Title :
Neural network based friction compensation in motion control
Author :
Ciliz, M.K. ; Tomizuka, M.
Author_Institution :
Electr. Eng. Dept., Bogazici Univ., Istanbul, Turkey
fDate :
6/10/2004 12:00:00 AM
Abstract :
Frictional uncertainties are known to be a major cause of performance degradation in motion control systems. Artificial neural network based modelling and compensation of nonlinear friction in direct drive servo mechanisms have been investigated. The proposed method is successfully tested experimentally on a direct-drive single-link arm and its performance is compared to different friction modelling techniques.
Keywords :
compensation; feedforward neural nets; friction; mechanical variables control; motion control; multilayer perceptrons; neurocontrollers; servomechanisms; artificial neural network based modelling; direct drive servo mechanisms; direct drive single link arm; friction modelling techniques; motion control systems; nonlinear friction compensation; performance degradation;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:20040500
