Prediction of Turkey's demand for transportation energy

Energy efficiency policies are among the issues that need to be taken into careful consideration due to their direct relevance to the sustainability of economic growth and social development objectives. Along with the growing population in the world and the growing economy of countries, the need and demand for energy is rapidly increasing. In addition, the increase in the gross domestic product (GDP) of the countries and the rising trend in living standards of the people increase the demand for in the transportation sector. The fact that transportation sector has a share of about 19% in terms of energy consumption in Turkey increases the significance of this situation. Future energy demand is of great importance in order to design, plan and use transportation systems more efficiently. However, reliable and accurate predictions are needed to be able to shape energy policies' future. In this study, modelling has been conducted using artificial neural networks to estimate the energy demand in transportation sector. In this study, in order to estimate the energy demand in Turkey's transport sector, GDP rates between the years 1975 and 2016, oil prices, population, vehicle-km and tonne-km and passenger-km data has been examined. Seven different models were set up for different scenarios in the study. As a result of the study, it has been determined that the first model has the lowest error value and the highest R2 value among the others. With this model, Turkey's demand for transport energy until the years 2020, 2023, 2025 and 2030 has been estimated. It was also compared with prediction studies previously done in the literature and the success of estimating the transport energy demand of the established model was tested. Keywords: Transportation, prediction of energy demand, Artificial neural network, Oil prices