Predicting Predicting Predicting Predicting T TT Thermal hermal hermal hermal C CC Conductivity of onductivity of onductivity of onductivity of GrapheneNanofluid with Artificial Neural Networks Multilayer Perceptron

The purpose of this study is to model and predict the thermal conductivity of graphenenano fluid through the artificial neural network of multilayer perceptron. Nano Fluid temperature, volumetric fractions and thermal conductivity of nanoparticles are considered as network inputs. According to the experimental data of previous experiments on thermal conductivity of graphene nanoparticles at the temperature of 25 to 50 ° C and the volume fraction of 0.055-0.066, the network performance tests were performed.In order to evaluate the accuracy of the model in predicting the thermal conductivity of the nanofluid, the root mean square error indices, detection coefficient and absolute error percentage are used. These values are0.04 W/ mK, 99% and 0.26% , respectively.The results of the indicators show the accuracy and reliability of the proposed model in comparison with experimental results and theoretical models.