Physics-Informed Model : REDUCEDHEATCARB Deliverable D1.2

Samenvatting

This deliverable is one of the results of the project Research on Energy Data Utilization for Carbon Emission Decline by Heat pumps and Efficiency, while Assuring Thermal Comfort Achievement in Residential Buildings (REDUCEDHEATCARB). This deliverable describes the REDUCEDHEATCARB model, a model used in physics informed machine learning of the heat performance signature of a residential building, heating installation and household comfort needs. Key improvements with respect to earlier models developed in the Twomes and Brains4Buildigns model include enhanced modeling of ventilation rates, infiltration, and heat distribution system parameters, along with separate learning stages for ventilation, thermal, and flow characteristics. Additionally, new case metrics for comfort, energy, and emissions have been introduced, as well as an improved gas-fired boiler model using GEKKO B-splines and an ON/OFF thermostat algorithm. Onderdeel van project: REDUCEDHEATCARB