Distributed resource allocation for multi-zone commercial buildings

Abstract

Advanced building structures are extremely complicated with a large number of dynamically interacting components and widely equipped with multi-zone air conditioning (ME A/C) systems. In this paper, multi-zone commercial building systems are built as second order Resistance-Capacitance (RC) models with individual zones communicating to the whole network through neighbors. The presented thermal dynamics are different from that in our previous work [8], which does not consider the wall’s thermal dynamics. The second order RC models are more accurate in capturing the thermal dynamics of a building in comparison with the first order model employed in [8]. We first propose a hierarchical distributed method using ME A/C systems in commercial building for demand response. The idea of the upper layer strategy is based on the balance of building’s energy demand and users’ comfort level to generate reference signals. The lower layer is designed as distributed model predictive controllers (DMPC) to track the reference signals whereas energy cost and demand are reduced. Simulation results are provided to verify the effectiveness of the proposed control.