This paper proposes a model-free real-time optimization and control strategy for CO2 transcritical refrigeration plants that assures covering the cooling demand and continuous tracking of conditions for maximum efficiency. Our approach obtains the feedback with only three measurements, and controls the opening degree of a back-pressure valve and the speed of the compressor. The strategy minimizes the power consumption of the compressor instead of maximizing the coefficient of performance, which avoids several sensors, and we demonstrate mathematically that both approaches are equivalent. We implemented the strategy with an algorithm that includes two independent auto tuned controllers, one devoted to regulate the high-pressure and another to regulate the outlet temperature of the secondary fluid of the evaporator. It also incorporates a real time perturb and observe procedure to locate on-line the optimum high-pressure that minimizes the compressor power consumption. The paper presents the experimental evaluation of the control strategy, verifying the stable operation of the algorithm and the energy optimization of the plant.
KEYWORDS