基于神经网络的不锈钢圆孔翻边件成型缺陷预测  

作　　者：赵博宁[1] 

机构地区：[1]柳州铁道职业技术学院装备制造学院,广西柳州545616

出　　处：《科技创新与应用》2022年第13期126-129,共4页Technology Innovation and Application

基　　金：柳州铁道职业技术学院2021年度校级立项项目(2021-KJB12)。

摘　　要：文章通过金属材料数值模拟操作、材料的塑性变形、神经网络技术等相关原理的实际应用,借助有限元数值模拟程序对典型圆孔翻边件翻边成形过程进行数值模拟,将BP神经网络的算法引入圆孔翻边件翻边成形过程中,应用BP神经网络模型预测圆孔翻边件翻边成形质量,后续通过相应的试验过程、建立神经网络系统等步骤对设计参数和模具实施优化改进,确保翻边件的成形结果达到最优化的结果,实验结果表明成形效果良好。In this paper, through the practical application of metal material numerical simulation operation, material plastic deformation, neural network technology and other related principles, the flanging forming process of typical round hole flanging parts is numerically simulated with the help of finite element numerical simulation program. The algorithm of BP neural network is introduced into the flanging forming process of round hole flanging parts, and the BP neural network model is used to predict the flanging forming quality of round hole flanging parts. Subsequently, the design parameters and die are optimized and improved through the corresponding test process and the establishment of neural network system to ensure that the forming result of the flanging part achieves the optimal result. The experimental results show that the forming effect is good.

关 键 词：神经网络算法 圆孔翻边件 优化参数 成形缺陷 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]