Design method of scrap steel group material for converter steelmaking

Abstract

The invention discloses a design method of scrap steel group materials for converter steelmaking, which comprises the following steps of: determining factors influencing smelting cost as independent variables and the cost as dependent variables, and designing a test scheme by adopting a quadratic regression orthogonal combination design method. And performing cost accounting on the test scheme, and based on cost data, fitting by adopting a least square method to obtain a quadratic regression equation reflecting the relationship between each variable and the cost. And performing variance analysis and significance test on the quadratic regression equation, and determining key factors and non-key factors which have significant influence on the cost. And the values of non-key factors are fixed, the influence of the key factors on the cost is analyzed, and a scrap steel proportioning scheme with the optimal cost is determined. According to the method, through systematic experimental design, a mathematical model, statistical analysis and a hierarchical optimization strategy, fundamental transformation of converter burdening from experience dependence to model driving and from local optimization to global optimization is achieved, and therefore the minimization of the steelmaking cost is achieved.

Description

Technical Field

[0001] This invention relates to the field of steel manufacturing technology, specifically to a design method for scrap steel batching in converter steelmaking. Background Technology

[0002] Scrap steel is one of the main metallic materials used in converter steelmaking. It provides iron-containing raw materials and serves as a coolant with stable cooling effects. Appropriately increasing the proportion of scrap steel can reduce the cost of converter steelmaking. However, scrap steel from different sources and from different enterprises varies significantly in impurities, elements, and quality, making scrap steel classification and benefit assessment difficult. Consequently, steel companies typically use simple material and heat balance methods to proportion scrap steel.

[0003] In traditional converter steelmaking, the amount of scrap added is typically calculated based on the furnace charge's heat balance, combined with factors such as molten iron temperature and composition. An empirical estimate of the total scrap amount is then fed back to the charging process, usually without specifying the exact proportions between different types of scrap. In practice, the proportions of various scrap types rely heavily on manual experience or historical charging habits, lacking scientific calculation models and failing to consider cost variations due to different scrap proportions. This empirical approach to batching can easily lead to problems such as unstable furnace heat balance, incomplete melting of some scrap, deviations in final temperature, and cost waste when there are significant fluctuations in molten iron composition or changes in production rhythm.

[0004] To address the aforementioned issues, existing technologies only provide information on the charging sequence and addition ratio, but do not address the specific sources of different scrap steel proportions, lack scientific data support, and do not address the impact of different scrap steel proportions on smelting and metallurgical effects, such as the amount of steel material, slag, slag-forming agent, and oxygen consumption. Existing process models cannot reflect the complex nonlinear relationship and interaction effects between scrap steel proportions and steel material costs. Decisions largely rely on experience or simple linear assumptions, resulting in proportioning schemes that cannot approximate the theoretical optimal solution. Furthermore, there is no contribution or regression analysis of scrap steel proportions to costs. Summary of the Invention

[0005] The purpose of this invention is to provide a design method for scrap steel batching in converter steelmaking. Through systematic experimental design, mathematical modeling and statistical analysis, as well as hierarchical optimization strategies, a fundamental transformation is achieved in converter batching from "experience-dependent" to "model-driven" and from "local optimization" to "global optimization," thereby minimizing steelmaking costs and addressing the shortcomings of existing technologies.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for designing scrap steel feedstock for converter steelmaking includes the following steps:

[0008] S1: The factors affecting smelting costs are identified as independent variables, and costs are identified as the dependent variable. The experimental scheme is designed using the quadratic regression orthogonal combination design method.

[0009] S2: Perform cost accounting on the experimental scheme in S1, and based on the cost data, use the least squares method to fit a quadratic regression equation reflecting the relationship between each independent variable and the cost. Let the quadratic term x in the quadratic regression equation be... j 2 Its corresponding encoding is x ij 2 The formula for centering the quadratic term is:

[0010] x ij =x ij 2 -0.843, then according to the encoding formula By performing back substitution, we obtain the natural variables. Regarding the cost of steel materials t The regression equation is given by, where, For encoding symbols, For varying spacing, Let t be the code for natural variables. As the cost of molten iron changes, different regression equations for the cost of steel materials t are obtained. By changing the level of each factor in the regression equation, the influence of a certain type of scrap steel on the cost of steel materials is studied. All other factors are taken as zero level. The influence of the amount of different types of scrap steel added on the cost of steel materials is analyzed.

