Experimental exergy analysis of low-GWP R290 refrigerant and derivation of exergetic performance equations with regression algorithms

Abstract

This study analyses the derivation of performance equations for a refrigeration system operating with R290 and R404A refrigerants using different regression models. Results of pace regression for coefficient of performance (COP), second law efficiency, and total exergy destruction showed mean absolute error (MAE) are 0.0993, 0.0159, and 0.0066, respectively. In all cases, the pace regression model made better predictions than elastic net regression. It was concluded that predictions made with regression models showed a good agreement with the actual experimental results. The derived equations can be utilised for refrigerants working in similar operational ranges.