Solar FaultNet: Advanced Fault Detection and Classification in Solar PV Systems Using SwinProba‐GeNet and BaBa Optimizer Models

ABSTRACT Detection of faults in solar photovoltaic (PV) systems is an important concern to guarantee that renewable energy source generation keeps its efficiency and reliability. It directly affects the development of sustainable development goals (SDGs), especially those on affordable and clean energy and climate action. However, most of the existing research studies related to fault detection methods have suffered from low accuracy, higher error rates, and time‐consuming processes when dealing with complex fault types in real‐time. The work is therefore targeted at trying to address these challenges in the development of an efficient and reliable model that could be applicable in the fault detection for solar PV systems. It also proposes the Solar FaultNet‐a novel deep learning‐based approach that significantly improves fault detection performance in solar PV systems and integrates the model with state‐of‐the‐art ML techniques like CNN and LSTM to capture inherent complex patterns and interdependencies of fault data. Besides, the proposed model outperforms conventional machine‐learning algorithms and state‐of‐the‐art deep‐ learning models for better performance by yielding higher accuracy, precision, recall, F1‐score, and low error rate on various fault types such as PV array faults, inverter faults, grid synchronization faults, and environmental faults. Experimental results show that Solar FaultNet outperforms the existing state‐of‐the‐art machine‐learning algorithms and deep‐learning architectures, achieving a precision of 99.1%, recall of 99%, F1‐score of 98.9%, and an error rate as low as 0.018%. In addition, the model enjoys high efficiency at 15 s training time and 25 s testing time in handling various types of faults such as PV array faults, inverter faults, grid synchronization faults, and environmental faults.