A Novel Design Ricker Wavelet Neural Networks for Heat Transfer in Maxwell Fluid Boundary Layer Flow with Viscous Dissipation over a Porous Stretchable Sheet

Abstract

The current research is a revolution in the field of neural computation as a quite new stochastic technique based on Ricker wavelet neural networks (RWNNs) is developed to analyze the Maxwell fluid (Max-F) boundary layer flow (BLF) with heat and mass transfer effects over an elongating surface. The global and local search solvers used with RWNNs are genetic algorithms (GAs) and sequential quadratic programming (SQP) respectively to design a new algorithm i.e. RWNNs-GASQP. The transformed nonlinear system of ODEs is acquired using the physical model represented by the flow and then solved using RWNNs-GASQP solver. The obtained numerical form results are successfully compared with reference results acquired through the Adams technique. The accuracy, convergence and effectiveness of the designed solver are identified using numerous statistical and performance analyses.