Short-Term Prediction of Wind Power Considering the Fusion of Multiple Spatial and Temporal Correlation Features

As the wind power penetration increases, short-term prediction accuracy of is great importance for safe and cost-effective operation grid in which integrated. Traditional farm uses numerical weather (NWP) information as an important input but does not consider correlation characteristics NWP from different farms. In this study, a convolutional neural network–long memory based model clusters proposed. Additionally, feature map established multiposition information, spatial farms fully explored, trained using spatiotemporal to obtain results clusters. Finally, case operational data cluster China are analyzed, proposed outperforms traditional methods terms accuracy.