CN115906538A - Method for predicting molten steel components in ladle refining furnace - Google Patents
Method for predicting molten steel components in ladle refining furnace Download PDFInfo
- Publication number
- CN115906538A CN115906538A CN202310026589.1A CN202310026589A CN115906538A CN 115906538 A CN115906538 A CN 115906538A CN 202310026589 A CN202310026589 A CN 202310026589A CN 115906538 A CN115906538 A CN 115906538A
- Authority
- CN
- China
- Prior art keywords
- molten steel
- steel
- components
- alloy
- predicting
- 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.)
- Pending
Links
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 124
- 239000010959 steel Substances 0.000 title claims abstract description 124
- 238000007670 refining Methods 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims abstract description 58
- 229910045601 alloy Inorganic materials 0.000 claims abstract description 77
- 239000000956 alloy Substances 0.000 claims abstract description 77
- 238000004519 manufacturing process Methods 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 31
- 239000000203 mixture Substances 0.000 claims abstract description 21
- 238000010079 rubber tapping Methods 0.000 claims abstract description 15
- 238000005070 sampling Methods 0.000 claims abstract description 12
- 239000002893 slag Substances 0.000 claims description 26
- 239000000306 component Substances 0.000 claims description 18
- 229910052782 aluminium Inorganic materials 0.000 claims description 14
- 238000003723 Smelting Methods 0.000 claims description 13
- 229910052799 carbon Inorganic materials 0.000 claims description 12
- PXHVJJICTQNCMI-UHFFFAOYSA-N Nickel Chemical compound [Ni] PXHVJJICTQNCMI-UHFFFAOYSA-N 0.000 claims description 10
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims description 9
- XKRFYHLGVUSROY-UHFFFAOYSA-N Argon Chemical compound [Ar] XKRFYHLGVUSROY-UHFFFAOYSA-N 0.000 claims description 8
- 229910052717 sulfur Inorganic materials 0.000 claims description 7
- 229910000592 Ferroniobium Inorganic materials 0.000 claims description 6
- 229910001200 Ferrotitanium Inorganic materials 0.000 claims description 6
- 239000000463 material Substances 0.000 claims description 6
- 235000008733 Citrus aurantifolia Nutrition 0.000 claims description 5
- 229910000604 Ferrochrome Inorganic materials 0.000 claims description 5
- 229910001309 Ferromolybdenum Inorganic materials 0.000 claims description 5
- 229910000519 Ferrosilicon Inorganic materials 0.000 claims description 5
- 229910000628 Ferrovanadium Inorganic materials 0.000 claims description 5
- 229910000720 Silicomanganese Inorganic materials 0.000 claims description 5
- 235000011941 Tilia x europaea Nutrition 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 5
- WUKWITHWXAAZEY-UHFFFAOYSA-L calcium difluoride Chemical compound [F-].[F-].[Ca+2] WUKWITHWXAAZEY-UHFFFAOYSA-L 0.000 claims description 5
- 239000010436 fluorite Substances 0.000 claims description 5
- PNXOJQQRXBVKEX-UHFFFAOYSA-N iron vanadium Chemical compound [V].[Fe] PNXOJQQRXBVKEX-UHFFFAOYSA-N 0.000 claims description 5
- 239000004571 lime Substances 0.000 claims description 5
- 229910052759 nickel Inorganic materials 0.000 claims description 5
- 239000002245 particle Substances 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 229910052786 argon Inorganic materials 0.000 claims description 4
- 238000007664 blowing Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 229910052758 niobium Inorganic materials 0.000 claims description 4
- 229910052720 vanadium Inorganic materials 0.000 claims description 4
- OYPRJOBELJOOCE-UHFFFAOYSA-N Calcium Chemical compound [Ca] OYPRJOBELJOOCE-UHFFFAOYSA-N 0.000 claims description 3
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 3
- NINIDFKCEFEMDL-UHFFFAOYSA-N Sulfur Chemical compound [S] NINIDFKCEFEMDL-UHFFFAOYSA-N 0.000 claims description 3
- 229910052791 calcium Inorganic materials 0.000 claims description 3
- 239000011575 calcium Substances 0.000 claims description 3
- WPBNNNQJVZRUHP-UHFFFAOYSA-L manganese(2+);methyl n-[[2-(methoxycarbonylcarbamothioylamino)phenyl]carbamothioyl]carbamate;n-[2-(sulfidocarbothioylamino)ethyl]carbamodithioate Chemical compound [Mn+2].[S-]C(=S)NCCNC([S-])=S.COC(=O)NC(=S)NC1=CC=CC=C1NC(=S)NC(=O)OC WPBNNNQJVZRUHP-UHFFFAOYSA-L 0.000 claims description 3
- 229910052751 metal Inorganic materials 0.000 claims description 3
- 239000002184 metal Substances 0.000 claims description 3
- 239000011593 sulfur Substances 0.000 claims description 3
