Spectral analysis of the stator current of a DTC control with a neuro-fuzzy flux estimator applied to an Induction Motor

One of the best control methods for managing the torque of an induction motor (IM) is direct torque control (DTC), because of its various advantages, this technique of control is the popular strategy used in the industrial sector, the key benefits of this technique are its simplicity and its only independence from the induction motor’s stator resistance. While direct torque control offers several advantages, it also has some notable drawbacks. These include variable switching frequency and torque and flux ripples caused by the discrete nature of the control algorithm, as well as sensitivity to variations in motor parameters, such as motor resistance. Because of these drawbacks, this paper presents a new study to promote its disadvantages and to increase its performance. We propose using neuro-fuzzy techniques to estimate the stator flux’s constituent parts, this proposal effectively removes the reliance of DTC control on stator resistance. Several enhancements in motor and control behavior were authorized by this unique technique that was simulated in Matlab/Simulink, such as the reduction of torque and flux ripples and the reduction of the THD value of the current.