Application of Machine Learning Algorithms in Load Management of Integrated Energy Systems

This paper applied the machine learning methods in the load management of integrated energy systems (IES). By introducing the variational mode decomposition (VMD) technique and the bidirectional long-short-term memory network (BiLSTM), this research proposes a novel load forecasting model, aiming to enhance the accuracy of multi-variable load forecasting. The results show that the proposed model demonstrates a high level of precision in the forecasting of power, cooling, and heating loads. Compared with the traditional BiLSTM and VMD-BiLSTM models, it exhibits superior forecasting performance. The research in this paper offers novel insights and methods for the load management of IES.