Roll force parameter self-learning method and device

Abstract

The invention discloses a roll force parameter self-learning method and device for determining a roll force self-learning parameter adaptively according to attribute of strip steel and a real-time working condition and improving the rolling precision of the strip steel. The method comprises the following steps: acquiring an attribute parameter of to-be-rolled strip steel; determining a first weight coefficient of a rolling model corresponding to the to-be-rolled strip steel according to thickness, width and finish rolling temperature of the to-be-rolled strip steel; determining the hereditaryform of the to-be-rolled strip steel based on the first weight coefficient, heat number, flow number and rolling roller number of the to-be-rolled strip steel as well as time interval, with the prevision strip steel, of the to-be-rolled strip steel; if the hereditary form of the to-be-rolled strip steel is a short hereditary form, determining the roll force self-learning coefficient correspondingto the to-be-rolled strip steel with a self-learning strategy corresponding to the short hereditary form; and if the hereditary form of the to-be-rolled strip steel is a long hereditary form, determining the roll force self-learning coefficient corresponding to the to-be-rolled strip steel with a self-learning strategy corresponding to the long hereditary form.

Description

technical field [0001] The invention relates to the technical field of hot rolling, in particular to a rolling force parameter self-learning method and device. Background technique [0002] In the industrial automatic control system, the production control system of hot continuous strip steel is relatively complicated. Improving the prediction accuracy of the model in the hot strip rolling process is the goal that the automatic control technology of hot strip rolling is constantly pursuing. In order to meet the strict requirements of strip size and shape accuracy, it is becoming more and more urgent to improve the setting accuracy of rolling parameters. Rolling force is one of the most important parameters in the production process of hot strip rolling, and its calculation model is the core of the setting model of the strip hot rolling finishing mill. The product thickness control level of hot-rolled strip depends largely on the prediction accuracy of the rolling force mod...

AI Technical Summary

Problems solved by technology

In the existing technology, the self-learning method of rolling force parameters is not accurate enough, resulting in a large deviation between the model rolling force and the actual rolling force, and the rolling precision of the strip is low

Images