Statistical Analysis and Accurate Prediction of Thermophysical Properties of ZnO-MWCNT/EG-Water Hybrid Nanofluid Using Several Artificial Intelligence Methods

Abstract

This paper presents a statistical analysis and modeling of the thermophysical properties of ZnO-MWCNT/EG-water hybrid nanofluid using three artificial intelligence models, including multilayer perceptron neural network, radial basis function neural networks, and least square support vector machine (LSSVM). The thermal conductivity of the nanofluid was modeled using experimental data, and statistical parameters such as R-squared (R²), average absolute relative deviation (AARD %), root mean squared error, and standard deviation were employed to investigate the accuracy of the proposed models. The R² values of 0.9926, 0.9951, and 0.9866 and AARD% values of 0.4996%, 0.3532%, and 0.6013% show the accuracy of the models for respective MLP, RBF, and LSSVM models. Among these models, the RBF model shows the best accuracy. The study demonstrates the potential of artificial intelligence methods in predicting the thermophysical properties of nanofluids, which can help minimize experimental time and cost for future work.