L-Band SBQP-based MPC for supermarket refrigeration systems

This paper presents an L-Band sequential binary quadratic programming based model predictive control (L-Band SBQP-based MPC) of a supermarket refrigeration system. This kind of system is of hybrid nature due to the presence of discretely switched actuators, as valves and compressors, and to the autonomous or forced changes of dynamics. A control issue for these systems is to maintain the outputs, namely air temperatures and pressure of liquid refrigerant, within the constraints and to reduce the switching of the compressors. The present work proposes a model predictive control strategy in which the control actions are the solutions of a SBQP optimization problem. The computational effort is reduced by means of the L-Band algorithm. In this paper, the hybrid model of the supermarket refrigeration system and the proposed control strategy are presented, providing some results and a comparison with traditional control performances.