Multi-class neuro-fuzzy classifier for photovoltaic array faults diagnosis

This paper proposes an adaptive multiclass neurofuzzy classifier (MC-NFC) for fault detection and classification in solar photovoltaic (PV) systems. The designed fuzzy classifier was optimized by seeking for the best numerical values of the parameters that tune its membership functions. The experiments have been conducted on the basis of collected data from a real time PV array emulator (namely array current and voltage, irradiance and module temperature) to classify three kinds of faults in the PV array. Results show the advantages of using the MC-NFC approach, and suggest that further improvements in terms of classification accuracy can be realized by the proposed neuro-fuzzy classification algorithm, and other type of faults can be considered for classification.