一种新型的基于最小二乘回归St14钢变形抗力模型的方法

来源期刊：钢铁2010年第6期

论文作者：杨景明 杨志芬 车海军 张青 郝瑞峰

文章页码：44 - 48

关键词：冷连轧； 变形抗力； 摩擦因数； 最小二乘

Key words：tandem cold rolling； deformation resistance； friction coefficient； least square

摘    要：冷连轧的主要工艺参数为轧制力和前滑,而轧制力和前滑设定计算的精度取决于轧件的变形抗力和摩擦因数的精度,变形程度是影响变形抗力的一个重要因素,将所选变形抗力回归模型经过取对数等变换成线性函数,以鞍钢生产的St14钢为例,利用现场实际数据通过最小二乘来逐次拟合变形抗力回归模型中的系数。在同一轧制条件下,摩擦因数用3种不同的获得方法,其中前2种是由实测前滑值反算得到的,以使所得模型能够很好地与实际生产的数据相吻合。用回归出的3种不同变形抗力模型,分别带入轧制力迭代公式进行计算,所得轧制力基本与实测轧制力相符,其中由斯通公式拟合出的变形抗力回归模型计算出的轧制力平均误差很小,由此可以选出最优的变形抗力模型来应用于实际生产中轧制力和前滑的预设定。

Abstract: The main process parameters of tandem cold rolling are both rolling force and forward slip,and the rolling force and the forward slip setting depend on the accuracy of the calculation of rolling deformation resistance and friction coefficient.The degree of deformation is an important factor of affecting the deformation resistance.The selected regression model of deformation resistance becomes a linear function after logarithmic transformation.Taking the St14 steel produced by An Shan Iron and Steel as an example,the coefficients of deformation resistance regression model were obtained using the actual data through least squares method.In the same rolled conditions,the friction coefficient obtained using three different methods,The first two of which are obtained from the anti-calculation measured slip values,so that model can be identical well with the actual production data.Three different deformation resistance regression models were put into the iterative calculation formula of rolling force,respectively.Results show that both rolling force and the measured rolling force meet basically with the linear relationship.Especially,the calculated rolling force of deformation resistance regression model from Stone formula can make the average error smaller.Therefore,the best model of deformation resistance is presented for actual production pre-set of rolling force and forward slip.