Modeling and Identification for HVAC systems

Heating, Ventilation and Air Conditioning (HVAC) systems consist of all the
equipment that control the conditions and distribution of indoor air. Indoor
air must be confortable and healthy for the occupants to maximize their
productivity. Moreover, HVAC energy consumption is between 20% and 40%
of the total energy consumption in developed countries and accounts around
33% of the global CO2 emissions. So the study of HVAC systems plays an
important role in building science.

The aim of this project is to identify mathematical models that will be
employed by intelligent control algorithms which guarantee human comfort
indoors, energy saving and less CO2 emissions at the same time. Three models,
based on first-principle physical knowledge, are proposed for CO2 concentration,
temperature and humidity, respectively, for a room in the Q-building at KTH.
Thermodynamic equations and an original estimation of the number of the
occupiers of the room are employed.

Validation shows that models have really good performances, even with a
short training dataset. Discussions on the obtained results are given and some
ideas for future work are proposed.