Utilizing Fuzzy Logic Control and Neural Networks Based on Artificial Intelligence Techniques to Improve Power Quality in Doubly Fed Induction Generator-Based Wind Turbine System

Abstract

This article presents novel artificial intelligence (AI)-based techniques for controlling wind energy conversion systems, specifically fuzzy logic control and neural networks, known as fuzzy hysteresis-direct power control (FH-DPC) and neural hysteresis-DPC (NH-DPC), respectively. The primary purpose is to overcome conventional DPC (C-DPC) limitations in doubly fed induction generator wind turbines (WT-DFIG), focusing on power quality improvement and enhanced system efficiency. The techniques aim to reduce power ripples and improve the quality of alternating current (AC) grid energy by improving current signal quality in all WTs’ operation modes with WT-DFIG and all compensation power modes. The suggested techniques are thoroughly examined using the MATLAB/Simulink environment under various wind scenarios, demonstrating a reduction in active power ripples by over 70%, a reduction in reactive power ripples of around 77% on average, a decrease in generated current total harmonic distortions (THDs) by over 70% compared to C-DPC. The performances of FH-DPC and NH-DPC are contrasted with C-DPC and other previously suggested methods, and it is concluded that the proposed control approaches perform more effectively than them regarding ripples in local reactive compensation and generated active powers as well as THD currents, with FH-DPC slightly outperforming NH-DPC. The research indicates that AI can enhance the effectiveness and quality of power generated by wind-power systems.