There have been dramatical increases in the demand of natural gas due to its superior consumption benefits and as a result of the industrialzation, urbanization and rapid population increase in Turkey. However, Turkey is not a natural gas country, nor is dependent on abroad in terms of providing its energy. That's why, it is of great significance that Turkey determine the amount of demand of natural gas. The more agreements on import of natural gas, the incrastructure investments across the country and the planning of the consumption. ln this study, the prediction for the natural gas demand of Turkey through ANFlS (Adaptive Neuro-FuzzyInference System), which is "Adaptif Ağ Tabanlı Bulanık Çıkarım Sistemi” or “Uyarlamalı Sinirsel Bulanık Denetim Sistemi". Though there have been many factors affecting the consumption of natural gas, only Gross Natural Product (GNP), temperatue and population has been taken into consideration as three input parameters. As the output paramater, the consumption of natural gas has been examined. The data set has been arranged monthly between the years of 2002-2014. The data of natural gas consumption predicted by ANFlS and the real natural gas consumption data have been compared and MAPE (Mean Absolute Percentage Error), used for the measurmeent of the performance of the model, has been calculated. ln the result calculated, the MAPE was low and a successful prediction was obtained through adaptive model. Finally, the demand for natural gas for 2015 was predicted.
Primary Language | Turkish |
---|---|
Journal Section | Research Articles |
Authors | |
Publication Date | June 25, 2020 |
Acceptance Date | June 26, 2020 |
Published in Issue | Year 2020 Issue: 3 |