CN113362903A - Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter - Google Patents
Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter Download PDFInfo
- Publication number
- CN113362903A CN113362903A CN202110614850.0A CN202110614850A CN113362903A CN 113362903 A CN113362903 A CN 113362903A CN 202110614850 A CN202110614850 A CN 202110614850A CN 113362903 A CN113362903 A CN 113362903A
- Authority
- CN
- China
- Prior art keywords
- percent
- phosphorus
- molten iron
- lime
- mass percent
- 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
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C10/00—Computational theoretical chemistry, i.e. ICT specially adapted for theoretical aspects of quantum chemistry, molecular mechanics, molecular dynamics or the like
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C5/00—Manufacture of carbon-steel, e.g. plain mild steel, medium carbon steel or cast steel or stainless steel
- C21C5/28—Manufacture of steel in the converter
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C7/00—Treating molten ferrous alloys, e.g. steel, not covered by groups C21C1/00 - C21C5/00
- C21C7/0037—Treating molten ferrous alloys, e.g. steel, not covered by groups C21C1/00 - C21C5/00 by injecting powdered material
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C7/00—Treating molten ferrous alloys, e.g. steel, not covered by groups C21C1/00 - C21C5/00
- C21C7/04—Removing impurities by adding a treating agent
- C21C7/064—Dephosphorising; Desulfurising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/045—Combinations of networks
-
- 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C60/00—Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C2300/00—Process aspects
- C21C2300/06—Modeling of the process, e.g. for control purposes; CII
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P10/00—Technologies related to metal processing
- Y02P10/25—Process efficiency
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Materials Engineering (AREA)
- Metallurgy (AREA)
- General Health & Medical Sciences (AREA)
- Organic Chemistry (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Manufacturing & Machinery (AREA)
- Carbon Steel Or Casting Steel Manufacturing (AREA)
Abstract
The invention relates to a method for intelligently adding lime in TSC (thyristor switched capacitor) stage of a large converter, belonging to the technical field of ferrous metallurgy. The technical scheme of the invention is as follows: predicting according to a metallurgical mechanism by adopting a mechanism model, training data, deeply learning artificial intelligence based on a BP neural network, carrying out standardization processing on input data based on the mechanism model, constructing internal relation between the input data and output data, and obtaining the output data through artificial intelligence calculation. The invention has the beneficial effects that: by adopting a mode of 'mechanism model + artificial intelligence prediction', on the basis of fully analyzing influence factors of the lime addition amount of the converter, the internal relation among the influence factors of the converter is constructed, the method for intelligently adding lime in TSC (thyristor switched capacitor) stages of the large-scale converter under different material conditions is realized, the problem of accurate addition of the lime of the converter is solved, quality accidents and production accidents are reduced, and the smelting cost is reduced.
Description
Technical Field
The invention relates to a method for intelligently adding lime in TSC (thyristor switched capacitor) stage of a large converter, belonging to the technical field of ferrous metallurgy.
Background
Lime is a main auxiliary material for converter smelting, and has the main function of forming a reasonable slag system in the smelting process and removing harmful elements such as steel slag P, S and the like through converter slag. Aiming at the converter controlled by the sublance method, lime is added mainly before the TSC of the sublance is measured, namely the TSC smelting stage of the converter. The general algorithm for the lime addition is to calculate through binary alkalinity, the method obtains the CaO content of lime through the detection and test of lime raw materials, and then calculates the lime addition according to the silicon content of molten iron.
Although the method can provide a certain guidance for the lime addition amount in the converter operation process, the method is not accurate, the dynamic change of lime components is not fully considered, the influence of other Ca-containing auxiliary materials such as dolomite, steel slag and other raw materials is not fully considered, the influence of lime melting at the TSC stage on the dephosphorization effect is not fully considered, the influence of lime on the converter smelting heat balance is not fully considered, and the difference between different devices and between shifts is not fully considered. Particularly, under the background of reducing iron consumption of the converter in the whole country at present, a large amount of self-produced cold materials such as steel slag and the like are added into each steel company, and the self-produced cold materials have poor component stability and cause great production fluctuation. All the factors can cause inaccuracy of lime addition, if the lime addition is too small, the converter smelting slag amount is small, the converter dephosphorization effect is influenced, the product components are not compatible, and the long service life of the converter lining is not facilitated; if the lime is added in too large amount, not only can the cost be wasted, but also serious production accidents such as large smelting slag amount and abnormal splashing in the smelting process can be caused, and steel can not be tapped at the end point of the converter.
