Performance prediction of a single-stage refrigeration system using R134a as a refrigerant by artificial intelligence and machine learning method

Abstract

In this study, COP and heat capacities of evaporator and condenser were calculated by artificial intelligence and machine learning method in a vapor compression mechanical refrigeration cycle using well-known R134a as a refrigerant. Dataset was obtained with CoolPack software to train the model. Evaporating, condensing, superheating and subcooling temperatures are selected as input data. COP, heat capacities of evaporator and condenser are included in the dataset as target values. Artificial Neural Network (ANN) model was created with Matlab R2018b software and validated with target data. The output files obtained were compared with the target files and it was found that the mean square error value was quite close to one. The results of this study show that the ANN method can be used to obtain cycle parameters in one stage refrigeration cycle with high accuracy.