A control strategy of heating system based on adaptive model predictive control

Space heating accounts for a large proportion of a building's energy consumption, and improving heating effi-ciency is a significant approach to reduce heating energy consumption and carbon emissions. To improve heating efficiency and meet the demand for thermal comfort, this paper proposes an adaptive model predictive control (AMPC) based heating control strategy to regulate the heating parameters of heat exchange stations in residential communities. A multi-input non-linear model combined with subspace identification algorithms is constructed using the heating data from the heat exchange stations and meteorological data. An AMPC control system, a model predictive control (MPC) and a proportional-integral-derivative (PID) control system, are then built to compare their control performances such as room temperature control accuracy, energy saving capacity, response speed, and robustness under extreme weather. The results show that the AMPC strategy outperforms the other two, with a 67.5% reduction in room temperature control deviation and a 20.3% reduction in energy consumption compared to the actual operation of the heating plant. Under extreme weather conditions, the AMPC strategy has 44% less deviation in indoor temperature control and a shorter response time than the least effective PID control strategy of the three. The AMPC system has broad application prospects in the heating field.