Patent application No. 201710074989.4 discloses a converter alkalinity dynamic control method, which corrects the alkalinity based on the setting and adjustment of the end point temperature, the end point carbon, the end point phosphorus, the dephosphorization efficiency and the converter slag amount.
Patent application number 201310449548.X discloses a method for determining lime using amount by calculating converter remaining slag amount, the method determines the remaining slag amount through molten iron silicon content, obtains the actual converter remaining slag weight through observing the furnace slag volume and calculating the slag density, and adds lime into a converter according to the proportion of the remaining slag amount to the lime 1: 1-0.5.
In a document [ application of a YaoHu, Yangzhou Chengzhou, Xushenglin, gray algorithm in calculation of lime addition amount in converter steelmaking, university of Hangzhou electronic technology, 2015], the deviation between the current heat of a 50t converter and the lime addition amount is forecasted through the gray algorithm, and a residual sequence is established for correction.
Although the above method involves prediction of lime addition, all factors affecting lime addition are not fully considered, and the inherent relationship between the factors is not considered. Such as: the influence of the stability of the team, the converter equipment and the materials is not considered, and the influence of the lime serving as an auxiliary raw material with a certain cooling effect on the heat balance of the converter is not considered. Due to the complex internal relation of the multiple factors, the current conventional mathematical statistical method cannot accurately predict and guide production. Therefore, a reliable method for guiding the addition of lime into the converter is needed.
Disclosure of Invention
The invention aims to provide a method for intelligently adding lime in TSC (thyristor switched capacitor) stage of a large converter. Mainly aiming at a 260t large-scale top-bottom combined blown converter, the slagging method is a single slag method, the charging system is quantitative charging, and sublance control is adopted. The method adopts a mode of 'mechanism model + artificial intelligence prediction', constructs the internal relation among all the influence factors of the converter on the basis of fully analyzing the influence factors of the lime addition amount of the converter, and realizes the intelligent addition of the lime at the TSC stage of the large converter under different material conditions.
The technical scheme of the invention is as follows: a method for intelligently adding lime into a large converter comprises the following steps:
(1) predicting according to a metallurgical mechanism by adopting a mechanism model, wherein the formula of the mechanism model is as follows:
wherein WshRepresenting the lime adding amount, and the data range is 22kg/t-49 kg/t;
ashcrepresenting the mass percent of CaO of lime, and the data range is 88-93 percent;
ashsrepresents lime SiO2Mass percent, data range 2% -5%;
Wqsrepresenting the addition amount of light-burned dolomite, and the data range is 7.5kg/t-30.1 kg/t;
aqscrepresenting the CaO mass percent of the light-burned dolomite, and the data range is 40-48 percent;
aqssstands for light-burned dolomite SiO2Mass percent, data range 0.1% -3.3%;
Wgzrepresenting the addition of steel slag and other cold materials, and the data range is less than 113 kg/t;
agzcrepresenting the CaO mass percent of the steel slag, and the data range is 20-45 percent;
agzsSiO for steel slag2Mass percent, data range 8% -23%;
Wtsrepresenting the adding amount of molten iron, and the data range is 900kg/t-1085 kg/t;
astrepresenting the mass percent of Si in the molten iron, and the data range is 0.15-0.60%;
2.14 conversion of Si to SiO2The conversion coefficient of (a);
r represents alkalinity which is determined by an alkalinity selection model, and the alkalinity is divided into four gradients, wherein the values are respectively 3, 3.5, 4 and 4.5;
