The subject of this investigation is the methods of switched reluctance motors controlling. The study objective is to review the existing methods of reducing shaft electromagnetic torque ripple in the switched reluctance motors. A comparative analysis of the existing approaches to valve drives control systems development is used to classify the electromagnetic torque control methods. The study results can be used when selecting the drive control method providing a balance between the complexity of the control system implementation and its effectiveness. On the basis of the comparative analysis, the conclusions on the neural network torque control systems effectiveness with offline learning are made; the control system enhancement using co-energy magnetic drive system as well as the inclusion of the residual magnetization material effects in the model is marked. It is noted that the existing systems aligning the magnetization of the actuator materials reduce the average torque for the work cycle.