Temperature and humidity model for predictive control of smart buildings

This paper addresses the energy efficiency of buildings in line with global climate policies aimed at reducing energy consumption and carbon emissions. It proposes a grey-box model for smart buildings equipped with Heating, Ventilation, and Air Conditioning (HVAC) systems and radiators. The model predicts temperature and humidity changes, essential for maintaining occupant comfort and optimizing energy efficiency through Model Predictive Control (MPC). The methodology involves developing a comprehensive model that incorporates various factors such as radiators, HVAC operation modes. The findings demonstrate that the model accurately predicts indoor temperature and humidity, with Root Mean Square Errors for indoor humidity ranging from 0.3 g/kg to 0.64 g/kg, HVAC output temperature from 0.4 °C to 0.75 °C, and room temperature prediction errors of 0.36 °C in winter and 0.11 °C in summer. The key contribution is the model’s ability to enhance energy efficiency and comfort prediction in smart buildings, providing a valuable tool for advanced control methods like MPC. The novelty lies in the integration of radiators and HVAC systems within a temperature and humidity predictive model, offering a practical solution for real-world applications, and accurately predicting for all scenarios. • Humidity prediction in various operating states, including when systems are off. • Radiator temperature depends on external temperature for district heating. • Extended systems state descriptions increase accuracy. • Multi-variant model accuracy results based on real building data.