A machine learning approach to the prediction of transport and thermodynamic processes in multiphysics systems - heat transfer in a hybrid nanofluid flow in porous media

• An artificial intelligence based predictive models for estimations of transport and thermodynamics processes. This study dealt with a hybrid nanofluid SPF past cylinder embedded in porous media. The computational results were then used to train an tool developed through supervised learning methods. validated was subsequently estimate the behaviours temperature fields, Nusselt Bejan number and, shear stress over cylinder. Comprehensive analyses phenomena complex multiphysics systems are laborious computationally intensive. Yet, such often required during design thermal process equipment. As remedy, this paper puts forward novel approach prediction systems, offering significant reductions time cost. is on machine techniques that utilize data generated by fluid dynamics training purposes. physical system under investigation includes stagnation-point flow (Cu?Al 2 O 3 /Water) blunt object problem further involves mixed convection, entropy generation, local non-equilibrium non-linear radiation within medium. SVR (Support Machine Vector) model employed approximate velocity, temperature, shear-stress as well generation functions. Further, PSO meta-heuristic algorithm applied propose correlations stress. effects number, fields surface blunt-body frictional analysed wide range parameters. it shown allow quantitative evaluation contribution large variables makes combined (AI) most suitable