Comprehensive self-adaptive control method of width of rough rolling band steel

Abstract

The invention provides a comprehensive self-adaptive control method of the width of a rough rolling band steel. The method comprises the steps of acquiring the actually measured rough rolling width data; combining the actually measured width data, vertical rolling force information and estimated width of the band steel in a full length; comprehensively calculating the natural width spread and dog bone width spread at all passes of performed rolling; carrying out multiplying learning calculation to update the model learning at all passes; performing learning for width-related pass after acquiring every actually measured width data each time later; learning global width target after acquiring the actual width value by finish rolling; starting rough rolling to re-calculate after calculated related to rough rolling is done each time, so as to rearrange a roll gap for subsequent vertical rolling. The comprehensive self-adaptive method for hot rolling width has the advantages that, the self-adaptive method of a width model control system is described from the perspective of the system, various rolling information can be fully utilized, and as a result, the hot rolling width control index precision is quickly improved; by adopting the technical method, the width control precision is up to 99.5% under 0-12.5mm, and thus the hot rolling width control precision is quickly improved.

Description

technical field [0001] The invention belongs to the technical field of steel rolling control. Specifically, the invention relates to a comprehensive self-adaptive control method for the width of rough-rolled strip steel. Length and width; comprehensively calculate the natural width and dog-bone width of all passes that have been rolled, and update the model learning of all passes through multiplicative learning calculations; and then carry out width-related passes every time the actual width measurement data is obtained After the actual value of the width is obtained in the finishing rolling, the global width target is learned; after the relevant calculation of each rough rolling is completed, the rough rolling width is recalculated, and the roll gap of the subsequent vertical roll rolling is re-set. Background technique [0002] In the field of hot rolling, the common configuration of rough rolling is two groups of rolling mills, and a width measuring instrument is configur...

AI Technical Summary

Problems solved by technology

This processing method will produce certain deviations under certain circumstances

[0004] (2) Learning the width model itself is difficult

[0007] Due to the natural width of rough rolling and the dog-bone model is difficult to optimize learning, many rough rolling systems cannot start and reset calculations, and manual intervention is also common

[0008] (3) Although the finishing width is obtained after finishing rolling, the value of the finishing width is only used for finishing rolling widening study or rapid correction of the vertical roll gap, and it is useless to make full use of this information; in addition,

[0009] (4) The start-up sequence of conventional finish rolling width learning is after finishing rolling and steel throwing. At this time, the width learning of finishing rolling often cannot work on the subsequent slabs, and can only be used for the subsequent strip slabs

[0028] Paper: Research and Optimization of Width Model Based on R2 Exit Width Measurement (Author: Liu Miao, "Baosteel Technology" Issue 5, 2011), this paper studies the problems existing in Baosteel 2050 hot rolling width model learning, and proposes to use R2 width data Carry out a learning strategy for the average distribution of the width deviation of each pass, but does not involve multiplicative learning and error security widening ratio distribution as mentioned in the present invention

Images