|
摘要：
|
|
运用神经网络技术实现材料性能参数的实时识别是智能化拉深的重要研究课题。由于训练样本数据的冗余使得BP网络收敛精度差、速度慢,直接影响到网络的识别精度。运用Rough集理论强大的数据分类简约功能,能够去掉多余属性的样本数据,从而优化了神经网络的拓扑结构。经过试验验证优化后的网络不仅收敛速度快、精度得到极大提高,而且网络预测相对误差精度都在6%以下。
|
|
The real-time identification via neural network is an important subject in intellectual deep drawing of sheet metal.Because of the redundancy of training data,it makes the convergence of BP neural network slow and imprecise.Using the data reduction and classify function of rough sets,the training data of surplus attribute can be(deleted) and the structure of the neural network can be optimized.Experiments prove that the optimized network converges more quickly and accurately.The relative prediction error precisions are all below 6%.
|
|
基金项目：
|
|
国家自然科学基金资助项目(50375136、59875074)
|
|
作者简介：
|
|
|
|
参考文献：
|
[1]Jackson A G.Rough sets applied to the discovery of materialsknowledge[J].Journal of Alloys and Compounds,1998,279(1 4):14 21.
[2]Pawan Lingras.Comparison of neofuzzy and rough neural net-works[J].Information Sciences,1998,110(3 4):207215.
[3]刘清.Rough集及Rough推理[M].北京:科学出版社,2001.
[4]王国胤.Rough集理论与知识获取[M].西安:西安交通大学出版社,2001.
[5]Tetsuya Yagami.Intelligent sheet stamping process using seg-ment blankholder modules[J].Journal of Materials Process-ing Technology,2004,155 156(30):2099 2105.
[6]Zhao J.Study on intelligent control technology for the deepdrawing of an axi-symmetric shell part[J].Journal of Materi-als Processing Technology,2004,151(1 3):98 104.
[7]赵军.圆锥形件拉深成形智能化研究[D].哈尔滨:哈尔滨工业大学,1997.
[8]赵军,秦泗吉,曹宏强,等.轴对称曲面件智能化拉深成形〗过程的解析定量描述[J].塑性工程学报,1998,5(4):47 58.
[9]姚华,陈军.BP神经网络在盒形件坯料外形预测中的应用[J].模具技术,1999(5):15 19,42.
[10]赵军,官英平,郑祖伟,等.基于神经网络的材料性能参数和摩擦系数的实时识别[J].塑性工程学报,2001,8(2):36 39.
[11]王凤琴,赵军,官英平,等.轴对称件智能拉深系统中的神经网络参数识别[J].锻压装备与制造技术,2003,(3):67 70.
|
|
服务与反馈：
|
|
【本网站尚未开通全文下载服务】【加入收藏】
|
|
|
|