- PWHULOQIROXLJO-UHFFFAOYSA-N Manganese Chemical compound [Mn] PWHULOQIROXLJO-UHFFFAOYSA-N 0.000 claims description 2
- 238000012360 testing method Methods 0.000 abstract description 3
- 238000012795 verification Methods 0.000 abstract 1
- 239000002699 waste material Substances 0.000 abstract 1
- 238000004364 calculation method Methods 0.000 description 12
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 11
- 229910052760 oxygen Inorganic materials 0.000 description 11
- 239000001301 oxygen Substances 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000009749 continuous casting Methods 0.000 description 4
- 238000007792 addition Methods 0.000 description 3
- 238000005275 alloying Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000012821 model calculation Methods 0.000 description 3
- 239000002436 steel type Substances 0.000 description 3
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000006477 desulfuration reaction Methods 0.000 description 2
- 230000023556 desulfurization Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000000087 stabilizing effect Effects 0.000 description 2
- 238000009628 steelmaking Methods 0.000 description 2
- 238000005406 washing Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009529 body temperature measurement Methods 0.000 description 1
- 238000005266 casting Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000010891 electric arc Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000012840 feeding operation Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- XWHPIFXRKKHEKR-UHFFFAOYSA-N iron silicon Chemical compound [Si].[Fe] XWHPIFXRKKHEKR-UHFFFAOYSA-N 0.000 description 1
- PYLLWONICXJARP-UHFFFAOYSA-N manganese silicon Chemical compound [Si].[Mn] PYLLWONICXJARP-UHFFFAOYSA-N 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 238000003756 stirring Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Landscapes
- Carbon Steel Or Casting Steel Manufacturing (AREA)
Abstract
The invention discloses a method for predicting molten steel components of a ladle refining furnace, which comprises the steps of taking an arrival parameter as an initial value and an departure parameter as a target value, obtaining tapping production parameters of a primary refining furnace, automatically forming an alloy feeding model by a system, obtaining production parameters in the refining process of an LF refining furnace, fully considering the influence of the weight change of molten steel caused by adding alloy and waste steel on the molten steel components, establishing a dynamic prediction model of the molten steel components, and realizing the dynamic prediction of the molten steel components in the LF refining process; the molten steel test value required by the normal process is used as the verification of the model precision, the dynamic prediction model is automatically corrected, and the hit rate of the terminal component is improved; the end point composition of the selected heat or the new steel grade is predicted. Under the technological conditions of steel ladle and steel scrap adding in the refining process, the historical mass production data is used for calculating the yield and the variable quantity of the elements to predict the content of the elements in the steel, and the sampling result is used for correcting the yield of the elements and finally predicting the content of the elements in the steel.
Description
Technical Field
The invention belongs to the technical field of LF refining, and particularly relates to a method for predicting molten steel components in a ladle refining furnace.
Background
In the modern steelmaking process adopting the continuous casting process, LF plays the effect of bearing down, therefore LF refining furnace is honored as the buffer in the steelmaking process. The molten steel after primary smelting in a high-power (or ultra-high-power) electric arc furnace or converter usually has large fluctuation, and the precise end point control of the molten steel through secondary refining in an LF refining furnace in a limited converter smelting period or continuous casting period is required. Therefore, the method researches and develops the molten steel composition end point control technology of the LF refining furnace, realizes the matching of the period of the LF refining process and the smelting period of the converter or the continuous casting period, and has important significance for improving and stabilizing the product quality and yield of iron and steel enterprises and reducing the production cost.
