Zusammenfassung
Die Ähnlichkeit der Matrix-Vorgabe beim Entwurf von beobachterfreien Ausgangs-Zustandsreglern mit dem Lernvorgang beim Backpropagation-Algorithmus wird aufgezeigt. Weiters wird der Lernprozess in künstlichen neuronalen Netzen mit und ohne Supervisor beschrieben, mit komplexen Lernvorgängen ausgestattet und an einem illustrativen Beispiel der automatischen Stabilisierung eines instabilen Prozesses angewendet.
Abstract
Simularities between matrix assignment in designing observer-free output controllers and the training phase in backpropagation algorithm are addressed. Learning processes in artificial neural networks with and without supervisor are emphasized regarding complex optimization, and demonstrated with an illustrative example of automatically stabilizing an unstable process.
Literatur
Cuthbert, T. R. (1987): Optimization using personal computers with application to electrical networks. Proc. Int. ICSC/IFAC Symposium on Neural Computation, Vienna. New York: Wiley, 854–860.
Fletcher, R. (1987): Practical methods of optimization. Chichester New York: Wiley.
Gill, P. E., Murray, W., Wright, M. H. (1981): Practical optimization. London: Academic Press.
Hagan, M. T., Demuth, H. B., Beale, M. H. (1996): Neural network design. Boston: PWS Publishing.
Hagan, M. T., Menhaj, M. (1994): Training feedforward networks with Marquardt algorithm. IEEE Trans. Neural Networks 5 (6).
Hecht-Nielsen, R. (1989): Neurocomputing. Reading: Addison-Wesley.
Nicoletti, G. M. (2000): An analysis of neural networks as simulators and emulators. Cybernetics and Systems: An International Journal 31: 253–282.
Papageorgiou, M. (1991): Optimierung. München: Oldenbourg.
Su, C.-T., Chang, H.-H. (2000): Optimization of parameter design: an intelligent approach using neural network and simulated annealing. Int. J. Systems Sci. 31: 1543–1549.
Wasserman, P. D. (1993): Advanced methods in neural computing. New York: Van Nostrand Reinhold.
Weinmann, A. (1999): Optimum system matrix assignment subject to limited controller Effort. IEEE-Conference on Decision and Control 1999, Phoenix, Arizona, December 1999, Paper WeA04-3, 1421–1426.
Weinmann, A. (2001): Matrix assignment subject to Lyapunov and ciosed-loop norm conditions. American Control Conference (ACC), Arlinton, June 2001, MP12-5, 1681–1686.
Zell, A. (1994): Simulation neuronaler Netze. Addison-Wesley.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Weinmann, A. Optimierung bei Reglerentwurf und neuronalem Lernen. Elektrotech. Inftech. 120, 172–175 (2003). https://doi.org/10.1007/BF03053937
Published:
Issue Date:
DOI: https://doi.org/10.1007/BF03053937