Distributed Energy Optimization for HVAC Systems in University Campus Buildings

Educational buildings consume about 2% of the total energy in a country, which leads to energy cost concerns for building operators. To reduce the building energy cost, an effective way is to intelligently schedule heating, ventilation, and air conditioning (HVAC) systems, which account for above 40% of the total energy consumption in educational buildings. In this paper, we investigate an energy optimization problem for HVAC systems in university campus buildings. To be specific, we first formulate a total cost optimization problem that minimizes the sum of energy cost related to HVAC systems and thermal discomfort cost associated with occupants considering zone occupancy pattern and thermal preference of occupants in each zone. Due to the existence of uncertain parameters as well as spatially and temporally coupled constraints, it is very challenging to solve the formulated problem. To this end, we propose a distributed model predictive control algorithm based on the alternating direction method of multipliers, which has high scalability with an increasing number of zones and can protect user privacy. Extensive simulations based on the real-world traces show that the proposed distributed algorithm could effectively reduce the total cost and offer a flexible tradeoff between energy cost and thermal discomfort.