基于神经网络的管件开式冷挤压相对伸长量预测

Prediction of relative Elongation of Thick-walled Tube based on ANN

为了预测各参数组合下的无芯棒开式冷挤压厚壁管件的伸长量，拓展开式冷挤压的实验结果，建立了无芯棒开式冷挤压厚壁管件的伸长量预测的神经网络模型。通过实验结果与预测结果的比较，二者吻合良好，为深入研究开式冷挤压工艺,建立开式冷挤压工艺设计专家系统提供了新的思路。

In order to predict relative elongation of thick-walled tu be without mandrel and expand the experimental results of open-die col d extrusion, a valid neural network model of relative elongation predi ction is established. By comparing experimental results with predictio n data, they fit well, which provides a new method to research open-di e cold extrusion.and to create expert system of its technological design.