At present, an empirical operation mode is generally adopted on the site of an LF refining furnace, the refining period of the LF refining furnace has large fluctuation, and as for molten steel components, the fluctuation of the components of incoming molten steel of the LF refining furnace is large, so that the factors such as component adjustment, desulfurization pressure, treatment period and the like of the furnace are greatly different, and repeated calculation is needed, so that the empirical operation mode is difficult to optimize the operation variables of the process according to the condition of specific furnace, the operation flexibility is lacked, the high energy consumption and material consumption level are caused, the endpoint control is difficult to be comprehensively considered during multi-parameter adjustment, the endpoint control precision of the molten steel components is influenced, and continuous casting and broken casting can be caused by the plan of the treatment period in serious cases.
As steel grades are increased, the original research methods such as a mechanism model, a statistical regression model and a neural network model mainly concentrate on the prediction of C, si, mn, P and S, and do not relate to the prediction of microalloy Nb, V, ti and B; the original prediction model has not been adapted to new production requirements:
1. the manual feeding depends on experience: at present, alloy adding of the LF furnace depends on manual experience, an operator calculates the type and the quantity of the alloy needing to be added according to the range requirements of operation key points on alloy element components, and when the production rhythm is tense, the operator can directly estimate according to the experience. The production of special steel often has the internal control standard of narrow composition, and the adding mode completely depending on manual experience often has the condition that the content of alloy elements exceeds the internal control standard and even exceeds the national standard, thereby causing the degradation and even the judgment of steel grade.
2. The cost control is relatively extensive: in the production process of the LF furnace, part of precious alloys such as ferrocolumbium, ferrotitanium and the like need to be added, and the adding model of the precious alloys is relatively extensive at present, so that the cost control space is large.
3. The lack of parameters self-learning function: the alloy yield is an important index for calculating the adding amount of the alloy, the yield is always fixed in a calculation formula used by an operator at present, but in an objective world, the alloy yield is influenced by molten steel dissolved oxygen content, top slag oxidability, alloy granularity, affinity of alloy elements and oxygen and the like, so that the alloy element receiving condition cannot be accurately analyzed through simple formula calculation, and the material adding efficiency is influenced.
Disclosure of Invention
In order to solve the above problems in the prior art, the present invention is directed to a method for predicting molten steel composition in a ladle refining furnace.
The technical scheme adopted by the invention is as follows:
a method for predicting the composition of molten steel in a ladle refining furnace comprises the following steps:
s1, obtaining station entering parameters of an LF refining furnace; the station-entering parameters comprise the component content of the station-entering molten steel and the weight of the station-entering molten steel;
s2, obtaining tapping production parameters of the primary smelting furnace; the tapping production parameters of the primary smelting furnace comprise tapping components, alloy and slag charge components and corresponding amount;
s3, acquiring station leaving parameters of the refining LF furnace; the station leaving parameters comprise the content of target components in the station-arriving molten steel;
s4, obtaining production parameters in the refining process of the refining LF furnace; the production parameters comprise argon blowing parameters, alloy adding, scrap steel adding, raw and auxiliary materials, and components and corresponding amounts of wire feeding;
s5, establishing a model; establishing a prediction model of molten steel components in the LF refining process by taking the inbound parameters as initial values and the outbound parameters as target values, and establishing a dynamic prediction model of the molten steel components and generating an alloy feeding model by combining the tapping production parameters of the primary smelting furnace;
s6, calculating; and calculating and training the production parameters by using the dynamic prediction model to obtain modified model parameters, and checking the precision of the dynamic prediction model.
S7, predicting; and predicting the end point component of the selected heat or steel grade by using the dynamic prediction model after the precision check is finished.
In step S1, the components of the inbound-side molten steel include mass percentages of C, si, mn, P and S and mass percentages of microalloy Nb, V, ti and B.
Preferably, in step S2, the obtaining of the tapping production parameters of the primary smelting furnace includes:
obtaining the components and the corresponding amount of added aluminum particles, silicomanganese, manganese metal, ferrosilicon, low-carbon ferrochromium, ferrocolumbium, ferrotitanium, ferrovanadium, ferromolybdenum, nickel plates and carburant;
and obtaining the components and the corresponding amount of the added lime, fluorite, pre-melted slag, high-aluminum pre-melted slag and returned slag.
