Short-term Load Prediction of Integrated Energy System Based on Random Matrix Coupling Analysis

With the increasing coupling degree among heterogeneous loads in IESs, a short-term load prediction method for IES based on random matrix coupling analysis was proposed in order to improve the accuracy of load prediction in application scenarios such as demand response. First, combined with engineering background knowledge, the random matrix theory was applied to analyze the coupling features of different energy systems and the coupling matrix was obtained. On this basis, the coupling matrix, load data, and external measurements were taken as inputs, a multitask learning load prediction model based on gate recurrent unit was established, and the coupling features between loads were fully mined and utilized through the shared layer. Finally, the feasibility and accuracy of the proposed method were verified by a real example of IES.