Turkey's long-term electricity consumption forecast using regression analysis

Abstract

Demand forecasting is essential primarily for planning. Although it is crucial in many sectors and issues, it has particular importance for electricity. Therefore, the issue of electricity consumption forecasting has recently become a prevalent topic. In light of the above, this study aimed to develop an appropriate model to estimate the long-term electricity consumption of Turkey. The study consists of three steps. In the first step, eight models were developed to separately investigate the effects of eight input variables frequently used in electricity consumption forecasting studies in the literature. In the second step of the study, two models consisting of input variables with high impact in the first step were developed, and the trained performances of the developed models were calculated by using the regression analysis. In the final step, the combined effect of eight variables on electricity consumption forecasting was investigated using regression analysis. It can be conclude that the model in the third step showed significant results, and the model performance was good. Finally, Turkey's electricity consumption forecast for the years 2020–2030 was performed using the model in the third step.