Preferably, in step S4, the obtaining of the production parameters in the refining process of the refining LF furnace includes:
obtaining the components and the corresponding amount of added aluminum particles, silicomanganese, metal manganese, ferrosilicon, low-carbon ferrochrome, ferrocolumbium, ferrotitanium, ferrovanadium, ferromolybdenum, nickel plates and carburant;
obtaining the components and the corresponding amount of added lime, fluorite, pre-melted slag, high-aluminum pre-melted slag and returned slag;
and obtaining components and corresponding amounts of the added aluminum wire, the added carbon wire, the added sulfur wire and the added pure calcium wire.
Preferably, in step S6, the calculating and training of the production parameters includes:
according to the element balance, ai.G = Bi.G + gj Cj-i Fi;
wherein G is the weight of molten steel at the time of pouring, and the unit is kg; ai is the target content percentage of the i element in the steel; bi is the initial content percentage of the element i in the molten steel after adding the scrap steel; gj is the amount of alloy j for adjusting the i element to reach the target content, and the unit is kg; cj-i is the percentage of i element in alloy j; fi is the average yield of the i element.
As a preference of the present invention, the added alloy does not affect the weight of molten steel, and each type of alloy has only one component finely adjusted.
Preferably, the average variation amount of the i element is Δ ω i;
calculating and finely adjusting gj of all the components of the alloy needing to be adjusted, and calculating the total amount of the added alloy: gadd = Σ gj;
calculating new molten steel weight G1= G + Gadd;
the effect of feed on endpoint composition Δ ω i-L = ρ iL/G1;
wherein rho i is the density of the type of the feeding wire, and the unit is kg/m; l is the length of the feed line and is expressed in m.
Preferably, in step S7, the end point component of the selected heat or steel grade is predicted as
Predicting the percentage of i element in steel:
Ai,cal＝(BiG1+Σgj*Cj-i*Fi)/G1+Δωi+Δωi-L
wherein Ai and cal are percentages of i elements in the predicted steel;
preferably, the average yield of the element i is sampled from the recent 10 furnaces of the steel of the furnace; and correcting the yield of the i element in the steel by using the sampling analysis result, and repeatedly calculating Ai and cal to give the percentage of the i element in the steel as a final predicted value of the i element in the steel.
The invention has the beneficial effects that: the method for predicting the molten steel components of the ladle refining furnace starts from actual production big data, establishes a prediction model of the molten steel components in the refining process of the LF furnace on the basis of optimizing a mechanism model, and corrects the model for multiple times by utilizing the traditional sampling analysis and approaches the actual production, so that the prediction precision of the model is ensured; the method and the system for predicting the molten steel components of the LF refining furnace are matched with a robot to automatically measure the temperature and sample, so that the centralized control of a plurality of refining furnaces can be realized, the refining period is effectively shortened, the production cost is reduced, the labor productivity and the product quality stability are improved, and the high and stable yield is realized by stabilizing the smelting period; the prediction method provided by the invention covers a plurality of intelligent models, and the current alloy yield is analyzed in real time through a self-learning algorithm; the tracking and prediction of components in the whole smelting process are adopted, so that smelting reference and early warning are provided for an operator; the whole application of the invention is to standardize the alloy process, finely add precious alloy and reduce the cost.
Drawings
The invention is described in further detail below with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a flowchart of calculation of a molten steel composition prediction model according to the present invention;
FIG. 2 is a flowchart for calculating the yield of alloying elements according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the detailed description and specific examples, while indicating the preferred embodiment of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention is described below with reference to fig. 1-2, and the method for predicting the molten steel composition of the ladle refining furnace comprise the following steps:
obtaining the inbound parameters of the refining process of the LF refining furnace, wherein the inbound parameters comprise inbound molten steel component content and molten steel weight;
acquiring primary furnace tapping production parameters, wherein the primary furnace tapping production parameters comprise tapping components, alloy and slag charge components and corresponding quantities;
obtaining off-station parameters of an LF refining furnace in a refining process, wherein the on-station parameters comprise the content of off-station molten steel components;
obtaining production parameters in the refining process of an LF refining furnace, wherein the production parameters comprise argon blowing amount, components and corresponding amount of added alloy, raw and auxiliary materials and fed wires;
obtaining a first sampling result in the refining process of the LF refining furnace, and comparing the first sampling result with tapping production parameters of the primary refining furnace;
establishing a dynamic prediction model of molten steel components in the refining process of the LF refining furnace by taking the inbound parameters as initial values, sampling results in the refining process of the LF refining furnace as target values and the outbound parameters as final target values, automatically forming an alloy feeding model, realizing automatic alloy feeding and dynamic component prediction, and verifying the model precision by taking a molten steel test value required by a normal process as a model precision;
calculating and training refined production parameters of the LF furnace by using the dynamic prediction model to obtain various modified model parameters, and finishing precision check of the dynamic prediction model by using a sampling result in the refining process;
using the dynamic prediction model, the first sampling result in the refining process and the final target value of the off-station component to give alloy feeding parameters;
and predicting the endpoint component of the selected heat or steel grade by using the dynamic prediction model for finishing the precision check.
