Embeddable adaptive model predictive refrigerator control for cost-efficient and sustainable operation

Emergent smart energy distribution systems offer new approaches in the sustainable management of energy consumption, even at household level. The smart grid and smart house paradigms provide the necessary infrastructure to improve the efficiency of the distribution system and at the same time offer benefits for the consumers, too. In this paper the control of an intelligent household refrigerator is introduced, which can autonomously react to hourly energy tariffs by shifting its operating periods to less expensive time slots, thus it greedily minimizes the energy bill of the consumer. As a “side effect”, the energy consumption is decreased during peak periods, which is the primary purpose of the utility provider, and also decreases the CO2 emission. The proposed controller adaptively learns the dynamic behavioral model of the appliance, and utilizes this model in a predictive control scheme. The proposed method is evaluated utilizing a realistic household refrigerator model and real pricing data. The operation is also validated by a real hardware implementation, which confirms that the control algorithm can be implemented using inexpensive embedded hardware devices. According to experiments using real pricing data, the proposed model predictive control can save 9% of the energy bill and can decrease the peak energy consumption by as much as 76%.