CN111112567A - Comprehensive judgment method for breakout prediction of slab caster - Google Patents
Comprehensive judgment method for breakout prediction of slab caster Download PDFInfo
- Publication number
- CN111112567A CN111112567A CN201911295769.XA CN201911295769A CN111112567A CN 111112567 A CN111112567 A CN 111112567A CN 201911295769 A CN201911295769 A CN 201911295769A CN 111112567 A CN111112567 A CN 111112567A
- Authority
- CN
- China
- Prior art keywords
- thermocouples
- breakout
- alarm
- thermocouple
- bonding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22D—CASTING OF METALS; CASTING OF OTHER SUBSTANCES BY THE SAME PROCESSES OR DEVICES
- B22D11/00—Continuous casting of metals, i.e. casting in indefinite lengths
- B22D11/16—Controlling or regulating processes or operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Mechanical Engineering (AREA)
- Continuous Casting (AREA)
Abstract
A comprehensive judgment method for breakout prediction of a slab caster belongs to the technical field of continuous casting production monitoring methods. The technical scheme is as follows: establishing a cyclic neural network model for each thermocouple on the crystallizer copper plate to judge whether the temperature curve of the thermocouple has the characteristic of adhesive breakout; establishing a logic judgment model according to a propagation rule of a casting blank bonding point on a space to judge whether a thermocouple of a temperature curve with bonding breakout characteristics has bonding breakout space propagation characteristics; all thermocouples with bonding breakout space propagation characteristics are alarm thermocouples, different weights are given to different alarm thermocouples according to the different positions of the alarm thermocouples, and a system gives a final bonding breakout alarm signal when the sum of the weights of the alarm thermocouples is greater than a specified threshold value. The invention utilizes the respective characteristics and advantages of the artificial intelligence model and the logic judgment model, and the miss-report rate and the false-report rate of the comprehensive model algorithm are obviously reduced compared with the single model, and are obviously superior to the traditional model algorithm.
Description
Technical Field
The invention relates to a method for comprehensively judging breakout prediction of a slab caster, belonging to the technical field of continuous casting production monitoring methods.
Background
The high-efficiency continuous casting is an integral technology for realizing high continuous casting rate and high operating rate on the premise of producing high-quality and defect-free casting blanks, and has the characteristics of high yield, low energy consumption, good quality, multiple varieties, high degree of mechanical automation and the like. With the development of the high-efficiency continuous casting technology, the heat load of the crystallizer is obviously increased, the occurrence probability of steel leakage accidents is greatly improved, and the normal production of a continuous casting machine is influenced. In order to accurately forecast the breakout accident, a logic judgment model or an artificial intelligence model is mainly adopted at present based on thermocouple temperature measurement. However, both have characteristics, and it is difficult to achieve zero missing report and zero false report at the same time.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a comprehensive judgment method for breakout prediction of a slab caster, which can simultaneously apply a comprehensive calculation method of a logic judgment model and an artificial intelligence model, can synthesize respective advantages of the two models in the breakout early warning process, and has the advantages of high prediction accuracy and timely prediction.
The technical scheme for solving the technical problems is as follows:
a method for comprehensively judging breakout prediction of a slab caster comprises the following steps:
(1) establishing a circulating neural network model for each thermocouple on the crystallizer copper plate, and judging whether the temperature curve of the thermocouple has the characteristic of adhesive breakout;
(2) establishing a logic judgment model according to a spatial propagation rule of a casting blank bonding point, and judging whether the thermocouples with the bonding breakout characteristic temperature curve in the step 1 have bonding breakout spatial propagation characteristics;
(3) all the thermocouples meeting the conditions in the step 2 are alarm thermocouples, different weights are given to different thermocouples according to different positions of the thermocouples, the weight of the alarm thermocouple at the position of the narrow surface of the crystallizer copper plate is 1.5, the weight of the alarm thermocouple at the position of the most lateral side of the wide surface of the crystallizer copper plate is 1.2, the weight of other alarm thermocouples at the position of the wide surface of the crystallizer copper plate is 1, and when the sum of the weights of the alarm thermocouples is greater than a specified threshold value, a final bonding breakout alarm signal is given.
According to the comprehensive judgment method for the breakout prediction of the slab caster, the cyclic neural network model in the step (1) is obtained by comprehensively analyzing the thermocouple temperature data when historical bonding occurs and is used for judging whether the temperature curve shape of a single thermocouple has bonding characteristics.
