振动力场作用下聚合物挤出制品结构与性能的质量控制与优化

利用实验数据，分析与研究振动场振动参数对聚合物挤出制品质量的影响，对聚合物挤出制品的熔体流动速率，微晶结构，拉伸强度等主要性能质量指标，建立以振动频率为主要控制变量的神经网络模型，并引入信息分配模型，探讨了一个网络输入节点下神经网络学习样本的特征提取与优化。实验结果表明，经过信息预处理的学习样本，可以使网络有更好的收敛效果。

THE QUALITY CONTROL AND OPTIMIZATION OF EXTRUDATES STRUC-TURE AND PROPERTY OF POLYMER UNDER VIBRATION FORCE FIELD

Using the datum of experiment, the influence factors of quality of extrusion structure and property of polymer under vibration force field were analyzed. The neural network model with input of vibration frequency about melt flow rate, crystallite structure as well as stretch strength of polymer products was established. It was investigated that the learning samples features of neural network are drawn and optimized in only one input of network by means of information distribution model. The test results show that the controlling model of network is satisfying in applying information distribution processing sample data.