Sliding‐mode Linearization Torque Control of an Induction Motor

Abstract

This paper proposes a sliding-mode linearization torque control (SMLTC) for an induction motor (IM). An ideal feedback linearization torque control method is firstly adopted in order to decouple the torque and flux amplitude of the IM. However, the system parameters are uncertainties, which will influence the control performance of the IM in practical applications. Hence, to increase the robustness of the IM drive for high-performance applications, this SMLTC aims to improve the immunity of those uncertainties. We modify the flux observer of Benchaib and Edwards [15] by means of the adaptive sliding-mode method. This not only eliminates the estimation of the uncertainty bounds, but also improves the performance of sliding control. In addition, a practical application of the proposed SMLTC, with a model reference adaptive control (MRAC) scheme incorporated as the inner and outer loop controller used for position control, is also presented. Some experiments are presented to verify the control theory and demonstrate the robustness and effectiveness of the proposed SMLTC.

Tensor product transformation‐based modeling of an induction machine

“…State space modeling of the IM is one of the most commonly used technique to describe an electrical machine behavior. As previous studies [6‐11] showed, the dynamic model of the IM practically can be formulated in different ways, depending on the choice of the reference frame. Each type of reference frames has its own advantage, and for control purposes, the reference frame is fixed to rotor flux direction using a field‐oriented control algorithm (DRFOC).…”

“…The second chapter presents the formulation of the state space model of the IM well‐known from the literature [6‐12]. From this, the IM TP model can be formulated with the chosen linear parameters.…”

Two time‐scale assignment with state extension for an autonomous helicopter

“…During the past decades, many approaches to the negligence of the aforementioned coupling have been proposed [3,4,[6][7][8][9]. Control methods such as feedback lineariza-tion [3,10], backstepping strategy [4,11,12], dynamic inversion [6,9], and neural-network based methods [8,13] were implemented on a simplified model. However, it has been demonstrated that such a simplification may cause the aforementioned techniques to result in limitations in the operating regimes, and inaccurately followed reference commands [5,14,15].…”