论文部分内容阅读
应用晶格能极小化技术计算了十九种骨架拓扑结构的高硅沸石的全硅骨架晶格能。采用一元线性回归分析、多元线性回归分析和人工神经网络误差反传算法（ＢＰ），将这十九种拓扑结构高硅沸石的全硅骨架晶格能计算值与它们的骨架拓扑结构参数相关。预测结果与计算晶格能符合良好，表明晶格能与配位序（Ｎ２～Ｎ４）具有优良的多元线性关系。利用多元线性回归方程对七种结构的高硅沸石的全硅骨架晶格能进行了预测。
The all-silicon framework lattice energy of 19 high-silica zeolites with framework topologies was calculated using the lattice energy minimization technique. Linear regression analysis, multivariate linear regression analysis and artificial neural network error back propagation algorithm (BP) were used to correlate the calculated values of the lattice energies of the all-silicon frameworks of the nineteen types of high-silica zeolites with their skeleton topology parameters. The predicted results are in good agreement with the calculated lattice, indicating that the lattice energy and the coordination (N2 ~ N4) have excellent multiple linear relationship. The multi-linear regression equation was used to predict the lattice energies of the all-silicon frameworks of seven types of high-silica zeolites.