Load forecasting analysis for regional and industry power systems-based on ARIMA model and LSTM model

The forecasting and analysis for the power system load profoundly affect security of electricity production daily life. Based on data 15- minute interval in a regional grid China, this paper forecasts 15 - next 10 days, maximum values 3 months, minimum months each industry region basis ARIMA model LSTM respectively, then compares effects two models. conclusions show that (1) has higher accuracy than general. (2) Daily levels are large commerce lower non-general general industry. (3) sudden changes occur mainly holidays days with rainy weather. And magnitude change industries is largest. Finally, innovative suggestions made different context “Double Carbon Plan”.