CN107626748A

CN107626748A - A kind of computer model control method of hot-rolled product width

Info

Publication number: CN107626748A
Application number: CN201610569128.9A
Authority: CN
Inventors: 李琳
Original assignee: Angang Steel Co Ltd
Current assignee: Angang Steel Co Ltd
Priority date: 2016-07-18
Filing date: 2016-07-18
Publication date: 2018-01-26
Anticipated expiration: 2036-07-18
Also published as: CN107626748B

Abstract

A kind of computer model control method of hot-rolled product width, this method are to spread learning value to roughing by steel grade, product width, roughing mill, width drafts and inlet thickness classify the method for reading and storage.Compared with prior art, the beneficial effects of the invention are as follows：A kind of computer model control method of hot-rolled product width, learning value is spreaded by steel grade, product width, roughing mill, width drafts and inlet thickness to roughing to classify, every factor in the sorting technique is obvious with spreading variable quantity correlation, and do not influenceed by roughing passage combination and product thickness, make roughing spread learning value fluctuation range substantially to narrow, it is adapted to the production of various incoming slab thickness, product width control accuracy can be effectively improved, reduces resulting waste product resection rate and cost allowance.

Description

A kind of computer model control method of hot-rolled product width

Technical field

The present invention relates to steel rolling automation field, more particularly to a kind of computer model controlling party of hot-rolled product width Method.

Background technology

With hot rolling line product development, specification extension and production schedule layout demand, certain line of equipment of Hot Strip Mill 1780 200mm thickness slab scheduling amounts continue to increase, because 1780 Hot Line convenient source slab thicknesses are 230mm, and original width mould Type does not have special counter-measure to the incoming slab of different-thickness, therefore causes width quality precise decreasing, width block and give up Product amount is substantially increased, and massive losses are brought to entreprise cost.

The principal element for the influence product width control accuracy that presently, there are is, same to steel grade, same specification product carry out flitch Base thickness is different, and usual 230mm thickness slabs are respectively back and forth rolled in two, roughing region roughing mill by 3 passages, hereinafter referred to as Combine and roll for 3+3 passages.When incoming slab thickness is less than 200mm, such as 170mm thickness incoming slabs are due to raw thickness It is smaller thus roughing region by 1+3 passages combine rolling can meet product demand.Due to incoming slab thickness and rolling The difference of passage combination, the crucial number such as its drafts, inlet port thickness in the same passes of same roughing mill According to that will change, developed width variable quantity is necessarily different, and the self-study of computer-controlled model is spreaded in current roughing Habit value is to become by what steel grade, product thickness, product width, roughing mill and each roughing mill passage were classified without supplied materials thickness The factor of change, therefore when incoming slab thickness changes, self study value is spreaded in roughing can occur larger fluctuation, it fluctuates table Relatively large deviation be present between bright roughing outlet developed width and desired value width, most directly affect product width precision at last.

The content of the invention

It is wide by steel grade, product it is an object of the invention to provide a kind of computer model control method of hot-rolled product width Degree, roughing mill, width drafts and inlet thickness spread learning value to roughing and classified, every factor in the sorting technique It is obvious with spreading variable quantity correlation, and do not influenceed by roughing passage combination and product thickness, roughing is spreaded study Value fluctuation range substantially narrows, and is adapted to the production of various incoming slab thickness, can effectively improve product width control essence Degree, reduces resulting waste product resection rate and cost allowance.

To achieve the above object, the present invention is realized using following technical scheme：

A kind of computer model control method of hot-rolled product width, this method be by steel grade, product width, roughing mill, Width drafts and inlet thickness spread the method that learning value carries out classification reading and storage to roughing；

Data setting moment startup roughing absolute spread learning value is carried out to roughing mill in hot rolling computer width control system model Classification read module, the module is collected and sentences to steel grade, product width, inlet thickness, width drafts information first Its disconnected correctness, the output error warning message if any one data have problem, while roughing to be read is spreaded into Habit value is set as initial value, ensures that follow-up roughing computer model calculates related setting data；Data above item is correctly read Afterwards, you can calculate steel grade call number, product width call number, inlet thickness call number, width drafts call number, root respectively Determine that the recording mechanism of learning value file is spreaded in roughing according to calculated each classified index number and roughing mill, final reading roughing is spreaded Measure learning value to be used to subsequently participate in the calculating of roughing computer model, obtain the every setting data of roughing mill；

