Solar Load Forecasting Based on Spatial Modeling and Artificial Neural Networks

Solar energy has the characteristics of strong randomness and volatility, which is a challenge for the safe operation of the power grid. It is necessary to forecast the solar load and plan the grid in advance. In this study, a new prediction method based on spatial modeling and artificial neural networks is proposed. This method does not need a huge data sample base, only based on the weather information data set in the four grid points of the photovoltaic power generation system, and can be accurately predicted. Compared with the traditional model, this model has the ability of self-learning and can be iterated continuously to reduce the error. The verification results show that the model can output high accuracy prediction data based on limited data.