Intelligent computing architecture for the analysis of Williamson Magneto hydrodynamic fluid model and heat transfer through a permeable stretching surface

Abstract

The Williamson fluid model's flow field properties have drawn researchers' attention in recent years. The objective of this research is to analyze the Williamson Magneto Hydrodynamic Fluid Model's unsteady flow field properties by applying the algorithm of Bayesian regularization with an artificial neural network of backpropagation. A magnetic field interacts with the nanosized particles, and the stretched porous surface causes the fluid to flow. Under similarity transformation, the non-linear partial differential equations of the Williamson fluid model are turned into ordinary differential equations by varying the thermophoresis parameter, the mass transfer parameter, the unsteadiness parameter and the Lewis number. A set of data is generated in Matlab for various Williamson fluid model scenarios using the Adam numerical technique for the proposed Bayesian regularization algorithm. The solution and error analysis plots are evaluated using the reference dataset. Furthermore, the performance of the algorithm of Bayesian regularization is clarified by mean square error results, error analysis plots, regression, and error histograms. The solution of the Williamson fluid model is exploited by the validation, training, and testing procedures. The validity of the obtained results is confirmed by the mean square error up to the level of E-13.