. The performances of the proposed method are evaluated through a continuous stirred tank heater system.

[1]  J. E. Jackson,et al.  Control Procedures for Residuals Associated With Principal Component Analysis , 1979 .

[2]  Sofiane Khadraoui,et al.  Improved detection of incipient anomalies via multivariate memory monitoring charts: Application to an air flow heating system , 2016 .

[3]  Sébastien Borguet,et al.  A Generalized Likelihood Ratio Test for Adaptive Gas Turbine Performance Monitoring , 2009 .

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

[5]  Gerald R. Benitz,et al.  A generalized likelihood ratio test for SAR CCD , 2012, 2012 Conference Record of the Forty Sixth Asilomar Conference on Signals, Systems and Computers (ASILOMAR).

[6]  Fouzi Harrou,et al.  Improved nonlinear fault detection strategy based on the Hellinger distance metric: Plug flow reactor monitoring , 2017 .

[7]  Fabrice Heitz,et al.  Generalized likelihood ratio tests for change detection in diffusion tensor images: Application to multiple sclerosis , 2012, Medical Image Anal..

[8]  Ying Sun,et al.  Amalgamation of anomaly-detection indices for enhanced process monitoring , 2016 .

[9]  Y. Chetouani.  Application of the Generalized Likelihood Ratio Test for Detecting Changes in a Chemical Reactor , 2006 .

[10]  Antonio Moschitta,et al.  Generalized Likelihood Ratio Test for Voltage Dip Detection , 2010, IEEE Transactions on Instrumentation and Measurement.

[11]  Sachin C. Patwardhan,et al.  Adaptive Predictive Control using GOBF-ARX Models: An Experimental Case Study , 2013 .

[12]  Mu Zhu,et al.  Automatic dimensionality selection from the scree plot via the use of profile likelihood , 2006, Comput. Stat. Data Anal..

[13]  Jie Zhang,et al.  Actuator fault monitoring and fault tolerant control in distillation columns , 2015, ICAC.

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

[15]  Rong Guo,et al.  Prediction of C3 Concentration in FCCU Using Neural Estimator Based on Dynamic PCA , 2009, 2009 International Conference on Computational Intelligence and Security.

[16]  Hazem N. Nounou,et al.  Detecting abnormal ozone levels using PCA-based GLR hypothesis testing , 2013, 2013 IEEE Symposium on Computational Intelligence and Data Mining (CIDM).

[17]  Michèle Basseville,et al.  Detection of abrupt changes: theory and application , 1993 .

[18]  Fouzi Harrou,et al.  Anomaly detection/detectability for a linear model with a bounded nuisance parameter , 2014, Annu. Rev. Control..

[19]  Steven X. Ding,et al.  A Survey of Fault Diagnosis and Fault-Tolerant Techniques—Part II: Fault Diagnosis With Knowledge-Based and Hybrid/Active Approaches , 2015, IEEE Transactions on Industrial Electronics.

[20]  Anna T. Lawniczak,et al.  Detection of stationary network load increase using univariate network aggregate traffic data by dynamic PCA , 2011, 2011 IEEE Symposium on Computational Intelligence for Security and Defense Applications (CISDA).

[21]  Hazem Nounou,et al.  Linear Inferential Modeling: Theoretical Perspectives, Extensions, and Comparative Analysis , 2012 .

[22]  Nina F. Thornhill,et al.  A continuous stirred tank heater simulation model with applications , 2008 .

[23]  Ying Sun,et al.  Improved data-based fault detection strategy and application to distillation columns , 2017 .

[24]  David M. Himmelblau,et al.  Sensor Fault Detection via Multiscale Analysis and Dynamic PCA , 1999 .

[25]  P. Tavella,et al.  Fault detection in atomic clock frequency standards affected by mean and variance changes and by an additive periodic component: the GLRT approach , 2008, 2008 IEEE Instrumentation and Measurement Technology Conference.

[26]  Christos Georgakis,et al.  Disturbance detection and isolation by dynamic principal component analysis , 1995 .

[27]  Jun Liang,et al.  The application of dynamic principal component analysis to enhance chunk monitoring of an industrial fluidized-bed reactor , 2004, Fifth World Congress on Intelligent Control and Automation (IEEE Cat. No.04EX788).

[28]  Anna T. Lawniczak,et al.  Feature extraction via dynamic PCA for epilepsy diagnosis and epileptic seizure detection , 2010, 2010 IEEE International Workshop on Machine Learning for Signal Processing.

[29]  Ping Zhang,et al.  A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process , 2012 .