According to the comprehensive judgment method for breakout prediction of the slab caster, the logic judgment model in the step (2) is as follows: and if the judgment of any two thermocouples in the same column with the characteristic temperature curve of the bonded breakout meets the following formula, judging the pair of thermocouples as alarm thermocouples with the space propagation characteristic of the bonded breakout:
wherein S is the arrangement distance of the thermocouple pair on the crystallizer copper plate, and the unit is mm; v is the drawing speed of the continuous casting machine, and the unit is mm/s; and delta t is the difference of the time of which the temperature curves of the pair of thermocouples respectively report the characteristic of bonding breakout, and the unit is s.
Any two thermocouples in the same row with the characteristic temperature curve of the bonded bleed-out are judged to satisfy the following formula:
wherein S is the arrangement distance of the thermocouple pair on the crystallizer copper plate, and the unit is mm; v is the drawing speed of the continuous casting machine, and the unit is mm/s; and delta t is the difference of the time of which the temperature curves of the pair of thermocouples respectively report the characteristic of bonding breakout, and the unit is s.
According to the comprehensive judgment method for breakout prediction of the slab caster, the threshold in the step (3) is set according to different steel types, the threshold of low-carbon steel and high-carbon steel is 8, the threshold of medium-carbon steel is 7, and the threshold of peritectic steel is 6.
The invention has the beneficial effects that:
the method is the first creation of the breakout prediction of the slab caster, solves the problem that zero breakout prediction and zero false prediction are difficult to achieve by the two groups of prediction methods at present, and has the advantages of high prediction accuracy and timely prediction.
The invention fully utilizes the respective characteristics and advantages of the artificial intelligence model and the logic judgment model, and the miss-report rate and the false-report rate of the comprehensive model algorithm are obviously reduced compared with the single model, thereby being obviously superior to the traditional model algorithm.
The invention adopts computer programming modeling to automatically acquire real-time thermocouple temperature data, automatically calculates and judges, and has simple and convenient operation and good practicability.
Drawings
FIG. 1 is a portion of a thermocouple temperature profile of the first embodiment;
FIG. 2 is a portion of the thermocouple temperature profile of the first embodiment;
FIG. 3 is a portion of a thermocouple temperature profile of the first embodiment;
FIG. 4 is a portion of the thermocouple temperature profile of example two;
FIG. 5 is a portion of the thermocouple temperature profile of example two;
FIG. 6 is a portion of the thermocouple temperature profile of example two;
FIG. 7 is a portion of the thermocouple temperature profile of example three;
FIG. 8 is a portion of the thermocouple temperature profile of example III;
FIG. 9 is a portion of the thermocouple temperature profile of example three.
Detailed Description
The invention is further illustrated by the following examples in conjunction with the accompanying drawings.
The thermocouple graphs in the attached figures are graphs of 3, 4 and 5 rows of thermocouples on the fixed side of the crystallizer, and each row of thermocouples is respectively provided with 3 rows from top to bottom and is displayed in the same graph. Wherein 3 columns are thermocouples at the most lateral side of the wide surface, and 4 and 5 columns are thermocouples at other positions of the wide surface.
Example 1
FIGS. 1-3 are graphs of thermocouple temperatures detected by the present invention, wherein FIGS. 1, 2, and 3 are graphs of thermocouple columns 3, 4, and 5, respectively, on the fixed side of the mold. And judging that the shape of the temperature curve graph of the circulating neural network does not have the bonding breakout characteristic, and not performing final bonding breakout alarm.
Example 2
FIGS. 4-6 are graphs of thermocouple temperatures detected by the present invention, wherein FIGS. 4, 5, and 6 are graphs of thermocouple columns 3, 4, and 5 on the fixed side of the mold, respectively. The temperature curve shapes of 2 rows of thermocouples in the temperature curves of 3, 4 and 5 rows of thermocouples of the circulating neural network are judged to have the characteristic of bonding breakout. For the thermocouples judged by the circulating neural network and having the characteristic of bonding breakout, the formula is not satisfied between the thermocouples in each column
The thermocouples in each row do not satisfy the formula
And (5) not performing final bonding breakout alarming.
Example 3
Fig. 7-9 are graphs of thermocouple temperatures detected by the present invention, wherein fig. 7, 8 and 9 are graphs of thermocouple columns 3, 4 and 5 on the fixed side of the crystallizer, respectively. The temperature curve shapes of 3 rows of thermocouples of the temperature curve graphs of 3 columns of thermocouples of the circulating neural network are judged to have the characteristic of bonding breakout. For the thermocouples judged by the recurrent neural network and having the characteristic of bonding breakout, the formula is satisfied between the thermocouples in each column
The thermocouples in each row satisfy the formula
Therefore, the 9 thermocouples all have the space propagation characteristic of the bonded breakout. The 3 columns are thermocouples at the most lateral side of the wide surface, the weight of each thermocouple is 1.2, the 4 columns and the 5 columns are thermocouples at other positions of the wide surface, and the weight of each thermocouple is 1. Thus, the sum of the alarm thermocouple weights is: 1.2 × 3+1 × 3+1 × 3 ═ 9.6. The produced steel grade is low-carbon steel, and the alarm threshold value is 8. And the sum of the weights 9.6 of the alarm thermocouples is greater than the alarm threshold value 8, so that the final bonding breakout alarm judgment is carried out.
