Neural Computing Approach to the Flow and Thermal Analysis of Ternary Hybrid Nanofluid Over Wedges

Abstract

The steady state motion of a ternary hybrid nanofluid across a  permeable wedge is investigated in the case when the wedge is stretching  or contracting. An applied magnetic field and the effect of radiation are  included in the analysis. An application of a similarity transformation to  the governing equations that describe the flow of ternary hybrid nanofluid  results in the conversion of those equations into ordinary differential equa
tions. All of these equations are solved numerically using the bvp4c solver  in MATLAB. Additionally, the flow of the ternary hybrid nanofluid is  governed by the nonlinear differential equations that are then solved by 
a hybrid approach combining artificial neural networks with Levenberg Marquardt backpropagation technique. The performance and hence the  comparison of algorithms are carried out using regression analysis, er
ror histograms, function fitting graphs, and mean squared error results. To  achieve the desired results, the considered parameters are varied systematically to be the stretching/shrinking parameter (λ), the magnetic parameter (M), the suction/injection parameter (S), the wedge angle (m), and the  radiation parameter (Rd). In terms of performance, the velocity and tem perature profiles for the proposed model demonstrated strong accuracy, as evidenced by the mean squared error values of [2.251 × 10−9, 5.5339 × 10−10, 1.2449 × 10−9, 1.7511 × 10−8, 3.1377 × 10−9]. According  to the investigation, solutions are found up to a particular suction and 
stretching/shrinking strength. An increase in the wedge angle parameter,  the critical value associated with the thermal response decreases. Additionally, the heat transfer rate provided by the ternary hybrid nanofluid is 
greater than those possessed by the conventional nanofluids. Moreover, an  increase in the radiation parameter enhances the heat transferring rate.