Identificación de Sistemas Dinámicos Utilizando Redes Neuronales RBF

Ricardo Valverde Gil, Diego Gachet Páez

Resumen

La identificación de sistemas complejos y no-lineales ocupa un lugar importante en las arquitecturas de neurocontrol, como por ejemplo el control inverso, control adaptativo directo e indirecto, etc. Es habitual en esos enfoques utilizar redes neuronales “feedforward” con memoria en la entrada (Tapped Delay) o bien redes recurrentes (modelos de Elman o Jordan) entrenadas off-line para capturar la dinámica del sistema (directa o inversa) y utilizarla en el lazo de control. En este artículo presentamos un esquema de identificación basado en redes del tipo RBF (Radial Basis Function) que se entrena on-line y que dinámicamente modifica su estructura (número de nodos o elementos en la capa oculta) permitiendo una implementación en tiempo real del identificador en el lazo de control.

Palabras clave

Identificación; sistemas no-lineales; redes neuronales; estimación de parámetros

Texto completo:

PDF

Referencias

Chen S., S.A. Billings y P.M. Grant (1990). Nonlinear system identification using neural networks. International Journal of Control, vol. 51(6), 1191-1214.

Chen, S. y S.A. Billings (1994). Neural Networks for Nonlinear Dynamic System Modelling and Identification. En: Advances in Intelligent Control (C.J. Harris (ed.)) 85-112. Taylor & Francis, London.

Chi S.R., R. Shoureshi y M. Tenorio (1990). Neural networks for system identification. IEEE Control Systems Magazine 10, 31-34.

Goodwin G.C. y K.S. Sin (1984) Adaptive filtering prediction and control. Prentice-Hall, Englewood Cliffs, NJ.

Kuschewski G.J, S. Hui y S.H. Zak (1993). Application of feedforward neural networks to dynamical system identification and control. IEEE Trans. Control Systems Technology 1(1), 37-49.

Le Cun, Y. (1985) Une procedure d’aprentissage pour reseau a sequil assymetrique. Proceedings of Cognitiva 85, 599-604.

Li, Y., N. Sundararajan, y P. Saratchandran (2000) Analysis of Minimal Radial Basis Function Network Algorithm for Real-Time Identification of Nonlinear Dynamic Systems. IEE Proc. on Control Theory and Applications 147(4), 476- 484.

Narendra, K.S. y K. Parthasarathy (1989). Backpropagation in dynamical systems containing neural networks. Technical Report 8905, Centre for Systems Science, Department of Electrical Engineering, Yale University, New Haven, CT.

Narendra K.S. y K. Parthasaraty (1990). Identification and Control of Dynamical Systems using Neural Networks. IEEE Transactions on Neural Networks 1(1), 4-27.

Obradovic, D. (1996). On-Line Training of Recurrent Neural Networks with Continuous Topology Adaptation. IEEE Trans. on Neural Networks 7(1), 222-228.

Panchapakesan, C., y M. Palaniswami (2002). Effects of Moving the Centres in an RBF Network. IEEE Transactions on Neural Networks 13(6), 1299- 1307.

Parker, D. B. Learning logic (1985). En: Technical Report TR-47, Massachusetts Institute of Technology, Cambridge, MA.

Peng, H. et al (2004) RBF-ARX model-based nonlinear system modeling and predictive control with application to a NOx decomposition process. Control Engineering Practice 12 (2), 191-203.

Poggio T. y F. Girosi. (1989). A Theory of Networks for Approximation and Learning. En: A.I. Memo No. 1140. Artificial Intelligence Laboratory, M.I.T.

Rumelhart, D.E., G.E. Hinton, y R.J. Williams (1986). Learning internal representations by error propagation. En: Parallel Distributed Processing: Explorations in the microstructure of cognition. D.E. Rumelhart and J.L. McClelland, eds Vol 1, Foundations. Cambridge, MA: Bradford Books/ MIT Press

Sanner R.M. y J.E. Slotine (1991) Stable Adaptive Control and Recursive Identification Using Radial Gaussian Networks. En: Proceedings 30th Conference on Decision and Control. Brighton, England.

Sjöberg, J., Q. Zhang, L. Ljung et al (1995). Nonlinear Black-box Modelling in System Identification: a Unified Overview. Automatica 31(12), 1691-1724.

Werbos P.J. (1974). Beyond regression: New tools for prediction and analysis in the behavioral sciences. En: Ph. D. dissertation., Harvard University, Cambridge, MA.

Widrow B. y S.D. Stearns (1985). Adaptive Signal Processing. Prentice-Hall, Englewood Cliffs, NJ.

Zemouri, R., D. Racoceanu y N. Zerhouni (2003). Recurrent radial basis function network for timeseries prediction. Engineering Applications of Artificial Intelligence 16, (5-6), 453-463.

Abstract Views

775
Metrics Loading ...

Metrics powered by PLOS ALM




Creative Commons License

Esta revista se publica bajo una Licencia Creative Commons Attribution-NonCommercial-CompartirIgual 4.0 International (CC BY-NC-SA 4.0)

Universitat Politècnica de València     https://doi.org/10.4995/riai

e-ISSN: 1697-7920     ISSN: 1697-7912