Prediction and Comparative Analysis of Thermal Conductivity of Jatropha Oil-based Hybrid Nanofluid by Multivariable Regression and ANN

In the present study, a multivariable regression (MR) and artificial neural network (ANN) method was used to predict the thermal conductivity of Jatropha oil- based ZnO-Ag hybrid nanofluid. Firstly, the ZnO-Ag hybrid nanoparticles were synthesized and mixed in the jatropha oil to prepare various nanofluids at different volume concentrations (F) ranging from 0.05 to 0.20%. The stability and thermal conductivity of the prepared nanofluids were investigated. Wide ranges of temperature and volume concentration as input data were used to predict the output parameters as thermal conductivity. In comparison with the multivariable regression, the artificial neural network (ANN) models, predict thermal conductivity values that are remarkably similar to the experimental values by offering a lower mean square error. This approach also aids with the issue of guesswork in figuring out the neural network layer ' s hidden structure.