A New direct power control method of the DFIG-DRWT system using neural PI controllers and four-level neural modified SVM technique

Abstract

Recently, direct power control (DPC) based proportional-integral (PI) controllers have become dominant for doubly fed induction generators (DFIGs) due to its simplicity and good dynamic response. But this strategy gives a high total harmonic distortion (THD) of the voltage and a large power ripple. In this paper, a new DPC method (called as NDPC-4L-NSVPWM) is proposed based on a four-level neural space vector pulse width modulation (4L-NSVPWM) and neural PI controllers for direct active and reactive powers command (DRAPC) of a DFIG integrated in a dual-rotor wind turbine (DRWT). This control scheme is based on direct reactive and active powers estimation. Theoretical principles of this strategy are presented along with simulation results. The advantages and robustness of NDPC-4L-NSVPWM control over other DPC with PI controllers (DPC-PI) are presented. The main advantages of NDPC-4L-NSVPWM strategy are relative to its good dynamic response, simplicity of the algorithm and operation at a constant switching frequency. Analysis of NDPC-4L-NSVPWM strategy based DFIG-DRWT has been done in MATLAB/Simulink software. Simulation results validate the performance of the NDPC-4L-NSVPWM strategy by evaluating the THD of current and the reduction of the ripples for the active/reactive power of DFIG-DRWTs.