Prediction Model of City Electricity Consumption

Abstract

Power load forecasting is a series of forecasting work for power load. From the point of view of forecasting objects, power load forecasting includes the prediction of future power demand and the prediction of load curve, which provides a reliable decision-making basis for power system planning and operation. This chapter focuses on the analysis and prediction of short-term electricity consumption and long-term electricity consumption data in urban areas. The short-term prediction of electricity consumption is carried out by using the time-series model, and the periodic characteristics of the long-term power consumption series are further mined by the seasonal time-series model. On this basis, the uncertainty interval prediction of electricity consumption data is realized by using heteroscedasticity model, and the distributed computing strategies of these methods under the framework of big data are provided.