Correo Electrónico
DNINFOA - SIA
Bibliotecas
Convocatorias
Identidad U.N.
Published
Fault diagnosis with neural networks. Part 1: Trajectory recognition
Diagnóstico de fallas con redes neuronales. Parte 1: Reconocimiento de trayectorias
DOI:
https://doi.org/10.15446/ing.investig.v27n1.14783Keywords:
fault diagnosis, artificial neural network, trajectory recognition, optimisation, noise tolerance (en)diagnóstico de fallas, redes neuronales, reconocimiento de trayectorias, optimización, tolerancia al ruido (es)
Downloads
The present investigation was focused on formulating a method for designing a fault diagnosis system for chemical plants by using artificial neural networks. Fault diagnosis is aimed at identifying a fault which affects a given process by analysing the signs supplied by process sensors. Neuronal networks are mathematical models which try to imitate the functioning of the human brain. A neural network is defined by its structure and the learning method used. The difficulty with diagnosing faults lies in recognising the trajectories (temporal series of data) followed by process variables when a fault affects the process; when trajectories are recognised, the associated fault is also identified. The theory so developed recommended an optimised structure and training method for the neural networks to use. Both the proposed structure and the training method were tested by carrying out comparative studies between traditional structures and a training method. The results showed the superiority of the neural networks designed and trained with the method proposed in this work. Except for simple processes, fault diagnosis is a more complex problem than simply identifying trajectories, because a fault may cause an infinite set of trajectories (i.e. flow). The fundaments established in this work are thus used in Part Il, where the analysis is extended to recognise flows.
La investigación realizada tuvo como objetivo la formulación de un método para el diseño de un sistema de diagnóstico de fallas para plantas químicas utilizando redes neuronales artificiales. El diagnóstico de fallas tiene como misión identificar la falla que está afectando a un proceso dado a través del análisis de las señales suministradas por los sensores del proceso. Las redes neuronales son modelos matemáticos que intentan reproducir la actividad cognoscitiva del cerebro humano. Estas se caracterizan por su estructura y el método de aprendizaje utilizado. El problema del diagnóstico de fallas se aborda a partir de la perspectiva de la identificación de las trayectorias (secuencias temporales de datos) que describen las variables del proceso al ser afectado por una falla. De esta forma, reconocidas las trayectorias, se habrá identificado la falla asociada. El desarrollo teórico realizado recomienda una estructura y un método de entrenamiento optimizado para las redes neuronales a emplear. Tanto la estructura como el método de entrenamiento propuesto fueron evaluados realizando estudios comparativos con estructuras y un método de entrenamiento tradicionales. Los resultados así obtenidos mostraron la superioridad de las redes neuronales diseñadas y entrenadas con el método propuesto en este trabajo. Salvo en procesos simples, el diagnóstico de fallas es más complejo que el reconocimiento de trayectorias porque cada falla puede provocar un conjunto infinito de trayectorias (flujo). Por ese motivo, los fundamentos establecidos en el trabajo son utilizados en la parte II, donde el análisis se extiende al reconocimiento de flujos.
References
Chen, B. H., Wang, X. Z., Yang, S. H. and Mcgreavy, C., Application of wavelets and neural networks to diagnostic system Development., I. Feature extraction, Computers and Chemical Engineering, 23 (7), 1999, pp. 899-906. DOI: https://doi.org/10.1016/S0098-1354(99)00258-6
Demuth, H. and Beale, M., Neural Network Toolbox for Use with MATLAB, The MathWorks, Inc. USA, 2000.
Garcez-Castro, A. R. and Miranda, V., An interpretation of neural networks as inference engines with application to transformer failure diagnosis., Electrical Power and Energy Systems, 27, 2005, pp. 620-626. DOI: https://doi.org/10.1016/j.ijepes.2005.08.002
Fan J., Nikolaou, M. and White, R., An approach to fault diagnosis of chemical processes via neural networks., American Institute of Chemical Engineers Journal, 39 (1), 1993, pp. 82-88. DOI: https://doi.org/10.1002/aic.690390109
Looney, C.G., Pattern Recognition Using Neural Networks: Theory and Algorithms for Engineers and Scientists., Oxford University Press, New York, 1997.
