摘要
借助Matlab神经网络工具箱的BP神经网络,建立了三辊Y型轧机钛合金棒材连轧轧制力BP神经网络模型。以φ18 TC4钛合金棒材为例,应用此BP神经网络轧制力模型来实现轧制力预报。结果表明,该BP网络模型轧制力预报精度高,且操作高效简洁,可代替计算过程繁杂的传统轧制力数学模型。
Three-roll Y-type mill rolling force BP neural network model was proposed with the help of MATLAB neural network toolbox. Taking titanium alloy TC4 18 for example,BP neural network model was applied to forecast rolling force. The result shows that three-roll Y-type mill rolling force BP neural network model has the advantages of high accuracy,simplicity and efficiency in operation,so this BP neural network model can replace traditional miscellaneous rolling force mathematical model.
引文
[1]宁兴龙,王国宏.钛合金棒线材生产及市场预测[J].稀有金属材料与工程,1998,27(4):248-252.
[2]高振莉,王海儒.Y型轧机轧制变形区的数学模型[J].太原重型机械学院学报,1997,18(4):338-343.
[3]曹喜发.KOCK棒线材轧机及轧制理论(一)[J].轧钢,1999(4):28-37.
[4]曹喜发.KOCK棒线材轧机及轧制理论(二)[J].轧钢,1999(5):40-60.
[5]帅美荣.钛合金棒材三辊热连轧过程变形机理与技术研究[D].太原:太原科技大学,2012:10-20.
[6]潘露.钛合金棒线材连轧孔型设计及优化[D].太原:太原科技大学,2012:3-20.