Estimation of diagnosis indicators for monitoring photovoltaic panels using artificial neural networks

Solar energy is widely recognized as one of the primary renewable energy sources. However, the efficiency and reliability of Photovoltaic (PV) systems can be significantly impacted by faults. For these reasons, it is paramount to Continuously monitor the PV health state to ensure optimal performance. In this context, this paper introduces a robust estimation model using an Artificial Neural Network (ANN) model to accurately predict three diagnosis indicators: power (P), current (I), and voltage (V). These indicators play a vital role in monitoring the behavior of PV systems considering different weather conditions. The computational algorithm establishes the mapping from PV electrical coordinates and temperature to the diagnosis indicators, without relying on an irradiation sensor. The performance of the proposed estimation model is evaluated via MATLAB/Simulink®, based on the real meteorological profiles for a typical year in Anglet, France.