Comparative analysis for the prediction of boiling heat transfer coefficient of R134a in micro/mini channels using artificial intelligence (AI)-based techniques

Abstract

ABSTRACTRefrigerant R134a has been extensively used in the past because of its zero ozone depletion potential (ODP). In the present research, artificial intelligence (AI)-based gene expression programming (GEP), artificial neural networks (ANN) and support vector regression (SVR) models have been developed for the prediction of heat transfer coefficient for the boiling of R134a in micro/mini channels. The performances of developed models were compared and evaluated against the experimental results in terms of statistical parameters such as coefficient of determination (R2) and average absolute relative error (AARE). The obtained results and findings from this research reveal that SVR is an effective technique for predicting the heat transfer coefficient of R134a, with lowest AARE value of 3.62% and a high R2 value of 0.9749 in comparison with other AI-based models. Furthermore, performance of the ε-SVR with four different kernels: linear, polynomial, sigmoid and radial basis functions (RBF) have also been assessed in this paper.Abbreviations: AARE: Average absolute relative error; AI: Artificial intelligence; ANN: Artificial neural networks; GEP: Gene expression programming; MRE: Mean relative error; RMSE: Root mean square error; SD: Standard deviation; SVM: Support vector machines; SVR: Support vector regression