Comparison of Artificial Neural Networks and Dynamic Principal Component Analysis for Fault Diagnosis

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6703)


Dynamic Principal Component Analysis (DPCA) and Artificial Neural Networks (ANN) are compared in the fault diagnosis task. Both approaches are process history based methods, which do not assume any form of model structure, and rely only on process historical data. Faults in sensors and actuators are implemented to compare the online performance of both approaches in terms of quick detection, isolability capacity and multiple faults identifiability. An industrial heat exchanger was the experimental test-bed system. Multiple faults in sensors can be isolated using an individual control chart generated by the principal components; the error of classification was 15.28% while ANN presented 4.34%. For faults in actuators, ANN showed instantaneous detection and 14.7% lower error classification. However, DPCA required a minor computational effort in the training step.


DPCA Artificial Neural Network Fault Detection Fault Diagnosis 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Venkatasubramanian, V., Rengaswamy, R., Kavuri, S., Yin, K.: A Review of Process Fault Detection and Diagnosis Part I Quantitative Model-Based Methods. Computers and Chemical Eng. 27, 293–311 (2003)CrossRefGoogle Scholar
  2. 2.
    Habbi, H., Kinnaert, M., Zelmat, M.: A Complete Procedure for Leak Detection and Diagnosis in a Complex Heat Exchanger using Data-Driven Fuzzy Models. ISA Trans. 48, 354–361 (2008)CrossRefGoogle Scholar
  3. 3.
    Astorga-Zaragoza, C.M., Alvarado-Martínez, V.M., Zavala-Río, A., Méndez-Ocaña, R., Guerrero-Ramírez, G.V.: Observer-based Monitoring of Heat Exchangers. ISA Trans. 47, 15–24 (2008)CrossRefGoogle Scholar
  4. 4.
    Morales-Menendez, R., Freitas, N.D., Poole, D.: State Estimation and Control of Industrial Processes using Particles Filters. In: IFAC-ACC 2003, Denver Colorado U.S.A, pp. 579–584 (2003)Google Scholar
  5. 5.
    Tan, C.K., Ward, J., Wilcox, S.J., Payne, R.: Artificial Neural Network Modelling of the Thermal Performance of a Compact Heat Exchanger. Applied Thermal Eng. 29, 3609–3617 (2009)CrossRefGoogle Scholar
  6. 6.
    Rangaswamy, R., Venkatasubramanian, V.: A Fast Training Neural Network and its Updation for Incipient Fault Detection and Diagnosis. Computers and Chemical Eng. 24, 431–437 (2000)CrossRefGoogle Scholar
  7. 7.
    Perera, A., Papamichail, N., Bârsan, N., Weimar, U., Marco, S.: On-line Novelty Detection by Recursive Dynamic Principal Component Analysis and Gas Sensor Arrays under Drift Conditions. IEEE Sensors J. 6(3), 770–783 (2006)CrossRefGoogle Scholar
  8. 8.
    Mina, J., Verde, C.: Fault Detection for MIMO Systems Integrating Multivariate Statistical Analysis and Identification Methods. In: IFAC-ACC 2007, New York U.S.A, pp. 3234–3239 (2007)Google Scholar
  9. 9.
    Detroja, K., Gudi, R., Patwardhan, S.: Plant Wide Detection and Diagnosis using Correspondance Analysis. Control Eng. Practice 15(12), 1468–1483 (2007)CrossRefGoogle Scholar
  10. 10.
    Tudón-Martínez, J.C., Morales-Menendez, R., Garza-Castañón, L.: Fault Diagnosis in a Heat Exchanger using Process History based-Methods. In: ESCAPE 2010, Italy, pp. 169–174 (2010)Google Scholar
  11. 11.
    Peña, D.: Análisis de Datos Multivariantes. McGrawHill, España (2002)Google Scholar
  12. 12.
    Tudón-Martínez, J.C., Morales-Menendez, R., Garza-Castañón, L.: Fault Detection and Diagnosis in a Heat Exchanger. In: 6th ICINCO 2009, Milan Italy, pp. 265–270 (2009)Google Scholar
  13. 13.
    Hotelling, H.: Analysis of a Complex of Statistical Variables into Principal Components. J. Educ. Psychol. 24 (1993)Google Scholar
  14. 14.
    Freeman, J.A., Skapura, D.M.: Neural Networks: Algorithms, Applications and Programming Techniques. Adisson-Wesley, Reading (1991)zbMATHGoogle Scholar
  15. 15.
    Korbicz, J., Koscielny, J.M., Kowalczuk, Z., Cholewa, W.: Fault Diagnosis Models, Artificial Intelligence, Applications. Springer, Heidelberg (2004)zbMATHGoogle Scholar
  16. 16.
    Woods, K., Bowyer, K.W.: Generating ROC Curves for Artificial Neural Networks. IEEE Trans. on Medical Imaging 16(3), 329–337 (1997)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  1. 1.Tecnológico de MonterreyMonterreyMéxico

Personalised recommendations