用模式识别轧机尾部跑偏的自适应控制

针对铝带热连轧轧制过程中影响跑偏的非线性因素众多,很难建立相应的跑偏控制模型的特点,基于现场测试数据进行分析,通过分析跑偏信号的数据特征,建立轧制过程跑偏量值神经网络辨识模型,找出了尾部跑偏模式与F_4机架敏感轧制参数的非线性关系,给出可应用于实际轧制生产的自适应纠偏控制模型,并以一组实测数据进行分析,纠偏效果良好.

The deviation of the tail adaptive control based on pattern recognition

For the characteristics of which it is difficult to set up the corresponding deviation control model that many nonlinear factors affect the deviation in the hot strip mill rolling process,set up neural network identification model of deviation in rolling process based on field test data analysis by analyzing data characteristics of the deviation signal,and find the non-linear relationship between the model of tail deviation and F_4 rolling rack-sensitive parameters,and give an adaptive correct control model which can be applied to actual production,and the effect of correction is good through analyzing a set of measured data.