Predicting the density of explosives by artificial neural network

摘要：

　　The fully optimized geometries and the vibrational analysis of 15 explosive molecules have been calculated by using the density functional theory(DFT) method at the B3LYP/6-31G*level.The quantum-chemical parameters calculated have been analyzed.The results indicate that the average of polarizability, The average of quadrupole moment and dipole moment all have certain correlation with density.The prediction model of density is created by ANN.The results from the model indicated that, the most relative error is 4.62% and the least is 0.47%.It was showed that the ANN is feasible to predict the properties of explosives.

关键词： artificial neural network quantum-chemical parameters density

作者: WANG Guodong LIU Guoqing GAO Xingyong DING Yukui

作者单位: Ammunition Engineering Department Ordnance Engineering College ShiJiaZhuang,China

会议类型: 国际会议

会议名称: 2013 International Autumn Seminar on Propellants, Explosives and Pyrotechnics(2013国际推进剂、炸药、烟火技术秋季研讨会)

会议地点: 成都

会议语种:英文

页码: 494-497

在线出版日期: 2013-09-24（万方平台首次上网日期，不代表论文的发表时间）