[0011] S3: Perform variance analysis and significance test on the quadratic regression equation in S2. The variance analysis shows that as the cost of molten iron changes, the impact of each factor, namely the type of scrap steel, on the cost of steel materials is ranked from largest to smallest. The key factors and non-key factors that have a significant impact on the cost are identified. Among them, key factors and non-key factors refer to the type of scrap steel.

[0012] S4: Fix the values ​​of non-critical factors, and set all other factors to zero. Analyze the impact of critical factors on costs and determine the cost-optimal scrap steel ratio scheme.

[0013] Furthermore, in S1, the factors affecting smelting costs are the type of scrap steel, the amount of molten iron and pig iron added; among them, the type of scrap steel is any combination of heavy scrap steel, sheared scrap steel, baled scrap steel, crushed material, rebar cut ends, and light and thin material.

[0014] Furthermore, when the independent variable in S1 is the type of scrap steel, the range of addition amount for different types of scrap steel is first determined based on the composition of molten iron, type of scrap steel, impurity content, price, amount of scrap steel loaded, amount of steel tapped, and the amount of addition based on experience in production practice.

[0015] Furthermore, the cost accounting steps in S2 are as follows:

[0016] S201: The material balance and heat balance calculations for converter steelmaking were performed using an EXCEL spreadsheet for the experimental scheme designed in S1.

[0017] S202: Based on the heat balance calculation results, determine the excess heat of the system, and adjust the amount of independent variables added to make the excess heat approach zero, and input the values ​​of various independent variables under this scheme;

[0018] S203: Based on the values ​​of various independent variables, determine the unit price of the independent variables and calculate the cost.

[0019] Furthermore, the significance test in S3 uses the F-test, which specifically includes calculating the F-value of each regression coefficient and comparing it with the critical F-value to determine whether it is significant at the α=0.01 level. By comparing the F-values ​​of the regression coefficients of each factor, the order of the impact of each scrap steel on the cost is determined, thereby determining the significance of each variable.

[0020] Furthermore, the cost-optimal scrap steel ratio scheme in S4 is determined based on minimizing the cost of the independent variable or the total cost.

[0021] Furthermore, the total cost is the sum of the independent variable cost, the slagging agent cost, and the oxygen consumption cost.

[0022] Furthermore, it also includes predicting smelting costs under specific scrap steel ratios using regression equations, providing real-time data support for production decisions.

[0023] Compared with the prior art, the beneficial effects of the present invention are:

[0024] The scrap steel batching design method for converter steelmaking of the present invention achieves a fundamental transformation from "experience-dependent" to "model-driven" and from "local optimization" to "global optimization" through systematic experimental design, mathematical modeling and statistical analysis, and hierarchical optimization strategy, thereby minimizing steelmaking costs. Attached Figure Description

[0025] Figure 1 A table or graph showing the percentage of the five types of scrap steel in the total amount of scrap steel in Case 1 of this invention;

[0026] Figure 2 This is a table showing the coding of five scrap steel factor levels in Case 1 of this invention;

[0027] Figure 3 The table shows the experimental formulations and orthogonal regression combination design of five types of scrap steel in Case 1 of this invention;

[0028] Figure 4 This is a table showing the proportions of five scrap steel raw materials in Case 1 of this invention;

[0029] Figure 5 This is a cost table for ton steel with a molten iron cost of 1880 yuan / t in Case 1 of the present invention;

[0030] Figure 6 This is a table showing five sets of raw material ratio combinations with lower cost per ton of steel when the cost of molten iron is 1880 yuan / t in Case 1 of this invention;

[0031] Figure 7 A table or graph showing the percentage of the six types of scrap steel in the total amount of scrap steel in Case 2 of this invention;

[0032] Figure 8 This is a table showing the coding of six scrap steel factor levels in Case 2 of this invention;

[0033] Figure 9 The table shows the experimental formulations and orthogonal regression combination design of six types of scrap steel in Case 2 of this invention;

[0034] Figure 10 This is a table showing 10 sets of raw material ratio combinations with lower steel material costs when the cost of molten iron is 1880 yuan / t in Case 2 of this invention;

[0035] Figure 11 This is a table showing the cost per ton of steel in Case 2 of this invention when the cost of molten iron is 1880 yuan / t;

