Renewable energy has advantages in environmental aspects such as greenhouse gas reduction and fine dust reduction, but the
stability of the power system is lowered due to the variability & uncertainty of renewable energy sources. Therefore, it is important
to predict the amount of generation of renewable energy, which can contribute to system stabilization. In order to predict the power
generation of floating photovoltaic(FPV), the generation amount of 500㎾ FPV and meteorological data were used to predict the
power generation through the Recurrent Neural Network(RNN). To perform appropriate prediction, identifying the correlation between
variables, removing multicollinearity, handling missing values properly are performed and the amount of power generation is
predicted through appropriate RNN. In addition, it is analyzed how the influence of wind affects the amount of power generation of
FPV