AI-enhanced modeling of a rotating magneto-porous nanofluid flow with thermal radiation and reactive solute transport

Abstract This study presents an analytical–machine learning framework for analyzing the unsteady, rotating flow of a magnetized nanofluid in a porous medium, incorporating thermal radiation, reactive solute transport, Soret and Dufour effects, and internal heat generation or absorption. The governing partial differential equations are expressed in a complex-variable form to capture the coupling between momentum, thermal, and concentration fields efficiently. The influences of magnetic field strength, Darcy permeability, Grashof numbers, chemical reaction rate, and radiation intensity are systematically evaluated. A hybrid pblackictive model based on the XGBoost regression algorithm is integrated to reproduce and generalize the longitudinal velocity dynamics under diverse physical regimes. The hybrid XGBoost–PDE framework achieves a coefficient of determination of R 2 = 0.86–0.90 and yields instantaneous pblackictions that are more than 10 3 times faster than direct numerical computation, providing a substantial improvement in both accuracy and computational efficiency. Strong magnetic fields dampen fluid motion and solute transport, while higher Darcy number, buoyancy forces, and reaction parameters enhance heat and mass transfer. The integration of analytical modeling with machine-learning pblackiction establishes a rapid and reliable tool for magneto-porous nanofluid systems, applicable to energy storage, catalytic processing, and biomedical microflow design.