[0036] Figure 12 This is a diagram showing five raw material ratio combinations with lower per-ton steel costs, assuming a molten iron cost of 1880 yuan / t in Case 2 of this invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] This invention provides a method for designing scrap steel feedstock for converter steelmaking, taking six different types of scrap steel (heavy scrap, bundled scrap, sheared scrap, rebar cuts, light scrap, and crushed scrap) as examples, including the following steps:

[0039] S1: The factors affecting smelting costs are identified as independent variables, and cost is the dependent variable. An experimental design using quadratic regression orthogonal combination design is employed; details are as follows:

[0040] (1) The above six types of scrap steel and the ratio (R) of molten iron / pig iron were selected as seven factors for research. In order to construct a mathematical model, the upper and lower limits of the amount of scrap steel added were first set based on information such as molten iron composition, scrap steel type, impurity content, price, and actual working conditions such as the amount of molten iron and scrap steel loaded and the amount of steel produced, combined with the scrap steel ratio adopted by the enterprise in actual production.

[0041] (2) In order to achieve standardized data processing and eliminate the influence of different factor dimensions, according to the relevant calculation formula of the quadratic regression orthogonal combination design method, the coding levels of each factor, such as upper asterisk arm, upper level, zero level, lower level, and lower asterisk arm, are determined, and the specific physical input quantities are converted into unified, dimensionless coding values.

[0042] (3) Due to the large number of experimental factors, 1 / 4 of these seven factors are implemented. By consulting the standard quadratic regression orthogonal combination design table, 47 experimental schemes containing the values ​​of the standard variables x1 to x7 are generated. This scheme matrix can ensure that the complete quadratic model parameters are estimated unbiasedly with the fewest number of experiments. It should be noted that the number of implementations in this embodiment can be selected according to the amount of experimental factors. For example, when there are 5 types of scrap steel and a six-factor orthogonal experiment is designed with the ratio of molten iron to pig iron, 1 / 2 of the implementation can be adopted.

[0043] S2: Perform cost accounting steps on the experimental schemes in S1, and based on the cost data, use the least squares method to fit a quadratic regression equation relating the independent variables to the cost. Specifically: Calculate the actual smelting cost of 47 schemes using a material and energy process model. This material and energy process model is an EXCEL calculation model established based on the company's actual molten iron conditions, endpoint composition, and slagging agent usage. Using this model, the proportions of the 47 scrap steel schemes can be substituted into it. The model will then output a surplus heat. With a surplus of 0 as the target, the amount of coolant, such as iron balls, is adjusted. Until the surplus heat is 0, each scheme can output the unit consumption per ton of steel for one material, such as molten iron and different types of scrap steel. Then, based on the unit price of molten iron and scrap steel, the total cost of steel materials (cost per ton of molten iron + cost per ton of scrap steel) is finally calculated. This cost is mainly obtained by summing the cost per ton of molten iron and different types of scrap steel. Based on the obtained steel material cost, the least squares method is used to calculate the total deviation sum of squares St, regression sum of squares Sr, residual sum of squares Se, and regression coefficients of each term in the regression equation. Based on this, a quadratic regression equation that can reflect the main effect, interaction effect, and secondary effect of each factor is fitted.

[0044] S3: Perform ANOVA and significance tests on the quadratic regression equation in S2 to identify key and non-key factors that have a significant impact on cost. Specifically, use ANOVA and F-significance tests to test the significance of the regression equation and each regression coefficient, including calculating the F-value of each regression coefficient and comparing it with the critical F-value to determine whether it is significant at the α=0.01 level (i.e., the reliability of the model exceeds 99%). By comparing the F-values ​​of the regression coefficients of each factor, determine the order of the magnitude of the impact of each scrap steel on cost.

[0045] S4: Fix the values ​​of non-critical factors and set all other factors to zero. Analyze the impact of critical factors on costs and determine the optimal scrap steel ratio scheme. Specifically, based on the above calculations and analysis, to simplify the decision-making process, factors with a small impact on costs are fixed at zero. The quadratic regression equation is simplified, and the influence curves of critical factors are analyzed to obtain several theoretically "lowest scrap cost" low ratio schemes, thus completing the transformation from a mathematical model to a specific ratio scheme.

[0046] The optimal solution for steel material cost was determined in the above implementation method. The total cost is calculated by adding the consumption of slag-forming agent and oxygen.

