Intelligent fault supervisory system applied on stand-alone photovoltaic system

Abstract

Fault supervision in stand-alone photovoltaic system is one of the most important task to increase the reliability, efficiencies and safety. This paper proposes a fault detection and identification of a stand-alone photovoltaic system based on feed forward Artificial Neural Network (ANN). The input network consist simply of current and voltage for PV, Battery and load. Two consequent ANN are developed respectively, for detection and diagnosis using a real data issued from experimental stand-alone photovoltaic system installed at LREA in the University of Médéa, Algeria. The results show a high rate of good classification for both detection and diagnosis network, which reveals the effectiveness of the proposed method.