Adaptive control of induction motor with unknown motor resistance

Abstract

In this paper, a new approach for induction motor rotor resistance variation estimate is presented by using adaptive neural networks. Indeed, the proposed neural networks are endowed with adaptive rules that allow them to estimate the true values of the necessary nonlinear state feedbacks for the input-output feedback linearization control of an induction motor. A comparison between the nonlinear state feedbacks provided by neural networks and those calculated through the nominal model of the induction motor with nominal parameters allows us to estimate the variation in the rotor resistance