BP神经网络在PP/POE-g-MAH增韧PA6研究中的应用  

Application of BP Neural Network on PP/POE-g-MAH Toughening PA6

作　　者：陈坤[1] 郑梯和[1] 宋克东[1] 田祥儒[1] 张瑜[1] 

机构地区：[1]株洲时代新材料科技股份有限公司,湖南株洲412007

出　　处：《中国塑料》2011年第12期55-58,共4页China Plastics

摘　　要：采用均匀设计法和BP神经网络研究了聚丙烯(PP)/乙烯-辛烯共聚物接枝马来酸酐(POE-g-MAH)对聚酰胺6(PA6)的增韧作用,并在此基础上建立了PA6/PP/POE-g-MAH共混物中各组分含量与共混物冲击强度关系的3层BP神经网络预测模型。结果表明,该模型与实验结果基本吻合,可信度较高;当POE-g-MAH含量为14.00%(质量分数,下同)、PP含量为9.00%时,共混物的缺口冲击强度达到92.12kJ/m2。The research of PP/POE-g-MAH toughening PA6 was conducted by BP neural network and homogeneous design in this paper. On this basis,a 3-layer BP neural network prediction model for the composites between impact strength and the composition was established. The prediction by the model agreed well with the experiments. The impact strength of the toughening PA6 reached 92. 12 kJ/m^2 when the content of POE-g-MAH was 14.00 % and PP content was 9.00 %.

关 键 词：神经网络 聚酰胺6 聚丙烯 增韧 马来酸酐 

分 类 号：TQ323.6[化学工程—合成树脂塑料工业]