(2) carrying out artificial intelligence prediction, adopting a BP neural network architecture as a basis for deep learning artificial intelligence, wherein the input of the artificial intelligence prediction is 16-dimensional data, the output of the artificial intelligence prediction is 1-dimensional data, the 16-dimensional input data are respectively the lime amount of a mechanism model, a smelting team, a smelting furnace base, steel grade requirements, scrap steel, molten iron temperature, molten iron weight, molten iron manganese, molten iron silicon, molten iron phosphorus, steel slag, dolomite addition amount, TSC phosphorus, TSC manganese, TSC carbon and TSC temperature, and the 1-dimensional output data are the lime amount; the input data is standardized based on a mechanism model, and the processing method is as follows:
y is a parameter value to be standardized, ystdIs the standard deviation of this set of parameters, yaveIs the average of this set of parameters;
constructing an internal relation between input data and output data, and obtaining the output data through artificial intelligence calculation;
(3) the correction error required for the calculation is reduced by means of the denormalization, which is carried out by means of the following formula
y=y*×yave+ystd
(4) The excitation function is of the form:
in the step (1), the principle of selecting a mechanism model is as follows:
the phosphorus content of the steel grade requires to establish 4 gradient gears, wherein the phosphorus mass percentage requirement is less than or equal to 0.015 percent, the phosphorus mass percentage requirement is more than 0.015 percent and less than or equal to 0.020 percent, the phosphorus mass percentage requirement is more than 0.020 percent and less than or equal to 0.025 percent, and the phosphorus mass percentage requirement is more than 0.025 percent;
for steel grades with the end point phosphorus element mass percentage less than 0.02% and the end point temperature more than 1650 ℃, the converter end point needs high-temperature steel tapping, the dephosphorization efficiency is lower than that of common steel grades, so the alkalinity is selected to be higher, the grade of the steel grade such as phosphorus mass percentage requires the first gear, and the alkalinity is selected to be 4.5; the requirement of phosphorus mass percent is two, and the alkalinity is selected to be 4;
for steel grades with the end point phosphorus element mass percentage more than 0.02% and the end point temperature more than 1650 ℃, as the steel grades require to keep partial phosphorus content and the subsequent ferrophosphorus adjustment is needed, the process alkalinity of the steel grade is low, which is beneficial to the decarbonization and phosphorus conservation of a converter, the sensitivity of the steel grade to the molten iron phosphorus mass percentage is low, and the alkalinity is 3;
for steel grades with the end point temperature of less than 1650 ℃, the phosphorus element requirement of the steel grade and the phosphorus element requirement of molten iron need to be comprehensively considered, so that the dephosphorization rate in the smelting process and the phosphorus element of tapping meet the control requirement, the phosphorus mass percentage of the molten iron is less than or equal to 0.15 percent, the phosphorus mass percentage requirement of the steel grade is that firstly, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is less than or equal to 0.15 percent and the mass percent of phosphorus in steel grade is required to be 2, the alkalinity is selected to be 3.5; the phosphorus mass percent of the molten iron is less than or equal to 0.15 percent, and the phosphorus mass percent of the steel grade is required to be 3.0 when the phosphorus mass percent of the steel grade is three and four; the mass percent of phosphorus in molten iron is more than 0.15 percent, the mass percent of phosphorus in steel grade is that firstly, the alkalinity is selected to be 4.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be two, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is ③, the alkalinity is 3.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be four, the alkalinity is selected to be 3.
In the step (2), the neural network model structure needs to adopt a structure with 3 hidden layers, the number of nodes of the first hidden layer is 3, the number of nodes of the second hidden layer is 50, and the number of nodes of the third hidden layer is 5, the model has strong nonlinear mapping capability, and is suitable for solving engineering problems such as converter smelting process with complex internal influence factors. Therefore, the optimal solution can be well found out by matching the intelligent model with the theoretical model, and the problem of inaccurate lime addition is solved.
