Multi-zone modeling and energy efficient control of shopping center cooling

In this paper we consider the problem of constructing a dynamical model for shopping center HVAC systems, suitable for proposing new high-level control designs to minimize energy consumption for the entire shopping center. We also propose a preliminary control design, to increase energy efficiency. The specific system considered in this paper, is a small section of a Danish shopping center, including three shops and their joint cooling system. The current control solution is investigated and described.
A dynamical model is constructed as a grey-box RC-equivalent model, a suitable modeling paradigm for control oriented models that also have to be scalable. Parameters for the model have been identified through a combination of measurement data from several days of live operation and table-lookup, calculating thermal properties based on shop dimensions.
The resulting model is used to propose a preliminary control solution, to increase efficiency by utilizing a higher forward temperature. This is achieved through a control design that seeks to drive valve openings closer to fully open, while still allowing headroom for disturbance rejection. One of the main benefits of this design, is the low implementation barrier, as it does not require alterations to shop-local temperature controllers.
Simulations show that the proposed control solution works as intended, without degrading the performance of the existing shop temperature control.