Genelleştirilmiş regresyon modelleri kullanılarak enerji talebi ve tüketimi için bir modelleme yaklaşımı

Abstract

Energy management is an important process for maintaining available energy resources and meeting basic household energy needs. Many studies seek to optimize the household energy consumption patterns to manage the load demand and minimize energy costs. Adopting such optimizations in conflict-affected countries is more beneficial due to limited energy sources. This study identifies the optimal energy consumption model for households in northern Syria. The objective is to identify the most cost-efficient energy sources while considering the prices, average monthly household income, the main source of electricity, battery storage capacity, and energy demands for space heating, water heating and cooking. One hundred and thirty-six (136) standardized surveys of residential households are collected and used as a test case. Statistical analysis of the data was carried out using the R-Studio software, where Poisson regression and negative binomial regression were employed. Findings revealed that the Negative Binomial (NB) model used has high explanatory power. In addition, the energy sources used for space heating and water heating have a direct impact on monthly expenditures. The produced model showed that the most cost-effective energy sources are coal for space heating and natural gas and kerosene for water heating.