Title :
Identification of a nonlinear motor system with neural networks
Author :
Sio, K.C. ; Lee, C.K.
Author_Institution :
Dept. of Electron. Eng., Hong Kong Polytech. Univ., Hung Hom, Hong Kong
Abstract :
In this paper, system identification with neural networks is studied. The insufficiency of a static feedforward neural network and a single-layer real-time recurrent neural network in system identification is discussed. Also, a modified recurrent network is proposed for a remedy. Derivations and simulations are provided to verify the beneficial properties of the proposed network
Keywords :
feedforward neural nets; identification; machine control; multilayer perceptrons; nonlinear control systems; recurrent neural nets; servomotors; modified recurrent network; neural networks; nonlinear motor system; single-layer real-time recurrent neural network; static feedforward neural network; system identification; Algorithm design and analysis; Backpropagation algorithms; Multi-layer neural network; Network topology; Neural networks; Neurofeedback; Power system modeling; Real time systems; Recurrent neural networks; System identification;
Conference_Titel :
Advanced Motion Control, 1996. AMC '96-MIE. Proceedings., 1996 4th International Workshop on
Conference_Location :
Mie
Print_ISBN :
0-7803-3219-9
DOI :
10.1109/AMC.1996.509420
