Prediction of energy and exergy performance for subcooled and superheated vapor compression refrigeration system working with new generation refrigerants

Abstract

In this study, an adaptive network-based fuzzy inference system is used to facilitate the thermodynamic performance analysis of a cooling system. Adaptive network-based fuzzy inference system can extract nonlinear relationships between variables using training data and has an advantage its speed and simplicity in modeling a multivariate problem. Since the evaporator, condenser, superheating, and subcooling temperatures affect the performance of the cooling system, thermodynamic efficiencies are estimated depending on these temperatures. New generation R516A, R515A, and R515B refrigerants were used in the refrigeration system. When the estimated and actual coefficient of performance values is compared, the mean absolute percentage error values were 7.3%, 5.7%, and 6.2% for these refrigerants respectively. The mean absolute percentage error values for exergy efficiency were 4.7%, 5.9%, and 5.5% for the same refrigerants respectively. The results show that the adaptive-network-based fuzzy inference system can be successfully used in complex processes such as estimating the thermodynamic performance of the refrigeration system. Thus, this model will help engineers to predict refrigeration system performance very accurately, quickly, and easily.