摘要
运用BP神经网络,建立了热镀锌各工艺参数对热镀锌钢板力学性能影响的数学模型,并与线性回归模型进行了比较。结果表明:BP神经网络预测均方根偏差明显比线性回归预测均方根偏差小,表明该BP神经网络模型用于热镀锌板力学性能预测是可行的,并具有一定的实用性。
A predictive model for mechanical properties of hot dip galvanized strip was established by BP neural network and compared with multi-variant linear regression model. The result shows that the root-meansquare deviation of BP neural network is less than that of multi-variant linear regression model, which proved that the predictive model for mechanical properties of hot dip galvanized strip is feasible and effective.
出处
《机械工程材料》
CAS
CSCD
北大核心
2007年第3期60-62,共3页
Materials For Mechanical Engineering
关键词
热镀锌钢板
BP神经网络
性能预测
hot-dip galvanization strip
BP neural network
property prediction
