Low-Complexity Hierarchical Control for Distributed Shopping Center HVAC

Abstract In this paper we present a low-complexity hierarchical control approach to fan-coil-based HVAC systems, applicable to shopping centers as exemplified through a case study of a Danish shopping center. Although Model Predictive Control remains the optimal approach performance-wise, we show that we can recover 66 % of the performance with the proposed approach, when considering no model-mismatch for the Model Predictive Controller. The recovered performance comes with the added benefits of increased reusablity and operator transparency, given no dependence on an accurate dynamical model and lower complexity.