Design of backpropagation networks for bioconvection model in transverse transportation of rheological fluid involving Lorentz force interaction and gyrotactic microorganisms

Abstract Exploration and exploitation of artificial intelligence (AI) techniques have growing interest for the research community investigating in engineering and technological fields to provide improved efficiencies and augmented human abilities in daily live operations, business strategies and society evolution. A novel application of AI based backpropagating networks (BPNs) was presented for bioconvection model in transverse transportation of rheological fluid involving Lorentz force interaction and gyrotactic microorganisms. The governing nonlinear PDEs for bioconvection rheological fluidic system (BRFS) was reduced to nonlinear system of ODEs by competency of similarity adjustments. A reference data of designed BPNs was constructed for variants of BRFS representing scenarios for thermophoresis parameter, Brownian motion, Prandtl numbers, magnetic variables, squeezing and Lewis numbers by applying the Adams numerical solver. The said data were segmented arbitrary in training, testing, and validation sets to execute BPNs to calculate the approximate solutions for variants of BRFS and comparison with standard solution to validate the consistent accuracy. The worthy performance of AI based BPNs was additionally certified by learning curve on MSE based fitness, histograms and regression metrics.