Low-cost and accurate computational system for efficiency measures over photovoltaic arrays

Abstract

Great efforts have been shown in the literature to improve the efficiency of photovoltaic (PV) systems. In this workpaper, authors present the development of a computational system based on free software that allows to estimate with an accuracy superior to 99.9 percent the performance of the PV array connected to a current inverter block, from the physical models behind. For this purpose, a solar radiation measures was made in the laboratory, and also voltage, current and temperature tests over the whole system, adding a storage interface and serial data transmission port to drive and process the data from a computer. Once the system was tested and calibrated in laboratory, the I-V experimental curves of the PV system were obtained in real time, in order to define the most important parameters: short-circuit current, open-circuit voltage, maximum current and voltage, in a natural environment. Those parameters were the input data to the computational tool based on Scilab, which allow estimate the performance of PV system described, solving the non-linear equations of the physical models used. The performance results were compared with a similar computational tool developed in Matlab, as well as with a commercial solar simulator used by the manufacturer of PV module. A highly accurate computational system based in Scilab was obtained allows to evaluate the efficiency and the filling factor of an array for different temperature and solar irradiance levels in real time.