论文部分内容阅读
计算出体系中烷基硫醇化合物的各种结构参数,以优选出的分子连接性指数和量化参数为结构描述符,首次采用反向传播算法(BP)人工神经网络、径向基函数网络(RBF)2种非线性方法建立了参数少且精度高的定量结构-色谱保留相关(QSRR)模型,预测了烷基硫醇在4种极性固定上的气相色谱保留指数(RJ)。结果表明:在4种固定相上建立的BP模型均优于RBF模型且非线性方法(BP、RBF)优于文献中多元线性回归(MLR)方法,所建定量结构保留关系(QSRR)模型具有良好的稳定性和预测能力。
Various structural parameters of the alkylthiol compounds in the system were calculated. Using the preferred molecular connectivity index and the quantitative parameters as the structural descriptors, BP artificial neural network and radial basis function network RBF), a quantitative structure-chromatographic retention correlation (QSRR) model with less parameters and higher precision was established. The gas chromatographic retention index (RJ) of alkylthiol on four polar fixtures was predicted. The results show that the BP model established on the four stationary phases is better than the RBF model and the nonlinear method (BP, RBF) is superior to the multiple linear regression (MLR) method in the literature. The quantitative structure retention relationship (QSRR) Good stability and predictive ability.