Grey Forecasting Models Optimized by Firefly Algorithm for Natural Gas Consumption Prediction in Turkey

Natural Gas is assumed to have a vital role as an energy source in all countries, serving as the primary fuel for industries, homes, and various sectors. The forecasting of natural gas consumption is very crucial for efficient energy management and the formulation of appropriate policies related to its production and usage, accurately. Forecasting has significant economic implications as it enables the implementation of cost-effective strategies based on reliable predictions of natural gas usage. This research focuses on predicting the consumption of natural gas in Turkey by employing Grey Forecasting Models (GF) Optimized by the Firefly Algorithm. The Firefly Algorithm optimizes the model parameters, while the GF models, namely GM (1,1) and NGBM (1,1), estimate the natural gas consumption in Turkey. The performance of these grey forecasting models is evaluated by comparing them with ARIMA and linear regression models. The calculations illustrate that the proposed NGBM (1,1) model, based on the Firefly Algorithm, surpasses other grey models such as OGM(1,1), GM(1,1) as well as statistical methods like ARIMA and linear regression, in terms of prediction accuracy.