A fundamental study of spectrum center estimation of solar spectral irradiation by statistical pattern recognition

Against a background of environmental problems and energy issues, it is expected that PV systems will be introduced rapidly and connected with power grids on a large scale in the future. For this reason, concerns about how PV power generation will affect supply and demand adjustment in electric power in the future arise, and the technique of correctly grasping PV power generation becomes increasingly important. PV power generation depends on solar irradiance, module temperature, and solar spectral irradiance. Solar spectral irradiance is the distribution of the light intensity for every wavelength. As the spectral sensitivity of solar cells depends on the kind of solar cell, it becomes important for an exact grasp of PV power generation. But the preparation of solar spectral irradiance data is not easy because observational instruments for solar spectral irradiance are expensive. Against this background, in this paper, we propose a new method based on statistical pattern recognition for estimating the spectrum center, which is a representative index of solar spectral irradiance. Some numerical examples obtained by the proposed method are also presented.