An Improved DTC Strategy for a DFIG using an Artificial Neural Network Controller

In this article, a control method called DTC is proposed for a Doubly Fed Induction Generator for power generation from variable speed wind. The objective is to control the torque and rotor flux of the DFIG directly through the inverters. Conventional DTC control (C-DTC) has the problem of a harmonic rate for the currents generated by the DFIG, because of the variable switching frequency. To remedy this problem, an Artificial Neural Network DTC (ANN-DTC) strategy is proposed. The proposed ANN-DTC technique reduces torque and flux ripples. In this command, the torque and rotor flux are regulated by a neural network type regulator, and the switching table and the hysteresis correctors have been eliminated. In order to better illustrate the effect of the C-DTC and ANN-DTC control on the signal quality provided by the DFIG, a spectral analysis of the stator and rotor currents has been carried out. Simulation results of the proposed controllers are compared for various step changes in the torque and rotor flux. The simulation results of the proposed ANN-DTC scheme are fully verified by using Matlab/Simulink.