ABSTRACT In this paper, an artificial neural network was utilized to examined the MHD flow of a Williamson nanofluid over a stretching porous sheet. Moreover, the effects of heat generation chemical reactions Brownian motion and thermophoresis is taken into account. The system of nonlinear PDEs, are converted into system of ODEs, with the help of pertinent similarity transformation. MATLAB bvp4c technique is used to solve the simplified ODEs numerically. The artificial neural network Levenberg–Marquardt backpropagation algorithm (ANN‐LMBPA) was used to study the model. Numerical dataset is generated from MATLAB bvp4c solver, whereas, dividing the dataset into three sets, 70%, 15%, and 15% (training, testing and validation). Graphical representations including, mean square error, training state, fit analysis, error histogram and regression analysis justify the proposed method showing best result. The outcomes illustrate that, velocity profile is decreases with increasing the Darcy number, Williamson, and magnetic parameter. The temperature and concentration profile are increases with increasing the Heat source (sink) parameter, thermophoresis parameter, ratio parameter and Darcy number. Also, the concentration profile decreases with increasing the Lewis number, chemical reaction parameter. Numerical results for skin friction and Nusselt number have been analyzed. The ANN model demonstrated exceptional accuracy, with error values ranging from , and regression values close to 1. The model validation shows strong agreement against existing literature, and confirming the correctness of the model.
Ullah et al. (Wed,) studied this question.