Neuro-evolution heuristics for Prandtl-Eyring nanofluid flow with homogenous/heterogeneous reaction across a linearly heated stretched sheet

Abstract

The present research-oriented study provides the application of a renowned numerical-based neuro-evolution heuristic technique by the utilization of feed-forward neural networks (FFNNs) trained with a hybrid composition of genetic algorithms (GAs) and sequential quadratic programming (SQP) i.e. FFNNs-GAs-SQP to provide a descriptive analysis on the dynamics of Prandtl-Eyring nanofluid flow over the stretchable sheet under the influence of heat and viscous dissipation. FFNNs-GAs-SQP approach is used to calculate numerical results based on the variation of temperature-difference parameter, material-parameter, stretching-parameter, heterogeneous-reactionparameter along with Biot-number, Prandtlnumber as well as the Schmidt-number. The effect of these parameters on velocity profile, temperature and nanofluid concentration is expressed in form of tables/graphs. Furthermore, uplifting the values of the temperature difference parameter and the stretching parameter enhances the nanofluid velocity. The larger values of the material parameter and Prandtl number diminish the temperature of nanofluid but this effect is reversed in the case of the Biot number and temperature difference parameter. It is also observed that a boost in the values of heterogeneous reaction parameter, Schmidt number and temperature difference parameter minimize the nanofluid concentration however this effect is reversed in Biot number and Prandtl number cases.