Neural network-based MPPT coupled with fuzzy logic control for DFIG wind energy conversion

Abstract Doubly Fed Induction Generators (DFIGs) are commonly integrated into variable-speed wind energy systems due to their enhanced controllability, high efficiency, and cost-effectiveness. This research introduces a neural network-based control approach to maximize wind energy capture and efficiently regulate DFIG speed by adapting in real time to varying wind conditions. The first stage introduces an intelligent MPPT controller leveraging neural networks to ensure optimal DFIG operation under fluctuating wind speeds. To control active and reactive power, the second part of the system employs Fuzzy Logic Control (FLC) and a PI-based Field-Oriented Control method. The effectiveness of both methods is evaluated through MATLAB/Simulink simulations on a 1.5 MW grid-connected DFIG. As a result, the proposed FLC achieves optimal aerodynamic efficiency and precise speed regulation via ANN-based MPPT, while offering faster response, reduced power overshoot, improved stability, and diminished oscillations compared to FOC-PI. Moreover, the FLC notably decreases stator current THD from 0.91% to 0.38%, surpassing IEEE-519 requirements and highlighting its superior ability to suppress harmonics.