Medium-to-Long Term City-level Electricity Consumption Forecasting Based on Cointegration-Granger Tesing and Error Correction Model

Abstract

With its economic society gradually entering the new normal, China's industrial structure has undergone great changes. Traditionally, predicting electricity demand only by historical data of electric energy consumption will result in the great deviation. It exists the highly coupling relationship between energy consumption and economic growth, by which the accuracy of electricity forecasting will be improved. Based on co-integration theory, Granger testing and error correction model, a novel methodology for medium-to-long term electricity forecasting is presented in this paper. Firstly, a longterm equilibrium model associated with electricity consumption and GDP is established by Granger and co-integration testing. Then, the error correction model is used to adjust the short-term fluctuation of the variables of the proposed prediction model. The analysis results in case studies verify the effectiveness of the prediction model, and demonstrate the short-term adjustment technique can improve the prediction accuracy.