@article { author = {Azari, A. and Marhemati, S.}, title = {Model for Thermal Conductivity of Nanofluids Using a General Hybrid GMDH Neural Network Technique}, journal = {International Journal of Nanoscience and Nanotechnology}, volume = {11}, number = {2}, pages = {71-82}, year = {2015}, publisher = {Iranian Nanotechnology Society}, issn = {1735-7004}, eissn = {2423-5911}, doi = {}, abstract = {In this study, a model for estimating the NFs thermal conductivity by using a GMDH-PNN has been investigated. NFs thermal conductivity was modeled as a function of the nanoparticle size, temperature, nanoparticle volume fraction and the thermal conductivity of the base fluid and nanoparticles. For this purpose, the developed network contains 8 layers with 2 inputs in each layer and also training algorithms of least squares regression. The obtained results of the model have shown good accuracy of hybrid GMDH-PNN for estimating the thermal conductivity of NFs. The RMSE of the model for 24 systems containing 211data sets was achieved 0.0224. MAPE for training and validation data setswere3.58 and 3.2%, respectively. Also, the proposed hybrid GMDH-PNN model was compared with different models from the literature. The results showed that the developed model can successively correlate and predict the thermal conductivity of different groups of NFs. Moreover, a remarkable agreement for the model with the experimental data was achieved with respect to the other models from the literature.}, keywords = {Artificial neural network,GMDH-PNN model,Nanofluids,Thermal conductivity}, url = {https://www.ijnnonline.net/article_13470.html}, eprint = {https://www.ijnnonline.net/article_13470_8ed3229a8920e8c46168f4c5ed758bb0.pdf} }