Real Time PV fault detection in solar PV using Ensemble learning algorithm

This This research proposes a technique to detect and classify defects in large-scale photovoltaic (PV) systems, Adaptive boosting ensemble learning algorithms to achieve higher levels of precision. Monitoring and maintenance should be performed on a regular basis to maintain the system efficiency. It may be tedious to manually inspect or monitor a large-scale PV system, which may result in a loss of solar PV. The PV system available at Thiagarajar College of Engineering(TCE) in Madurai, was utilized for extraction of specifications such as voltage and current. The voltage and current readings were analysed, to classify the fault in the solar panels. The proposed ensemble algorithm was trained and validated using the real-time dataset. The algorithm achieves an accuracy of 99.8%for validation. The trained classification model was deployed in the real time environment using MATLAB/SIMULINK.