Model predictive control, inverter air conditioner, real-time electricity price, smart grid
Real-time electricity prices provide great opportunities for residential consumers to participate in demand response (DR) programs in smart grids. One of the biggest challenges faced by the DR program participants are the lack of advanced DR-enabled residential appliances which can automatically respond to real-time electricity prices, especially the lack of DR-enabled air conditioners which are the major contributors to electricity bills. In this paper, we aim to use model predictive control (MPC) techniques to control the inverter air conditioners. Instead of simply adjusting temperature set-points in the conventional DR control strategies, the MPC controllers are used to directly control the operating frequencies of inverter ACs. In contrast to the conventional control methods for inverter-driven ACs, the MPC approach can systematically integrate the predictions of weather, occupancy and real-time electricity prices into the optimization problems to achieve the energy-and-cost saving tasks. For computational efficiency, a simple-structured room thermal model and steady-state performance maps of inverter ACs are developed and integrated. Two types of MPC controllers with different level of complexities are designed for comparison. Simulation results show that the MPC of inverter ACs can effectively improve the thermal comfort at the beginning of occupation, shift the peak power demands, reduce the total energy consumptions and reduce the electricity costs.