Controle intelligent et systemes autonomes : application a la robotique

Abstract

The predictive control has become currently a precious tool for control in various domains. It is well known and largely studied in the case of linear systems. The extension of this technique for the control of non linear systems has recently been the subject of many researches, where several algorithms were proposed. The objective of this work is to consider the application of neural networks to this technique of predictive control. This work deals with the most significant difficulties met during the development of these structures of nonlinear predictive control; that is obtaining a nonlinear model and a fast implementation of the control algorithm for real time applications. Thus, a part of this work is devoted to the identification of nonlinear systems using neural networks. The other part, deals with the problem of the implementation of the predictive control law when using a neural model for the prediction of a future behaviour of the system. In order to highlight the contribution of the proposed methods, simulation examples were considered.