Abstarct: This thesis presents several approaches for the speed control of induction motors (IMs) using the approximation capability of neural networks and fuzzy systems. The uncertainties including parametric variations, the external load and input voltage disturbances and unmodeled dynamics are estimated and compensated by designing proper neuro-fuzzy controllers. The main contribution of this thesis is developing speed controllers for medium sized IMs with guaranteed stability. Also, the structure of the proposed neuro-fuzzy systems is simple and leads to less computational load in practical implementation and easier tuning of controller parameters is needed. Another advantage of this method is compensating of the reconstruction error in neuro-fuzzy estimators in order to guarantee the asymptotic convergence of the speed tracking error. In addition, the boundedness of all signals in the closed loop system is obtained. The control systems are proposed baxsed on the indirect adaptive fuzzy and dynamic sliding mode strategies and in order to simulate these systems, the proper mathematical models of induction motor are presented. Finally, simulation results show that the proposed controllers provide high-performance characteristics and are robust with regard to plant parameter variations, external load and input voltage disturbance.