Artificial neural network-based study of unsteady MHD radiative hybrid nanofluid flow with activation energy in a porous medium

This study investigates the unsteady heat and mass transfer characteristics of water-based hybrid nanofluid flow over a stretching surface embedded in a porous medium under mixed convection conditions. The hybrid nanofluid consists of titanium dioxide (TiO₂) and copper (Cu) nanoparticles dispersed in water. The analysis incorporates the effects of activation energy, thermal radiation, Brownian motion, and thermophoresis. The governing partial differential equations are transformed into ordinary differential equations using similarity transformations and solved numerically with a fourth-order Runge-Kutta method coupled with the shooting technique. The influence of key dimensionless parameters including magnetic field strength, Prandtl number, buoyancy ratio, unsteadiness parameter, thermal radiation, chemical reaction rate, and activation energy on velocity, temperature, and concentration profiles is examined. To improve predictive capability, an artificial neural network (ANN) is employed to estimate the skin friction coefficient using SCG, LM, CGB, and CGF algorithms. Performance evaluation through mean square error (MSE), regression plots, and error analysis indicates that the Levenberg–Marquardt (LM) algorithm provides superior accuracy, making it the most efficient method for predicting heat transfer characteristics. This hybrid numerical AI framework demonstrates strong potential for accurately modeling magnetohydrodynamic hybrid nanofluid transport phenomena. The findings contribute to improved design and thermal management strategies in advanced energy and engineering systems.