Low computational cost technique for predictive management of thermal comfort in non-residential buildings

Antoine Garnier (1, 2) , Julien Eynard (2) , Matthieu Caussanel (2) , Stéphane Grieu (2)

1 Pyrescom
2 PROMES - Procédés, Matériaux et Energie Solaire

In France, non-residential buildings account for a significant part of energy con-sumption. A large part of this consumption is due to HVAC (Heating, Ventilation and Air-Conditioning) systems, which are in most cases poorly handled. The present work deals with an efficient approach allowing energy consumption to be minimized while still ensuring thermal comfort. We propose a predictive con-trol strategy for existing zoned HVAC systems and consider the PMV (Predicted Mean Vote) index as a thermal comfort indicator. In order to test this strategy, we modelled a non-residential building located in Perpignan (south of France) us-ing the EnergyPlus software. The twofold aim is to limit the times during which the HVAC sub-systems are turned on and to ensure a satisfactory thermal comfort when people are working in the considered building. This predictive approach, computationally tractable, allows thermal comfort requirements to be met without wasting energy.

Mots clés: non-residential building; energy consumption; zoned HVAC system; thermal comfort; artificial neural networks; model predictive control

Domaines: Automatique / Robotique

Antoine Garnier, Julien Eynard, Matthieu Caussanel, Stéphane Grieu. Low computational cost technique for predictive management of thermal comfort in non-residential buildings. Journal of Process Control, 2014, 24 (6), pp.750-762. ⟨10.1016/j.jprocont.2013.10.005⟩. ⟨hal-01098219⟩