Integrated wind and solar power forecasting in China

Abstract

The renewable power forecasting is very crucial for large-scale renewable energy integration to the electric grid. In this paper, a novel integrated wind and solar power forecasting is proposed. Different with previous systems, the proposed system can predict the power of wind and solar electric farms by combination of the high-resolution predictions of their generating equipments, such as wind turbines and photovoltaic panels. Therefore, the proposed system can better capture the power characteristic of renewable electric farms, and achieve the better forecasting performance. Firstly, the proposed system makes high-resolution numerical weather prediction (NWP) for single generating equipment by leveraging the real-time weather monitoring data. Secondly, it uses a combination of different statistical models to achieve the short-term and very short-term predictions of wind turbines and photovoltaic panels, and then lead to the predictions of wind and solar electric farms. A real-world case in China shows that the system can accurately predict the wind power and photovoltaic power for the next day and the next four hours. The average monthly accuracies of short-term and very short-term forecast are 92% and 94% respectively, which largely outperform the requirement for the state grid.