Fuzzy neural network approach for heat transfer optimization in Williamson hybrid nanofluid

Abstract This research examines the rheological and thermal properties of a Williamson hybrid (Al 2 O 3 and Cu/EO) nanofluid flow across Riga wedge under influence of heat source, velocity slip, stagnation point, and thermal effect. This study combines fuzzy logic and an artificial neural network approach to model dynamic behavior of the hybrid nanofluid that involves complicated interactions and predicts its performance under different conditions. Triangular fuzzy number (TFN) is used to analyze the importance of fuzzy volume fractions of nanomaterials. TFN is used to compare the thermal properties of nanofluids including Cu/EO, Al 2 O 3 /EO, and hybrid nanofluid (Al 2 O 3 and Cu/EO), using triangular fuzzy number the concentration of nanoparticles swaying the thermophysical properties are considered as uncertain parameters. The ANN framework integrates two optimization techniques: LMS (Levenberg‐Marquardt scheme) used for quick training and BRS (Bayesian regularization scheme) used for improved generalization, is implemented to investigate the physical properties and initial data. Results indicate that the hybrid nanofluid improves heat transfer efficiency compared to conventional fluids, making it a promising candidate for advanced thermal management systems. Best validation performance of LMS and BRS for skin friction and Nusselt number is 3.5665e‐07 and 4.0682e‐10 at epoch 272 and 63, respectively. Potential applications include cooling technologies, heat exchangers, electronic device cooling, and industrial thermal regulation. These findings provide a robust foundation for optimizing nanofluid formulations in energy‐efficient engineering applications.