“…The combined model can improve the model performance based on the advantages of the sub-model (Wang et al, 1210;Zhou et al, 2019;Wang et al, 2020a;Wang et al, 2020b;Qiao et al, 2021;Zhang et al, 2021). This method can decompose the forecasting error of the forecasting model into multiple modal components by using the EMD method (Yu et al, 2008;Xu et al, 2019;Xia and Wang, 2020), build the ARIMA model (Conejo et al, 2005;Karabiber and Xydis, 2019) for each modal component for forecasting, and add up the forecasted values of all the modal components to compensate error for the original forecasting model.…”