Training data of artificial intelligent prediction are screened by a data preference rule model, data acquisition errors or no typicality data are screened out, the model calculation precision is improved, the interference of error data on model calculation is reduced, and a specific preference rule is as follows:
the weight of molten iron and the weight of scrap steel are selected preferably from 280t to 320 t. For a 260t large converter, the loading capacity is about 300t generally, and if the loading capacity does not meet the requirement, the problem of inaccurate data acquisition exists;
the silicon content of the molten iron is less than 0.6 percent, the selection is preferred, the silicon content is too high, the problem of laboratory inspection deviation exists, meanwhile, the too high silicon content is a special event, special operation of a converter is needed, and the universality and the representativeness are avoided;
thirdly, the lime addition amount is more than 6t, the selection is preferred, and for a large converter, the problem of data acquisition error due to too small lime addition amount is solved;
the addition amount of the light-burned dolomite is more than 2t, the light-burned dolomite is selected preferentially, the light-burned dolomite is mainly added, the requirements of slag melting of a large-sized converter and protection of a furnace lining are mainly met, the addition amount is too small, and the problem of data acquisition errors exists;
the adding amount of the ore is mainly related to the heat balance in the furnace, therefore, when the temperature of the molten iron is more than 1320 ℃, the amount of the scrap steel is less than 30t, and the mass percent of silicon in the molten iron is more than 0.3 percent, the ore is added and selected preferentially, and the rest of the ore has the problem of ore mining omission.
The invention has the beneficial effects that: by adopting a mode of 'mechanism model + artificial intelligence prediction', on the basis of fully analyzing influence factors of the lime addition amount of the converter, the internal relation among the influence factors of the converter is constructed, the method for intelligently adding lime in TSC (thyristor switched capacitor) stages of the large-scale converter under different material conditions is realized, the problem of accurate addition of the lime of the converter is solved, quality accidents and production accidents are reduced, and the smelting cost is reduced.
Drawings
FIG. 1 is an artificial intelligence algorithm neural network architecture diagram of the present invention;
FIG. 2 is a flow chart of the present invention;
FIG. 3 is a scatter plot of the effectiveness of use of an embodiment of the present invention;
fig. 4 is a histogram of the effects of use of an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions of the embodiments of the present invention with reference to the drawings of the embodiments, and it is obvious that the described embodiments are a small part of the embodiments of the present invention, rather than all embodiments, and all other embodiments obtained by a person of ordinary skill in the art without creative work based on the embodiments of the present invention belong to the protection scope of the present invention.
The implementation steps are as follows:
(1) predicting according to a metallurgical mechanism by adopting a mechanism model, wherein the formula of the mechanism model is as follows:
wherein WshRepresenting the lime adding amount, and the data range is 22kg/t-49 kg/t;
ashcrepresents lime CaO mass percent, data range 88% -93%;
ashsrepresents lime SiO2Mass percent, data range 2% -5%;
Wqsrepresenting the addition amount of light-burned dolomite, and the data range is 7.5kg/t-30.1 kg/t;
aqscrepresenting the CaO mass percent of the light-burned dolomite, and the data range is 40-48 percent;
aqssstands for light-burned dolomite SiO2Mass percent, data range 0.1% -3.3%;
Wgzrepresenting the addition of steel slag and other cold materials, and the data range is less than 113 kg/t;
agzcrepresenting the CaO mass percent of the steel slag, and the data range is 20-45 percent;
agzsSiO for steel slag2Mass percent, data range 8% -23%;
Wtsrepresenting the adding amount of molten iron, and the data range is 900kg/t-1085 kg/t;
astrepresenting the mass percent of Si in the molten iron, and the data range is 0.15-0.60%;
2.14 conversion of Si to SiO2The conversion coefficient of (a);
r represents alkalinity which is determined by an alkalinity selection model, and the alkalinity is divided into four gradients, wherein the values are respectively 3, 3.5, 4 and 4.5;
the principle of mechanism model selection is as follows:
the phosphorus content of the steel grade requires to establish 4 gradient gears, wherein the phosphorus mass percentage requirement is less than or equal to 0.015 percent, the phosphorus mass percentage requirement is more than 0.015 percent and less than or equal to 0.020 percent, the phosphorus mass percentage requirement is more than 0.020 percent and less than or equal to 0.025 percent, and the phosphorus mass percentage requirement is more than 0.025 percent;
if the final temperature of the steel grade is more than 1650 ℃, the grade required by phosphorus mass percentage is first, and the alkalinity is selected to be 4.5; the requirement of phosphorus mass percent is two, and the alkalinity is selected to be 4; the phosphorus mass percentage requirement is that the phosphorus mass percentage is;
for steel grades with the end point temperature of less than 1650 ℃, the mass percentage of phosphorus in molten iron is less than or equal to 0.15 percent, the requirement of the mass percentage of phosphorus in the steel grades is that firstly, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is less than or equal to 0.15 percent and the mass percent of phosphorus in steel grade is required to be 2, the alkalinity is selected to be 3.5; the phosphorus mass percent of the molten iron is less than or equal to 0.15 percent, and the phosphorus mass percent of the steel grade is required to be 3.0 when the phosphorus mass percent of the steel grade is three and four; the mass percent of phosphorus in molten iron is more than 0.15 percent, the mass percent of phosphorus in steel grade is that firstly, the alkalinity is selected to be 4.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be two, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is ③, the alkalinity is 3.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be four, the alkalinity is selected to be 3.