The actual value collection of roughing region is completed in hot rolling computer width control system model and the absolute spread study moment starts slightly The classification storage module of absolute spread learning value is rolled, the module is believed steel grade, product width, inlet thickness, width drafts first Breath is collected and judges its correctness, the output error warning message if any one data have problem, while by roughing The renewal mark for spreading learning value is set as 0, represents renewal failure, is used for subsequent processing logic and data retrospective analysis；With After upper data item is correctly read, you can calculate steel grade call number, product width call number, inlet thickness call number, width respectively Drafts call number is spent, determines that the record of learning value file is spreaded in roughing according to calculated each classified index number and roughing mill Number, the roughing being most calculated at last is spreaded study updated value and is stored in the corresponding book of final entry of learning value file, and confession is next Secondary roughing data setting uses.

Compared with prior art, the beneficial effects of the invention are as follows：

A kind of computer model control method of hot-rolled product width, depressed by steel grade, product width, roughing mill, width Amount and inlet thickness spread learning value to roughing and classified, and every factor in the sorting technique is with spreading variable quantity correlation Substantially, and not influenceed by roughing passage combination and product thickness, roughing is spreaded learning value fluctuation range and substantially narrow, It is adapted to the production of various incoming slab thickness, product width control accuracy can be effectively improved, reduces resulting useless Product resection rate and cost allowance.

Brief description of the drawings

Fig. 1 is the workflow diagram that learning value classification read module is spreaded in roughing.

Fig. 2 is the workflow diagram that learning value classification storage module is spreaded in roughing.

Fig. 3 is that former sorting technique spreads study instantaneous value changing trend diagram.

Fig. 4 is that the sorting technique of the present invention spreads study instantaneous value changing trend diagram.

Fig. 5 is former sorting technique product width precision statisticses figure.

Fig. 6 is the sorting technique product width precision statisticses figure of the present invention.

Embodiment

Embodiments of the present invention are further illustrated with reference to specific embodiment：

As Figure 1-Figure 2, a kind of computer model control method of hot-rolled product width, this method are by steel grade, production Product width, roughing mill, width drafts and inlet thickness spread the method that learning value carries out classification reading and storage to roughing；

Data setting moment startup roughing absolute spread learning value is carried out to roughing mill in hot rolling computer width control system model Classification read module, the module is collected and sentences to steel grade, product width, inlet thickness, width drafts information first Its disconnected correctness, the output error warning message if any one data have problem, while roughing to be read is spreaded into Habit value is set as initial value, ensures that follow-up roughing computer model calculates related setting data；Data above item is correctly read Afterwards, you can calculate steel grade call number, product width call number, inlet thickness call number, width drafts call number, root respectively It can determine that the recording mechanism of learning value file is spreaded in roughing according to calculated each classified index number and roughing mill, finally read roughing Absolute spread learning value is used to subsequently participate in the calculating of roughing computer model, obtains the every setting data of roughing mill.Involved by this module And part core code it is as follows：

The actual value collection of roughing region is completed in hot rolling computer width control system model and the absolute spread study moment starts slightly The classification storage module of absolute spread learning value is rolled, the module is believed steel grade, product width, inlet thickness, width drafts first Breath is collected and judges its correctness, the output error warning message if any one data have problem, while by roughing The renewal mark for spreading learning value is set as 0, represents renewal failure, is used for subsequent processing logic and data retrospective analysis；With After upper data item is correctly read, you can calculate steel grade call number, product width call number, inlet thickness call number, width respectively Drafts call number is spent, can determine that the note of learning value file is spreaded in roughing according to calculated each classified index number and roughing mill Record number, the roughing being most calculated at last is spreaded study updated value and is stored in the corresponding book of final entry of learning value file, under One time roughing data setting uses.Part core code involved by this module is as follows：

Embodiment 1：

It is shown in Table the 1 original scheme data for being the present embodiment：

Table 1：Original scheme data implement the cold chi value (unit of sample table:mm)

