Comparison of PCA and FDA for monitoring of coupled liquid tank system

This paper deals with the implementation of data driven techniques, Principal component analysis (PCA) and Fisher Discriminant analysis (FDA), for fault detection and identification in coupled liquid tank system (CLTS). A CLTS is used as a non-linear benchmark in control engineering. PCA transforms the higher dimensional data to a lower dimension, while FDA extracts the discriminant information for fault diagnosis. Actuator and sensor faults are detected. Multiple faults are detected simultaneously by using T2 - statistics and Q - statistics(SPE). Component fault is not detected by common Fault detection and isolation (FDI) schemes, but data driven techniques are applied to detect such faults as well. PCA has limitations for fault identification purpose, while FDA performance for component fault detection and fault identification is significantly better then PCA.

[1]  M. Tufail,et al.  On fault detection in coupled liquid three tank system using subspace aided data driven design , 2012, 2012 International Conference on Emerging Technologies.

[2]  H. Hotelling.  Analysis of a complex of statistical variables into principal components. , 1933 .

[3]  J. Edward Jackson,et al.  A User's Guide to Principal Components. , 1991 .

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

[5]  Maria J. Fuente,et al.  Fault detection and identification method based on multivariate statistical techniques , 2009, 2009 IEEE Conference on Emerging Technologies & Factory Automation.

[6]  Duarte Valério,et al.  Laboratory Software for the Three-Tank Benchmark System: From PID to Multi-Agent Fault-Tolerant Fractional Control , 2012 .

[7]  S. Joe Qin,et al.  Statistical process monitoring: basics and beyond , 2003 .

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

[9]  Karl Pearson F.R.S..  LIII. On lines and planes of closest fit to systems of points in space , 1901 .

[10]  Q. Peter He,et al.  A New Fault Diagnosis Method Using Fault Directions in Fisher Discriminant Analysis , 2005 .

[11]  Lamiaa M. Elshenawy,et al.  Fault Detection of Nonlinear Processes Using Fuzzy C-means-based Kernel PCA , 2015 .

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

[13]  Jing-Yu Yang,et al.  Optimal fisher discriminant analysis using the rank decomposition , 1992, Pattern Recognit..

[14]  S. Joe Qin,et al.  Multivariate process monitoring and fault diagnosis by multi-scale PCA , 2002 .

[15]  Muhammad Abid,et al.  Exponential discriminant analysis for fault diagnosis , 2016, Neurocomputing.

[16]  Zhiwei Gao,et al.  From Model, Signal to Knowledge: A Data-Driven Perspective of Fault Detection and Diagnosis , 2013, IEEE Transactions on Industrial Informatics.

[17]  Muhammad Abid,et al.  Sensor Fault Detection in Coupled Liquid Tanks System , 2012, 2012 10th International Conference on Frontiers of Information Technology.