(2) Carrying out artificial intelligence prediction, adopting a BP neural network architecture as a basis for deep learning artificial intelligence, wherein the input of the artificial intelligence prediction is 16-dimensional data, the output of the artificial intelligence prediction is 1-dimensional data, the 16-dimensional input data are respectively the lime amount of a mechanism model, a smelting team, a smelting furnace base, steel grade requirements, scrap steel, molten iron temperature, molten iron weight, molten iron manganese, molten iron silicon, molten iron phosphorus, steel slag, dolomite addition amount, TSC phosphorus, TSC manganese, TSC carbon and TSC temperature, and the 1-dimensional output data are the lime amount; the input data is standardized based on a mechanism model, and the processing method is as follows:
y is a parameter value to be standardized, ystdIs the standard deviation of this set of parameters, yaveIs the average of this set of parameters;
constructing an internal relation between input data and output data, and obtaining the output data through artificial intelligence calculation;
the correction error required for the calculation is reduced by means of the denormalization, which is carried out by means of the following formula
y=y*×yave+ystd
The excitation function is of the form:
the BP neural network structure is a structure with 3 hidden layers, the number of nodes of the first hidden layer is 3, the number of nodes of the second hidden layer is 50, and the number of nodes of the third hidden layer is 5.
Training data of artificial intelligent prediction are screened by a data preference rule model, data acquisition errors or no typicality data are screened out, the model calculation precision is improved, the interference of error data on model calculation is reduced, and a specific preference rule is as follows:
selecting the weight of molten iron and the weight of scrap steel in a range of 280t to 320 t;
the silicon content percentage of the molten iron is less than 0.6 percent, and the selection is preferred;
thirdly, selecting lime with the addition amount of more than 6t preferentially;
fourthly, the addition amount of the light-burned dolomite is more than 2t, and the selection is preferred;
fifthly, when the temperature of molten iron is higher than 1320 ℃, the amount of scrap steel is lower than 30t, and the mass percent of silicon in the molten iron is higher than 0.3%, adding ore, and selecting preferentially.
Example (b):
the implementation unit has three 260t top-bottom combined blown converters, and the converters are controlled by sublance guns and are respectively a converter No. 1, a converter No. 2 and a converter No. 3, and the staff work in four-shift three-operation mode and are respectively four shifts, groups A, B, C and D.
Under the raw material conditions that the content of CaO in the sintered ore is 7 percent and the content of SiO in the sintered ore is 7 percent2The content of the lime is 3 percent, the content of the lime CaO is 88 percent, and the lime SiO is22 percent of light-burned dolomite CaO, 40 percent of light-burned SiO20.1 percent, 20 percent of CaO content in the steel slag and SiO2The content is 8%. By the method, the lime addition is predicted, the actual production on site is guided, and no component accident or production accident occurs in the heat of the embodiment.
The raw material conditions are that the CaO content of the sintered ore is 10 percent and the SiO content is26 percent of lime CaO, 93 percent of lime SiO25 percent of light-burned dolomite CaO and 48 percent of light-burned SiO23.3 percent, the CaO content of the steel slag is 45 percent, and SiO2The content was 23%. By the method, the lime addition is predicted, the actual production on site is guided, and no component accident or production accident occurs in the heat of the embodiment.
