Published Paper Details:

This project presents a smart and efficient system for detecting faults in solar panels using deep learning techniques. Solar energy is a clean and sustainable power source, but faults in photovoltaic (PV) panels--such as short circuits, shading, or panel degradation--can reduce energy output and efficiency. The proposed system uses artificial neural networks (ANNs) trained on historical voltage, current, and temperature data, as well as thermal images, to automatically identify different types of faults in real-time. The model was tested using a dataset collected under varying environmental conditions and achieved high accuracy in classifying normal and faulty panel behavior. By enabling early fault detection, this system helps reduce maintenance time, minimize power losses, and improve the overall performance of solar installations. This work supports the move toward smarter, more reliable, and sustainable energy solutions.