Prediction and comparison of urban electricity consumption based on grey system theory: A case study of 30 southern China cities

Abstract

Rapid urbanization has put heavy pressure on urban energy supply over the past decades in China and such a trend is expected to continue. Therefore, it is critical to capture the impacts of key factors on urban energy consumption and further predict its future growth. This study proposes a framework to predict urban-scale electricity consumption which incorporates an original urban information database with 18 macroeconomic factors and 2 electricity consumption indicators. The grey prediction model is the core for prediction, whose performance is investigated in the established database. Meanwhile, multiple strategies, e.g., variable screening and optimal data allocation, have been conducted to improve the accuracy of the prediction. A case study of 30 cities in southern China indicates that the established database can well represent the relation between urban development and energy consumption. The proposed prediction framework performs well in forecasting urban-scale electricity consumption with less than 10 % of mean absolute percentage error, showing a positive correlation between model accuracy and data amount. Overall, the proposed prediction framework generates valuable insights into the interaction between urban development and electricity consumption, which can quantitatively inform urban designers when planning urban energy systems.