In this work, three control techniques of minimization of torque ripple in SRM are proposed and compared to the classical control technic names as Hysteresis control proposed in [1]. The principle of these methods is based on the Torque Sharing Function (TSF) fixing the reference torque in each phase. In the literature review, the classical method used in the reduction of torque ripples in SRM is the TSF-Hysteresis Controller (TSF-H). The first proposed method is the TSF-Predictive control (TSF-P), here the objective is to find the voltage to be applied to the machine minimizing the error of the current through a prediction algorithm on the supply current. The second proposed method is the sliding mode control (TSF-SMC-PSO) whose parameters are optimized by the particle swarm algorithm (PSO). Finally, the third proposed method is the sliding mode controller whose parameters are optimized by fuzzy logic controller (TSF-SMFC). The finite element method (FEM) through the magnetic field calculation software FEMM was used to calculate the flux and torque in static system and to take into account magnetic saturation circuit of the SRM. The results obtained show that the TSF-SMC-PSO control methods have better performance compared to the other proposed methods mentioned above.