Enhancing COP in R600a Refrigerators by dispersing CuO/SiO2 Nanolubricants: A machine Learning Prediction

The performance evaluation of the R600a refrigerator was enhanced through the application of machine learning methods.In the present investigation, major performance parameters (COP, refrigeration effect and compressor work) in R600a refrigerators utilizing CuO/SiO nanolubricants have been predicted by machine learning algorithms (XGBoost, Linear Regression and Random Forest).Atmospheric temperature, R600a mass flow rate and CuO/SiO2 concentrations are the experimental data used as input factors.The corresponding performance variables were analyzed as outputs.XGBoost demonstrated significantly better accuracy than Linear Regression and Random Forest, achieving a low mean squared error (MSE) of 0.00465 for COP prediction, compared to an average MSE of 26.671 for the other models.This study highlights XGBoost's capability to accurately predict and optimize refrigeration system performance, particularly under conditions involving advanced nanolubricants.It introduces a novel approach for enhancing the refrigeration effect while reducing compressor energy consumption, thereby significantly improving the performance of R600a-based systems and contributing to the development of more environmentally sustainable cooling technologies.