A Grey-box Model Based on Unscented Kalman Filter to Estimate Thermal Dynamics in Buildings

Buildings are responsible of about 40% of primary energy consumption. The widespread diffusion of Internet-of-Things devices provide allow collecting large amount of energy related data such as indoor air-temperature and power consumption of heating/cooling systems. Collected information can be used to develop data-driven models to learn building characteristics and to forecast indoor temperature trends. In this paper, we present a Grey-box model to estimate thermal dynamics in buildings based on Unscented Kalman Filter and thermal network representation. The proposed methodology has been applied to different implementation of building thermal networks to test their accuracy in temperature prediction. Results show the accuracy of the proposed methodology in predicting indoor temperature trends up to next 24-hours with a maximum error of 1.50°C.