[0047] Based on the optimal or several cost-optimal scrap steel ratios obtained from the above experimental scheme, material consumption is calculated using a material and energy process model. This model (i.e., the EXCEL mentioned above) directly outputs the calculation results of the effects of different scrap steel ratios on smelting and metallurgy, such as steel material, slag quantity, slagging agent, and oxygen consumption. Simultaneously, by adding the consumption of slagging agent and oxygen to the cost of steel material, the total cost can be calculated. The optimal total cost schemes are then ranked, ultimately providing steel companies with several cost-optimal scrap steel ratio schemes for selection based on actual site conditions.

[0048] To further verify the feasibility of the method of the present invention, the following specific implementation examples are provided for illustration:

[0049] Case 1: The optimal composition design method for five types of scrap steel is as follows:

[0050] 1. Scrap steel proportioning and factor coding:

[0051] This case study optimizes a 120t furnace charge, investigating the impact of the amounts of heavy scrap, sheared scrap, bundled scrap, rebar cuts, slag steel, molten iron, and pig iron added on smelting costs, and exploring the lowest-cost furnace charge ratio. The percentage of a single scrap piece relative to the total scrap is set as follows: Figure 1 As shown in Table 1.

[0052] The experimental design incorporated six factors, with each type of scrap steel and slag steel representing one factor. Let R = (amount of molten iron added / amount of pig iron added), and take R as the sixth factor. A quadratic regression orthogonal combination design method (1 / 2 implementation) was employed. The zero-level experiment number m0 = 1, and γ = 1.724 was calculated using the asterisk arm length formula. A total of 45 experiments were conducted, and the range of variation for each factor was adjusted according to the actual smelting conditions in the converter steelmaking process. The natural variables were designated Z1, Z2, Z3, Z4, Z5, Z6. To facilitate data processing and calculation, the natural variables were transformed into standardized variables x1, x2, x3, x4, x5, x6 using the coding formula. Factor level coding was performed, resulting in the factor level codes as follows: Figure 2 As shown in Table 2.

[0053] Let x be the quadratic term in the quadratic regression equation. j 2 Its corresponding encoding is x ij 2 The formula for centering the quadratic term is as follows: x ij =x ij 2 -0.843.

[0054] Quadratic regression orthogonal combination design: Experimental formula and orthogonal regression combination design table are shown below. Figure 3 .

[0055] 2. Optimization of furnace charge structure based on steel material cost:

[0056] Quadratic regression equation: Based on the experimentally obtained steel cost, the data is processed to calculate the regression coefficients, total sum of squares (St), regression sum of squares (Sr), and residual sum of squares (Se) of each term in the regression equation, thereby fitting a quadratic regression equation based on normalized variables.

[0057] y t =2066.15+11.26x1+5.35x2+5.85x3+12.28x4-8.41x5-3.54x6+0.03x1x2+0.01x1x3+0.04x1x4-0.07x1x5+0.05x1x6 +0.01x2x3+0.03x2x4-0.04x2x5+0.05x2x6+0.02x3x4+0.01x3x5+0.04x3x6-0.07x4x5+0.04x4x6+0.03x5x6-0.01x1 2 -0.02x2 2 -0.02x3 2 -0.01x4 2 +0.04x5 2 +0.87x6 2

[0058] Among them, y t The standard variable is the cost of steel materials (yuan / t). The standard variable x1 is the code value of the amount of heavy scrap steel added, the standard variable x2 is the code value of the amount of sheared scrap steel added, the standard variable x3 is the code value of the amount of baled scrap steel added, the standard variable x4 is the code value of the amount of rebar cut ends added, the standard variable x5 is the code value of the amount of slag steel added, and the standard variable x6 is the code value of molten iron / pig iron (R).

[0059] The quadratic regression equation for these normalized variables is primarily used for model fit and significance testing during the experimental phase. The regression equations for the coded normalized variables x1, x2, x3, x4, x5, and x6 are derived according to the coding formula. By performing back substitution, we obtain the natural variables Z1, Z2, Z3, Z4, Z5, Z6 with respect to the cost y of steel. t The regression equation is:

[0060] y t =2024.21+7.02z1+3.96z2+4.32z3+9.23z4-15.78z5-1.36z6+0.01z1z2+0.01z1z3+0.02z1z4-0.09z1z5+0.004z1z6+0 .01z2z3+0.02z2z4-0.05z2z5+0.004z2z6+0.01z3z4+0.02z3z5+0.004z3z6-0.10z4z5+0.004z4z6+0.01z5z6-0.002z1 2 -0.01z2 2 -0.01z3 2 -0.01z4 2 +0.13z5 2 +0.01z6 2

