In this study, the impact of behavioral actions of a building occupant on energy performance and thermal sensation are investigated. The study focuses on the six following actions: use of blinds, lighting system, windows, fan, thermostat and clothing adjustments. Eight types of buildings, classified among three criteria (air-conditioning, thermal inertia and climate), are studied. Simulation of the occupant's actions, building performance and thermal sensation have been carried out by using TRNSYS 17. Impact on energy demand and thermal sensation of each action has been investigated with a Design Of Experiments methodology coupled with the use of Yate's algorithm. This study shows that for a given building, the occupant's actions have a significant impact on energy demand. Building simulation in literature typically does not model human activity in energy consumption, yet our study demonstrate a strong correlation. Results from the design of experiments methodology are compared to conventional French design strategy. It appears that conventional French design strategy, which does not take into account occupants' actions, tends to strongly underestimate building energy demand.

Occupants have influence on buildings performances due to their presence and their behavior towards indoor environmental conditions controls. However, most building energy models consider occupants in an over-simplified way. Many experiences feedbacks have shown that this assessment leads to huge differences between simulation results and actual energy consumption. In this paper we propose a new method aiming at reducing this uncertainty. It embeds mainly two models: a two-node thermophysiological model that calculates thermal sensation of a human being and an occupant behavior model based on artificial intelligence. Results of this new method are then presented and compared with actual data from field studies.