|
基于神经网络的板料毛坯设计方法研究
|
|
英文标题：Research on methods of sheet blank design based on ANN |
|
作者：张燕琴 李龙根 |
|
单位：东莞职业技术学院机电工程系 |
|
关键词：毛坯设计 反向模拟 神经网络模型 |
|
分类号：TG386 |
|
出版年,卷(期):页码：2009,34(6):62-64 |
|
摘要：
|
|
研究了一种冲压毛坯设计的方法——基于神经网络模型的坯料优化设计技术,即利用神经网络反映映射关系,实现对同一类型零件坯料的快速预测。文章阐述了该方法的设计原理、流程,解释了径向基函数神经网络、坯料形状转化成神经网络的输出参数、训练样本采用修正坯料来提高设计精度等关键技术。实际生产表明,一旦神经网络训练成功,可快捷得到合理的坯料设计结果,具有很大的经济效益。
|
|
A method for designing the blank was researched:blank optimization design based on ANN model,using mapped relation reflected by ANN to achieve the forecast blank of congeneric parts. The design principle and process were expatiated,and some key technologies ,such as radial basis function neural network,transforming blank contour to the output parameters of ANN and using modified blank as training sample to increase design precision were expatiated. The results show that,if the ANN is trained successfully,the suitable blank can be obtained rapidly,which has great economic benefits.
|
|
基金项目：
|
|
|
|
作者简介：
|
|
|
|
参考文献：
|
[1]吴建军,张萍.基于板料理想成形的有限元方法研究[J].锻压技术,2008,33(1):131-132.
[2]Chung K,Richmond O.Ideal formingⅡ:sheet forming with opti mumdeformation[J].International Journal of Mechanic Science,1992,34(8):617-633.
[3]周军,马闯,李光耀,等.基于有限元网格映射方法的精确反算求解毛坯[J].机械工程学报,2002,38(12):103-106.
[4]谢慧超,廖代辉,龚英菊.基于滑移约束面方法的毛坯反求技术[J].锻压技术,2008,33(2):134-137.
[5]罗亚军,王静怡,张永清,等.板料成形过程中合理毛坯形状的确定[J].塑性工程学报,2001,8(3):63-66.
[6]L Xuchan,L Bingwen.Modeling of flange deformation of ir-regular drawn cups using a fluid analogy[J].International Journal of Mechanic Science,1986,28(8):491-497.
[7]罗四维.大规模人工神经网络理论基础[M].北京:清华大学出版社,2004.
[8]扶名福,范洪春,张庭芳,等.BP神经网络在镁合金流变应力预测中的应用[J].锻压技术,2008,33(2):157-159.
[9]Zi wu Ren,Ye San.Designing for RBF networks based on particle swarm opti mization and regularized orthogonal least squares[A].Proceeding of the6th congress on intelligent control and automation[C].Dalian,2006.
|
|
服务与反馈：
|
|
【文章下载】【加入收藏】
|
|
|
|
