Neural network analysis of bioconvection effects on heat and mass transfer in Non-Newtonian chemically reactive nanofluids

Using artificial neural networks, this study sought to investigate the magneto Williamson two-phase nanofluid, taking into account chemical reactions and the motion of gyrotactic motile microorganisms. Fluid flow behavior is influenced by chemical reactions, magnetic effects, Brownian motion, and thermophoresis, according to the study. Thermal transmission is enhanced in non-Newtonian fluids as a result of their propensity to thin under shear, increased turbulence, and superior convective heat transfer. As a result of the fluid's increased thermal conductivity, the incorporation of nanoparticles enhances heat conduction. Additionally, epidermis friction, Nusselt and Sherwood numbers, and the quantity of motile microorganisms were assessed in the study. The overall Absolute Errors lies in the range of 10−2to10−10.The mean squared error generated by Neural Networks lies in the range of 10−02−10−10, and 10−02−10−09 respectively. Suction or injection parameter and Prandtl number have an inverse relation with fluid temperature, while Thermophoretic parameter have a direct relation. Thermophoretic parameter, Schmidt number and suction or injection parameter have an inverse relation with the concentration of nanofluid and gyrotactic microorganisms' density, while micro-organisms density have a direct relation with the microorganisms. Engineering and medicine have utilized bioconvection, a process involving heat transfer and microorganism motion, in the development of nanomedicine, pharmacokinetics, drug delivery, and biosensors, among others. Solvers utilizing stochastic numerical computing include nonlinear networks, atomistic physics, thermodynamics, astrometry, fluid mechanics, nanobiology. As a result, variant scenarios are then tested, trained, and validated, in order to prove its accuracy.