Multi-layer artificial neural network modeling of entropy generation on MHD stagnation point flow of Cross-nanofluid

Abstract

This research article explores the influence of the magnetic field and Joule heating on the 2D stagnation point flow of a Cross-nanofluid. Viscous dissipation, multiple slips, thermophoresis and Brownian motion are embedded in the current model. The artificial neural network (ANN) is one of the notable alternative approaches to solve fluid flow problems since it effectively reduces the processing time. The ANN model is best suited for predicting the data after training, validation and testing with existing data. ANN results are verified via feed-forward neural networks with the Levenberg-Marquard Scheme-based Backpropagation Technique (NN-BLMS). Data was collected for training, certification, and testing in the ANN model. To obtain this data, we solved the nonlinear coupled ordinary differential equations using the MATLAB software bvp4c. ANN model is used to select data, create and train a network, and evaluate its performance using mean square error and regression analysis. ANN data got a good agreement with the numerical data. Velocity, temperature, concentration, entropy generation rate and Bejan number outlines are revealed and deliberated for innumerable cases, namely outer convex, uniform thickness, and inner convex stretching sheet.