高聚物流变性能预测与研究  

Prediction and Research of High Polymer Rheological Properties

作　　者：高志生[1] 马万珍[1] 张晗[1] 张磊[1] 

机构地区：[1]内蒙古工业大学材料科学与工程学院,内蒙古呼和浩特010051

出　　处：《上海塑料》2007年第2期31-34,共4页Shanghai Plastics

摘　　要：根据影响高聚物流变性能的各因素之间的关系,运用Matlab软件建立一个人工神经网络。对5种高聚物材料进行实验;用实验数据对网络进行训练,并对高聚物流变性能进行预测;将预测和实测的结果进行分析对比。结果表明,用人工神经网络的方法预测高聚物的流变性能是可行的,预测精度较高,平均误差小于5%。In accordance with the relations among the factors which affect high polymer rheological properties, an artificial neural networks is established based on Matlab software. Sample data are obtained by experiments of five kinds of high polymer. The networks is trained by the obtained data and used to predict the rheological properties of high polymer. Comparison and analysis are made between the predicted and the actual results. The resuits show that it is feasible to predict the high polymer rheological properties using the artificial neural networks method. The prediction accuracy is higher, the average error margin is less than 5 %.

关 键 词：高聚物 流变性能 预测 人工神经网络 

分 类 号：TQ317.3[化学工程—高聚物工业]