Faults diagnosis in Stand-Alone photovoltaic system using Artificial Neural Network

Diagnosis process has an important role to improve the efficiency, reliability and safety in Stand-Alone Photovoltaic system. This Paper proposes new faults diagnosis method in Stand-Alone Photovoltaic System based on feedforward Artificial Neural Network. The voltage and current of photovoltaic modules, Battery and Load are used as input for developing and testing the Artificial Neural Network. This approach has been validated using an experimental measurements issued from Standalone photovoltaic system installed at LREA in the University of Médéa, Algeria. The same system is also simulated in Simulink/MATLAB environment which take the irradiation and temperature as input captured from the site, the simulated model used to compare the results with the experiment. The proposed algorithm shows a high accuracy of faults identification in both experiment and simulation.