Experimental demonstration of penalty-based distributed MPC for peak load and cost reduction in a dual-zone building test bed

초록

This study proposes a penalty-based distributed MPC framework to simultaneously reduce electricity cost and mitigate peak load in buildings. The framework employs a resistance-capacitance (RC)-network-based grey-box thermal model developed from measured data. The framework also enables non-intrusive, automatic control through real-time data acquisition and application programming interface (API)-based integration with a commercial heating, ventilation, and air conditioning (HVAC) system. In the proposed framework, each zone independently solves an MPC problem to maintain indoor comfort, reduce operating costs, and satisfy operational constraints. At the same time, information exchange among zones enables coordinated load distribution. The proposed strategy was validated through both simulation and experimental implementation in a twinchamber test bed. Simulation results show that the framework reduces electricity cost by 11.23% relative to the baseline feedback control and lowers the peak load by 45.68% relative to basic MPC. Experimental results confirm that the proposed method maintains indoor comfort while reducing the peak load by 43.7% and electricity cost by 11.0% relative to the baseline feedback control. These findings indicate that the proposed control framework provides a practical approach for coordinated peak-load management among independent HVAC units and offers a basis for future extension to larger multi-zone building systems.

키워드