节点文献

基于神经网络和遗传算法的拉深成形工艺优化

Processing Optimization of Sheet Metal Drawing Based on Neural Network and Genetic Algorithm

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 纪良波周天瑞

【Author】 JI Liangbo1,2,ZHOU Tianrui1(1.School of Mechanical & Electronic Engineering,Nanchang University,Nanchang Jiangxi 330031,China;2.School of Mechanical & Materials Engineering,Jiujiang University,Jiujiang Jiangxi 332005,China)

【机构】 南昌大学机电工程学院九江学院机械与材料工程学院

【摘要】 论述人工神经网络和遗传算法在板料拉深成形工艺优化中的应用,利用人工神经网络建立拉深工艺参数与零件厚度相对误差之间关系的数学模型,用遗传算法对工艺参数进行优化。其中由正交法设计得到实验样本,由数值模拟软件计算得出的厚度同实际成形件厚度进行比较,得到零件厚度相对误差,将其作为优化目标。按优化后的工艺参数进行实验,获得较高质量的拉深制件,从而为优化拉深模工艺参数提供了一种行之有效的方法。

【Abstract】 The application of artificial neural network and genetic algorithm on processing optimization of sheet metal drawing was discussed.The mathematics model between the process parameters of drawing and the relative thickness errors of the part was set up with neural network.The process parameters were optimized with genetic algorithm.Experimental samples were got by orthogonal plan method,the relative thickness errors of the parts were got by comparing the thickness of numerical simulation and the real thickness of the parts and they were taken as the goal of optimizing.Tests were made with optimizing parameters,the superior quality part was obtained.It provides a kind of effectual means for setting up and controlling parameters of drawing.

【基金】 江西省科技支撑项目(2007BG09604);南昌大学学科交叉基金项目(ncu0718)
  • 【文献出处】 机床与液压 ,Machine Tool & Hydraulics , 编辑部邮箱 ,2010年05期
  • 【分类号】TG386.3
  • 【被引频次】11
  • 【下载频次】153
节点文献中: 

本文链接的文献网络图示:

本文的引文网络