Modeling and Energy Cost Optimization of Air Conditioning Loads in Smart Grid Environments

Electricity for heating, ventilation, and air condition (HVAC) machines takes up a large percentage of energy consumption in the buildings and thus in turn, a large portion of the energy monetary cost. Optimization of air conditioners use throughout the day will reduce energy consumption and expenditure. This study introduces a second-order differential equation model to capture the indoor temperature dynamics of a building. An experimental test bed is developed to collect a set of indoor/outdoor temperature and sunlight data. Using a least-squares-based system identification process, the model parameters are identified and checked through simulation. Optimization of the room temperature is then determined by solving a mixed-integer quadratic programming problem in relation to the hourly-updated energy prices. Mixed-integer quadratic programming solution is compared to a two-point thermostatic control system. A hybrid solution compromising the quadratic programming algorithm and the conventional thermostatic control scheme is proposed as a tractable approach for the near-optimal energy management of the system.