Machine learning analysis for radiative boger nanoliquid with porous medium and thermophoresis effects

This article describes an intelligent Levenberg Marquardt scheme using artificial neural networks (ILMS-ANNs) to evaluate the effect of thermal radiation on Boger nanofluid flow and reduce entropy formation. The effect of thermophoresis and Brownian motion is also discussed in the current study. This model has significant applications in sophisticated temperature and fluid engineering systems where exact heat transportation and irreversibility are required. The Boger nanofluid model with Brownian motion and thermophoresis, optimized by artificial neural networks, can be used to design high-efficiency heat exchangers, micro- and nano-scale cooling systems, polymer processing, and biomedical devices like targeted drug delivery and hyperthermia treatments. By assessing entropy generation, the model aids in identifying optimal operating parameters that reduce energy losses and improve system efficiency, making it particularly effective in renewable energy systems, electronic cooling, and industrial processes involving viscoelastic fluids and nanoparticles. The numerical solutions have been obtained using the MATLAB‘s bvp4c method and the shooting method. The approximate responses for multiple scenarios have been identified using ILMS-ANN‘s testing, training, and validation methods to ensure the suggested model‘s accuracy. Subsequently, regression analysis, MSE, and histogram assessments are used to study and show the suggested ILMS-ANN. The rate of heat transfer increases as the thermal radiation parameter‘s values increase.