Alloy feeding model: according to the requirements of initial components and target components of molten steel, the alloy input combination and the input amount for minimizing the total alloy input cost are determined.
The model algorithm is accurate and strict; for the condition that no solution exists in mathematical computation due to certain limiting conditions, the model adopts an intelligent algorithm to give a sub-optimal solution;
the design of the model has good interactivity, and operators can realize specific functions through flexible setting of parameters. Such as element target range, initial value, element yield, alloy input limit, etc.;
the component constraint of the specified element can be conveniently shielded, and the component requirement is ignored, so that the component constraint can be adapted to the solution under the specific condition.
Target adjustment amounts required for elements in molten steel:
element adjustment amount = (target component-current component) × molten steel amount/element yield
Calculation of the total amount of elements contained in each input alloy:
total amount of elements put = sigma alloy put amount of total element content of alloy
The core algorithm of the model can be summarized as a linear programming problem: and establishing a mathematical model by taking the content of each element as an independent variable and the yield of the element as limiting conditions, and solving the optimal solution under different smelting processes and target components. Due to the complexity of the actual production data, solution-free alloy models based on linear programming often occur. In order to overcome the defect, the linear programming and the artificial intelligence technology are combined, so that an alloy input formula meeting the feeding requirement can be ensured, no solution can be generated under any condition, and the situation that the solution of a simple mathematical model is trapped in the dead end is avoided.
The intelligent component analysis and feeding system of the refining furnace obtains the current production information of the steel ladles entering the station through an operation plan, and constructs a material distribution calculation formula by combining the contents of metallurgical specifications, alloy addition experience formulas, mechanism models, alloy addition influence factors, field sampling and testing results and the like. And the calculation result is sent to a secondary system, and after the secondary system acquires the information of the types and the amounts of the added alloys, an operator issues the information to a PLC (programmable logic controller) to perform feeding operation.
Molten steel component prediction model: calculating the components of the molten steel according to the actual input amount of the existing sample components and the alloy;
the molten steel composition is calculated according to the existing sample composition and the alloy input amount set by an operator, and a confirmation and checking means for alloying according to own experience is provided for the operator.
The component forecast model data item comprises three parts: model setting data items, model calculation output data items, and process display data items.
The model setting data item refers to data that can be input by an operator through a model operation screen, such as element yield, alloy yield, and the like.
The model calculation output data item refers to a data item output by the model calculation. The method mainly comprises the predicted values of the element components, the weight increment of the molten steel and the total alloy investment cost.
The process display data items comprise steel numbers, steel grades, free oxygen, element target components, element standard yield, element sample values and the like. These data are not modifiable and are only available for operator reference.
Further, the component content of the inbound molten steel is predicted to comprise the mass percentage of C, si, mn, P and S and the mass percentage of microalloy Nb, V, ti and B.
