Surface tension gradient effects on hybrid nanofluid with activation energy using machine learning technique: Stefan blowing phenomena.

The combined effects of Stefan blowing and Arrhenius activation energy bioconvection flow of hybrid nanofluid with microorganisms and thermophoresis create a highly nonlinear transport problem that necessitates precise prediction modeling. The reference dataset is created by using similarity transformations to convert the governing PDEs into a system of nonlinear ODEs, which are then resolved using the HAM (Homotopy Analysis Method). The obtained numerical data are then used to validate, train, and test a Bayesian Regularization neural network (BR-NN) using a feed forward single layer architecture. The ANN's performance is assessed using EH (error histograms), regression analysis, MSE (mean squared error), and function fit curves, which show great agreement with the HAM solutions and high predicted accuracy. This model has uses in numerous advanced industrial and engineering processes where effective heat and mass transport are important. It can be utilized in cooling systems for electronic devices, solar thermal collectors, nuclear reactors, chemical processing equipment, and microfluidic devices, were hybrid nanofluid boost thermal performance. The model's incorporation of surface tension gradient (Marangoni) effects, Stefan blowing, activation energy, and thermophoretic particle deposition makes it appropriate for studying evaporation, coating technologies, drying processes, fuel cells, and biological transport phenomena.