BP神经网络方法预测改性聚乙烯醇的吸水率

用人工神经网络法来预测高分子化学反应产物的性能,通过研究顺丁烯二酸酐与聚乙烯醇反应制备吸水材料过程中反应物的配比、反应体系的温度、pH值以及反应时间等条件和产物吸水率之间的关系,得出了制备最大吸水率的反应条件。并通过人工神经网络方法建立BP神经网络模型,采用Matlab语言编写神经网络程序,再以实验测得的数据对网络进行训练,然后用训练好的网络对实验条件和吸水率之间的关系进行仿真,并将仿真结果和实验数据加以比较。结果表明,BP神经网络用于化学反应产物性能的预测是可行的.为神经网络在化学反应控制预测领域的应用奠定了基础.

Prediction of water absorption of modification polyvinyl alcohol using BP Neural network

Artificial neural network method has been used to predict the properties of as-prepared product of polymer's chemical reactions,for instance,the preparation of water-adsorbing materials from Maleic Anhydride and poly(vinyl alcohol).The experimental parameters for materials with maximal absorption rate were obtained by studying the reaction parameters,such as ratio of the reagents,temperature and pH value and so on.BP neural network was developed and written by matlab languages according to the method of manual neural network,and then the network was trained according to the data obtained. Then,we proceeded the emulation between the experimental parameters and adsorption rate by using the well trained network and compared the results with the experimental data.All these indicate that BP neural network can predict the properties of the product of chemical reaction.This laid be the foundation of the neural network's extensive application in the field of prediction and control of reactions.