摘要
针对宝山钢铁 (集团 )公司 2 0 5 0热连轧层流冷却系统 ,采用神经网络与数学模型相结合的方法 ,给出优化的层流冷却对流换热系数 ,以实现准确地预报卷取温度的目的。结果表明 ,采用神经网络计算出的对流换热系数后 ,卷取温度的计算值与实测值的标准差降低了 2 2 .84% 。
An optimized convection heat transfer coefficient for laminar cooling has been obtained by using BP neural network combined with mathematic model to predict coiling temperature accurately on 2050 hot strip mill at Baosteel. The results indicated that mean standard deviation of difference between the calculated and measured temperature is decreased by 22 84 % after using heat transfer coefficient calculated by BP neural networks.
出处
《钢铁》
CAS
CSCD
北大核心
2002年第8期37-40,共4页
Iron and Steel
基金
国家自然科学基金资助项目 ( 50 10 4 0 0 4 )
关键词
人工神经网络
层流冷却
热轧带钢
竖流换热系数
卷取温度
neural network, laminar cooling, hot rolled steel strip, convection heat transfer coefficient