Forecasting residential energy demand in China

Abstract

China is undergoing rapid economic development, and experiencing increased energy consumption. An accurate prediction of residential energy demand is beneficial to both energy supply decision-making at the local level and energy policy makers at the national level. It provides the most likely trend of residential energy demand in the specified areas and how the trend may be controlled by technologies and policies. Complexity and difficulty exist regarding the forecasting of energy demand because there are too many variables and uncertainties that may have significant impact, and also because essential historical data regarding residential energy consumption is in most cases inadequate. Unlike most existing models, we have developed a multiple-level forecasting model, with a focus on the impacts of technologies. Essentially, there are four levels in this forecasting system: the household model, community model, city model, and national model. Each level of the model has its own focused variables so that other variables can be isolated to reduce the complexity and difficulty of model implementation. This paper outlines the framework of this forecasting model and details the two lowest levels: household and community level models.