Paper Conference

Proceedings of BSA Conference 2017: Third Conference of IBPSA-Italy


Temporal and Spatial Predictability of Occupants’ Presences in a Library Building

Ki Uhn Ahn, Won Jun Suh, Cheol Soo Park

Abstract: Although it has been widely acknowledged that occupants play a critical role in building energy consumption, the characteristics of the occupants’ presences and actions are not accurately represented in building simulation tools. In this study, the authors aim to identify the temporal and spatial difference in the characteristics of the occupants’ presences in a library building located in South Korea. The data of each individual entering and leaving rooms in the building were recorded for weeks from 15 to 31 January by Radio-Frequency Identification (RFID) tags. Based on the data, the periodicity and predictability of the occupants’ presences were investigated in terms of (1) spatial variation (individual room vs. entire floor vs. entire building), and (2) temporal variation (5 minutes vs. 30 minutes vs. 1 hour sampling times). The periodicity and predictability of the occupants’ presences were quantified using a Normalized Cumulative Periodogram (NCP). The occupants’ presences in an individual room were less periodic or less predictable than those in the entire building. The predictability of the occupants’ presences was influenced by the sampling time. The greater the sampling time (5 minutes vs. 1 hour), the better predictable the occupants’ presences were. In other words, it is much easier to predict the aggregated number of occupants over the entire building with a longer sampling time (e.g. a couple of hours) than to predict the number of people in individual rooms with a shorter sampling time (e.g. 10 minutes). These findings could be further applied to an energy simulation study of the building and its relevance to energy prediction will be studied.
Pages: 505 - 512