Set-point temperature representation in a resistance–capacitance network model to predict both heating rates and room temperatures

This study proposes a novel control-oriented Resistance–Capacitance (RC) model that incorporates a set-point temperature representation to enhance compatibility with real Heating, Ventilation, and Air-Conditioning (HVAC) systems and to improve predictive control accuracy. Model Predictive Control (MPC) is increasingly adopted in HVAC applications, where the effectiveness of control relies on accurate physical modeling. Conventional RC models typically calculate the required heating rate by controlling the room temperature, whereas actual HVAC systems operate based on set-point temperatures. This mismatch between the control input in the model and actual system behavior can lead to prediction and control errors. The proposed model introduces a time-variable thermal resistance between the set-point and room temperatures and calculates the heating rate based on the temperature difference. This structure enables the direct use of the set-point as the control input and allows the simultaneous prediction of both the room temperature and heating rate while preserving physical interpretability. Because the set-point temperature is the target in actual Air Handling Unit (AHU) operations, the model improves both the stability and accuracy. Validation using real building data showed that the proposed model achieved a 3.6% improvement in room temperature prediction over the reference model. For the AHU heating rate prediction, which conventional RC models do not explicitly address, the model demonstrated acceptable performance. The proposed RC model provides a practical and accurate framework for MPC-based HVAC control implementation by capturing the natural convergence of room temperature towards the set point and explicitly modeling its influence.