ABSTRACT

This work aims to estimate the experimental heat transfer coefficients of a circular channel using artificial neural network. The experiments are carried out at a forced turbulent flow regime of 10,000 < Re <50,000. The obtained experimental Nusselt numbers are compared using the ANN (Artificial Neural Network). In the developed ANN structure are showed mean square error (MSE), average relative deviation (ARD %), and correlation coefficient (R²) in modeling of overall experimental datasets of Nusselt number. As a result, it is observed that the heat transfer correlation predicted by ANN are sufficiently consistent with the experimental results.

摘要

这项工作旨在使用人工神经网络估计圆形通道的实验传热系数。实验在 10,000 < Re < 50,000 的强制湍流状态下进行。使用 ANN（人工神经网络）比较获得的实验 Nusselt 数。在开发的 ANN 结构中，在对 Nusselt 数的整体实验数据集进行建模时显示了均方误差 (MSE)、平均相对偏差 (ARD %) 和相关系数 (R ² )。结果，可以看出，人工神经网络预测的传热相关性与实验结果完全一致。