Adaptive Neuro-Fuzzy Reinforcement Learning Control for Grid Stability and LVRT Enhancement in PMSG-Based WECS

In this work, an Adaptive Neuro-Fuzzy Reinforcement Learning (ANF-RL) control framework is developed to improve the grid-integration performance of Permanent Magnet Synchronous Generator (PMSG)-based Wind Energy Conversion Systems (WECS). The proposed controller integrates neural-network-based nonlinear approximation, fuzzy reasoning for uncertainty handling, and reinforcement-learning-based adaptive optimization. It is coordinated with a Dynamic Voltage Restorer (DVR) and a   31-level Cascaded H-Bridge Multilevel Inverter (CHBMLI) to address voltage sags, DC-link instability, torque ripple, harmonic distortion, and Low Voltage Ride Through (LVRT) requirements. The system is formulated for MATLAB/Simulink evaluation under variable wind profiles and grid disturbances, with performance assessed using LVRT capability, Total Harmonic Distortion (THD), DC-link voltage regulation, torque ripple, grid synchronization, voltage recovery, and settling time. The study also establishes a system-level baseline for a broader postdoctoral programme that subsequently extends intelligent LVRT control toward physics-informed reference-current generation and converter-level control in DFIG-based WECS.