Title: Decentralised neural control of active suspensions

Authors: N. Lagraa, D. Boukhetala, A. Guessoum, G. Bloch

Addresses: Laboratoire des Sciences Fondamentales, Universite de Laghouat, Algeria ' Laboratoire de Commande des Processus (LCP), Ecole Nationale Polytechnique, 10, avenue Pasteur BP 182, El Harrach, Algiers, Algeria. ' Laboratoire de Systemes Electriques et Telecommunications, Universite de Blida, Algeria. ' Centre de Recherche en Automatique de Nancy CRAN-UMR CNRS 7039, ESSTIN 2 rue Jean Lamour 54519 Vandoeuvre, Les Nancy, France

Abstract: This work deals with the decentralised supervised control of active suspension using neural networks. Firstly, a non-linear control law is synthesised via the backstepping control design approach by using a non linear half-car model. In order to improve the dynamic performances of the closed loop system, a neural network is used to supervise the developed control law. To demonstrate the effectiveness of the proposed control strategy, several simulations are performed on the non-linear system. The obtained results show that the performances are improved by the decentralised supervised controller.

Keywords: active suspension; backstepping control; decentralised control; half-car model; neural controller; vehicle suspension; neural networks; nonlinear control; simulation.

DOI: 10.1504/IJVAS.2005.008238

International Journal of Vehicle Autonomous Systems, 2005 Vol.3 No.2/3/4, pp.292 - 305

Published online: 25 Nov 2005 *

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