摘要
在收集的 2 5 2个钢种的连续冷却转变曲线图 (CCT)的基础上 ,用不同的人工神经网络模型预测了钢的贝氏体开始转变临界冷却速度 ,与实测值比较证明 ,不同网络模型的预测精度不同。此外 ,用预测精度较高的人工神经网络模型计算了合金元素Si和B的含量对贝氏体开始转变临界冷却速度的定量影响 ,计算结果与试验结果相符合。
The critical cooling rates of bainite start point R cbs were predicted using various artificial neural network(ANN) models based on 252 types of collected CCT diagram.Compared with experimental results,the precision of prediction varies from the models of artificial neural network.The quantitative influences of alloying elements content of Si and B on critical cooling rate of bainite transformation are calculated with the specific artificial neural network model,and the predicted results accord well with experimental ones.
出处
《金属热处理》
CAS
CSCD
北大核心
2004年第1期58-62,共5页
Heat Treatment of Metals
关键词
贝氏体开始转变临界冷却速度
人工神经网络
合金元素
critical cooling rate of bainite start point
artificial neural network
alloying element