Obtaining argon blowing flow of arrival station starting, electrifying, strong desulfurization, temperature measurement sampling, alloying, wire feeding, soft stirring and waiting position in the LF furnace refining process;
obtaining the components and the corresponding amount of added aluminum particles, silicomanganese, metal manganese, ferrosilicon, low-carbon ferrochrome, ferrocolumbium, ferrotitanium, ferrovanadium, ferromolybdenum, nickel plates and carburant;
obtaining the composition and the corresponding amount of added lime, fluorite, pre-melted slag, high-aluminum pre-melted slag and returned slag;
obtaining components and corresponding amounts of an aluminum wire, a carbon wire, a sulfur wire and a pure calcium wire;
the calculation of the combined dynamic prediction model on the molten steel components comprises the following steps:
selecting usable alloy types according to the element limitation conditions of the steel types;
firstly, supposing that the added alloy does not influence the weight of molten steel and each type of alloy only finely adjusts one component;
there are Ai · G = Bi · G + gj × Cj-i × Fi according to the element balance;
wherein G is the weight of molten steel at the time of pouring, and the unit is kg; ai is the target content percentage of the i element in the steel; bi is the initial content percentage of the element i in the molten steel after adding the scrap steel; gj is the amount of alloy j for adjusting i element to reach the target content, and the unit is kg; cj-i is the percentage of i element in alloy j; fi is the average yield of the i element (the average yield of the i element of 10-furnace steel with similar recent process conditions of the steel is taken);
according to historical mass production data, data of 10 furnaces of steel with similar recent process conditions are obtained:
the average variation of the i element is delta omega i;
the amount gj of each alloy added to fine-tune all the components to be adjusted was calculated, and the total amount of alloy added was determined: gadd = Σ gj;
calculating new molten steel weight G1= G + Gadd;
considering the effect of feed line on endpoint composition Δ ω i-l = ρ iL/G1; wherein rho i is the density of the type of the feeding wire, and the unit is kg/m; l is the length of the feed line in m.
Predicting the percentage of i element in steel:
Ai,cal＝(BiG1+Σgj*Cj-i*Fi)/G1+Δωi+Δωi-l
wherein Ai and cal are percentages of i elements in the predicted steel;
and correcting the yield of the i element in the steel by using the sampling analysis result, and repeating the calculation to obtain the percentage of the i element in the steel as the final predicted value of the i element in the steel.
Element yield: and calculating the yield of different alloy elements according to the components and the states of the current molten steel and the components and the states of the slag.
Influence factors of alloy yield are as follows: molten steel dissolved oxygen content, top slag oxidability, alloy granularity and affinity of alloy elements and oxygen. When the dissolved oxygen in the steel is higher, the added alloy elements react with the dissolved oxygen to generate oxides; when the oxidability of the top slag is high, the added alloy elements are oxidized and lost in a slag layer before entering molten steel; when the alloy granularity is larger, the alloy can be mixed in the molten steel insufficiently to influence the molten steel quality, and errors can be brought to the calculation of the alloy yield; different alloy elements have different affinities for oxygen, and after the alloy elements are added, the elements with high oxygen affinity are oxidized before other elements or are reduced by the elements with high oxygen affinity after other elements are oxidized, so that the adding sequence of the alloy elements needs to be adjusted according to the oxygen affinity in order to improve the element yield.
The yield solving method comprises the following steps: reference heat-self learning, support vector machine, grammatical Evolution (GE), neural network (deep learning, big data drive based), etc.
And (4) referring to the heat-self learning solution, namely selecting the average value of the yield of the alloy elements of the molten steel of a plurality of heats which is the same as the steel type of the current heat, has the shortest time interval and has the closest production condition from an actual production database as a prediction value of the yield of the alloy elements of the heat. After refining, all the operation information of the molten steel in the furnace is stored in a database and used as the reference information to be selected later, and the self-learning of the model is completed.
The method is characterized in that a big data driven neural network calculation model is used for performing prediction calculation on the element yield under the current refining condition by using the previous refining data of the furnace and the refining information data of other furnaces under the same parameter condition as a database, and a large amount of previous successful refining data is required to be used as a core support.
The system acquires the steel grade information of the current heat, the operation key points, the slagging mode (slag washing/non-slag washing), and the converter feed amount of silicon iron, silicon manganese and aluminum blocks. And starting the one-key refining system to obtain empirical value data of the total slag charge amount of the same steel type.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and the like are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be directly connected or indirectly connected through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
The foregoing is merely exemplary and illustrative of the present invention and various modifications, additions and substitutions may be made by those skilled in the art to the specific embodiments described without departing from the scope of the invention as defined in the following claims.
