Özet:

Energy demand forecasts are needed for developing and emerging countries to be able to determine sustainable energy policies. Grey prediction models can make successful predictions with limited data without the need for any prior information. In this study, Grey prediction models are considered for the total energy demand of Turkey that shows significant economic and social development in recent years. Grey prediction includes forecasting models based on time series and cause-effect relationships. Four different grey prediction models including GM (1,1) and Grey Verhulst from the time series models, and GM (0,N) and GM (1,N) based on the cause-effect relationships were discussed. The purpose of using two types of forecasting structures is to achieve reliable and strong forecasts by capturing the recent trend with the time series and obtaining the change in energy demand by cause-effect relationship. GM (1,1) and Grey Verhulst model has been established with past total energy consumption data. By using independent variables of GDP, population, import, export and building surface area for GM (0, N) and GM (1, N) models, GM (0,6) and GM (1,6) models were formed. All applied models have been compared according to performance criteria, GM (1,1) and GM (1,6) have been designated to be superior models that performed successful predictions. As a result; Turkey’s total energy demand has been forecasted up to 2025 with GM (1,1) based on time series and GM (1,6) including high and low scenarios.

Anahtar Kelimeler: