以理论溶剂化变色参数来定量表达有机分子结构 ,探讨 42个有机探针分子的结构及其与聚合物相互作用的Flory Huggins参数 χ12 之间的定量关系。采用改进的神经网络方法进行建模和预测 ,并与基于理论线性溶解能关系的结果进行了比较 ,结果表明神经网络预测结果优于理论线性溶解能关系的结果 The theoretical solvatochromic parameters were used to quantify the structure of organic molecules. The structures of 42 organic probe molecules and their relationship with the Flory Huggins parameter χ12 were investigated. The improved neural network method was used to model and predict the results. The results were compared with the results based on the theoretical linear solution energy. The results show that the neural network prediction results are superior to the theoretical linear solution energy