The raw material conditions are that the CaO content of the sintered ore is 8 percent and the SiO content is2The content of the lime is 4 percent, the content of the lime CaO is 90 percent, and the lime SiO is24 percent of light-burned dolomite CaO, 44 percent of light-burned SiO 22 percent, 35 percent of CaO content in the steel slag and SiO2The content is 15%. By the method, the lime addition is predicted, the actual production on site is guided, and no component accident or production accident occurs in the heat of the embodiment.
As can be seen from the above examples, the predicted lime content value follows the actual addition amount well, and the average deviation amount is-0.154 ton.
Claims (4)
1. A method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter is characterized by comprising the following steps:
(1) predicting according to a metallurgical mechanism by adopting a mechanism model, wherein the formula of the mechanism model is as follows:
wherein WshRepresenting the lime adding amount, and the data range is 22kg/t-49 kg/t;
ashcrepresenting the mass percent of CaO of lime, and the data range is 88-93 percent;
ashsrepresents lime SiO2Mass percent, data range 2% -5%;
Wqsrepresenting the addition amount of light-burned dolomite, and the data range is 7.5kg/t-30.1 kg/t;
aqscrepresenting the CaO mass percent of the light-burned dolomite, and the data range is 40-48 percent;
aqssstands for light-burned dolomite SiO2Mass percent, data range 0.1% -3.3%;
Wgzrepresenting the addition of steel slag and other cold materials, and the data range is less than 113 kg/t;
agzcrepresenting the CaO mass percent of the steel slag, and the data range is 20-45 percent;
agzsSiO for steel slag2Mass percent, data range 8% -23%;
Wtsrepresenting the adding amount of molten iron, and the data range is 900kg/t-1085 kg/t;
astrepresenting the mass percent of Si in the molten iron, and the data range is 0.15-0.60%;
2.14 conversion of Si to SiO2The conversion coefficient of (a);
r represents alkalinity which is determined by an alkalinity selection model, and the alkalinity is divided into four gradients, wherein the values are respectively 3, 3.5, 4 and 4.5;
(2) carrying out artificial intelligence prediction, adopting a BP neural network architecture as a basis for deep learning artificial intelligence, wherein the input of the artificial intelligence prediction is 16-dimensional data, the output of the artificial intelligence prediction is 1-dimensional data, the 16-dimensional input data are respectively the lime amount of a mechanism model, a smelting team, a smelting furnace base, steel grade requirements, scrap steel, molten iron temperature, molten iron weight, molten iron manganese, molten iron silicon, molten iron phosphorus, steel slag, dolomite addition amount, TSC phosphorus, TSC manganese, TSC carbon and TSC temperature, and the 1-dimensional output data are the lime amount; the input data is standardized based on a mechanism model, and the processing method is as follows:
y is a parameter value to be standardized, ystdIs the standard deviation of this set of parameters, yaveIs the average of this set of parameters;
constructing an internal relation between input data and output data, and obtaining the output data through artificial intelligence calculation;
(3) the correction error required for the calculation is reduced by means of the denormalization, which is carried out by means of the following formula
y=y*×yave+ystd
(4) The excitation function is of the form:
2. the method for intelligently adding lime at TSC (TSC-period) stages of large-scale converters according to claim 1, which is characterized in that: in the step (1), the principle of selecting a mechanism model is as follows:
the phosphorus content of the steel grade requires to establish 4 gradient gears, wherein the phosphorus mass percentage requirement is less than or equal to 0.015 percent, the phosphorus mass percentage requirement is more than 0.015 percent and less than or equal to 0.020 percent, the phosphorus mass percentage requirement is more than 0.020 percent and less than or equal to 0.025 percent, and the phosphorus mass percentage requirement is more than 0.025 percent;
if the final temperature of the steel grade is more than 1650 ℃, the grade required by phosphorus mass percentage is first, and the alkalinity is selected to be 4.5; the requirement of phosphorus mass percent is two, and the alkalinity is selected to be 4; the phosphorus mass percentage requirement is that the phosphorus mass percentage is that the phosphorus mass percentage is not less than that the phosphorus mass percentage;
for steel grades with the end point temperature of less than 1650 ℃, the mass percentage of phosphorus in molten iron is less than or equal to 0.15 percent, the requirement of the mass percentage of phosphorus in the steel grades is that firstly, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is less than or equal to 0.15 percent and the mass percent of phosphorus in steel grade is required to be 2, the alkalinity is selected to be 3.5; the phosphorus mass percent of the molten iron is less than or equal to 0.15 percent, and the phosphorus mass percent of the steel grade is required to be 3.0 when the phosphorus mass percent of the steel grade is three and four; the mass percent of phosphorus in molten iron is more than 0.15 percent, the mass percent of phosphorus in steel grade is that firstly, the alkalinity is selected to be 4.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be two, the alkalinity is selected to be 4; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is ③, the alkalinity is 3.5; when the mass percent of phosphorus in molten iron is more than 0.15 percent and the mass percent of phosphorus in steel grade is required to be four, the alkalinity is selected to be 3.
