Akademik Veri Yönetim Sistemi
Journal of Thermal Analysis and Calorimetry, cilt.150, sa.6, ss.4735-4745, 2025 (SCI-Expanded, Scopus)
This study investigates the energetic and exergetic performance of subcooled and superheated vapor compression refrigeration system utilizing data mining techniques. New-generation refrigerants R457A and R459B, considered alternatives to R404A, were utilized in the system. The analysis focuses on the impact of temperatures in the subcooling, superheating, condenser and evaporator. Data mining methods including multilayer perceptron (MLP), linear regression, M5 rules, M5P model tree (M5P), random committee and decision table models were used to estimate both energy and exergy efficiencies. The MLP model proved to be the most effective approach for predicting the energy (COP) and exergy efficiencies of R457A and R459B. When the predicted and actual COP values were compared, R-squared (R2) values of 0.9997 and 0.9994 were obtained for R457A and R459B, respectively. Similarly, the R2 values for exergy efficiency were 0.9984 and 0.9989 for the same refrigerants. These results demonstrate the successful application of data mining, in particular the MLP model, in evaluating the complex processes involved in refrigeration system performance analysis. This approach provides engineers with a fast, accurate and user-friendly method for predicting the behavior of refrigeration systems.