Seasonality of Human Behavior in Smart Buildings
Marc Le Goc, Fabien Barthelot, Éric Pascual · 2015
The aim of the presented work is to give an operational solution to model and to monitor the behavior of peoples in smart buildings together with the power consumption optimization of the environment resources. The proposed solution is based on the discovery, the modeling and the validation of behavioral knowledge with an unsupervised learning process included in a whole Knowledge Discovery in Database process. The particularity of the proposed approach is that the learning process has been design to work on the timed data directly provided by a smart environment. This paper describes the approach and its application to a real world building of offices. The application puts on the light one of the most difficult problem in this kind of approach, the seasonality of human behavior, that changes the behavior models over time, making the usual approaches difficultly practicable.