论文部分内容阅读
以炸药分子的结构描述符和爆轰性能等参数,利用主成分分析(PCA)结合BP神经网络的方法,建立了炸药分子结构与爆速之间的定量关系预测模型,并对20种炸药的爆速进行了预测,其相对误差均低于9%。结果表明,所建立的模型较好地反映了炸药分子结构与爆速之间的关系,具有较高的预测精度。该方法为设计新型炸药分子时正确估算其爆速提供了一条新的途径。
Based on the structural descriptors and detonation performance parameters of explosives, PCA and BP neural network were used to establish a quantitative prediction model of explosive structure and detonation velocity. The detonation velocity of 20 explosives The prediction was made with relative errors less than 9%. The results show that the model established can well reflect the relationship between explosive molecular structure and detonation velocity and has high prediction accuracy. This method provides a new way to correctly estimate the detonation velocity of new explosive molecules.