铝热粗轧机自适应辊缝动态设定

Dynamic Setting of Adaptive Roll Gap in Aluminum Hot Roughing Mill

针对某现场铝热粗轧板带头部厚度偏差较大现象,应用自适应模糊神经网络控制技术,建立了热粗轧辊缝动态设定系统。该系统将轧制力预报误差及弹跳方程误差作为输入,运用模糊神经网络预测下一道次的辊缝设定补偿值,并以实际数据对系统进行校验,结果显示该方法可以大幅提高铝热粗轧板带头部厚度精度。

In view of the big deviation in head thickness of aluminum hot roughing strip, a dynamic setting system for roll gap was established using an adaptive fuzzy neural network control technology. This system adopts the prediction error of rolling force and the error of spring equation as inputs and uses fuzzy neural network to predict the set compensation of roll gap in the next pass, finally was verified by actual data. The results show that this method can greatly improve the head thickness accuracy of aluminum hot roughing strip.