Research on prediction of temperature-sensitive loads based on multi-factor correlation analysis

Abstract

Amidst the backdrop of global warming， the unstable climate presents challenges to the secure operation of power systems. Therefore， it is particularly crucial to accurately and scientifically forecast the electricity demand of temperature-sensitive loads such as heating and cooling. Conventional methods for predicting temperature-sensitive load fail to encompass the comprehensive influence of factors such as climate， geography， and society. Therefore， a method of forecasting the increase of cooling and heating loads and electricity based on the 3T （temperature， territory and time） model. Based on the decomposition of temperature-sensitive load and electricity， the temperature and social factors affecting electricity consumption for cooling and heating are first sorted out. Then， the influencing factors with greater correlation with the temperature-sensitive load are selected based on historical data， and the load and electricity forecasting function is determined with the help of the proposed optimization model. Then， the predicted data of temperature and social conditions of the corresponding time period are substituted in the function to obtain the prediction results. Finally， the efficacy of the method is validated through case studies. Results indicate that considering multiple factors in temperature-sensitive load prediction yields estimations closer to reality， enhancing the adaptation of electricity demand to future temperature trends.

Keywords

• temperature-sensitive load
• correlation analysis
• electricity demand
• forecasting
• temperature factors
• social factors