PV Performance Assessment through Calibrated Modeling and Experimental Characterization

The theoretical methods for modeling the photovoltaic systems available in the literature describes the perfect behavior of the solar component _cellimoduleistring_ as a generator working in the environmental conditions regarding the temperature of the cell and the emitted irradiance. This last is symbolized in the named PV characteristics. This way of modeling passes over many parameters that might influence on the electrical output of the photovoltaic unit. In the recent year other methods start using the experimental measurements to characterize the losses. In the present paper an upgraded method of modeling the photovoltaic unit will be presented. As the current-voltage characteristic curve do not still remain the same during the years of operation due to aging. Moreover, sometimes the overall shape is deviated due to the occurrence of temporary defaults; this method of modeling will permit a comparison between the in-situ measurements and a previous saint reference curve of the same system, as long as a perfect matching is expected. In the case of non-matching, which refer to system failure, a diagnosis algorithm is developed based on key parameters comparison, to detect malfunctions on the system based on curve deviation analysis.