NEURAL NETWORKS PARADIGM FOR ROTATING FLOW OF JEFFREY FLUID WITH HEAT TRANSFER BETWEEN TWO PARALLEL HORIZONTAL PLATES

Abstract

The current research explores the analysis of heat transfer in the rotating Jeffrey fluid between two parallel plates, employing the BLMS-ANN approach based on the back propagation Levenberg-Marquardt scheme. The fluid-flow is initially described by a system of PDE, which is subsequently transformed into a system of ODE through appropriate correspondence transformations and boundary conditions. Eventually, the equations are rendered dimensionless using a boundary-layer approximation. By changing parameters including the radiation parameter, Rd, Deborah number, lambda, viscosity parameter, R, Prandtl number, Reynolds number, and Rotation parameter, kr, using the differential transform method, a group of data for suggested BLMS-ANN is constructed for several cases. To assess the predicted outcomes for specific scenarios, the BLMS-ANN methodology undergoes testing validation and training. Subsequently, the proposed model is scrutinized for confirmation. The validity of the suggested BLMS-ANN approach is confirmed through regression analysis, examination of mean square error, and histogram studies. The suggested method differs from both the proposed and reference results with an accuracy level between.