0
TECHNICAL PAPERS: Gas Turbines: Oil and Gas Applications

A Demonstration of Artificial Neural-Networks-Based Data Mining for Gas-Turbine-Driven Compressor Stations

[+] Author and Article Information
K. K. Botros

Nova Research and Technology Corporation, Calgary, Alberta, Canada

G. Kibrya, A. Glover

TransCanada Pipelines, Ltd., Calgary, Alberta, Canada

J. Eng. Gas Turbines Power 124(2), 284-297 (Mar 26, 2002) (14 pages) doi:10.1115/1.1414130 History: Received November 01, 1999; Revised February 01, 2000; Online March 26, 2002
Copyright © 2002 by ASME
Your Session has timed out. Please sign back in to continue.

References

Bishop, C. M., 1995, Neural Networks for Pattern Recognition, Clarendon Press, Oxford, UK.
Jain,  A. K., , 1996, “Artificial Neural Networks—A Tutorial,” Computer, 29, No. 3, pp. 31–44.
DePold, H. R., and Gass, F. D., 1988, “The Application of Expert Systems and Neural Networks to Gas Turbine Prognostics and Diagnostics,” ASME Turbo-Expo, Stockholm, Sweden, June 2–5.
MacIntyre, J., Tait, J., Kendal, S., Smith, P., Harris, T., and Brason, A., 1994, “Neural Networks Applications in Condition Monitoring,” Applications of Artificial Intelligence in Engineering, Proceedings of the 9th International Conference, Pennsylvania, July 19–21, pp. 37–48.
Kim, D. S., Shin, S. S., and Carison, D. K., 1991, “Machinery Diagnostics for Rotating Machinery Using Backpropagation Neural Network,” Proceedings of the 3rd International Machinery Monitoring and Diagnostics Conference, Las Vegas, NV, Dec. 9–12, pp. 309–320.
Lombardo, G., 1996, “Adaptive Control of a Gas Turbine Engine for Axial Compressor Faults,” Proceedings of the 1996 International Gas Turbine and Aeroengine Congress & Exhibition, Burmingham, UK, June 10–13, ASME, New York.
Zhang,  S., and Ganesan,  R., 1997, “Multivariable Trend Analysis Using Neural Networks for Intelligent Diagnostics of Rotating Machinery,” ASME J. Eng. Gas Turbines Power, 119, pp. 378–384.
Lo,  C. F., and Shi,  G. Z., 1992, “Wind-Tunnel Compressor Stall Monitoring Using Neural Networks,” J. Aircr., 29, No. 4, pp. 736–738.
Gibson, S., 1994, “Detection of Surge Utilizing Magnetic Bearings,” Revolve 1994, Sponsored by ASME, National Petroleum Show.
Chbat, N. W., Rajamani, R., and Ashley, T. A., 1996, “Estimating Gas Turbine Internal Cycle Parameters Using a Neural Network,” Proceedings of the 1996 International Gas Turbine and Aeroengine Congress & Exhibition, Burmingham, UK, Jun 10–13, ASME, New York.
Botros, K. K., and Glover, A., 1999, “Neural Networks & Fuzzy Logic—Overview of Pipeline Applications (Part I and II),” 18th International Conference on Offshore Mechanics and Arctic Engineering, St. John’s, Newfoundland, July 11–16.
Hunt,  K. J., Sbarbaro,  D., Zbikowski,  R., and Gawthrop,  P. J., 1992, “Neural Networks for Control Systems—A Survey,” Automatica, 28, No. 6, pp. 1083–1112.
Kohonen,  T., 1989, “The Self-Organizing Map,” Proc. IEEE, 78, No. 9, pp. 1464–1480.

Figures

Grahic Jump Location
Schematic of a three-spool gas turbine driving a gas compressor
Grahic Jump Location
(a) MLP (7-14-8) architecture, (b) MLP (7-28-8) architecture, (c) MLP (7-14-14-8) architecture, (d) RBF (7-128-8) architecture, (e) GRNN (7-1844-9-8) architecture  
Grahic Jump Location
Results of testing the RBF network architecture of Fig. 2(d) using 1996 data set for training and then testing
Grahic Jump Location
Prediction of Qse for a period in 1997 versus measurements
Grahic Jump Location
Detection of CDP sensor failure up to Case 720
Grahic Jump Location
Estimate of the fuel gas flow for 1998 when the flow sensor failed in the entire year
Grahic Jump Location
(a) Comparison between predicted versus measured power turbine inlet pressure for 1997 year based on 1996 trained RBF NN; (b) comparison between predicted versus measured GG exhaust temperature for 1997 year based on 1996 trained RBF NN; (c) comparison between predicted versus measured power turbine exhaust temperature for 1997 year based on 1996 trained RBF NN; (d) comparison between predicted versus measured shaft power for 1997 year based on 1996 trained RBF NN; (e) comparison between predicted versus measured HP spool speed for 1997 year based on 1996 trained RBF NN; (f) comparison between predicted versus measured HP fuel gas flow for 1997 year based on 1996 trained RBF NN
Grahic Jump Location
Hybrid NN–FPM scheme for estimation of parameters that cannot be measured
Grahic Jump Location
RBF architecture used in TIT training/prediction
Grahic Jump Location
NN predicted versus estimated TIT from FPM for the same year
Grahic Jump Location
NN predicted versus estimated TIT from FPM for a subsequent year
Grahic Jump Location
Kohenen SOM of 15 input units and 3×15 output units
Grahic Jump Location
Winning unit activation level for each case of 1996 data set
Grahic Jump Location
Activation levels of the input units for cases 326 and 327 of 1996 data set
Grahic Jump Location
Activation levels of the input units for cases 410 and 415 of 1996 data set
Grahic Jump Location
Activation levels of the input units for cases 561 and 562 of 1996 data set
Grahic Jump Location
Winning unit activation level for each case of 1997 data set
Grahic Jump Location
Winning unit activation level for each case of 1998 data set

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In