Application of Distributed Model Predictive Approaches to Temperature and $CO_2$ Concentration Control in Buildings

Shalika S. W. Walker (1) , Warody Lombardi (2) , Suzanne Lesecq (2) , Samira Roshany-Yamchi (3)

1 TU/e - Eindhoven University of Technology [Eindhoven]
2 CEA-LETI - Commissariat à l'énergie atomique et aux énergies alternatives - Laboratoire d'Electronique et de Technologie de l'Information
3 NIMBUS - NIMBUS Centre [Cork]

In the context of energy consumption reduction, this paper focuses on the application of Model Predictive Control to occupants' thermal comfort together with indoor air quality control while improving the whole building energy efficiency. First, an open-space office split in three zones, located in Cork Institute of Technology, is modeled. A centralized MPC is designed to control the temperature and $CO_2$ concentration in the three zones. Then, a distributed version of the MPC, with three separate local controllers, is considered. Finally, simulation results show that the distributed MPC solution achieves control performance quite close to the centralized version with less computing effort.

Mots clés: Model Predictive Control; centralized; distributed; RC model; Energy management; temperature; CO$_2$ concentration

Domaines: Automatique / Robotique

Shalika S. W. Walker, Warody Lombardi, Suzanne Lesecq, Samira Roshany-Yamchi. Application of Distributed Model Predictive Approaches to Temperature and $CO_2$ Concentration Control in Buildings. IFAC world Congress 2017, Jul 2017, Toulouse, France. ⟨cea-01564058⟩