Accurate simulation and prediction of occupants’ energy use behavior are crucial in building energy consumption research. However, few studies have focused on household energy use behavior in severely cold regions that have unique energy use patterns because of the low demand of cooling in summer and the use of central heating system in winter. Thus, we developed an agent-based model to simulate the household electricity use behavior in severely cold regions, according to data for Harbin, China. The model regards apartments, residents, household appliances, and energy-management departments as agents and generates the household electricity consumption with respect to time, temperature, and energy-saving events. The simulation parameters include basic information of the residents, their energy-saving awareness, their appliance use behaviors, and the impact of energy-saving management. Electricity use patterns are described by decision-making mechanisms and probabilities obtained through a questionnaire survey. In the end, the energy-saving effects of different management strategies are evaluated. The results indicate that the model can visually present and accurately predict the dynamic energy use behavior of residents. The energy-saving potential of household electricity use in severely cold regions is mainly concentrated in lighting and standby waste, rather than cooling and heating, since the cooling demand in summer is low and the heating in winter mainly relies on central heating system of the city, not on household electricity appliances. Energy-saving promotion can significantly reduce the amount of energy waste (41.89% of lighting and 97.79% of standby energy consumption), and the best frequency of promotional events is once every four months. Residents prefer incentive policies, in which energy-saving effect is 57.7% larger than that of increasing electricity prices. This study realized the re-presentation of the changes of energy consumption in a large number of households and highlighted the particularity of household energy-saving potential in severely cold regions. The proposed model has a simple structure and high output accuracy; it can help cities in severely cold regions formulate energy-saving management policies and evaluate their effects.
Exploratory study on clustering methods to identify electricity use patterns in building sector