Analysis of thermal diffusion and diffusion thermo effects under the condition of pollutant concentration and bio-convection: An artificial neural network approach

The current research explores the influence of bio-convection, thermophoresis and Brownian motion (TBM) on the Casson liquid flow across a stretchable surface subjected to porous media. Additionally, the consequence of pollutant concentration, heat source/sink, melting effect, and magnetic field on the fluid motion is considered in the analysis. To evaluate the fluid flow's mass and heat transport characteristics, the Soret and Dufour effects are taken into account. Using the proper similarity transformations, the governing partial differential equations (PDEs) are converted into dimensionless ordinary differential equations (ODEs). The resultant ODEs are solved with the aid of Runge Kutta Fehlberg’s fourth-fifth order (RKF-45) approach. Further, the Levenberg-Marquardt artificial neural network (LM-ANN) is employed to assess the heat transport rate for several parameters. Additionally, the values obtained by RKF-45 method are compared with the outcomes of LM-ANN approach. The graphs illustrate the effect of numerous parameters on the temperature, concentration, microorganism, and velocity profiles. The upsurge in the magnetic and porosity parameters diminishes the velocity profile. The rise in the Dufour number and heat source/sink parameter improves the thermal profile. The rise in the values of the Soret number and the pollutant external source parameter increases the concentration profile.