Özet:

Natural gas demand forecasting is of great importance especially for decision-makers of national economies with high energy consumption, the industrial sector, and all players in the natural gas market, in particular. This study aims to present Turkey's monthly natural gas demand forecast model according to meteorological parameters. In the study, models were created with four current metaheuristic algorithms of Artificial Bee Colony Algorithm (ABC), Charged System Search Algorithm (CSS), Crow Search Algorithm (CSA), and Harmony Search Algorithm (HSA) were compared. In the research, three mathematical models, linear, exponential, and quadratic, were developed and the performances of the models were evaluated with six different global error measurement metrics (AE, MAE, R2, MAPE, RMS, MARNE). In the study, average temperature, pressure, humidity, wind, and precipitation meteorological data were used as input parameters. The data between 2010-2017 was applied as training data, and the data between 2018-2020 was applied as test data. The most successful forecasting model for the natural gas demand forecast training data set is the quadratic model of the CSS algorithm, while the most successful forecasting model for the test data is the quadratic model of the ABC algorithm.

Anahtar Kelimeler: