Employment of a Progressive Learning Neural Network for Identification and Control
In this paper a direct inverse control scheme is presented, which is based on a clustering neural network, called Progressive Learning Network (PLN) because of its inherent capacity of learning on-line. .
After describing the PLN, the generalised and specialised inverse control schemes are introduced and then a method for using the PLN in this kind of control is shown. In particular a new version of this PLN is developed for the on-line control with specialised learning. This approach can control the whole system without having to use a very rich training set; moreover it is able to adapt itself on-line to new working conditions as it is based on an algorithm capable of varying the number of neurons of the hidden layer in order to learn examples that had not been presented previously or to forget rare situations. Numerical tests then follow to validate the control strategy.
KeywordsHide Layer Specialise Learning Load Torque Neural Controller Merging Algorithm
Unable to display preview. Download preview PDF.
- Miller, Sutton,Werbos editors, Neural networks for Control,, MIT Press, Cambridge MA, 1992Google Scholar
- M. Cirrincione, Diagnostica e controllo degli azionamenti elettrici con reti neurali, Ph. D. Thesis, Palermo 1996.Google Scholar
- Sorensen O.: Neural Networks in Control Applications Ph.D Thesis, Aalborg University, Dept. Control Engineering, 1994Google Scholar
- Widrow B., Walach E.: Adaptive Inverse Control, Prentice-Hall, Englewood Cliffs, N.J., 1996Google Scholar
- Renders Jean-Michel:Algorithmes génétiques et réseaux de neurones Hermès Paris 1995Google Scholar
- Cirrincione G., Cirrincione M. Piglione F.: A Clustering Neural Network for On-Line Learning in Power System., Neurap ’95, Marseille, 13–15 December 1995Google Scholar
- G. Cirrincione, M.Cirrincione,F.Piglione:Applicazione di una rete neurale ad apprendimento progressivo alio studio dei sistemi di potenza, 96° riunione annuale AEI, Roma, 25–27 September 1995Google Scholar