[0061] Among them, y t The value is the cost of steel scrap (yuan / t), the natural variable Z1 is the amount of heavy scrap added (%), the natural variable Z2 is the amount of sheared scrap added (%), the natural variable Z3 is the amount of baled scrap added (%), the natural variable Z4 is the amount of rebar cut ends added (%), the natural variable Z5 is the amount of slag steel added (%), and the natural variable Z6 is the amount of molten iron / pig iron (R, dimensionless).

[0062] The independent variables (Z1-Z6) of this natural variable equation are the actual physical quantities in production. In practical applications, operators do not need to perform variable coding conversion. They can directly substitute the actual values ​​(Z1-Z6) such as the planned scrap steel ratio into this equation to quickly predict the corresponding steel material cost (yt), thereby providing real-time and intuitive data support for production decisions.

[0063] Based on the optimal ratio of steelmaking materials based on cost: 10 combinations of raw material ratios with lower steelmaking material costs were determined when the cost of molten iron was 1880 yuan / t, such as... Figure 4 As shown in Table 3, in the several steel material ratio schemes with the lowest cost, the proportions of various scrap steels are relatively low, while the proportion of molten iron + pig iron is relatively high. This indicates that under the price system of molten iron at 1880 yuan / t, increasing the proportion of molten iron is beneficial to reducing costs.

[0064] 3. Optimization of furnace charge structure based on cost per ton of steel:

[0065] The cost of steelmaking materials only includes the cost of scrap steel, molten iron, pig iron, etc. required in the steelmaking process. To comprehensively consider the impact of oxygen, auxiliary raw materials, and other factors on smelting costs and provide a more reasonable reference for actual production, experiments were conducted using 45 schemes obtained through a quadratic regression orthogonal combination design method. This yielded a steel cost per ton of molten iron of 1880 yuan / t. Figure 5 As shown in Table 4.

[0066] Furthermore, based on the optimal ratio for the cost per ton of steel: combining the cost per ton of steel for each scheme, and based on the optimal ratio combination for the cost of steel materials, the following five raw material ratio combinations with lower cost per ton of steel were obtained when the cost of molten iron was 1880 yuan / t, such as... Figure 6 As shown in Table 5.

[0067] Case Study 2: The optimal composition design method for six types of scrap steel is as follows:

[0068] 1. Scrap Steel Proportioning and Factor Coding: This case study optimizes a 120t furnace charge. The experiment investigates the impact of the addition of heavy scrap, sheared scrap, bundled scrap, crushed material, rebar cuts, light scrap, molten iron, and pig iron on smelting costs, aiming to determine the lowest-cost furnace charge proportioning scheme. The percentage of a single scrap steel item in the total scrap steel is set as follows: Figure 7 As shown in Table 6.

[0069] This experiment incorporated seven factors, with each type of scrap steel representing a separate factor. Let R = (amount of molten iron added / amount of pig iron added), and take R as the seventh factor. A quadratic regression orthogonal combination design method was employed (1 / 4 implementation). The zero-level experiment number m0 = 1, and γ = 1.841 based on the asterisk arm length calculation formula. A total of 47 experiments were conducted, adjusting the range of each factor according to the actual smelting conditions in the converter steelmaking process. The natural variables were designated Z1, Z2, Z3, Z4, Z5, Z6, Z7. To facilitate data processing and calculation, the natural variables were transformed into standardized variables x1, x2, x3, x4, x5, x6, x7 using the coding formula. Factor level coding was performed, resulting in the factor level coding table, as follows: Figure 8 As shown in Table 7.

[0070] Let x be the quadratic term in the quadratic regression equation. j 2 Its corresponding encoding is x ij 2 The formula for centering the quadratic term is as follows: x ij =x ij 2 -0.825.

[0071] Quadratic regression orthogonal combination design: experimental formulation and orthogonal regression combination design, such as Figure 9 As shown.

[0072] 2. Optimization of furnace charge structure based on steel material cost:

[0073] Quadratic regression equation: Based on the experimentally obtained steel cost, the data is processed to calculate the regression coefficients, total sum of squares (St), regression sum of squares (Sr), and residual sum of squares (Se) of each term in the regression equation, thereby fitting a quadratic regression equation based on normalized variables.

