Non-linear predictive control of a vapour compression cycle

A potential energy saving can be obtained by a global optimal setting of each actuator in a vapour compression cycle. This paper describes a predictive optimal control algorithm applying this strategy. At each step, an optimal command profile is computed, upon predictions of variables system evolution, using a simple non-linear model of a vapour compression cycle. Choice of this profile is based on a multiple criterion including cycle efficiency and technological constraints. The approach is experimentally validated on a pilot scale refrigeration plant with variable speed compressor. Trials have been performed and showed a meaningful energy saving.