Claims (9)
1. A method for predicting the composition of molten steel in a ladle refining furnace is characterized by comprising the following steps of:
s1, obtaining station entering parameters of an LF refining furnace; the station-entering parameters comprise the component content of the station-entering molten steel and the weight of the station-entering molten steel;
s2, obtaining tapping production parameters of the primary smelting furnace; the tapping production parameters of the primary smelting furnace comprise tapping components, alloy and slag charge components and corresponding amount;
s3, acquiring off-station parameters of the refining LF furnace; the station leaving parameters comprise the content of target components in the station-arriving molten steel;
s4, obtaining production parameters in the refining process of the refining LF furnace; the production parameters comprise argon blowing parameters, alloy adding, scrap steel adding, raw and auxiliary materials, and components and corresponding amounts of wire feeding;
s5, establishing a model; establishing a prediction model of molten steel components in the LF refining process by taking the inbound parameters as initial values and the outbound parameters as target values, and establishing a dynamic prediction model of the molten steel components and generating an alloy charging model by combining tapping production parameters of the primary refining furnace;
s6, calculating; calculating and training the production parameters by using the dynamic prediction model to obtain modified model parameters, and checking the precision of the dynamic prediction model;
s7, predicting; and predicting the end point component of the selected heat or steel grade by using the dynamic prediction model after the precision check is finished.
2. The method for predicting the composition of molten steel in a ladle refining furnace according to claim 1, wherein the method comprises the following steps: in step S1, the composition contents of the inbound molten steel comprise the mass percentages of C, si, mn, P and S and the mass percentages of the microalloy Nb, V, ti and B.
3. The method for predicting molten steel composition in a ladle refining furnace according to claim 1, wherein in step S2, the obtaining of the tapping production parameters of the primary refining furnace comprises:
obtaining the components and the corresponding amount of added aluminum particles, silicomanganese, manganese metal, ferrosilicon, low-carbon ferrochromium, ferrocolumbium, ferrotitanium, ferrovanadium, ferromolybdenum, nickel plates and carburant;
and obtaining the components and the corresponding amount of the added lime, fluorite, pre-melted slag, high-aluminum pre-melted slag and returned slag.
4. The method for predicting the composition of molten steel in a ladle refining furnace according to claim 1, wherein in the step S4, the obtaining of the production parameters in the refining process of the refining LF furnace comprises:
obtaining the components and the corresponding amount of added aluminum particles, silicomanganese, metal manganese, ferrosilicon, low-carbon ferrochrome, ferrocolumbium, ferrotitanium, ferrovanadium, ferromolybdenum, nickel plates and carburant;
obtaining the components and the corresponding amount of the added lime, fluorite, pre-melted slag, high-aluminum pre-melted slag and returned slag;
and obtaining components and corresponding amounts of the added aluminum wire, the added carbon wire, the added sulfur wire and the added pure calcium wire.
5. The method of claim 1, wherein the calculating and training of the production parameters in step S6 comprises:
there are Ai · G = Bi · G + gj × Cj-i × Fi according to the element balance;
wherein G is the weight of molten steel at the time of pouring, and the unit is kg; ai is the target content percentage of the i element in the steel; bi is the initial content percentage of the element i in the molten steel after adding the scrap steel; gj is the amount of alloy j for adjusting i element to reach the target content, and the unit is kg; the Cj-i is the percentage of the i element in the alloy j; fi is the average yield of the i element.
6. The method for predicting the composition of molten steel in a ladle refining furnace according to claim 5, wherein: the added alloy does not influence the weight of the molten steel, and each type of alloy only finely adjusts one component.
7. The method for predicting the composition of molten steel in a ladle refining furnace according to claim 6, wherein the method comprises the following steps:
the average variation of the i element is delta omega i;
calculating and finely adjusting gj of all the components of the alloy needing to be adjusted, and calculating the total amount of the added alloy: gadd = Σ gj;
calculating new molten steel weight G1= G + Gadd;
effect of feed on endpoint composition Δ ω i-L = ρ iL/G1;
wherein rho i is the density of the type of the feeding wire, and the unit is kg/m; l is the length of the feed line in m.
8. The method of claim 7, wherein the end point component of the selected heat or steel grade is predicted as
Predicting the percentage of i element in steel:
Ai,cal＝(BiG1+Σgj*Cj-i*Fi)/G1+Δωi+Δωi-L
wherein Ai and cal are the percentages of the i elements in the predicted steel.
