Forseeing energy photovoltaic output determination by a statistical model using real module temperature in the north east of France

The goal of the work carried out is firstly to achieve a model of parameters influencing the performance of photovoltaic (PV) energy output under real working conditions and secondly implementing the model into a single Field Programmable Gate Array (FPGA) that should be applied to energy management for future smart building design. Only the former aspect is discussed in this paper to guide the undergoing research laboratory theme of the latter as the GREEN platform is equipped with several renewable energy technologies for modeling, managing and optimization of energy consumption. It's well known that PV nominal specifications such as power output or energy yield of recent modules are evaluated by manufacturers under Standard Test Conditions (STC) that are generally not representative of real working conditions of solar modules. In this study, we showed high correlation between the operating cell temperature (Tc) and the measured back surface temperature (Tback) of polycrystalline PV modules in real experimental conditions. Methods used in previous study referred mainly to ambient temperature. An accurate relationship is derived for power output between solar irradiance φ (W/m2) and the back surface temperature and the calculated power output is compared to measured power output data with a high correlation coefficient. This power output is introduced in the Weibull probability density function to model the energy yield which is finally compared to real conditions recorded data. Good concordance is shown between the proposed model and measured data of energy yield for the middle year (2013) and last year (2016) of the seven years range experimental data of a semi continental zone. The model is ready to be implemented on FPGA chip to be compared to real working high temperature environment conditions.