论文部分内容阅读
进行碳－１３ＮＭＲ波谱模拟的一种常用方法为线性回归，其数学模型则由已知结构所测得的化学位来建立。显然，每一数学模型仅可用于某种相似的化学环境。因此，为了进行化学环境的分类，本文提出权重层次位图法和分子连接性指数法两种方法以进行一些参数的计算，同时采用多元分析手段，如主成分分析和聚类分析，以进行多维空间数据点的显示，所得结果比较满意，从而可为碳－１３波谱模拟中回归方程的选择提供理论依据。
A common method for performing carbon-13 NMR spectroscopy is linear regression, the mathematical model of which is established from the chemical sites measured by known structures. Obviously, each mathematical model can only be used in a similar chemical environment. Therefore, in order to classify the chemical environment, this paper proposes two methods of weight hierarchy bitmap and molecular connectivity index to calculate some parameters. At the same time, using multiple analytic methods, such as principal component analysis and cluster analysis, Spatial data points show the results are satisfactory, which can provide a theoretical basis for the choice of regression equation in the carbon-13 spectrum simulation.