3. The method for intelligently adding lime into a large converter according to claim 1, wherein the method comprises the following steps: in the step (2), the BP neural network structure is a structure of 3 hidden layers, the number of nodes of a first hidden layer is 3, the number of nodes of a second hidden layer is 50, and the number of nodes of a third hidden layer is 5.
4. The method for intelligently adding lime into a large converter according to claim 1, wherein the method comprises the following steps: in the step (2), training data of artificial intelligent prediction are screened by a data preference rule model, data acquisition errors or no typicality data are screened out, the model calculation precision is improved, the interference of error data on model calculation is reduced, and the specific preference rule is as follows:
selecting the weight of molten iron and the weight of scrap steel in a range of 280t to 320 t;
the silicon content percentage of the molten iron is less than 0.6 percent, and the selection is preferred;
thirdly, selecting lime with the addition amount of more than 6t preferentially;
fourthly, the addition amount of the light-burned dolomite is more than 2t, and the selection is preferred;
fifthly, when the temperature of molten iron is higher than 1320 ℃, the amount of scrap steel is lower than 30t, and the mass percent of silicon in the molten iron is higher than 0.3%, adding ore, and selecting preferentially.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110614850.0A CN113362903B (en) | 2021-06-02 | 2021-06-02 | Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110614850.0A CN113362903B (en) | 2021-06-02 | 2021-06-02 | Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113362903A true CN113362903A (en) | 2021-09-07 |
CN113362903B CN113362903B (en) | 2022-05-20 |
Family
ID=77531390
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110614850.0A Active CN113362903B (en) | 2021-06-02 | 2021-06-02 | Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113362903B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115522012A (en) * | 2022-08-25 | 2022-12-27 | 邯郸钢铁集团有限责任公司 | Method for controlling TSC (thyristor switched capacitor) stage oxygen blowing amount of large converter |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101463407A (en) * | 2008-11-22 | 2009-06-24 | 大连理工大学 | Method for calculating converter steel melting lime adding amount |
CN103276136A (en) * | 2013-01-02 | 2013-09-04 | 邯钢集团邯宝钢铁有限公司 | Converter-steelmaking molten steel phosphorus-determination method based on sublance system |
CN103397140A (en) * | 2013-07-19 | 2013-11-20 | 东北大学 | System and method for predicating amount of refining slag required during refining and desulfuration of LF (Ladle Furnace) on line |
CN106119458A (en) * | 2016-06-21 | 2016-11-16 | 重庆科技学院 | Converter steelmaking process cost control method based on BP neutral net and system |
CN110362044A (en) * | 2019-06-03 | 2019-10-22 | 齐鲁工业大学 | A kind of the lime stone additive amount Predictive Control System and method of copper ore floatation device |
CN112036081A (en) * | 2020-08-26 | 2020-12-04 | 北京科技大学 | Method for determining addition amount of silicon-manganese alloy in converter tapping based on yield prediction |
US20210142253A1 (en) * | 2019-11-13 | 2021-05-13 | Aktana, Inc. | Explainable artificial intelligence-based sales maximization decision models |
-
2021
- 2021-06-02 CN CN202110614850.0A patent/CN113362903B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101463407A (en) * | 2008-11-22 | 2009-06-24 | 大连理工大学 | Method for calculating converter steel melting lime adding amount |
CN103276136A (en) * | 2013-01-02 | 2013-09-04 | 邯钢集团邯宝钢铁有限公司 | Converter-steelmaking molten steel phosphorus-determination method based on sublance system |
CN103397140A (en) * | 2013-07-19 | 2013-11-20 | 东北大学 | System and method for predicating amount of refining slag required during refining and desulfuration of LF (Ladle Furnace) on line |
