Title: Sliding mode control and neuro-fuzzy network observer for induction motor in EVs applications

Authors: Moez Ghariani, Mohamed Radhouan Hachicha, Arafet Ltifi, Ibrahim Bensalah, Moez Ayadi, Rfik Neji

Addresses: Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia. ' Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia. ' Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia. ' Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia. ' Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia. ' Electric Vehicle and Power Electronics Group (VEEP), Laboratory of Electronics and Information Technology (LETI), National School of Engineers of Sfax, B.P. 1173 3038 Sfax, Tunisia

Abstract: In order to deal with the major problems of electric vehicle (EV) propelled by an induction motor (IM), a neuro-fuzzy sliding mode control was designed. The proposed scheme uses an adaptive observer that is based on a full order model of the IM in indirect vector-controlled drive. Moreover, it is evaluated on an EV global model taking into account the vehicle dynamics, the thermal effect and the parameters variations. The neuro-fuzzy network is used to adaptively adjust the parameters prediction. However, the sliding mode control can offer many good properties such as good performance against unmodelled dynamics, insensitivity to external disturbance rejection and fast dynamic. Simulations results carried out on a test vehicle, show that the proposed controller is superior to PID one at response speed, steady-state tracking error and resisting perturbation whenever driving or braking. This shows that the proposed scheme is good candidate for EV|s propulsion.

Keywords: decoupling mechanism; induction motors; recurrent neural networks; speed sensorless IM drives; electric vehicles; electric propulsion; sliding mode control; neuro-fuzzy networks; adaptive observers; simulation; response speed; tracking errors; perturbation.

DOI: 10.1504/IJEHV.2011.040471

International Journal of Electric and Hybrid Vehicles, 2011 Vol.3 No.1, pp.20 - 46

Published online: 30 May 2011 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article