Application of Neural Networks in Direct Torque Control of Induction Motors by Reference Flux Vector Method

In this paper, an approach to DTC control scheme for induction machine is developed which makes DTC more applicable while it has lower torque ripple, lower flux ripple and almost fixed switching frequency rather than other DTC methods. Comparisons between simulation and practical results confirm that not any noticeable advantage of other developed DTCs is lost while it becomes more applicable. The possibility of utilizing high speed chips like FPGAs or low cost DSPs is the priority of the presented method.