论文部分内容阅读
稠环烃有机半导体的禁带宽度与其分子构型以及π电子数有关，将可表征分子几何结构特征的拓扑指数一顶点复杂度信息指数和π电子数作为人工神经网络的输入特征量，利用经已知样本集训练的人工神经网络对稠环烃有机半导体的禁带宽度进行预报，预报结果与实测结果符合较好，表明人工神经网络是进行定量结构—性质或定量结构—活性相关性研究的一种有效方法。
The forbidden band width of polycyclic organic semiconductors is related to their molecular configuration and π electron number. Using topological index-vertex complexity index and π electron number, which can characterize molecular geometrical structure, as the input characteristic quantity of artificial neural network, Artificial neural networks trained by the known sample set are used to predict the forbidden band width of polycyclic organic semiconductors. The prediction results are in good agreement with the measured results. It indicates that artificial neural network is used to conduct quantitative structure-property or quantitative structure-activity correlation studies An effective method.