Fault Detection and Classification in PY Systems using MRA DWT and AdaBoost classifier

The increasing integration of photovoltaic (PV) systems into the global energy grid emphasizes the need for robust fault detection and classification methodologies to ensure optimal performance and reliability. This research presents a novel approach for fault identification in PV arrays by combining the Multiresolution Analysis (MRA), Discrete Wavelet Transform (DWT), and AdaBoost classifier. The proposed methodology involves capturing the unique frequency and time-domain characteristics associated with various fault types within PV arrays. MRA is employed to extract relevant features from the acquired data, followed by DWT to enhance the discriminative power of the features. The processed data is then fed into an AdaBoost classifier, leveraging its ensemble learning capabilities to improve fault classification accuracy. Simulation and Experimental validation is conducted using real-world 3X3 PV array data under different fault scenarios, including shading, module failures, and electrical anomalies. The results demonstrate the effectiveness of the proposed approach in accurately detecting and classifying faults, even in the presence of noise and changing environmental conditions. Comparative analyses with existing fault detection methods highlight the superiority of the proposed MRA-DWT-AdaBoost framework in terms of accuracy, robustness, and computational efficiency.