Implementation of Robust Fractional-Order Neural Modified Sliding Mode Controls for Managing the Power Output of Doubly Fed Induction Generators

Abstract

This paper presents a robust fractional-order neural-modified sliding mode control (FONMSMC) strategy for enhancing the performance of dual-rotor wind turbines (DRWTs) driven by doubly-fed induction generators (DFIGs). Conventional direct power control methods often exhibit unstable power output, high electrical noise, and limited adaptability under variable wind conditions. The proposed FONMSMC integrates fractional calculus for precise dynamic tuning, neural networks for adaptive adjustment, and a modified sliding mode control framework to improve robustness, combined with pulse-width modulation for efficient power conversion. Simulation results demonstrate significant performance improvements, including reductions of up to 94% in active power steady-state error, 87.2% in active power fluctuations, and 71.79% in total harmonic distortion. The controller also maintains stable operation across a wide range of wind speeds, ensuring enhanced grid stability and reduced mechanical stress. The proposed method offers a reliable and efficient control solution for DFIG-based DRWT systems, contributing to improved sustainability and robustness in modern wind energy applications.