A TIME ENCODED SIGNAL PROCESSING/NEURAL NETWORK APPROACH TO FAULT CLASSIFICATION OF AN ELECTROHYDRAULIC CONTROL SYSTEM
Keywords:
fault diagnosis, fluid power systems, time encoded signal processing, artificial neural networksAbstract
A fault classification approach is presented which considers the advantages of Time Encoded Signal Processing (TESP) of dynamic signals combined with the ability of Artificial Neural Networks (ANNs) to classify changes in TESP codes. This is demonstrated using a new TESP code approach applied to a pressure control system exhibiting both leak-age at the actuator and a servovalve fault. It was found that the use of both pressure transducer voltage and servovalve drive voltage, when entered into the ANN in a parallel data structure manner, resulted in an excellent fault classification capability. In addition the inherent classification approach gave very good leakage discrimination for arbitrarily-set, and low, levels of 0, 2, 4, 6 l/min. A range of 16 different ANNs were investigated and the classification results indicate a preferred topology for this application.
Downloads
References
Atkinson, R. M. et al 1992. Automated fault tree anal-ysis for hydraulic systems, Part 1: fundamentals, Proc IMechE, Part I. Journal of Systems and Con-trol Engineering, Vol 206, pp. 207-214.
Crowther, W. J. et al 1998. Fault diagnosis of a hy-draulic actuator circuit using neural networks, Proc IMechE, Part I. Journal of Systems and Control Engineering, Vol 212, pp. 57-68.
Darling, R. and Tilley, D. G. 1993. Progress towards a general purpose technique for the condition moni-toring of fluid power systems. Proc IMechE Con-ference on Aerospace Hydraulics and Systems, London, pp. 47-55.
Edge, K. A. et al 1995. An approach to automated fault diagnosis of hydraulic circuits using neural net-works. ASME Fluid Power Systems and Technolo-gy, Vol 2, pp. 37-43.
Freebody, N. and Watton, J. 1999. A time encoded signal processing approach to fault classification of an electrohydraulic pressure control system and its application to a hot strip rolling mill. Proc IMechE, Part I. Journal of Systems and Control Engineering, Vol 213, pp. 407-426.
Holbeche, J., Hughes R. D. and King, R. A. 1986. Time encoded speech (TES) descriptors as a symbol feature set for voice recognition systems. Proc IEE International Conference on Speech Input/Output Techniques and Applications, London, pp. 310-315.
Le, T. T., Watton, J. and Pham, D. T. 1997. An arti-ficial neural network based approach to fault diag-nosis and classification of fluid power systems. Proc IMechE, Part I. Journal of Systems and Control Engineering, Vol 211, pp. 307-317.
Le, T. T., Watton, J. and Pham, D. T. 1998. Fault classification of fluid power systems using a dynam-ics feature extraction technique and neural net-works. Proc IMechE, Part I. Journal of Systems and Control Engineering, Vol 212, pp. 87-97.
Ramden, T. 1995. On Condition Monitoring of Fluid Power Pump and Systems. Linköping Studies in Science and Technology Thesis No 474, Linköping University.
Ramden, T., Weddfelt, K. and Palmberg Jan-Ove 1993. Condition Monitoring of Fluid Power Pumps by Vibration Measurement. 10th International Con-ference on Fluid Power-The future for Hydraulics, Brugge, Belgium, MEP Publications Ltd, pp. 263-276.
Rodwell, G. M. and King, R. A. 1996. TESPAR/FANN architectures for low power, low cost condition monitoring applications. Proc COMADEM 1996, pp. 1089-1101.
Watton, J., Lucca-Negro, O. and Stewart J. C. 1994. An on-line approach to fault diagnosis of fluid pow-er cylinder drive systems. Proc IMechE, Part I, Vol 208, pp. 249-262.
Watton, J. and Stewart, J. C. 1996. Co-operating ex-pert knowledge and artificial neural networks for fault diagnosis of electrohydraulic cylinder position control systems. Proc 3rd JHPS International Sym-posium, Yokohama, pp. 217-222.
