Prediction method for wind - solar power supply guarantee capacity based on energy storage time - series characteristic simulation

Accurate prediction of wind-solar power generation is vital for improving the power supply guarantee capability of high proportion of wind-solar power, and the use of energy storage is the key to further guarantee the power balance. However, the existing power prediction methods have not fully considered the charging and discharging characteristics of energy storage, limiting efficient characterization of wind-solar-storage power supply capability. Therefore, this paper proposes a wind-solar power prediction method based on the simulation of energy storage timing characteristics, so that the prediction model can be adapted to the charging and discharging characteristics of energy storage during the training process, and the overall reliab le power supply capacity of wind-solar-storage can be improved. Firstly, the wind-solar-storage power prediction evaluation index based on the storage charge-discharge simulation model is proposed to quantitatively characterise the tracking ability of the wind-solar-storage power prediction results to the actual power output. On this basis, a loss function for wind-solar power prediction considering energy storage support is established to achieve highly accurate wind-solar power prediction under the condition of energy storage participation in the supply protection scenario. Finally, the effectiveness of the proposed metho d is verified based on the actual data of a wind power cluster.