Logic Embedded NMPC of a Solar Air Conditioning Plant

This paper presents a hybrid controller for a solar air conditioning plant, located at the University of Seville, Spain, and used as a benchmark for the HYCON NoE of the European Union (EU). The plant uses two sources of energy: solar and gas, plus a set of accumulation tanks and an absorption tower to provide conditioned air to a university building. The hybrid character comes from the discrete decisions that have to be taken in relation with the energy source in use, which is selected by means of on/off valves, plus the continuous nature of the process and other manipulated variables. The control aim is to operate the conditioned air system using the smaller possible amount of energy from gas and maintaining other variables close to their set points. The hybrid control is based on a model predictive control strategy developed with the objective of dealing with the mixed discrete-continuous nature of the process in an efficient way. A novel approach incorporating an internal model with embedded logic control is used to transform the hybrid problem in a continuous-nonlinear one where NMPC can be applied. The paper presents results obtained both, in simulation and in the real system.