金属塑性变形抗力的BP神经网络预报模型  

BP NEURAL NET PREDICTION SYSTEM FOR METAL PLASTIC DEFORMATION STRESS

作　　者：张建华[1] 高燕[1] 刘战英[2] 

机构地区：[1]唐山钢铁公司热轧薄板厂,河北唐山063016 [2]河北理工大学,河北唐山063016

出　　处：《河北冶金》2006年第2期17-18,共2页Hebei Metallurgy

摘　　要：采用BP神经网络程序,训练用G leeb le-1500热模拟实验机测试验钢种在三个温度阶段的变形抗力数据,实测值与BP神经网络模型预报值误差较小,得到了精度较高的训练预报模型。With BP neural net program the Gleeble - 1500 hot simulating experimental machine is trained to do the simulation test and check the deformation stress values of experimental steel grade under three steps of different temperature. The difference between the real measured value and predicated one by BP neural net model is small, so the training and predicating model with high precision got.

关 键 词：金属 塑性变形 抗力 BP神经网络 模型 

分 类 号：TG115.5[金属学及工艺—物理冶金]