Response surface methodology optimization of dynamical solutions of Lie group analysis for nonlinear radiated magnetized unsteady wedge: Machine learning approach (gradient descent)

Abstract

When a consistent transverse magnetic field and heat radiation are present, the flow of boundary layer, Over porous wedge nanofluids, hybrid nanofluids, and ternary hybrid nanofluids have been studied., Water as base fluid, Hybrid nanofluid, Ternary Hybrid nanofluid, and nanofluid cases containing Case-1 Polyethylene Glycol-Water + AA7072, Case-2 Zirconium oxide + AA7072 + Polyethylene Glycol-Water Case-3 Magnesium oxide + Polyethylene Glycol-Water + AA7072 + Zirconium oxide is taken into consideration. Runge-Kutta (4th order) with the Shooting technique is used to solve the governing equations expressed in terms of Odes. For various values of the relevant parameters, the approximate relationship between temperature, velocity, rate of heat transfer, and shear stress at the wedge is depicted visually. It is found that the Nusselt number transfer rate is more in Ternary hybrid nanofluids than Hybrid and Nanofluids and the Skin friction rate is more in Hybrid nanofluids than in Ternary and Nanofluids. Table 9 shows the comparative study with recently published paper numerically having a good agreement of results. RSM method is useful to find the optimization conditions values based on the key factors that impact the Response Function. Simple linear regression machine learning with the Gradient descent method has been applied to some of the dimensionless parameters this method predicts the truth values accurately.

Keywords

• Nanofluid
• Response surface method
• Radiation
• Hybrid nanofluid
• Ternary hybrid nanofluid
• Lie group transformations