Short-Term Load Forecasting of Power System Based on VMD-CNN-BIGRU

Abstract

In order to improve the accuracy of load prediction, taking into account the internal laws and external influencing factors of historical load itself, a kind of variational modal decomposition (VMD) -convolutional neural networks (CNN) -bi-directional gated recurrent units are proposed. BIGRU) short-term load prediction method for hybrid networks, improving training duration and prediction results. The effectiveness of the proposed method is verified by simulation analysis, and the method has higher load prediction accuracy and stronger robustness than other models, which can improve the accuracy of short-term load prediction of power system.