Experimental investigation and modeling of thermal radiative properties of f-CNTs nanofluid by artificial neural network with Levenberg–Marquardt algorithm

• Published in: International communications in heat and mass transfer Vol. 78; pp. 224 - 230
• Main Authors: Vakili, M., Karami, M., Delfani, S., Khosrojerdi, S.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2016
• Subjects: Artificial neural networks; Nanofluids; Prediction; Thermal radiative properties; Artificial neural networks; Prediction; Nanofluids; Thermal radiative properties
• ISSN: 0735-1933; 1879-0178
• DOI: 10.1016/j.icheatmasstransfer.2016.09.011

The aim of this study is to predict the thermal radiative properties such as transmittance and extinction coefficient of nanofluids containing carbon nanotubes against sun radiation with the help of a multilayer artificial neural network of perceptron. To check the network performance, the optical properties of nanofluids were measured with the help of an experimental method in volume fractions of 5, 10, 25, 50, 100 and 150ppm at radiation wavelengths of 200 to 1400. The number of measured data was 798; 560 were chosen for training and the rest was for testing and validating the network. To check the accuracy of the model in predicting the optical properties of nanofluids, the indicator root mean square error (RSME), the mean absolute percentage error (MAPE), coefficient of determination (R2) and mean bias error (MBE) were used; these amounts were in the order of 0.019, 0.009%, 99.8% and 6.94×10−5. Hence, the results from the indicators show a highly accurate and reliable model compared with the experimental results.