Decentralized Control Design for User Comfort and Energy Saving in Multi-zone Buildings.

Abstract This paper proposes novel decentralized control design methods for occupant comfort and energy saving in smart multi-zone buildings. First, the graph-based modeling technique for buildings is utilized to obtain a simplified lumped model for thermal dynamics of zones. Second, the graph theory is employed to derive features of the model useful for the control design. Next, two methods are proposed based on the pole placement in linear matrix inequality (LMI) region technique and the Lyapunov inequality to synthesize decentralized controllers to track the reference temperatures set by users for their thermal comfort and to save energy. The effectiveness of the proposed designs are then illustrated in a numerical example. [ABSTRACT FROM AUTHOR]