Akademik Veri Yönetim Sistemi
INTERNATIONAL JOURNAL OF LOW-CARBON TECHNOLOGIES, cilt.21, ss.1-19, 2026 (SCI-Expanded, Scopus)
Accurately predicting the coefficient of performance (COP) in vapor compression refrigeration systems demands accounting for numerous thermodynamic and operational variables, frequently resulting in intricate and protracted analytical expressions. To mitigate this complexity, this research explores the utility of Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) models for predicting the performance of an R600a-based domestic refrigerator employing Al2O3/SiO2 hybrid nanolubricants. The experimental findings reveal that the baseline system, running without nanolubricants, demonstrates a COP of 2.5; this value rises to 2.79 upon the incorporation of 0.4 g/l of Al2O3/SiO2 hybrid nanolubricants. Under identical operating conditions and refrigerant charge, the ANN model predicts an improved COP of 3.4 accompanied by mitigated compressor power consumption. In comparison, the ANFIS model further enhances the COP to 3.5 while simultaneously achieving a lower compressor work of similar to 100 W as well as a refrigeration influence of 215 W at the optimal nanolubricant concentration of 0.4 g/l and a refrigerant charge of 70 g. Additionally, ANFIS predictions suggest a decrease in compressor work to roughly 100 W. The maximum predicted COP of 3.5 is attained through the ANFIS methodology, exceeding both ANN predictions and experimental findings. In addition, the ANFIS model presents a considerably lower training error of 0.29901, hence validating its superior accuracy and durability in predicting cooling system performance.