Hybrid Neural-Numerical Modeling of Bio-convective Casson MHD Nanofluid Flow with Thermophoresis and Activation Energy

The paper presents a hybrid framework that integrates three distinct approaches through analytical techniques and numerical methods and Artificial Neural Network (ANN) techniques to investigate how Casson magnetohydrodynamic (MHD) nanofluids behave under the combined effects of thermophoretic forces and activation energy and density variations. The physical model describes mixed convective flow over an exponentially stretching surface through its implementation of nonlinear dissipative processes to show how velocity and temperature and nanoparticle concentration and microorganism density fields behave. The study solves nonlinear partial differential equations through dimensionless transformation using similarity transformations and combines it with semi-analytical methods and ANN-based predictive modeling. The neural network learns to recognize nonlinear connections between controlling parameters and flow and thermal characteristics which improves prediction accuracy. The analysis shows that three factors-including magnetic field strength, Casson fluid parameter, and nonlinear convection parameter-lead to decreased velocity distribution, while thermophoresis and Brownian motion create significant increases in both thermal profiling and nanoparticle spreading abilities. The alteration of activation energy and density parameters leads to different effects on the distribution patterns of microorganism density and concentration levels. The connection between semi-analytical solutions and ANN-based predictions demonstrates how the system functions correctly while the hybrid modeling framework operates at high efficiency. The results provide essential knowledge that can enhance heat and mass transfer performance for biomedical cooling devices, nextgeneration energy systems, and industrial applications.