[0074] y t =2041.16+10.24x1+6.50x2+6.89x3+7.21x4+11.82x5+4.39x6-2.94x7+1.9 1x1x2+1.90x1x3+1.91x1x4+0.59x1x5+1.90x1x6+0.61x1x7+1.90x2x3+1.90 x2x4+0.58x2x5+1.90x2x6+0.61x2x7+1.91x3x4+0.62x3x5+1.94x3x6+0.62x 3x7+0.58x4x5+1.90x4x6+0.61x4x7+0.62x5x6+1.94x5x7+0.62x6x7+0.44x1 2 +0.44x2 2 +0.44x3 2 +0.44x4 2 +0.45x5 2 +0.44x6 2 +1.30x7 2

[0075] Among them, y t The cost of steel scrap (yuan / t) is defined by the following variables: x1 is the coded value for the amount of heavy scrap added; x2 is the coded value for the amount of sheared scrap added; x3 is the coded value for the amount of baled scrap added; x4 is the coded value for the amount of crushed scrap added; x5 is the coded value for the amount of rebar cut ends added; x6 is the coded value for the amount of light and thin scrap added; and x7 is the coded value for molten iron / pig iron (R).

[0076] The quadratic regression equation for these normalized variables is primarily used for model fit and significance testing during the experimental phase. The regression equations for the coded normalized variables x1, x2, x3, x4, x5, x6, and x7 are derived according to the coding formula. By performing back substitution, we obtain the natural variables Z1, Z2, Z3, Z4, Z5, Z6, and Z7 related to the cost y of steel. t The regression equation is:

[0077] y t =2145.53-10.25z1-14.32z2-13.16z3-13.75z4-4.24z5-16.24z6-3.05z7+1 .28z1z2+1.27z1z3+1.28z1z4+0.40z1z5+1.27z1z6+0.06z1z7+1.27z2z3+1.2 7z2z4+0.39z2z5+1.27z2z6+0.06z2z7+1.28z3z4+0.42z3z5+1.30z3z6+0.06 z3z7+0.39z4z5+1.27z4z6+0.06z4z7+0.42z5z6+0.19z5z7+0.06z6z7+0.30z1 2 +0.29z2 2 +0.30z3 2 +0.29z4 2 +0.30z5 2 +0.29z6 2 +0.02z7 2

[0078] Among them, y t The value is the cost of steel scrap (yuan / t). Natural variable Z1 is the amount of heavy scrap added (%), natural variable Z2 is the amount of sheared scrap added (%), natural variable Z3 is the amount of baled scrap added (%), natural variable Z4 is the amount of crushed scrap added (%), natural variable Z5 is the amount of rebar cut ends added (%), natural variable Z6 is the amount of light and thin scrap added (%), and Z7 is the amount of molten iron / pig iron (R, dimensionless).

[0079] The independent variables (Z1-Z7) of this natural variable equation are the actual physical quantities in production. In practical applications, operators do not need to perform variable coding conversion. They can directly substitute the actual values ​​(Z1-Z7) such as the planned scrap steel ratio into this equation to quickly predict the corresponding steel material cost (yt), thereby providing real-time and intuitive data support for production decisions.

[0080] Based on the optimal ratio of steelmaking materials based on cost: 10 combinations of raw material ratios with lower steelmaking material costs were determined when the cost of molten iron was 1880 yuan / t, such as... Figure 10 As shown in Table 8.

[0081] 3. Optimization of furnace charge structure based on cost per ton of steel:

[0082] Cost per ton of steel for each scheme: The cost of steel materials only involves the cost of scrap steel, molten iron, pig iron, etc. required in the steelmaking process. In order to comprehensively consider the impact of oxygen, auxiliary raw materials, etc. required for smelting on the smelting cost and provide a more reasonable reference for actual production, experiments were conducted on 47 schemes obtained by the quadratic regression orthogonal combination design method to obtain the cost per ton of steel when the cost of molten iron is 1880 yuan / t. Figure 11 As shown in Table 9.

[0083] Optimal proportions based on cost per ton of steel: Combining the cost per ton of steel for each scheme, and based on the optimal proportions of steel raw materials, the following five raw material proportion combinations with lower cost per ton of steel were obtained when the cost of molten iron was 1880 yuan / t, such as... Figure 12 As shown in Table 10.

