针对传统板对板的自学习模型不能修正轧件个体差异的缺陷,提出了道次对道次的道次修正模型。该模型利用上道次的实测轧制数据,如实测轧制力、轧制速度、测温仪温度等,通过道次出口厚度再计算模块、钢板温度修正模块和轧制力短期修正模块精确计算出当前钢板的实际状态,然后通过剩余道次再计算模块对后续未轧道次的辊缝进行修正。实际应用表明,该道次修正模型投入使用后,轧制力预报精度和成品厚度精度有了很大提高。 Aiming at the defect that the self-learning model of traditional board-to-board can not correct the individual differences of rolling materials, this paper proposes a pass modification model of pass to pass. The model makes use of the actual measured rolling data of the previous pass, such as the measured rolling force, the rolling speed, the temperature of the thermometer, etc., through the outlet thickness recalculation module, the steel temperature correction module and the rolling force short-term correction module The actual state of the current steel plate, and then through the remaining pass recalculation module on the subsequent un-pass roll gap correction. Practical application shows that after the pass correction model is put into operation, the precision of rolling force forecasting and finished product thickness accuracy have been greatly improved.