Research on Electricity Consumption Model of Library Building Based on Data Mining

With the exponential development of Chinese population,the massive energy consumption of buildings has recently become an interest subject.Although much research has been conducted on residential buildings,heating ventilation and air conditioning(HVAC),little research has been conducted on the relationship between student’s behavior,campus buildings,and their subsystems.Using classical seasonal decomposition,hierarchical clustering,and apriori algorithm,this paper aims to provide an empirical model for consumption data in campus library.Smart meter data from a library in Beijing,China,is adopted in this paper.Building electricity consumption patterns are investigated on an hourly/daily/monthly basis.According to the monthly analysis,electricity consumption peaks each year around June and December due to teaching programs,social exams,and outdoor temperatures.Hourly data analysis revealed a relatively stable consumption pattern.It shows three different types of daily load profiles.Daily data analysis demonstrated a high relationship between HVAC consumption and building total consumption,with a lift value of 5.9.Furthermore,links between temperature and subsystems were also discovered.Through a case study of library,this study provides a unique insight into campus electricity use.The results could help to develop operational strategies for campus facilities.

Modelling method of inter-building movement for campus-scale occupancy simulation:A case study

As an important factor in the investigation of building energy consumption,occupant behavior(OB)has been widely studied on the building level.However so far,studies of OB modelling on the district scale remain limited.Indeed,district-scale OB modelling has been facing the challenges from the scarcity of district-scale data,modelling methods,as well as simulation application.This study initiates the extrapolation of occupancy modelling methodology from building level to district scale through proposing modelling methods of inter-building movements.The proposed modelling methods utilize multiple distribution fittings and Bayesian network to upscale the event description methods from inter-zone movement events at the building level to inter-building movement events at the district level.This study provides a framework on the application of the proposed modelling methods for a university campus in the suburbs of Shanghai,taking advantages of data sensing,monitoring and survey techniques.With the collected campus-scale occupancy data,this paper defines five patterns of inter-building movement.One pattern represents the dominated inter-building movement events for one kind of students in their daily campus life.Based on the quantitative descriptions for various inter-building movement events,this study performs the stochastic simulation for the campus district,using Markov chain models.The simulation results are then validated with the campus-scale occupancy measurement data.Furthermore,the impact of inter-building movement modelling methods on building energy demand is evaluated for the library building,taking the deterministic occupancy schedules suggested by current building design standard as a baseline.