Artificial neural network modelling for Casson hybrid nanofluid flow with entropy generation analysis
摘要
In this study, a combination of a numerical solver and an artificial neural network (ANN) is employed to analyze the hybrid nanomaterial flow of Casson fluid. A feedforward neural network is trained using the Levenberg-Marquardt backpropagation algorithm to model the solution for various parameter variations based on a reference dataset generated by the numerical solver. Under the influence of viscous dissipation and heat source/sink, the flow characteristics, entropy generation, and heat transmission properties are investigated through a vertical channel. We utilized appropriate transformations to reduce the complexity of the fully developed Casson fluid flow’s governing equations from a dimensional system of non-linear ODEs to a simple non-dimensional system. This study offers a conceptual understanding of engineering and biological applications, particularly in analyzing blood samples, processing pharmaceutical fluids, administering drug delivery via blood flow, and refining crude oil.