DETERMINATION WITH ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM AND ARTIFICIAL NEURAL NETWORK APPROACHES OF THERMODYNAMIC PROPERTIES OF Al2O3-WATER NANOFLUID

Nanofluids are used to increase heat transfer performance and have an important application area in energy systems. In this study, thermodynamic properties of Al<sub>2</sub>O<sub>3</sub>-water nanofluid, such as thermal conductivity and dynamic viscosity, were estimated using adaptive neuro-fuzzy inference system (ANFIS) and artificial neural networks (ANN) approaches. Data obtained from five different correlations were used to estimate the properties of Al<sub>2</sub>O<sub>3</sub>-water nanofluid, such as dynamic viscosity (DV) and thermal conductivity (TC). The performance of the models was evaluated in terms of accuracy by comparing them with the correlation results. The results showed that both methods can estimate the nanofluid properties with high accuracy, but ANFIS exhibited superior performance in some cases. The R<sup>2</sup> value for the dynamic viscosity values of the Al<sub>2</sub>O<sub>3</sub>-water nanofluid was found to be 0.99965. The R<sup>2</sup> value for the thermal conductivity values of the Al<sub>2</sub>O<sub>3</sub>-water nanofluid was found to be 0.9994. For the ANFIS model, while the MAPE value for thermal conductivity is 0.94006, the R<sup>2</sup> value is 0.9996. For viscosity, the MAPE value is 0.87537, and the R<sup>2</sup> value is 0.9997. This study has shown that ANFIS and ANN methods can be used successfully to accurately predict the thermodynamic properties of Al<sub>2</sub>O<sub>3</sub>-water nanofluid instead of complex calculations, and provided a new perspective for more efficient designs in energy systems.