9. The method for predicting the composition of molten steel in a ladle refining furnace according to claim 8, wherein the method comprises the following steps: the average yield of the i element is sampled from 10 recent furnace steels of the present furnace steels; and correcting the yield of the i element in the steel by using the sampling analysis result, and repeatedly calculating Ai and cal to give the percentage of the i element in the steel as a final predicted value of the i element in the steel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310026589.1A CN115906538A (en) | 2023-01-09 | 2023-01-09 | Method for predicting molten steel components in ladle refining furnace |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310026589.1A CN115906538A (en) | 2023-01-09 | 2023-01-09 | Method for predicting molten steel components in ladle refining furnace |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115906538A true CN115906538A (en) | 2023-04-04 |
Family
ID=86484212
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310026589.1A Pending CN115906538A (en) | 2023-01-09 | 2023-01-09 | Method for predicting molten steel components in ladle refining furnace |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115906538A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116469481A (en) * | 2023-06-19 | 2023-07-21 | 苏州方兴信息技术有限公司 | LF refined molten steel composition forecasting method based on XGBoost algorithm |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110955956A (en) * | 2019-11-07 | 2020-04-03 | 北京科技大学 | Method and system for joint prediction of molten steel temperature and components based on LF (ladle furnace) refining process |
| CN113088628A (en) * | 2021-03-31 | 2021-07-09 | 山东钢铁股份有限公司 | LF refining method of low-carbon steel |
-
2023
- 2023-01-09 CN CN202310026589.1A patent/CN115906538A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110955956A (en) * | 2019-11-07 | 2020-04-03 | 北京科技大学 | Method and system for joint prediction of molten steel temperature and components based on LF (ladle furnace) refining process |
| CN113088628A (en) * | 2021-03-31 | 2021-07-09 | 山东钢铁股份有限公司 | LF refining method of low-carbon steel |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116469481A (en) * | 2023-06-19 | 2023-07-21 | 苏州方兴信息技术有限公司 | LF refined molten steel composition forecasting method based on XGBoost algorithm |
| CN116469481B (en) * | 2023-06-19 | 2023-08-29 | 苏州方兴信息技术有限公司 | LF refined molten steel composition forecasting method based on XGBoost algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110955956A (en) | Method and system for joint prediction of molten steel temperature and components based on LF (ladle furnace) refining process | |
| CN104962800B (en) | Smelting method for stainless steel material | |
| CN112560218B (en) | LF refining slagging lime addition amount prediction method and system and LF refining method | |
| CN110991916A (en) | Casting blank quality judgment system and method | |
| CN115906538A (en) | Method for predicting molten steel components in ladle refining furnace | |
| CN110766452B (en) | Method for measuring and calculating metal yield of scrap steel | |
| CN115659709B (en) | Molten steel temperature prediction method in refining process of LF refining furnace | |
| CN101592650B (en) | Method for continuously measuring carbon content of molten steel in electric steelmaking furnace | |
| CN102373310B (en) | Method for guiding converter reblowing process operation | |
| CN106811570A (en) | Smelting method of medium-carbon high-manganese steel | |
| CN103898391B (en) | A kind of loading control method of converter high alloy steel scrap | |
| CN106811685B (en) | Smelting method of low-carbon high-manganese steel | |
| CN113805537B (en) | Smelting alloy composition control system and control method | |
| CN113430333B (en) | Intelligent stable smelting control method for ladle refining furnace | |
| CN114737020B (en) | High-precision control method for effective titanium component in steelmaking process | |
| CN107586915A (en) | The alloyage process of manganese element in a kind of middle potassium steel | |
| CN105242545B (en) | A kind of Forecasting Methodology of the coverter pig oxygen flow based on process batch model | |
| CN104060019A (en) | Steelmaking method with semi-steel | |
| Kovačič et al. | Modelling and optimization of sulfur addition during 70MnVS4 steelmaking: An industrial case study | |
| CN118256668A (en) | Method for smelting low-silicon aluminum steel without taking steel sample in whole process | |
| CN109609724B (en) | Adding system and method for alloy-containing cold steel | |
| CN117925941A (en) | CO utilization2Method for smelting ultra-clean Cr-Ni-Mo-V molten steel by Ar | |
| KR20130126210A (en) | Method for refining molten steel | |
| CN114713640A (en) | Production control method of hot-rolled straight HRB400 steel bar containing Nb component | |
| CN116987838A (en) | Manganese content control method for smelting ultra-low manganese steel by large converter |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20230404 |
|
| RJ01 | Rejection of invention patent application after publication |