Optimal Coordination of HVAC Scheduling for Commercial Buildings

Commercial buildings are spending tens of millions of dollars on energy every year. Part of the energy cost can be saved by optimizing the operation of these buildings. This paper proposes an optimal scheduling solution for the heating ventilation and air conditioning (HVAC) systems that serve a university campus. The proposed solution aims at minimizing the total energy cost of the campus by coordinating the operation of HVAC systems in multiple buildings under the time-of-use (TOU) tariff framework. The proposed methodology maintains the temperature comfort level while taking into consideration the distribution feeder constraints and power losses. To obtain the optimal scheduling, the power consumption model and the building temperature prediction model are obtained using a data-driven method. Then, the models are embedded into a Mixed Integer Quadratic Programming (MIQP) problem that solves the optimal scheduling of HVAC. The data-driven model is trained and calibrated using actual historical data and the optimization is done under TOU framework.