Claims (4)
1. A method for comprehensively judging breakout prediction of a slab caster is characterized by comprising the following steps: it comprises the following steps:
(1) establishing a cyclic neural network model for each thermocouple on the crystallizer copper plate to judge whether the temperature curve of the thermocouple has the characteristic of adhesive breakout;
(2) establishing a logic judgment model according to a spatial propagation rule of a casting blank bonding point to judge whether the thermocouples with the bonding breakout characteristic temperature curves in the step 1 have bonding breakout spatial propagation characteristics;
(3) and (3) all the thermocouples meeting the conditions in the step (2) are alarm thermocouples, different weights are given to different alarm thermocouples according to different positions of the alarm thermocouples, the weight of the alarm thermocouple at the narrow surface position of the crystallizer copper plate is 1.5, the weight of the alarm thermocouple at the most lateral position of the wide surface of the crystallizer copper plate is 1.2, the weights of other alarm thermocouples at the wide surface position of the crystallizer copper plate are 1, and when the sum of the weights of the alarm thermocouples is greater than a specified threshold value, a final bonding breakout alarm signal is given.
2. The slab caster breakout prediction comprehensive judgment method according to claim 1, characterized in that: and (2) the cyclic neural network model in the step (1) is obtained by comprehensively analyzing the thermocouple temperature data when historical bonding occurs and is used for judging whether the temperature curve shape of a single thermocouple has bonding characteristics.
3. The slab caster breakout prediction comprehensive judgment method according to claim 1, characterized in that: the logic judgment model in the step (2) is as follows: and if the judgment of any two thermocouples in the same column with the characteristic temperature curve of the bonded breakout meets the following formula, judging the pair of thermocouples as alarm thermocouples with the space propagation characteristic of the bonded breakout:
wherein S is the arrangement distance of the thermocouple pair on the crystallizer copper plate, and the unit is mm; v is the drawing speed of the continuous casting machine, and the unit is mm/s;
respectively reporting the time difference of the temperature curve with the characteristic of bonding breakout for the pair of thermocouples, wherein the unit is s;
any two thermocouples in the same row with the characteristic temperature curve of the bonded bleed-out are judged to satisfy the following formula:
wherein S is the arrangement distance of the thermocouple pair on the crystallizer copper plate, and the unit is mm; v is the drawing speed of the continuous casting machine, and the unit is mm/s;
the difference in time in s for the temperature curves with the characteristic of sticking breakout is reported for the pair of thermocouples.
4. The slab caster breakout prediction comprehensive judgment method according to claim 1, characterized in that: the threshold value in the step (3) is set according to different steel types, the threshold values of the low-carbon steel and the high-carbon steel are 8, the threshold value of the medium-carbon steel is 7, and the threshold value of the peritectic steel is 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911295769.XA CN111112567B (en) | 2019-12-16 | 2019-12-16 | Comprehensive judgment method for breakout prediction of slab caster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911295769.XA CN111112567B (en) | 2019-12-16 | 2019-12-16 | Comprehensive judgment method for breakout prediction of slab caster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111112567A true CN111112567A (en) | 2020-05-08 |
| CN111112567B CN111112567B (en) | 2021-09-21 |
Family
ID=70499283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911295769.XA Active CN111112567B (en) | 2019-12-16 | 2019-12-16 | Comprehensive judgment method for breakout prediction of slab caster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111112567B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115446276A (en) * | 2022-10-05 | 2022-12-09 | 大连理工大学 | Continuous casting breakout early warning method for recognizing V-shaped bonding characteristics of crystallizer copper plate based on convolutional neural network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101850410A (en) * | 2010-06-22 | 2010-10-06 | 攀钢集团钢铁钒钛股份有限公司 | Continuous casting breakout prediction method based on neural network |
| CN102825234A (en) * | 2012-09-25 | 2012-12-19 | 鞍钢股份有限公司 | Bonded bleed-out judging and alarming method |
| KR101242086B1 (en) * | 2011-11-16 | 2013-03-11 | 주식회사 포스코 | Injection nozzle for mold flux |
| CN105328155A (en) * | 2015-10-08 | 2016-02-17 | 东北电力大学 | Steel leakage visualized characteristic forecasting method based on improved neural network |
