Occupant-location-catered control of IOT-enabled building HVAC systems

Catered control strategies for building HVAC systems, which react to the changing location of an occupant of the building, are studied. A structured Markov Jump-Linear System (MJLS) model is developed which captures the diffusive heat-exchange process in the building, stochastic occupant movement, and feedback control enacted by the HVAC system. A statistical analysis of the closed-loop dynamics is undertaken, and used to evaluate stability and control performance. The tuning of simple proportional-integral-derivative control schemes to optimize comfort is also discussed, and demonstrated in an example.