Intelligent computing Levenberg Marquardt approach for entropy optimized single-phase comparative study of second grade nanofluidic system

Abstract

In the presented research study, the novel intelligence based numerical computation by artificial neural networks backpropagated with Levenberg Marquardt method (ANN-BLMM) has been designed for the comparative study of second grade nanofluid flow model (SG-NFM) and examined the effects of parameters of interest associated with the proposed fluid flow system on its velocity and temperature profiles. The designed SG-NFM initially represented by system of PDEs which can be converted into system of non-linear ODEs through the subsequent corresponding transformation. The reference dataset for the SG-NFM is obtained by state of the art Adams numerical method in Mathematica Software for the different scenarios of SG-NFM by variation of unsteadiness parameter, parameters of velocity slip, Biot number, porous media parameter, relaxation time, thermal radiation, volume fraction of nanoparticles and suction/injection. The approximated solutions are interpreted for designed SG-NFM by testing, training and validation process of ANN-BLMM. Moreover, the comparative studies and performance analysis of ANN-BLMM are validated through regression analysis, histogram studies, correlation index and results of MSE. Furthermore, the effects on skin friction, Nusselt and number entropy generation are also analyzed.