Thermal optimization of electrically conducting Ag/Al2O3 Ostwald fluid across solar surface using ANN for solar thermal application

This study develops an artificial neural network framework to model the thermo-hydrodynamic behavior of an electrically conducting Ostwald–de Waele hybrid nanofluid over a solar-heated surface under mixed convection, nonlinear thermal radiation, Joule heating, and dissipative effects. The governing pressure-dependent, time-varying flow equations are solved using a finite difference method, and the resulting data are employed to train and validate the artificial neural network model. The analysis shows that velocity increases with a higher Grashof number and temperature decreases with the radiation parameter. The ANN demonstrates excellent predictive accuracy, with mean squared error values on the order of regression coefficients above 0.996, and absolute errors confined within across training, testing, and validation phases. These findings confirm the strong correlation between numerical and ANN-based predictions, validating the robustness of the proposed framework. This work is significant for solar thermal applications by accurately predicting non-Newtonian nanofluid behavior, with novelty in coupling ANN and finite difference simulations to deliver an efficient, precise, and practical tool for optimizing next-generation collectors.