关于含能材料硝胺撞击感度的研究

Neural Networks Study on the Correlation between Impact Sensitivity and Molecular Structures for Nitramines

作者在利用Gaussian 98DFTB3P86/6-31G**水平研究了36种炸药分子的撞击感度与其分子结构关系并取得较好结果的基础上,又对33种硝胺炸药分子,仍采用人工神经网络方法进一步研究了撞击感度与分子结构特征量的关联关系.结果表明,所实验的硝胺其含特征量ΔE(原子化能)的输入方案预测结果最理想,即ΔE与撞击感度的关联度最强,说明ΔE可作为预测其撞击感度的标志性指标.

In the authors＇ former research, backpropagation neural networks were used to study about the correlation between impact sensitivity and molecular properties of 36 explosive molecules and good results were obtained. Based on this, the authors again utilize the above method to study the correlation order between impact sensitivity and 33 nitramine explosives the molecular properties via B3P86/6-31G The training and testing results show that the input vector with the descriptor AE can obtain the relatively better outcomes. It further indicates that with the same net structure and training parameters, molecular descriptor AE has the strongest correlation with impact sensitivity of explosives, which indicates that AE can be a symbolic index to predict the impact sensitivity of nitramines.