Control Methods for Energy Management of Refrigeration Systems

In this study, a high fidelity model for supermarket refrigeration systems is developed. The problem of energy management is investigated under two different demand-side management schemes: indirect and direct load control. The former is based on the electricity price incentive, and in the latter the consumption is directly managed by a third party based on a contract. For the indirect load control, we propose two different supervisory control approaches using model predictive control (MPC) technique to address the problem of electricity cost minimization. For the direct load control, three different control methods are proposed. The first method is the least complex one including decentralized proportional-integral control loops that can be applied to a general class of supermarket refrigeration systems. In the second method, an MPC is formulated for which a high fidelity model specific for each supermarket system is required for the controller implementation. The advantage is that a higher performance, in terms of following the reference load, is achieved at the price of increasing the complexity. In the third method, we utilize a data-driven approach using the subspace method. Here the complexity is decreased as the method does not need an explicit model of the system and, at the same time, the desired load following performance is attained.

Recent research findings indicate that the refrigeration system commonly employed in food transportation can account for 40% of the total greenhouse gas emissions from the corresponding vehicle engines. Finally, the problem of optimization of a hybrid transport refrigeration system is addressed here. The hybrid refrigeration system is made by the integration of conventional refrigeration technology with thermal energy storage devices.