| CN105689675A (en) * | 2015-07-24 | 2016-06-22 | 安徽工业大学 | Cure control method for continuous casting steel breakout by sticking |
| CN106980729A (en) * | 2015-07-24 | 2017-07-25 | 安徽工业大学 | A kind of continuous casting breakout prediction method based on mixed model |
-
2019
- 2019-12-16 CN CN201911295769.XA patent/CN111112567B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101850410A (en) * | 2010-06-22 | 2010-10-06 | 攀钢集团钢铁钒钛股份有限公司 | Continuous casting breakout prediction method based on neural network |
| KR101242086B1 (en) * | 2011-11-16 | 2013-03-11 | 주식회사 포스코 | Injection nozzle for mold flux |
| CN102825234A (en) * | 2012-09-25 | 2012-12-19 | 鞍钢股份有限公司 | Bonded bleed-out judging and alarming method |
| CN105689675A (en) * | 2015-07-24 | 2016-06-22 | 安徽工业大学 | Cure control method for continuous casting steel breakout by sticking |
| CN106980729A (en) * | 2015-07-24 | 2017-07-25 | 安徽工业大学 | A kind of continuous casting breakout prediction method based on mixed model |
| CN105328155A (en) * | 2015-10-08 | 2016-02-17 | 东北电力大学 | Steel leakage visualized characteristic forecasting method based on improved neural network |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115446276A (en) * | 2022-10-05 | 2022-12-09 | 大连理工大学 | Continuous casting breakout early warning method for recognizing V-shaped bonding characteristics of crystallizer copper plate based on convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111112567B (en) | 2021-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | Mold breakout prediction in slab continuous casting based on combined method of GA-BP neural network and logic rules | |
| CN105108097B (en) | Hybrid model for continuous casting breakout prediction | |
| US11105758B2 (en) | Prediction method for mold breakout based on feature vectors and hierarchical clustering | |
| CN102886504B (en) | Crystallizer bleed-out visual forecasting method based on machine vision | |
| CN105328155B (en) | A kind of bleed-out visual feature forecasting procedure based on improvement neutral net | |
| CN103336507B (en) | Automatically the statistical modeling divided based on the multi-modal collaborative period and on-line monitoring method | |
| CN109365769A (en) | A kind of crystallizer bleedout prediction electric thermo method based on mixed model judgement | |
| CN108580827B (en) | A method of Crystallizer bleed-out is forecast based on Agglomerative Hierarchical Clustering | |
| CN111112567B (en) | Comprehensive judgment method for breakout prediction of slab caster | |
| CN104181196A (en) | On-line detection method of continuous casting billet surface longitudinal crack | |
| CN102527971A (en) | Online forecasting method for internal crack defect of casting blank | |
| CN105522133A (en) | Judgment method of bonding behavior between continuous casting slab and crystallizer copper plate | |
| CN111570748B (en) | Crystallizer bleed-out forecasting method based on image processing | |
| CN107096899A (en) | The crystallizer bleedout prediction electric thermo system that a kind of logic-based judges | |
| CN104550801B (en) | Control method and control device for continuous bloom casting soft reduction | |
| Duan et al. | Integrated approach to density-based spatial clustering of applications with noise and dynamic time warping for breakout prediction in slab continuous casting | |
| Zhang et al. | Breakout prediction for continuous casting using genetic algorithm-based back propagation neural network model | |
| CN112926622B (en) | Crystallizer breakout prediction method for generating countermeasure network based on feature vector and SWGAN-GP | |
| CN111496211B (en) | Method for tracking and identifying bonding point on surface of casting blank | |
| CN113935416A (en) | Continuous casting breakout prediction method for multi-state crystallizer | |
| CN113319259B (en) | Bonding breakout logic judgment method based on space-time sequence characteristics | |
| CN107252882A (en) | Method and system for Direct Rolling high temperature billet continuous casting process breakout prediction | |
| CN113505785A (en) | Crystallizer steel leakage forecasting method based on eigenvector and logistic regression model | |
| Zhang et al. | Breakout prediction based on BP neural network of LM algorithm in continuous casting process | |
| CN115330996A (en) | Method for detecting interested area of continuous casting bleed-out based on spatial characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Gao Yu Inventor after: Li Yang Inventor after: Liu Zhanli Inventor after: Wang Peng Inventor after: Zhang Shihui Inventor after: Li Huiya Inventor after: Zhao Xiaohu Inventor before: Gao Yu Inventor before: Li Yang Inventor before: Liu Zhanli Inventor before: Zhang Peng Inventor before: Zhang Shihui Inventor before: Li Huiya Inventor before: Zhao Xiaohu |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |