摘要
采用BP神经网络原理对1350mm铝箔轧机轧制数据重新处理,建立了基于人工神经网络的轧制力模型.结果表明,用人工神经网络轧制力模型的计算值与实测值相比较偏差<3%.该模型较真实地反映了轧制过程的特征.
Based on the principle of BP neural networks, the rolling force model is established after thoroughly analyzing and reprocessing the data of 1350 mm aluminium foil mill. It states that the difference between the output of artificial neural networks rolling force model and the real value is in the order of 3 percent. The model reflects the real feature of process.
出处
《北京科技大学学报》
EI
CAS
CSCD
北大核心
1997年第2期173-177,共5页
Journal of University of Science and Technology Beijing
基金
国家"八五"攻关项目
关键词
神经网络
铝箔
轧制力模型
轧机
neural networks,aluminium foil, rolling force model