Persina, S. and Tovornik, B., Real-time implementation of fault diagnosis to a heat exchanger., Control Engineering Practice 13, 2005, pp. 1061-1069. DOI: https://doi.org/10.1016/j.conengprac.2004.12.005
Principe, J., Dynamic Neural Networks and Optimal Signal Processing, Capítulo 6 en Handbook of Neural Network Signal Processing, CRC Press, USA, 2002. DOI: https://doi.org/10.1201/9781420038613.ch6
Rengaswamy, R. and Venkatasubramanian, V., A fast training neural network and its updation for incipient fault detection and diagnosis., Computers and Chemical Engineering, 24, (2-7), 2000, pp. 431- 437. DOI: https://doi.org/10.1016/S0098-1354(00)00434-8
Russell, S.J. and Norvig, P., Artificial Intelligence - A Modern approach., Prentice Hall, New Jersey, 1995.
Tarifa, E.E. y Martínez, S.L., Diagnóstico de fallas con redes neuronales. Parte II: Reconocimiento de flujos, Ingeniería e Investigación, In press, 2007.
Tarifa, E.E., Humana, D., Franco, S., Martínez, S., Nuñez, A. and Scenna, N., Fault diagnosis for a MSF using neural networks., Desalination, 152, 2002, pp. 215-222. DOI: https://doi.org/10.1016/S0011-9164(02)01066-4
Venkatasubramanian, V., Rengaswamy, R., Kavuri, S. and Yin, K., A review of process fault detection and diagnosis. Part III: Process history based methods., Computers and Chemical Engineering, 27, 2003, pp. 327-346. DOI: https://doi.org/10.1016/S0098-1354(02)00162-X
Wah, B. and Qian, M., Constraint-Based Neural Network Learning for Time Series Predictions., Department of Electrical and Computer Engineering and the Coordinated Science Laboratory University of Illinois, Urbana-Champaign, USA, 2002.
Wang, Z., Zhang, Y., Li, C. and Liu, Y, ANN-Based Transformer Fault Diagnosis., 59th American Power Conference, Chicago, Vol. 59-I, 1997, pp. 428-432.
Witczaka, M., Korbicza, J., Mrugalskia, M. and Patton, R.J., A GMDH neural network-based approach to robust fault diagnosis: Application to the DAMADICS benchmark problem., Control Engineering Practice, 14, 2006, pp. 671-683. DOI: https://doi.org/10.1016/j.conengprac.2005.04.007
Zhang, J., Improved on-line process fault diagnosis through information fusion in multiple neural networks., Computers and Chemical Engineering, 30, 2006, pp. 558-571. DOI: https://doi.org/10.1016/j.compchemeng.2005.11.002
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
CrossRef Cited-by
Dimensions
PlumX
Article abstract page views
Downloads
License
Copyright (c) 2007 Enrique Eduardo Tarifa, Sergio Luis Martínez
This work is licensed under a Creative Commons Attribution 4.0 International License.
The authors or holders of the copyright for each article hereby confer exclusive, limited and free authorization on the Universidad Nacional de Colombia's journal Ingeniería e Investigación concerning the aforementioned article which, once it has been evaluated and approved, will be submitted for publication, in line with the following items:
1. The version which has been corrected according to the evaluators' suggestions will be remitted and it will be made clear whether the aforementioned article is an unedited document regarding which the rights to be authorized are held and total responsibility will be assumed by the authors for the content of the work being submitted to Ingeniería e Investigación, the Universidad Nacional de Colombia and third-parties;
2. The authorization conferred on the journal will come into force from the date on which it is included in the respective volume and issue of Ingeniería e Investigación in the Open Journal Systems and on the journal's main page (https://revistas.unal.edu.co/index.php/ingeinv), as well as in different databases and indices in which the publication is indexed;
3. The authors authorize the Universidad Nacional de Colombia's journal Ingeniería e Investigación to publish the document in whatever required format (printed, digital, electronic or whatsoever known or yet to be discovered form) and authorize Ingeniería e Investigación to include the work in any indices and/or search engines deemed necessary for promoting its diffusion;
4. The authors accept that such authorization is given free of charge and they, therefore, waive any right to receive remuneration from the publication, distribution, public communication and any use whatsoever referred to in the terms of this authorization.
