摘要
对带钢厚度偏差和轧制力偏差进行BP智能预测 ,并提出变辊缝法偏差校正算法实现偏差校正。离线模拟与在线应用的结果表明 ,提高机组带钢厚度偏差命中率 2 0 %和轧制力偏差命中率 8%。
The strip thickness deviation and the rolling force difference were predicted by BP neural networks,and a correction algorithm was carried out to rectify these deviations by changing the roll gap.The results of off line simulation and on line application showed that the hit rate of strip thickness deviation in ±50μm was 20% while that of the rolling force deviation in ±500kN was 8%.
出处
《上海金属》
CAS
2003年第4期28-32,共5页
Shanghai Metals
