• building thermal modeling
• carbon emissions reduction
• energy and cost optimization
• grey-box modeling
• lumped models
• model predictive control (MPC)
• parameter estimation

Model Predictive Control strategies represent a promising approach to enhancing the performance of HVAC systems in buildings. Key benefits include reduced electricity consumption, lower CO₂ emissions and significant cost savings in heating and cooling operations. This project first focuses on developing a grey-box model for a residential case study, employing low-order thermal equivalent circuits based on electrical analogies, namely 1R1C and 3R2C models. After evaluating the accuracy of indoor temperature prediction and the physical consistency of the parameters estimated through optimization algorithms, the selected model is integrated into an MPC framework. Due to the absence of a fully validated white-box model of the building, the 3R2C model is used as a reference to simulate the impact of the predicted control actions. Based on this setup, the 1R1C model was used within the MPC as a predictive model to compute the optimal control actions. This configuration enables the performance evaluation of MPC strategies under various operating scenarios, assessing their potential in improving thermal efficiency and minimizing operational costs leveraging the building's thermal inertia. The most significant results show that implementing a predictive control strategy can lead to an economic saving of approximately 11%. In this scenario, optimized for cost reduction, the approach also achieves a 5% decrease in CO2 emissions and a 3.5% reduction in electricity consumption.