In this paper, some identification methods for a piece wise linear (PWL) ci
rcuit model of hysteresis phenomenon are presented and compared. The model
provides a close prediction of static hysteresis and exhibits realistic dyn
amic features. The parameter identification problem can be solved by means
of a suitable procedures based on representation theorem, on optimisation m
ethods and on neural networks. Basic features are addressed and discussed.
(C) 2000 Elsevier Science B.V. All rights reserved.