LAUSR

ABSTRACT This paper proposes a data acquisition and analysis workflow and framework to clarify the daily residential behaviour of occupants based on the UWB indoor positioning system. The daily activities are analysed in different scales, namely grid level, convex polygon level and room level. At the first level, how long occupants stay in each 1 m*1 m grid was calculated, so that areas occupants stay longer was obtained. In the convex polygon level, the structure of the spatial network was clarified through analysis of the properties of each zone and weight of their connections. Although the network dynamically changes when the users and time differ, general patterns can be captured through analysis of large amounts of data. In the room level, which is the common analysing units, we get the movements between rooms and analysed the room-based behavioural networks of family members. Based on the findings in three levels, new variables were introduced in a regression model, and mutual visibility is found to significantly affect the household’s behaviour. By introducing the high-resolution indoor positioning system and the network analysis most used in the urban and social studies, this research provides a new prospect of data collecting and analysis of occupant indoor behaviour.