Short-Term Industrial Load Forecasting Based on VMD-KELM-Xgboost

To enhance the precision of short-term industrial load prediction, this study presents a method for short-term industrial load forecasting grounded in VMD-KELM-XGBoost. Firstly, the original industrial load signal is decomposed by variational mode decomposition (VMD) to obtain the modal components of each frequency. Secondly, the kernel extreme learning machine (KELM) is used to perform nonlinear mapping on each mode to extract high-dimensional features. Finally, the gradient boosting algorithm (XGBoost) is used to predict the short-term characteristics of different modes, and the predicted values of each modal component are added to realize the short-term prediction of industrial load.