论文部分内容阅读
在通过实验获得大量样本数据的基础上,利用Matlab语言与VB语言混合编程,开发了聚合物挤出胀大比预测软件。并利用BP网络建立不同口模挤出条件下挤出胀大比与剪切应力之间的关系来实现人工神经网络的训练。经过反复训练满意后,即可输入一系列剪切应力值来预测挤出胀大比。结果表明:网络的训练精度可控制在10-2以下,预测点与实测点吻合得较好,实现了由理论成果向应用技术的转化。
Based on a large number of sample data obtained through experiments, a software program of polymer extrusion swell ratio was developed by mixed programming with Matlab language and VB language. The BP network is used to establish the relationship between the extrusion expansion ratio and the shear stress under different die extrusion conditions to realize the artificial neural network training. After repeated training satisfaction, you can enter a series of shear stress values to predict the extrusion expansion ratio. The results show that the training accuracy of the network can be controlled below 10-2, and the predicted point is in good agreement with the measured point, which realizes the transformation from theoretical result to applied technique.