Forecasting net energy demand of Turkey using artificial neural networks


Thesis Type: Postgraduate

Institution Of The Thesis: Gazi Üniversitesi, Fen Bilimleri Enstitüsü, Turkey

Approval Date: 2013

Student: HÜSEYİN AVNİ ES

Supervisor: FATMA YEŞİM KALENDER ÖKSÜZ

Abstract:

Energy, having vital importance in countries' development policies, is a strategic issue. Rapidly increment of energy demand with population, industry and developing technology increases more importance of energy policies. Reliable and accurate predictions are needed in order to determine energy policy for the future. Within this framework, energy demand forecast based on artificial neural networks (ANN) in this thesis was carried out with the cause-and-effect relationships. By doing research on ANN which are widely used in the field of energy, prediction models were investigated. ANN model has been established with GDP, population, import, export, building surface area and the number of vehicles for Turkey net energy demand forecast between 1970-2010 years. Regression and time series models were developed in order to evaluate success and predict performance of ANN model and comparisons are made. As a result of the analyzes and comparisons, ANN performs seen more successful predictions and Turkey net energy demand forecast between 2011-2025 years were estimated.