Rejeb S.,Unit C3S |
Ali R.B.H.,University of Monastir |
Hmida F.B.,Unit C3S |
Chaari A.,Unit C3S |
Messaoud H.,University of Monastir
Scientific Research and Essays | Year: 2011
This paper deals with the recursive parameter identification of the Hammerstein ARMAX (HARMAX) model with General model of discontinuous and asymmetric Nonlinearity (GNL) containing hysteresis, saturation, preload and dead-zone. This model contains different parameters the choice of which may generate several different nonlinearities. The linear and nonlinear structures are known, the linear block parameters are unknown. However the nonlinear block contains nine parameters only one of them is known. The estimation of the nonlinearity parameters, the internal variables relative to the nonlinearity as well as the linear model parameters are ensured by Recursive Extended Least Squares algorithm (RELS). The provided model is validated by simulation results. © 2011 Academic Journals.
Regaya C.B.,Unit C3S |
Zaafouri A.,Unit C3S |
Chaari A.,Unit C3S
Acta Polytechnica Hungarica | Year: 2014
In this paper, the speed of the induction machine is controlled by a variable structure controller. To eliminate speed sensor we use a sliding mode observer based on fuzzy logic "FSMSO". The control algorithm and observation is emphasized by simulation tests and it is comapred with the classical sliding mode speed observer"SMSO". Analysis of the results shows the characteristic robustness to disturbances of the load and the speed variation for the FSMSO than SMSO.