Artificial neural network analysis of MHD Maxwell nanofluid flow over a porous medium in presence of Joule heating and nonlinear radiation effects

This study utilized a novel Artificial Neural Network methodology to examine the magnetohydrodynamic flow of a Maxwell nanofluid over a stretching surface situated in a porous medium, accounting for various physical phenomena including electromagnetic forces, nonlinear thermal radiation, heat generation/absorption, viscous dissipation, and Joule heating. The effects of Brownian motion and thermophoresis on the temperature and concentration fields of the nanofluid were analyzed. We have utilized the local non-similarity methodology to formulate the boundary layer equations to the second level of truncation for the purpose of calculating the solutions of non-similar boundary layer equations. The system is solved using an innovative Levenberg Marquardt Backpropagation neural network. The Artificial Neural Network based model surpassed conventional numerical methods by delivering accurate solutions for velocity, temperature, and nanoparticle concentration distributions across diverse physical parameters, thereby advancing accuracy and efficiency. The investigation demonstrated that electromagnetic fields, Darcy-Forchheimer effects, and thermal radiation substantially influenced flow dynamics and heat transfer rates. The best validation performance ranges from E−09 to E−10.The value of mu and gradient lie between E−08-E−09. The Error histogram ranges from E−05−E−07. The total absolute error is within the range of E−03 to E−10. This Artificial Neural Network based approach has shown significant promise in addressing intricate fluid dynamics challenges, providing expedited computations and enhanced accuracy relative to traditional methods, hence facilitating advanced modelling of nanofluid flows in various heat management systems. The porosity parameter, thermal radiation parameter, heat generation/absorption parameter and ratio temperature all have a direct relation with the temperature profile. A higher value of thermophoresis parameter results in an increased concentration profile, while concentration declines as the Brownian motion