application of a new grey model to forecast the relation of supply and demand of natural gas in turkey


Eren M., Gültekin S.

CIEP 2017, Isparta, Turkey, 5 - 07 October 2017, pp.147-148

  • Publication Type: Conference Paper / Summary Text
  • City: Isparta
  • Country: Turkey
  • Page Numbers: pp.147-148
  • Ataturk University Affiliated: Yes

Abstract

APPLICATION OF A NEW GREY MODEL TO FORECAST THE RELATION OF SUPPLY AND DEMAND OF NATURAL GAS IN TURKEY Miraç Eren* Sena Gültekin ** ABSTRACT Consequently, in this study, Turkey’s supply and demand’s prediction and insights tackle to ensure Turkey’s supply and demand equilibrium, and to encourage economy’s sustainable growth by authority, and to implement reasonably energy consumption plans and policy precautions. Keywords: Grey prediction model, parameter estimation, supply and demand conditions of natural gas in Turkey. INTRODUCTION AND RESEARCH QUESTION Natural gas is one of the productive, clean and cheap fossil fuels. However, at the present time, natural gas becomes more and more popular because it burns cleaner than other fossil fuel types and it has high heating power. Therefore, in recent years, lots of countries prefer natural gas instead of other fuels such as especially coal and gasoline. It is beneficial to estimate scientifically and effectively scale of production and consumption of natural gas, because natural gas’s both supply and demand sides has to accord international “take or pay” rule. CONCEPTUAL FRAMEWORK According to previous research results, to predict natural gas consumption, lots of models were developed and these models can be classified two main branches as intelligent and statistical models. Statistical models are frequently simple like logisticsbased models[1,2], the Bayesian Model Averaging [3], polynomial model [4] or even linear logarithmic function which identifies the relation between energy consumption and relevant factors. As regards intelligent models, researchers generally choose present regression models such as artificial neural networks-based models and support vector regression. METHODOLOGY To estimate Turkey’s natural gas consumption, according to a new polynomial grey prediction model named as TDPGM (1, 1), Turkey’s natural gas output and consumption is envisaged by the year 2025. FINDINGS AND DISCUSSION In this study, findings show that suggested TPGM (1,1) model has the best simulation and prediction performance. * Assist. Prof. Dr., Ondokuz Mayıs University, Faculty of Economic and Administrative Sciences, Department of Economics, mirac.eren@omu.edu.tr ** Master Candidate, Ataturk University, Faculty of Economic and Administrative Sciences, Department of Economics, sena.gultekin@atauni.edu.tr 2nd Congress on International Economic and Administrative Perspectives: Sustainable Global Competition, 5-7 October 2017, Isparta 148 RESULTS AND RECOMMENDATIONS According to results, in future, because of the increase of demand on the natural gas, Turkey will be contingent upon importation. REFERENCES: M. Forouzanfar, A. Doustmohammadi, M.B. Menhaj, S. Hasanzadeh, Modelingandestimation of thenaturalgasconsumptionforresidentialandcommercialsectors in Iran, Appl. Energy 87 (1) (2010) 268–274. F. Shaikh, Q. Ji, Forecastingnaturalgasdemand in China: Logisticmodellinganalysis, Int. J. Electr. PowerEnergySyst. 77 (2016) 25–32. W. Zhang, J. Yang, Forecastingnaturalgasconsumption in ChinabyBayesian Model Averaging, EnergyRep. 1 (2015) 216–220. G. Xu, W. Wang, ForecastingChina’snaturalgasconsumptionbased on a combination model, J. Natur. GasChem. 19 (5) (2010) 493–496. V. Bianco, F. Scarpa, L.A. Tagliafico, Scenarioanalysis of nonresidentialnaturalgasconsumption in Italy, Appl. Energy 113 (6) (2014) 392– 403. A. Azadeh, S.M. Asadzadeh, M. Saberi, V. Nadimi, A. Tajvidi, M. Sheikalishahi, A Neuro-fuzzy-stochasticFrontieranalysisapproachforlong-termnaturalgas consumptionforecastingandbehavioranalysis: Thecases of Bahrain, SaudiArabia, Syria, and UAE, Appl. Energy 88 (11) (2011) 3850–3859. O. Kaynar, I. Yilmaz, F. Demirkoparan, Forecasting of naturalgasconsumptionwithneural network andneurofuzzysystem, EnergyEduc. Sci. Technol.26 (2) (2011) 221–238. Y. Bai, C. Li, Daily naturalgasconsumptionforecastingbased on a structurecalibratedsupportvectorregressionapproach, EnergyBuild. 127 (2016)571–579.