用分子连接性指数关联和预测链烷烃的pVT性质

CIESC Journal

用分子连接性指数关联和预测链烷烃的pVT性质

刘华,蒋文华,韩世钧

浙江大学化学系!杭州310027,浙江大学化学系!杭州310027,浙江大学化学系!杭州310027

出版日期:1999-10-25 发布日期:1999-10-25

CORRELATION AND PREDICTION THE pVT PROPERTIES OF LINEAR ALKANES BY MOLECULAR CONNECTIVITY INDEX

Liu Hua, Jiang Wenhua and Han Shijun(Department of Chemistry, Zhejiang University, Hangzhou 310027)

Online:1999-10-25 Published:1999-10-25

摘要/Abstract

摘要： <正>流体的pVT性质在化工理论基础和应用研究中都具有重要的意义,通常采用经验关系式或基于一定理论模型的状态方程来描述.而流体性质与其结构存在着内在的、本质的联系,因此用于描述分子结构的分子拓扑指数可进行结构/性质的定量关联.本文基于分子拓扑指数,采用人工神经网络技术来研究烷烃的pVT性质.并假定物质的pVT性质与分子拓扑指数呈以下函数关系V=f(p,T,~mX_t)(1)函数f无统一的解析表达式且为一非线性函数.将分子拓扑指数对pVT性质的影响视为一个由拓扑指数空间到pVT性质空间的一个映射,即V=N(p,T,~mX_t)(2)在此映射关系中,N为神经网络,其权重参数可由部分物质的已知实验数据经神经网络训练后抽提得到,进而可对其他物质的pVT性质进行预测.1 分子拓扑指数的选择和计算

Abstract: The pVT properties of linear alkanes are studied by molecular connectivity index and artificial neural network (ANN) . It is found that the properties can be directly calculated from the molecular structure. Here 25 compounds are treated as a training set to extract the weight factor by ANN. According to the training results of ANN, the pVT data of other 15 compounds were predicted. The calculated values are satisfactory. Using the ANN and the molecular connectivity index can provide a convenient and effective method to calculate the pVT data.

中图分类号:

O621.1

刘华,蒋文华,韩世钧. 用分子连接性指数关联和预测链烷烃的pVT性质 [J]. CIESC Journal.

Liu Hua, Jiang Wenhua and Han Shijun(Department of Chemistry, Zhejiang University, Hangzhou 310027). CORRELATION AND PREDICTION THE pVT PROPERTIES OF LINEAR ALKANES BY MOLECULAR CONNECTIVITY INDEX[J]. .