An artificial intelligence-based prediction way to describe flowing a Newtonian liquid/gas on a permeable flat surface

Abstract

The purpose of this study is to utilize artificial neural network (ANN), as one of the most powerful artificial intelligence methods, for modeling stream function (f) and the dimensionless temperature (θ) for the considered problem. The problem that is investigated here is flowing a Newtonian fluid on a permeable flat surface. The Homotopy Perturbation Method (HPM) recently developed by the authors for this problem is utilized to provide enough number of the input data. The best ANN is found for each of the two indicated outputs. Then, the best ANN model for each output is utilized to investigate the impact of changing the similarity variable in the range 0.0 to 10.0 on prediction error of the two mentioned outputs. Four values for porosity, which are 0.2, 0.5, 0.8, and 1.0, are investigated. According to the findings, an almost quadratic relation for changes prediction error of f as a function of η is seen, whereas after a sudden drop, the error in prediction of θ declines linearly. Moreover, for the whole range, and for both outputs, the error remains in an acceptable range, which verifies the good accuracy of ANN.