针对广泛应用的Bland-Ford-Hill冷轧轧制力工艺模型,通过挖掘现场实际数据隐含的规律,对其变形抗力和摩擦因数的模型参数进行优化,以提高轧制力计算精度。首先,推导由轧制力计算变形抗力和摩擦因数的逆计算算法,采用L-M非线性多项式回归方法对变形抗力和摩擦因数的模型参数进行优化回归计算,建立轧制力优化算法;然后,根据现场海量的实际数据,采用数据挖掘的方法,使用上述优化方法计算更加符合现场实际的变形抗力和摩擦因数的模型参数。优化结果在线运行后,轧制力精度明显提高。 Aiming at the widely used Bland-Ford-Hill cold rolling force process model, the model parameters of deformation resistance and friction coefficient are optimized by mining the implicit rules of the actual field data to improve the calculation accuracy of the rolling force. Firstly, the inverse calculation method of calculating the deformation resistance and the friction coefficient by the rolling force is deduced. The LM nonlinear polynomial regression method is used to optimize the regression calculation of the model parameters of the deformation resistance and the friction coefficient, and the rolling force optimization algorithm is established. Then, Large amounts of actual data, the use of data mining methods, using the above optimization method to calculate more in line with the actual field deformation resistance and friction factor model parameters. After the optimization result is online, the precision of the rolling force is obviously improved.