Simple and Efficient Direct Torque Control of Induction Motor Based on Artificial Neural Networks

The simplicity and the effectiveness are the main merits of the proposed control. The paper presents simulation and experimental results of implementing a simple and an effective ANN-based DTC for the IM drive. Neural networks with a simple architecture are designed and implemented in the influential points of the DTC scheme of the three-phase IM in order to improve its performances while preserving the DTC structure simplicity. In the conventional DTC the band limits of flux and torque hysteresis comparators are defined beforehand and remain fixed whatever the operating regime. In the proposed control, neural controllers can give the appropriate logic outputs for the torque and flux by adjusting online the bandwidth on the basis of their inputs and feedback outputs. Thus, reduce the ripples content of the torque, flux and the stator currents which are basically affected by the width of this band. Hence to improve motor dynamic performance under transient and steady state conditions. The proposed control performances are highlighted by comparing to the conventional DTC control. Simulation and experimental results show the feasibility, the easiness of implementation and qualitative improvement in performances of the proposed control.