A weather temperature methodology on the Italian electricity demand

Predicting Energy Demand Using Machine Learning: Exploring Temporal and Weather-Related Patterns, Variations, and Impacts

“…The primary driving force for load forecasting has been identified as temperature [6], [21], and a strong negative correlation is found to exist in winter, and a positive in summer between demand and temperature [10], [21], [22]. During the winter season, a noticeable demand for lighting and heating occurs simultaneously with decreasing temperatures.…”

“…It significantly affects energy demand by potentially altering the extent and intensity of daily and seasonal heating and cooling needs and posing considerable threats to infrastructure and communities [5]. This emphasized the need to accurately establish the correlation between load and temperature [6].…”

“…Electricity demand is influenced by several critical factors including population, households, wealth, culture, climatic circumstances, economic variables [9], [10], system operators, and market participants [6]. Given the anticipated increase in demand in the forthcoming period, it is essential to establish a correlation between prospective demand changes and the required generation and network capacity from a system planning perspective [11].…”

“…The study by MacMackin et al [13] highlighted the significant impact of weather as a vital determinant of both daily and seasonal fluctuations in electricity demand, and the absence of comprehensive data on end-use consumption for space cooling and heating poses an important challenge in accurately predicting the impact of weather on electricity demand. Various heating and cooling technologies are used across diverse geographic regions, and when a nation uses electricity for cooling and/or heating, even a slight fluctuation in temperature can result in significant variations in the electricity demand [6].…”

“…The electricity demand on an hourly basis was projected for both a 1-week and 1-day period, achieving a 3% mean absolute percentage error (MAPE) [27]. Contu et al [6] explored temperature-related changes in electricity demand in Italy with a proposed approach customizable by geography and time, and the results illustrate sensitivity coefficients vary by Italian region, and the reasons are linked to domestic demand such as building climate and technology. Dahl.…”