Maintaining thermal comfort in buildings: feasibility, algorithms, implementation, evaluation

Limiting peak-loads and reduction in energy consumption are two important considerations in the design of smart-home control systems. A smoother load profile benefits both utilities and the consumers, in terms of improved grid stability and QoS (fewer occurrences of load-shedding, brownouts and blackouts). Building energy loads are dominated by thermostatically controlled electrical devices (TCEDs) such as air-conditioners and heaters and these loads must be scheduled to deliver the desired thermal comfort by maintaining the temperature of a given environment within a band. Even though the periodic duty-cycle of TCEDs appears to make real-time scheduling algorithms suitable for the scheduling of TCEDs, we find that existing policies are not suitable for the coordinated control of TCED loads. Further, the loads must be managed taking into account important practical issues, especially, (i) considering mandatory restart-delay in scheduling compressor-driven TCEDs, (ii) avoiding undesirable switching (ON/OFF) of electrical appliances (to improve efficiency of the equipment and reduce failures), and (iii) accounting for the effect of periodic scheduling decisions, taken in discrete time, on the maintenance of thermal-comfort. We present a new Thermal Comfort-band Maintenance algorithm whose design is motivated by the above considerations. We also show how the approach leads to energy-efficiency and adaptive demand–response control by adapting the comfort-band. Results from simulation and real-life implementation demonstrate that our algorithm is superior to the existing algorithms for building electrical load scheduling in terms of maintenance of thermal comfort and reduced number of undesirable switching.