Direct Torque Control Based on Artificial Neural Network of a Five-Phase PMSM Drive

Abstract

Direct torque control (DTC) based on artificial neural network (ANN) of a five-phase permanent magnet synchronous motor drive (PMSM) is presented in this paper. Using the mathematical model of the five-phase motor, DTC control strategy is developed, and the corresponding controllers are properly designed in order to provide independent torque and flux control. In order to improve the performance of the DTC, a neural network based DTC scheme is adopted instead of the DTC based on the look-up table. The employed neural network uses the Levenberg-Marquardt back propagation algorithm for the adjustment of weights to increase the learning process accuracy. The efficacy of the proposed method is verified by simulation for various dynamic operating conditions, and the system's performance is compared with conventional DTC.