网站首页期刊简介编委会过刊目录投稿指南广告合作征订与发行联系我们English
基于BP神经网络的预切冲裁断面质量的仿真预测
英文标题:Simulation and prediction of crosssection quality for precut blanking based on BP neural network
作者:张良 
单位:江阴职业技术学院 机电工程系 
关键词:预切冲裁 冲裁断面质量 BP神经网络 正交试验 有限元仿真 
分类号:TG385
出版年,卷(期):页码:2018,43(12):175-179
摘要:

 以汽车上冲压零件为研究对象,利用DEFORM-2D有限元软件对QSTE460板料进行预切冲裁过程的有限元仿真,通过板料冲裁试验得出,零件冲裁断面质量的试验值为0.557 mm,模拟值与试验值之间的相对误差为9.72%,验证了有限元仿真的正确性。基于板料预切冲裁正交试验设计,运用BP神经网络对板料预切冲裁断面质量进行仿真预测。以预切深度、落料冲裁间隙、冲裁速度、预切冲裁间隙以及模具刃口圆角半径为输入层,利用光亮带的长度作为输出层,建立了用于冲裁断面质量预测的5-12-1的3层BP人工神经网络结构。通过BP神经网络的训练与测试得出,BP神经网络的预测值与有限元仿真值之间的最大相对误差为1.44%,从而为板料冲裁断面质量的预测提供一种更为可靠的预测方法。

 For stamping part of automobile, the precut blanking process of QSTE460 sheet metal was simulated by finite element software DEFORM2D, and the experimental value of blanking crosssection quality was 0.557 mm by the blanking experiment. Then, the relative error between simulated value and experimental value is 9.72% which verifies the correctness of finite element simulation, and the crosssection quality of precut blanking was simulated and predicted by the precut blanking orthogonal test and BP neural network. Furthermore, the three layer BP neural network structure of 5-12-1 was established by taking the precut depth, blanking clearance, blanking speed, precut blanking clearance and edge radius of punch as the input layer and taking the length of bright band as the output layer. After training and testing of BP neural network, the results show that the maximum relative error between prediction value of BP neural network and simulation value of finite element is 1.44%, which provides a more reliable prediction method for the prediction of blanking crosssection quality.

基金项目:
作者简介:张良(1973-),男,硕士,副教授 Email:zlthzl@hotmail.com
作者简介:
参考文献:

 参考文献:


 


[1]武万斌,年雪山. 汽车轻量化技术发展趋势
[J]. 汽车工程师,2017,(1):15-17.

 

Wu W B,Nian X S.Development of vehicle lightweight technology
[J].Automotive Engineer,2017,(1):15-17.

 


[2]刘超,王磊,刘杨.汽车用先进高强钢的发展及其在车身设计中的应用
[J].特钢技术,2012,18(2):1-4.

 

Liu C,Wang L,Liu Y.Development of advanced highstrength steel used for automobile and its application in body design
[J].Special Steel Technology,2012,18(2):1-4.

 


[3]郭玉琴,姜虹,王小椿.板料冲压加工数值模拟中接触摩擦问题的研究
[J].机械工程学报,2004,40(11):174-177.

 

Guo Y Q, Jiang H, Wang X C. Research on contact friction in sheet forming numerical simulation
[J].Chinese Journal of Mechanical Engineering,2004,40(11):174-177.

 


[4]苏春建,王朋,王清,等.厚板双侧齿圈压边精冲参数对成形影响规律研究
[J].机械设计与制造,2017,(6):160-163.

 

Su C J,Wang P,Wang Q,et al.The research of fineblanking parameters on the forming of thickplate bilateral gearring blankholder
[J].Machinery Design & Manufacture,2017,(6):160-163.

 


[5]徐晓,王二冬,夏琴香,等.某汽车底梁加固件冲压成形全工序数值模拟分析
[J].现代制造工程,2017,(8):57-62.

 

Xu X,Wang E D,Xia Q X,et al.Numerical simulation analysis of complete stamping process for stiffener part of automotive beam
[J].Modern Manufacturing Engineering,2017,(8):57-62.

 


[6]张慧,沈兴全,王步奎.基于Deform薄板冲裁模具间隙有限元模拟
[J].煤矿机械,2015,36,(1):138-140.

 

Zhang H,Shen X Q,Wang B K.Finite element simulation of die clearance in sheet blanking based on Deform
[J].Coal Mine Machinery,2015,36,(1):138-140.


[7]张余.基于DEFORM的预减振轴套板预切冲裁
[J].锻压技术,2017,42(5):34-37.

 

Zhang Y.Precutting blanking of intermediate disc based on DEFORM
[J].Forging & Stamping Technology,2017,42(5):34-37.

 


[8]王晓莉, 穆瑞, 张咏琴. 基于BP神经网络的薄板成形回弹仿真预测
[J].锻压技术,2016, 41(6):146-149.

 

Wang X L,Mu R,Zhang Y Q.Numerical prediction of springback in sheet metal forming based on BP neural network
[J]. Forging & Stamping Technology,2016,41(6):146-149.

 


[9]黄珍媛,蔡志兴,阮锋,等.高速精密级进冲压中的冲裁断面质量实验研究
[J].塑性工程学报,2009,16(3):9-12.

 

Huang Z Y, Cai Z X, Ruan F, et al.Experiment investigation on blanking fracture surface quality in highspeed & precision progressive stamping
[J].Journal of Plasticity Engineering,2009,16(3):9-12.

 


[10]王静,鲁世红,于长生.基于ANN的冲裁合理间隙的预测研究
[J].机械科学与技术,2006,25(8):891-894.

 

Wang J,Ru S H,Yu C S.Prediction of optimum blanking clearance based on artificial neural networks
[J].Mechanical Science and Technology,2006,25(8):891-894.

 


[11]吴炎林,向华,庄新村,等.基于BP神经网络精冲塌角高度的预测
[J]. 塑性工程学报,2017,24(6):32-37.

 

Wu Y L,Xiang H,Zhuang X C,et al.Prediction of fine blanking dieroll height based on BP artificial neural networks
[J].Journal of Plasticity Engineering,2017,24(6): 32-37.

 


[12]韩茂盛,陶欢.B10铜合金高温流变行为及BP神经网络本构模型
[J].锻压装备与制造技术,2016,51(6):112-115.

 

Han M S,Tao H.High temperature flow stress behavior of B10 copper alloy and BP neural network constitutive model
[J].China Metalforming Equipment & Manufacturing Technology,2016,51(6):112-115.
服务与反馈:
文章下载】【加入收藏
《锻压技术》编辑部版权所有

中国机械工业联合会主管  中国机械总院集团北京机电研究所有限公司 中国机械工程学会主办
联系地址:北京市海淀区学清路18号 邮编:100083
电话:+86-010-82415085 传真:+86-010-62920652
E-mail: fst@263.net(稿件) dyjsjournal@163.com(广告)
京ICP备07007000号-9