User Behavior Prediction in Energy Consumption in Housing Using Bayesian Networks

This paper deals with the problem of the user behavior prediction in a home automation system. Anticipating the needed energy for a service is based on the available prediction (like user requests) which contains the uncertainties. When the future users requests are not available in a home automation system thanks to programmatic, it is interesting to predict it to anticipate the energy needed in order to avoid some problems like peak consumption. A general method to predict users requests for services in energy consumption is proposed. The method relies on Bayesian networks to predict and diagnose user's behavior in housing. Some results and perspectives are presented in this paper.