Multi-model prediction and simulation of residential building energy in urban areas of Chongqing, South West China

Abstract

Energy simulation and prediction plays a vital role in energy policy and decision making. This study has been conducted to predict the future energy demand in the urban residential buildings of Chongqing a city in south west China. The comparative study adopts and compares the results of different demand models to improve estimation efficiency for future projections. A structured questionnaire survey was undertaken to collect primary household energy consumption data for inclusion in the annual energy consumption simulation model. An ANN model, two Grey models, a Regression model, a Polynomial model and a Polynomial regression model were used to forecast and compare demand. The precision of the models have been used statistical methods. The predicted results show that the total residential building energy and electricity consumption in urban areas of Chongqing is increasing rapidly. Based on MRPE (%) and the statistical tests, the study concluded that an ANN model is the most acceptable forecasting method of the six models. Hence, based on ANN model, urban residential building energy consumption will be at 1005 × 10 4 SCE and electricity consumption will be at 264.81 × 10 8 kW h in 2025 which is about three times and four times higher than that of the 2012, respectively. © 2014 Elsevier B.V.