论文部分内容阅读
针对某冷轧厂1 730mm连退机组六辊CVC平整机组出现张力设定表数值不准确或缺失的问题,建立了基于BP神经网络的平整张力预设定模型,对平整机张力设定表进行了完善并实现了张力设定值的在线预报。在线运行结果表明:张力预报值与实测值的相对误差在±8%以内,实现了平整机组张力的高精度预报。进一步离线测试表明:采用本文所建立的平整张力神经网络预设定模型后,平整机出口成品板形质量得到较大提升,平直度横向分布均值降低约78%,满足生产要求,并提高了轧制的稳定性。
Aiming at the problem of inaccurate or missing values of the tension setting table of a 6-roller CVC leveling unit of a 1 730mm continuous rolling mill in a cold-rolling mill, a presetting model of the leveling tension based on the BP neural network is established, It has been perfected and realized the online forecast of tension setting value. The results of online operation show that the relative error between the predicted value of tension and the measured value is within ± 8%, and the high-precision forecast of the tension of the flat unit is achieved. Further off-line tests show that after the pre-set model is established by using the tensioning neural network established in this paper, the flatness of the finished product at the exit of the flattening machine is greatly improved, the average of the straightness distribution is reduced by about 78%, meeting the production requirements and improving The stability of the rolling.