Affected by solar radiation intensity

geographical latitude

climates and other factors

the output power of the photovoltaic power generation systems has large fluctuations

while the current prediction models for the photovoltaic power generation have low accuracy. In order to solve this problem

this paper proposes a prediction method for photovoltaic power generation

on the basis of the Krige model. In this method

the surface radiation model is established firstly

and the relationship between the radiation intensity and the photovoltaic power generation is investigated to ascertain the interpolation methods and constraints. Secondly

the weight coefficients are calculated according to the original sample. Finally

this Krige forecast model is applied to predict the next few days' photovoltaic power capacity. The experimental results show that

compared with the BP neural network and grey theory model

the accuracy of the model is improved

which provides the scientific research value for the development of photovoltaic power prediction method.