自适应模糊神经网络预测金属离子水解常数的研究

Studies on Prediction of Hydrolysis Constants pK1 of Metal Ions by Using Adaptive Fuzzy Neural Network

采用自适应模糊神经网络的方法,以金属离子价电子结构因子、电荷-半径比、适配价轨道数因子等为参数,关联金属离子水解常数.利用减法聚类算法确定模糊神经网络的结构,并结合模糊推理系统进行该网络参数的调整.网络仿真的结果是满意的.自适应模糊神经网络可望成为元素和化合物构效关系研究的辅助手段.

An adaptive fuzzy neural network was applied to study the relationships between the structural parameters of metal ions and their hydrolysis constants pK1 of metal ions, with the parameters of valence electronic structural factor of metal ions（S）, electric charge-radius ratio（Z^2/r）, adaptive valence＇orbit number（w）. Subtractive clustering algorithm is used to confirm the structure of fuzzy neural network, and combined fuzzy inference systems to process regulation of the network parameters. The simulation results are satisfactory. Therefore, the authors can expect that adaptive fuzzy neural network might be used as an effective assistant technique for the investigation of quantitative structure-property relationship（QSPR） of the elements and compounds.