Application of M5 tree regression, MARS, and artificial neural network methods to predict the Nusselt number and output temperature of CuO based nanofluid flows in a car radiator

• Published in: International communications in heat and mass transfer 2020-07, Vol.116, p.104667, Article 104667
• Main Authors: Kahani, Mostafa, Ghazvini, Mahyar, Mohseni-Gharyehsafa, Behnam, Ahmadi, Mohammad Hossein, Pourfarhang, Amin, Shokrgozar, Motahareh, Zeinali Heris, Saeed
• Format: Article
• Language: English
• Subjects: Artificial neural network; Car radiator; CuO nanoparticles; M5 tree regression; Multi-Variate adaptive regression splines (MARS); Nusselt number
• ISSN: 0735-1933; 1879-0178
• DOI: 10.1016/j.icheatmasstransfer.2020.104667

In the current study, CuO nanoparticles were dispersed in a mixture of Ethylene Glycol-Water (60/40 wt. %) to prepare stable nanofluid in different concentrations (0.05 − 0.8 vol. %). The samples were used as the coolant fluid in a specific car radiator to evaluate the thermal performance of nanofluid and base fluid in the system. Five different and novel Machine-learning methods were applied over experimental data to predict the Nusselt number and output temperature of the coolant in the system. These methods are M5 tree regression, Linear and Cubic Multi-Variate Adaptive Regression Splines (MARS), Radial Basis Function (RBF), and Artificial Neural Network-Levenberg Marquardt Algorithm (ANN-LMA). Although all studied methods show acceptable accuracy in predicting experimental data, the ANN-LMA method in output temperature modeling and the MARS-Linear method in Nusselt number modeling has more precision.