Mathematical and artificial neural network modeling to predict the heat transfer of mixed convective electroosmotic nanofluid flow with Helmholtz-Smoluchowski velocity and multiple slip effects: An application of soft computing

The growing popularity of artificial neural networks (ANNs) is due to their exceptional capability in handling complex and highly nonlinear mathematical problems. ANNs provide a flexible computational framework that is extremely useful in complex fields such as biological computation, fluid dynamics, and biotechnology. Consequently, this study utilizes machine learning techniques to examine the heat transfer in mixed convective Electroosmotic Nanofluid flow, considering Helmholtz-Smoluchowski velocity and multiple slip effects on an exponentially stretching surface. Different influential factors, including, thermal radiation, viscous dissipation, joule heating, linear thermal heat source/sink and exponential thermal heat source/sink are considered in this investigation. In the current study, graphene oxide and magnesium oxide are used as nanomaterials, with ethylene glycol serving as the base fluid. The bvp4c solver in Matlab is employed to solve nonlinear versions of established mathematical problems, including those related to the skin friction coefficient, heat transfer rates, velocity, and temperature. The artificial neural network model is structured to handle several tasks: data selection, network construction, network training, and evaluating performance employing the mean square error metric. The significance of parameter on subjective profiles is demonstrated though tables and graphical representation. The Numerical results and artificial neural network results have been compared. The results showed that there is excellent validation between numerical results and artificial neural network results.