Optimization and application of control strategies for hybrid heat pump heating systems in public buildings

Gas-electric hybrid heat pump heating systems exhibit superior heating stability and energy efficiency under cold conditions, representing a crucial technical pathway for the low-carbon transition of building heating. However, research on this type of hybrid heat pump is still relatively limited. In this paper, the gas-electric hybrid heat pump heating system was studied and developed its component-level dynamic model. The hybrid heat pump unit couples a gas boiler in series with an air source heat pump, incorporating flue gas heat recovery on the refrigerant side. The model is established in Modelica/Dymola and validated against laboratory data with a maximum relative error of 9.2%. A supervisory model predictive control strategy is then developed via Modelica-Python co-simulation using NSGA-II multi-objective optimization, simultaneously minimizing operating cost, carbon emissions, and thermal discomfort over a 24-h prediction horizon. For a representative hotel in Tai'an, the proposed strategy reduces primary energy consumption by 4.4% to 5.6%, annualized cost by 3.0% to 7.3%, and CO₂ emissions by 4.7% to 5.9% relative to baseline control and maximizing off-peak electricity utilization control strategies, while outperforming pure air source heat pump and pure gas boiler systems by 9.1% and 12.9% in primary energy, respectively. Regional analysis of six representative heating cities in China shows that hybrid heat pump system consistently outperforms single-source systems. Compared with baseline control, MPC delivers additional energy savings of 3.8–7.1%, annualized cost reductions of 5.3–9.9%, and carbon mitigation of 3.8–7.3%. © 2026. Published by Elsevier Ltd.