Optimization of the rolling force self-learning for specifications changing in the hot strip rolling

In the hot strip rolling process, the prediction precision of the rolling force which is largely dependent on the rolling force self-learning directly affected the thickness precision of strips. In the view of the rolling force prediction precision decreased with the specifications changing, and through the analysis of the reasons for generating rolling force deviation, this paper introduced the concept of steel grade deformation resistance parabolic deviation curve, the equipment standers self-learning coefficient and equipment state effective coefficient to solve the problems. The practice application results showed that the relative error of the rolling force prediction for the first piece strip after changing specifications decreased by 4% compared with the conventional prediction method which satisfied the automatic thickness control system. The product quality of the strip was enhanced and good economic value was obtained, which indicated that the new prediction method was suitable for the industrial production promotion. © 2015, Editorial Department of Journal of Northeastern University. All right reserved.