[0084] In summary, the present invention provides a design method for scrap steel batching in converter steelmaking. Through systematic experimental design, mathematical modeling and statistical analysis, as well as hierarchical optimization strategies, it achieves a fundamental transformation in converter batching from "experience-dependent" to "model-driven" and from "local optimization" to "global optimization," thereby minimizing steelmaking costs.

[0085] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A design method for scrap steel feedstock for converter steelmaking, characterized in that, Includes the following steps: S1: The factors affecting smelting costs are identified as independent variables, and costs are identified as the dependent variable. The experimental scheme is designed using the quadratic regression orthogonal combination design method. S2: Perform cost accounting on the experimental scheme in S1, and based on the cost data, use the least squares method to fit a quadratic regression equation reflecting the relationship between each independent variable and the cost. Let the quadratic term x in the quadratic regression equation be... j 2 Its corresponding encoding is x ij 2 The formula for centering the quadratic term is: x ij =x ij 2 -0.843, then according to the encoding formula By performing back substitution, we obtain the natural variables. Regarding the cost of steel materials t The regression equation is given by, where, For encoding symbols, For varying spacing, Let t be the code for natural variables. As the cost of molten iron changes, different regression equations for the cost of steel materials t are obtained. By changing the level of each factor in the regression equation, the influence of a certain type of scrap steel on the cost of steel materials is studied. All other factors are taken as zero level. The influence of the amount of different types of scrap steel added on the cost of steel materials is analyzed. S3: Perform variance analysis and significance test on the quadratic regression equation in S2. The variance analysis shows that as the cost of molten iron changes, the impact of each factor, namely the type of scrap steel, on the cost of steel materials is ranked from largest to smallest. The key factors and non-key factors that have a significant impact on the cost are identified. Among them, key factors and non-key factors refer to the type of scrap steel. S4: Fix the values ​​of non-critical factors, and set all other factors to zero. Analyze the impact of critical factors on costs and determine the cost-optimal scrap steel ratio scheme.

2. The design method for scrap steel feedstock for converter steelmaking as described in claim 1, characterized in that: In S1, the factors affecting smelting costs are the type of scrap steel, the amount of molten iron and pig iron added; among them, the type of scrap steel can be any combination of heavy scrap steel, sheared scrap steel, baled scrap steel, crushed material, rebar cut ends, and light and thin material.

3. The design method for scrap steel feedstock for converter steelmaking as described in claim 2, characterized in that: When the independent variable in S1 is the type of scrap steel, the range of addition amount for different types of scrap steel is first determined based on the composition of molten iron, type of scrap steel, impurity content, price, amount of scrap steel loaded, amount of steel tapped, and the amount of addition based on experience in production practice.

4. The design method for scrap steel feedstock for converter steelmaking as described in claim 1, characterized in that: The cost accounting steps in S2 are as follows: S201: The material balance and heat balance calculations for converter steelmaking were performed using an EXCEL spreadsheet for the experimental scheme designed in S1. S202: Based on the heat balance calculation results, determine the excess heat of the system, and adjust the amount of independent variables added to make the excess heat approach zero, and input the values ​​of various independent variables under this scheme; S203: Based on the values ​​of various independent variables, determine the unit price of the independent variables and calculate the cost.

5. The design method for scrap steel feedstock for converter steelmaking as described in claim 1, characterized in that: In S3, the significance test uses the F-test, which specifically involves calculating the F-value of each regression coefficient and comparing it with the critical F-value to determine whether it is significant at the α=0.01 level. By comparing the F-values ​​of the regression coefficients of each factor, the order of the impact of each scrap steel on the cost is determined, thereby determining the significance of each variable.

6. The design method for scrap steel feedstock for converter steelmaking as described in claim 1, characterized in that: The cost-optimal scrap steel ratio scheme in S4 is determined based on minimizing the cost of the independent variable or the total cost.

7. The design method for scrap steel feedstock for converter steelmaking as described in claim 6, characterized in that: The total cost is the sum of the independent variable cost, the slagging agent cost, and the oxygen consumption cost.

8. The design method for scrap steel feedstock for converter steelmaking as described in claim 1, characterized in that: It also includes predicting smelting costs under specific scrap steel ratios using regression equations, providing real-time data support for production decisions.

Images