摘要
为研究嘧啶类含能化合物的爆轰性能与其分子结构的构效关系,根据嘧啶类含能化合物分子中原子相互连接的空间特性,提出了一种新的分子结构指数——定位校正指数B;此外还计算了19个含能化合物分子的电拓扑态指数,并优化筛选出其中的E 13作为分子结构描述符,将E 13与定位校正指数B结合,与嘧啶类含能化合物的爆轰性能进行回归分析;将B和E 13这两种结构指数作为BP神经网络的输入变量,神经网络结构采用2-3-1的结构方式,构建了预测嘧啶类含能化合物爆轰性能好的神经网络法模型。结果表明,每个模型的总相关系数均超过0.99,达优级相关,计算得到的氧平衡、爆热、爆速和爆压的预测值与文献值的平均相对误差分别为1.59%、1.05%、0.37%和1.28%;如将指数扩展到与噁三唑稠环类含能材料分子、含氟唑类含能化合物分子的爆轰性能等进行分析预测,同样能得到令人满意的结果。说明嘧啶类含能材料分子的爆轰性能与定位校正指数、电拓扑态指数有良好的非线性关系。
In order to study the quantitative structure-property relationship(QSPR)between the detonation performance of pyrimidine energetic compound and its molecular structure,a novel structure index,namely,positioning correction index B,was derived based on the spatial characteristics of atomic interconnections in pyrimidine energetic compound molecules.Moreover,the electrical topology state indexes of 19 energetic compound molecules were calculated,and the electrical topology state index E 13 was selected as the structural descriptor.The electrical topology state index E 13 was combined with positioning correction index B,which was introduced to the regression analysis of the detonation performance of pyrimidine energetic compound.The neural network models predicting the detonation performance of pyrimidine energetic compound were constructed,where the structural indexes B and E 13 were used as the input variables of BP neural network,and 2-3-1 as the network structure.The results show that total correlation coefficient for each model is more than 0.99,which means the excellent level of correlation.The average relative errors between the predicted values and the literature values of oxygen balance(OB),detonation heat(Q),detonation velocity(D)and detonation pressure(P)are 1.59%,1.05%,0.37%and 1.28%,respectively.Satisfactory results can also be obtained when introducing the index to analyze and predict the detonation performance of oxtriazole fused ring energetic material molecules and fluorinated azole energetic compound molecules.It is proved the there is a good non-linear relationship between the positioning correction index,electrical topology state index and the detonation performance of pyrimidine-based energetic material molecules.
作者
堵锡华
宋明
徐艳
DU Xi-hua;SONG Ming;XU Yan(School of Materials and Chemical Engineering,Xuzhou University of Technology,Xuzhou Jiangsu 221018,China)
出处
《火炸药学报》
EI
CAS
CSCD
北大核心
2024年第10期890-898,I0002,共10页
Chinese Journal of Explosives & Propellants
基金
江苏省自然科学基金(No.BK20171169)
江苏省高校自然科学基金重大项目(No.19KJA430020)。
关键词
人工智能
嘧啶
噁三唑稠环
含氟唑类化合物
爆轰性能
定位校正指数
电拓扑态指数
神经网络法
artificial intelligence
pyrimidine
oxtriazole fused ring
fluorinated azole compounds
detonation performance
positioning correction index
electrical topology state index
neural network method
