Typical occupancy profiles in buildings based on big data of mobile positioning

Occupant behavior in buildings has significant impacts on building energy consumption. Occupancy profiles are important input parameters for building consumption simulation. Current methods to obtain the occupancy profiles, such as infrared sensing methods and manual counting methods, have low accuracy and systematic error, and are often time-consuming and labor-consuming, which may not be widely used in real projects. With the advances of social media software, the mobile positioning data can reflect building occupancy schedule. Based on the cluster analysis, this research proposes a series of descriptive indexes reflecting the characteristics of daily and weekly profiles. Taking the occupancy profiles of a hospital in Beijing as a case study, this research discovers that the occupancy schedules based on mobile positioning data have significant difference from those in energy codes. The former can reflect occupancy characteristics of buildings more objectively.