CN106119458A (en) * | 2016-06-21 | 2016-11-16 | 重庆科技学院 | Converter steelmaking process cost control method based on BP neutral net and system |
CN110362044A (en) * | 2019-06-03 | 2019-10-22 | 齐鲁工业大学 | A kind of the lime stone additive amount Predictive Control System and method of copper ore floatation device |
US20210142253A1 (en) * | 2019-11-13 | 2021-05-13 | Aktana, Inc. | Explainable artificial intelligence-based sales maximization decision models |
CN112036081A (en) * | 2020-08-26 | 2020-12-04 | 北京科技大学 | Method for determining addition amount of silicon-manganese alloy in converter tapping based on yield prediction |
Non-Patent Citations (2)
Title |
---|
XIANLIANG DONG ET AL.: "The Converter Steelmaking End Point Prediction Model Based on RBF Neural Network", 《APPLIED MECHANICS AND MATERIALS》 * |
刘远洋 等: "基于机理和数据驱动的转炉输入—输出混合模型", 《有色金属科学与工程》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115522012A (en) * | 2022-08-25 | 2022-12-27 | 邯郸钢铁集团有限责任公司 | Method for controlling TSC (thyristor switched capacitor) stage oxygen blowing amount of large converter |
CN115522012B (en) * | 2022-08-25 | 2024-01-12 | 邯郸钢铁集团有限责任公司 | Method for controlling TSC stage oxygen blowing amount of large converter |
Also Published As
Publication number | Publication date |
---|---|
CN113362903B (en) | 2022-05-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103361461B (en) | Online prediction control method for phosphorus content of low-carbon steel smelted by converter | |
CN102344986B (en) | Method, device and system for controlling steel-making endpoint of converter | |
CN103103309B (en) | Method of supplementarily forecasting steelmaking finishing point of converter | |
CN110885912B (en) | Automatic steelmaking method and system based on data analysis | |
CN113362903B (en) | Method for intelligently adding lime in TSC (thyristor switched capacitor) stage of large converter | |
CN106834593A (en) | A kind of method that RH refining furnace decarbonization process data are determined with reference heats method | |
CN108504815A (en) | Automatic control system and its method is added in a kind of converter alloy | |
CN108647407A (en) | A kind of pneumatic steelmaking flue gas analysis carbon determination method | |
CN115522012B (en) | Method for controlling TSC stage oxygen blowing amount of large converter | |
CN102399933A (en) | Automatic control method for converter blowing low-carbon steel oxygen lance | |
CN114854929A (en) | Converter blowing CO 2 Method for dynamically predicting molten steel components and temperature in real time | |
CN106636530A (en) | Method for predicting steel-making temperature of converter and server | |
CN116469481A (en) | LF refined molten steel composition forecasting method based on XGBoost algorithm | |
CN103695593B (en) | Automatic steelmaking and slag-leaving operation method for converter | |
CN102373310A (en) | Method for guiding converter reblowing process operation | |
CN110502781A (en) | A kind of ferroalloy production blending optimization method based on priori knowledge | |
CN1603424A (en) | Bessemerizing control method based on intelligent compound dynamic model with sublance converter | |
CN113961865A (en) | Method for accurately controlling addition amount of temperature regulator in TSC (thyristor switched capacitor) stage of large converter | |
CN111518980B (en) | Correction method and system for converter end point carbon content prediction model | |
WO2014167982A1 (en) | Correction device, correction method and steel refining method | |
CN115927784B (en) | Based on CO 2 Converter steelmaking end point control method by dynamic mixed blowing | |
CN116741322A (en) | Steelmaking alloy batching method based on low-cost measurement and calculation | |
CN202401093U (en) | Control system for model steel making | |
CN113846204B (en) | Steel grade smelting cost control method suitable for 120t top-bottom combined blown converter | |
CN111518981B (en) | Method and system for forecasting phosphorus content of converter endpoint |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |