神经网络在钢管孔型设计中的应用  

Application of Nerual Network in Pass Design During Rolling of Steel Tube

作　　者：胡建华[1] 双远华[1] 王世杰[1] 王俊芳[2] 

机构地区：[1]太原科技大学,太原030024 [2]沈阳重型机械集团有限责任公司,沈阳110025

出　　处：《太原科技大学学报》2007年第3期202-205,共4页Journal of Taiyuan University of Science and Technology

基　　金：山西省自然基金项目(20021066)

摘　　要：文章将人工神经网络与有限元嵌合技术用于钢管孔型参数的预测,运用BP网络建立孔型参数与钢管尺寸精度之间的非线性关系,实现了对孔型参数的优化。解决了长期以来靠经验试凑的问题,为实际生产提供了理论依据,并给钢管生产带来了很大的便利。Both artificial neural network and finite element method are integrated to predict the pass parameters during rolling of steel tube in the paper. The non-line relationship is developed between pass parameters and precision of steel tube with BP network, which makes optimizing the pass parameters come ture. It resolves the problem that pass parameters are decided only by the method of trial and error according to experience for a long time. It pro- vides the theoretic bases and brings great convenience for production of steel tube.

关 键 词：人工神经网络 有限元 BP网络 孔型参数 

分 类 号：TG335.71[金属学及工艺—金属压力加工]