Adaptive Neuro-Fuzzy Inference System Application for The Identification of a Photovoltaic System and The Forecasting of Its Maximum Power Point

Maximum Power Point Tracking (MPPT) techniques play an extremely important role in standalone photovoltaic (PV) systems or connected to distribution grids. PV, delivering power dependent weather conditions (temperature and insolation), need a mechanism to deliver permanently the optimal power available. Thus, several techniques have been developed in the literature. A new approach based on Artificial Intelligence (AI) is involved in this paper for the research of the Maximum Power Point (MPP) of a PV. It is based on the Adaptive Neuro-Fuzzy Inference System (ANFIS). This adaptive MPPT controller consists of two inputs (voltage and current) and an output (optimal power of the PV). The proposed ANFIS command is composed of five layers with nine fuzzy rules.The implementation of ANFIS under the Matlab / Simulink environment shows that this technique is more efficient, more powerful and even more robust than the InC command. This is due to the low Response time (Rt) of 2.4 ms and excellent performance of 99.94% for the proposed method and contrary to the classical control which has a Rt of 401 ms and a yield of 94.41%. Using the proposed technique, the efficiency of the solar cells is improved by 5.53%.