Forecast of power generation for grid-connected photovoltaic system based on Pawlak Attribute Importance Algorithm of Rough Sets

Rough sets can deal with imprecise, inconsistent, incomplete problems, which is due to the maturity mathematical basis, without prior knowledge and its ease of use. Currently, there are two methods of PV system power generating forecast. One is based on radiative transfer, energy conversion, DC/AC conversion and AC grid. Another is establishment of a variety of mathematical models. Considering the environmental factors, the forecasting model of grid-connected PV system based on Pawlak Attribute Importance Algorithm of Rough Sets was established, which is based on the 5.6kW grid-connected PV system in Beijing Institute of Architectural Engineering. Compared the forecasting result with acture operation results of the weekly, monthly, seasonal and annual generation, it shows that former is more correct and accuracy. So the research proved that the forecasting model of rough set is practical.