The switched reluctance motor (SRM) have attracted many researchers over the last decade. This is due to its many advantages such as simple and robust construction, low-inertia, high-speed, high temperature performance, low costs, and fault tolerance control capabilities. However, several disadvantages like acoustic noise generation, torque ripple, nonlinear characteristics and the strong dependence on rotor position to achieve a good control, are limiting its utilization compared to other kind of motors. Torque ripple in low speed Cause to fluctuation in motor speed and fatigue. Thus, a lot of methods have been proposed so far for torque-ripple minimization in switched reluctance motors. There are two methods to improve the performances of SRM drive systems. The design procedure for SRM has been explored in detail to try minimizing the torque ripple. A lot of control techniques proposed for ripple reduction. The nonlinear factors are the doubly salient structure and extreme saturation, back-EMF so these motors are highly nonlinear plant for the controller design. The purpose of this thesis is to propose a method for controlling the motor speed with reduced electromagnetic torque ripple. First, a review on research opportunities and torque ripple reduction methods is given, and then the operation principle and modeling of the motor is discussed. For this purpose to be achieved, a proper model able to express the nonlinearity and saturation of the motor is required. In this thesis the required model is based on the data points of a 4KW 6/8 motor. These data points are obtained by the means of finite element analysis (FEA) and their validity is confirmed by experimental tests. Afterwards, a review and comparison is done on Different switch reluctance motor’s converters and the specification of the converter applied in this thesis’s simulations and experimental tests is discussed. Initially a normal PI controller has been used to control the speed in matlab simulink software, then to counter the uncertainties and nonlinearities in the model; an emotional learning controller is applied to reduce the torque ripple of the PI controller. Being Non model based, the BELBIC (Brain Emotional Based Intelligent Controller) controller is a good choice for controlling non linear systems. Also this controller has a large freedom factor to reach the final desirable answer (in terms of overshoot, settling time, steady state error and soft starting). This Controller has shown reasonable robustness and uncertainty handling abilities While it is simple and can be implemented easily and being unaffected by noise has made this controller an ideal Keywords : Switched Reluctance Motor, Speed Control, Torque Ripple Reduction, Brain Emotional Learning Based Intelligent Controller