Ardashir Mohammadzadeh

This thesis studies developing adaptive control techniques for a class of nonlinear systems. The dynamic of a nonlinear systems is usually perturbed by several factors, such as measurement errors, external disturbance, physical device destruction, and so on. In such cases, conventional non-adaptive controllers cannot result in reliable performance. To cope with dynamic perturbation, adaptive controllers and especially adaptive fuzzy controllers have attracted attention in recent years. In this research, we address induction motors (IMs) and chaotic systems (CSs) for the case studies, and we develop and design several adaptive control methods. These methods are summarized as follows: Part I of the thesis concentrates on induction motor control. In accordance with their advantages, induction motors are recognized as the best candidate for mechanical drives in automotive applications. Nevertheless, precise control of these motors has been found to be very difficult, mainly due to their complex nonlinear dynamics. In addition, the inability to directly measure rotor variables and the variations of physical parameters under different operating conditions have further increased the difficulties of the controller design process, causing induction motor control to be a challenging topic. In the present study, control problem of the induction motors is investigated in two different cases. In the first case, a number of parametric uncertainties including load torque and time-varying rotor resistance are taken into account and the dynamics of the induction motors (IM) are considered to be identified; therefore, a novel Immersion and Invariance-based adaptive controller and a non-adaptive robust Immersion and Invariance (I&I) controller are proposed. In the second case, to estimate uncertainties, a type-2 fuzzy system based on rough neural network (T2FRNN) and Group Method of Data Handling (GMDH) neural networks are proposed and the dynamics of the induction motor are considere