A new model to predict the densities of nanofluids using statistical mechanics and artificial intelligent plus principal component analysis

In this work, some thermodynamic properties of nanofluids such as Sb_2O_5;SnO_2/(EG + H_2O), ZnO/(EG + H_2O),Al_2O_3/(EG + H_2O), ZnO/(PEG + H_2O), ZnO/PEG, and TiO_2/EG were estimated from the extended Tao–Mason equation of state, together with the Pak and Cho equation in various temperature, pressure, and volume fractions.The equations of state using minimum input data and density at room temperature as scaling constants, were developed to estimated densities of the above mentioned nanofluids. Furthermore, the artificial neural network plus principal component analysis(PCA) technique(with 20 neuron in hidden layer) was performed over the whole range of available conditions. The AADs of the calculated molar densities of all nanofluids using the EOS and ANN at various temperatures and volume fractions are 1.11% and 0.48%, respectively. In addition, the heat capacity and isentropic compressibility of the above mentioned nanofluids are predicted using obtained densities of EOS with the Pak and Cho equation.

基于ANN的动态系统状态方程辨识建模仿真

对系统辨识原理、基于神经网络(ANN)的动态系统辨识进行了分析,针对动态系统辨识模型描述的复杂性,为简化ANN辨识建模的输入／输出关系的表达,提高算法的简洁性,采用了状态方程辨识模型,并给出了基于ANN的动态系统状态方程辨识模型。为比较分析不同网络结构的辨识建模效果及网络模型泛化能力,针对三种不同网络结构方案进行了辨识建模仿真研究。仿真结果显示,基于ANN的动态系统状态方程模型的辨识建模是有效的,并且简单合理的网络结构方案,可提高网络辨识模型的泛化能力。