Abstract
In this paper, the authors discuss a new synthesis approach to train associative memories, based on recurrent neural networks. They propose to use soft margin training for associative memories, which is efficient when training patterns are not all linearly separable. On the basis of the soft margin algorithm used to train support vector machines, the new algorithm is developed in order to improve the obtained results via optimal training algorithm also innovated by the authors, which are not fully satisfactory due to that some times the training patterns are not all linearly separable. This new algorithm is used for the synthesis of an associative memory implemented by a recurrent neural network with the connection matrix having upper bounds on the diagonal elements to reduce the total number of spurious memory. The scheme is evaluated via a full scale simulator to diagnose the main faults occurred in fossil electric power plants and taking into account three different cases.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Liu, D., Lu, Z.: A new synthesis approach for feedback neural networks based on the perceptron training algorithm. IEEE Trans. Neural Networks 8, 1468–1482 (1997)
Michel, A.N., Farrel, J.A.: Associative memories via artificial neural networks. IEEE Contr. Syst. Mag. 10, 6–17 (1990)
Cortes, C., Vapnik, V.N.: Support Vector Networks. Machine Learning 20, 273–297 (1995)
Luemberger, D.: Linear and Non Linear Programming. Addison Wesley Publishing Company, USA (1984)
Casali, D., Constantini, G., Perfetti, R., Ricci, E.: Associative Memory Design Using Support Vector Machines. IEEE Transactions on Neural Networks 17, 1165–1174 (2006)
Ruz-Hernandez, J.A., Sanchez, E.N., Suarez, D.A.: Designing and associative memory via optimal training for fault diagnosis. In: Proceedings of International Joint Conference on Neural Networks, Vancouver, B. C., Canada, pp. 8771–8778 (2006)
Ruz-Hernandez, J.A., Sanchez, E.N., Suarez, D.A.: Optimal training for associative memories: application to fault diagnosis in fossil electric power plants, Book Chapter of Hybrid Intelligent Systems Analysis and Design. In: Castillo, O., Melin, P., Kacprzyc, J., Pedrycz, W. (eds.) International Series on Studies in Fuzzyness and Soft Computing, vol. 208, pp. 329–356 (2007); ISBN: 3-540-37419-1
Ruz-Hernandez, J.A.: Development and application of a neural network-based scheme for fault diagnosis in fossil electric power plants (In Spanish). Ph. D. Thesis, CINVESTAV, Guadalajara Campus (2006)
Frank, P.M.: Diagnostic procedure in the automatic control engineering. Automatic Control Engineering 2, 47–63 (1994)
Patton, R.J., Frank, P.M., Clark, R.N.: Fault Diagnosis in Dynamic Systems: Theory and Application. Prentice Hall, New York (1989)
Chen, J., Patton, R.J.: Robust Model Based Fault Diagnosis for Dynamic Systems. Kluwer Academic Publishers, Norwell (1999)
Köppen-Seliger, B., Frank, P.M.: Fault detection and isolation in technical processes with neural networks. In: Proceedings of the 34th Conference on Decision & Control, New Orleans, USA, pp. 2414–2419 (1995)
Comisión Federal de Electricidad, Manual del Centro de Adiestramiento de Operadores Ixtapantongo, Módulo III, Unidad 1, México (1997)
Ruz-Hernandez, J.A., Suarez, D.A., Shelomov, E., Villavicencio, A.: Predictive control based on an auto-regressive neuro-fuzzy model applied to the steam generator startup process at a fossil power plant. Revista de Computación y Sistemas 6(3), 204–212 (2003)
Noorgard, M., Ravn, O., Poulsen, N.K., Hansen, L.K.: Neural Networks for Modelling and Control of Dynamic Systems. Springer, London (2000)
Marquardt, D.: An algorithm for least-squares estimation of nonlinear parameters. SIAM Journal Appl. Mathematics 11(2) (1963)
Levenberg, K.: A method for solution of certain nonlinear problems in least squares. Quart. Appl. Mathematics 2, 164–168 (1944)
Nøorgard, M.: Neural Network based System Identification Toolbox (NNSYSID TOOLBOX), Tech. Report 97 E-851, Department of Automation, DTU, Lyngby, Denmark (1997)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Ruz-Hernandez, J.A., Sanchez, E.N., Suarez, D.A. (2008). Soft Margin Training for Associative Memories: Application to Fault Diagnosis in Fossil Electric Power Plants. In: Castillo, O., Melin, P., Kacprzyk, J., Pedrycz, W. (eds) Soft Computing for Hybrid Intelligent Systems. Studies in Computational Intelligence, vol 154. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-70812-4_12
Download citation
DOI: https://doi.org/10.1007/978-3-540-70812-4_12
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-70811-7
Online ISBN: 978-3-540-70812-4
eBook Packages: EngineeringEngineering (R0)