Advanced Photovoltaic Panel Fault Diagnosis and Maximum Power Enhancement through Neural Network Modelling

The objective of this study is to detect power loss issues in photovoltaic panels resulting from fluctuations in temperature and solar irradiance, using artificial neural networks (ANN). The proposed approach involves creating a baseline model that reflects the panel’s normal operating behaviour under fault-free conditions. By comparing this reference with the actual power output, the system computes a residual signal, which is then analysed to identify potential anomalies. The ANN is trained on real-world input data, including temperature and irradiance values, to predict power output accurately. The results demonstrate that this intelligent diagnostic method can detect faults with high precision, even in the presence of non-linear system responses. By leveraging real measurement data, the method enables early fault detection without relying on complex physical modelling, thus streamlining maintenance and improving the reliability of PV system monitoring.