Steel grade

Rolling volume number

Product thickness

Product width

Slab thickness

Width of plate slab

Slab length

General carbon

Volume 200

3.50

<950

230、170

1050

9640

First, by taking certain equipment of Hot Strip Mill No. 2 roughing mill of 1780 line as an example, the classification to roughing absolute spread learning value is read with depositing Store up module file record and carry out comparative illustration：

Table 2：Sorting technique of the present invention spreads learning value file content (unit:mm)

The classification reading of roughing absolute spread learning value with memory module according to steel grade call number, product width call number, enter Mouth thickness call number, width drafts call number spread learning files to roughing, and record is read out and stored operation accordingly Process can more intuitively be showed by table 2, and the data cells in table 2 represent the record that learning files are spreaded in roughing.

Table 3：It is original to spread learning value file content (unit:mm)

By the contrast of table 2 and table 3 it can be found that original learning value document classification that spreads can not meet and tackle come flitch The situation that base thickness changes or the combination of roughing passage changes, because it does not consider the changing factor of inlet thickness The factor changed with width drafts.

2nd, new and old roughing spreads study classification method to spreading the influence of learning value：

See Fig. 3, by taking the 1st and the 3rd passage of certain equipment of Hot Strip Mill No. 2 roughing mill of 1780 line as an example, illustrate original classification The fluctuation situation of study instantaneous value is spreaded in roughing during method, it can be seen that it is fluctuated between -10mm to+10mm, fluctuation range Up to 20mm, the more big then product width precision of fluctuation range is lower.

It is that No. 2 roughing mills of sorting technique of the present invention spread the fluctuation situation of study instantaneous value to see Fig. 4, and its transverse axis is width Drafts, it can be seen that except width drafts is less than 10mm and more than two sections of 35mm because amount of rolling is less, spread study wink Duration fluctuation is larger outer, and rest interval spreads the study very intensive fluctuation range of instantaneous value between -3mm to+3mm, thus It can be seen that new sorting technique is fairly obvious for product width control effect.

3rd, influence of the study classification method to product width precision is spreaded in new and old roughing：

See Fig. 5, Fig. 6, two figure abscissas are coil of strip count number, it is seen that study classification method is spreaded in the roughing of the present invention After implementation, product width precision is increased dramatically, and product width precision fluctuation scope is obviously reduced.The implementation solution of the present invention Influence problem of the incoming slab thickness change of having determined to hot-rolled product width accuracy, is greatly reduced width quality blocking rate, reduces Resulting waste product resection rate and cost allowance.

Claims

1. a kind of computer model control method of hot-rolled product width, it is characterised in that this method is wide by steel grade, product Degree, roughing mill, width drafts and inlet thickness spread the method that learning value carries out classification reading and storage to roughing；

Data setting moment startup roughing absolute spread learning value is carried out to roughing mill in hot rolling computer width control system model to divide Class read module, the module are collected to steel grade, product width, inlet thickness, width drafts information and judge it first Correctness, the output error warning message if any one data have problem, while learning value is spreaded into roughing to be read It is set as initial value, ensures that follow-up roughing computer model calculates related setting data；After data above item is correctly read, i.e., Steel grade call number, product width call number, inlet thickness call number, width drafts call number can be calculated respectively, according to Each classified index number and roughing mill calculated determines that the recording mechanism of learning value file is spreaded in roughing, final to read roughing absolute spread Habit is worth to be calculated for subsequently participating in roughing computer model, obtains the every setting data of roughing mill；

The actual value collection of roughing region is completed in hot rolling computer width control system model and the startup roughing of absolute spread study moment is wide The classification storage module of exhibition amount learning value, the module are entered to steel grade, product width, inlet thickness, width drafts information first Row is collected and judges its correctness, the output error warning message if any one data have problem, while roughing is spreaded The renewal mark of learning value is set as 0, represents renewal failure, is used for subsequent processing logic and data retrospective analysis；Above number After correctly being read according to item, you can calculate steel grade call number, product width call number, inlet thickness call number, width pressure respectively Lower amount call number, determine that the recording mechanism of learning value file is spreaded in roughing according to calculated each classified index number and roughing mill, most The roughing being calculated at last is spreaded study updated value and is stored in the corresponding book of final entry of learning value file, for roughing next time Data setting uses.