Direct torque control of three-level inverter using neural networks as switching vector selector

In this paper, based on the detailed study on the characteristic of direct torque control (DTC) and output vector of the three-level inverter, a simplified method to choose the output vector for DTC of the three-level inverter fed induction motor is proposed. A novel switching vector selector using the ANN (artificial neural network) is trained under the tutor of the method mentioned above. By the usage of the ANN, when the error of the torque and stator flux is made certain, the output vector can be expediently acquired. The validity of the proposed vector selector is confirmed by the simulative results.