Abstract
A Bayesian method for determining occupancy in a buildings using motion and presence detectors. This was done by gathering data from different locations and then analyzing the data to determine occupancy within a certain location. The data was collected from a set of motion detectors and then compared against a presence detector to verify the results of the motion events. These motion events were then used to build the conditional tables to generate the belief from the Bayesian Network. The results were then again compared with the presence detector, along with a comparison between a clean data set and a data set with errors to view the change in the belief of the network. The preliminary results demonstrate that the Bayesian theory held true within the beliefs and that with errors, the beliefs maintained their accuracy. This is an ongoing research project in building environments.
Presenters
Avery SchwerProfessor, Architectural Engineering and Construction, University of Nebraska - Lincoln, Nebraska, United States Donald Levi Tryon
University of Nebraska, Nebraska, United States Dale Tiller
University of Nebraska
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
BUILDING ENVIRONMENTS, OCCUPANCY DETECTION, BAYESIAN METHOD, HIGH PERFORMANCE BUILDINGS