Identication of the cooling system of a large business center

This paper deals with the development of a control oriented model for a real cooling station serving a large business/commercial area with about 65,000 square meters of spaces and 3,000 permanent workplaces. The model is derived by focusing on one of the subsystems at a time, eventually simulating the physical closed-loop present in the plant. Since first principle equations would be impractical in view of the dimension and uncertainty of the system, the model largely relies on black-box relations estimated from the available measurements. The final model can be used for the design of an optimization-based model predictive control algorithm.