![Sample our Built Environment journals, sign in here to start your access, Latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5926/TOC-Subject-Sample-2021-Built-Environment.jpg"})
ABSTRACT
The world’s highest energy consumer (HC) countries currently constitute around 62% of the world energy consumption. Therefore, it is highly important to model their energy consumption to obtain an estimated profile of future world energy consumption. In this study the HCs’ energy consumptions are modeled using artificial neural networks (ANNs). The models are developed based on economic and demographic variables, which are gross domestic product, population, import, and export of the countries selected. Performance of the derived models is assessed using mean absolute percentage error (MAPE), mean absolute error (MAE) and root mean square error (RMSE) for the testing data. The contribution rate of each variable to the HCs’ energy consumption are also determined to demonstrate the governing variables on the energy consumption. The results show that the correlation coefficients between the ANN predictions and actual energy consumptions are higher than 90%. This indicates a high reliability of the models for forecasting future energy consumption of the HC. Additionally, MAPE, MAE and RMSE values indicate that the ANN models can give adequate forecasting for the HCs’ energy consumption. Furthermore, contribution rates of input variables on energy consumption also indicate that energy consumption of each country studied is governed by different variables. It is expected that this study will be helpful for developing highly applicable energy policies for the HC countries. Furthermore, the results of this study can also be used for determining future trends in the global energy demand.
Funding
The authors would like to thank the Scientific and Technological Research Council of Turkey (TUBITAK) for the financial support supplied to Prof. Dr. Erkan Topal for his sabbatical visit.