Using a simple computational tool with a very high connection and the determining role of connections between neurons in identifying network function are the two similarities between natural and Artificial Neural Networks (ANNs). In this article, the very significant subjects of nanofluids efficiency and the thermal performance factor of these fluids operating as heat transfer have been investigated. To model the data in this article, ANNs are used. This modeling is presented for different $\varphi$and modeling results have been compared with experimental data for MgO-water nanofluids flow. The data relating to the efficiency of these nanofluids use complicated patterns, so a type of ANNs has been used with the ability to distinguish the number of neurons required for modeling and following the data pattern. To evaluate the accuracy of the modeling using ANN, the experimental data have been compared with ANN results in different volume fraction of nanoparticles ($\varphi )$. The results show that ANN has a better agreement with experimental data in estimating the data with a higher Reynolds number.

使用具有非常高连接的简单计算工具以及神经元之间连接在识别网络功能中的决定作用是自然和人工神经网络 (ANN) 之间的两个相似之处。在这篇文章中，研究了纳米流体效率和这些流体作为传热的热性能因素的非常重要的主题。为了对本文中的数据建模，使用了人工神经网络。这种建模是针对不同的$\phi$并将建模结果与 MgO-水纳米流体流动的实验数据进行了比较。与这些纳米流体的效率相关的数据使用复杂的模式，因此已经使用了一种 ANN，能够区分建模和遵循数据模式所需的神经元数量。为了评估使用 ANN 建模的准确性，将实验数据与 ANN 在不同纳米颗粒体积分数下的结果进行了比较（$\phi )$. 结果表明，人工神经网络在估计雷诺数较高的数据时与实验数据有较好的一致性。