OPTIMIZED ENSEMBLE LEARNING FOR FAULT DETECTION IN WIND FARMâINTEGRATED POWER SYSTEMS USING GREY WOLF OPTIMIZATION

This study presents an intelligent fault detection technique for wind farm-integrated power systems, contributing to the advancement of clean energy reliability and efficiency, a key focus of clean energy integration. The proposed method integrates discrete wavelet transform (DWT) with an ensemble-based bagged decision tree (BDT) classifier, optimized using the grey wolf optimization (GWO) algorithm. DWT extracts critical time-frequency features from fault signals, enabling accurate characterization under diverse operating conditions. GWO enhances BDT performance by selecting the optimal number of decision trees, improving classification accuracy while reducing computational complexity. The results demonstrate high accuracy and robust fault detection, ensuring stable wind power integration into the grid. By minimizing power losses, enhancing grid stability, and reducing system downtime, this method strengthens the sustainability of renewable energy systems, aligning with global efforts toward a cleaner and more efficient power infrastructure.