Modeling the effects of triple diffusivity on magnetohydrodynamic periodically accelerating flow in porous channels via artificial neural networks: Biomedical engineering applications

Peristaltic motion in magnetohydrodynamic (MHD) flow has gained a lot of attention because of its uses in industrial fluid transport systems. In this study, the impact of triple diffusivity on Casson fluid flow through a permeable channel is examined using a computer method. The lubrication approximation is used to simplify the governing nonlinear equations, which are then numerically solved using a MATLAB-based approach. The results not only validate the numerical technique but also show strong agreement with the existing literature. The fluid's temperature and concentration profiles are greatly impacted by changes in diffusivity-related parameters, such as the Grashof and Brinkman numbers. Furthermore, it is discovered that the size of the trapped bolus is significantly influenced by the porosity parameter. It is demonstrated that the interaction between thermophoresis and Brownian motion greatly affects the movement of chemical species. The Levenberg–Marquardt backpropagation neural network (LMBP-NN) technique is uniquely implemented in MATLAB to provide precise numerical answers. Across all calculated profiles, the neural model exhibits good convergence, low error levels, and high regression accuracy. These discoveries advance our understanding of the peristaltic transport of MHD in porous media and have implications in real-world biomedical applications such as the passage of ova through fallopian tubes, the transport of bile in biliary ducts, and other fields of biomedical engineering.