JP4915316B2 - Converter blowing control method and converter blowing control system - Google Patents

Description

  The present invention relates to a converter blowing control method for determining a blown oxygen amount and an auxiliary material input amount necessary for appropriately controlling an end point component temperature in converter blowing, and the converter blowing control. The present invention relates to a converter blowing control system used in carrying out the method.

  In converter blowing, blown control that combines static control and dynamic control based on sublance measurement is performed in order to target the molten steel component concentration and molten steel temperature at the time of blowing to the target values. In static control, the amount of blown oxygen and various sub-levels necessary to bring the molten steel component concentration and molten steel temperature at the time of blowing to the target values using a mathematical model based on the mass balance and heat balance before the start of blowing. The raw material input amount is determined, and blowing is performed according to this. On the other hand, in dynamic control, the temperature and carbon concentration of the molten steel are actually measured by sublance during blowing, and the blown oxygen amount determined by static control using a mathematical model based on the mass balance and heat balance. In addition, the amount of input of various auxiliary materials has been optimized to improve control accuracy. Naturally, if the controllability of the static control is high, the performance of the dynamic control is also improved.

  In the static control in the converter blowing described above, the amount of blown oxygen and the amounts of various auxiliary materials charged are determined before the start of blowing in order to bring the molten steel component concentration and molten steel temperature to the target values. As the molten steel component to be controlled, components that affect product quality such as phosphorus other than carbon can be cited, and this phosphorus is removed from the molten steel (hereinafter, removing phosphorus from the molten steel is referred to as “de-P”. ), It is necessary to input auxiliary raw materials such as quicklime containing a CaO component. Hereinafter, description will be made with reference to FIG. 1 showing an outline of the reaction in the converter furnace as appropriate.

  As shown in FIG. 1, unlike carbon, the de-P reaction is a reversible reaction that occurs between molten steel and slag. If a sufficient amount of quicklime is added to increase the CaO concentration in the slag, the de-P reaction will proceed, but the operating costs such as the cost of the quicklime itself and the cost of the heat source required for hatching (melting) of the quicklime will increase. Not only does this increase, but the amount of slag generated increases, and its processing costs also increase.

  Furthermore, depending on other reaction conditions (such as temperature), the de-P reaction may not proceed sufficiently even if the CaO concentration in the slag is high. For this reason, conventionally, a method for determining the input amount of auxiliary raw materials such as quick lime with reference to a table created based on empirical knowledge has been often used.

  As a technique relating to control of converter blowing, for example, Patent Document 1 discloses an auxiliary material input amount calculation device for a converter that can be economically operated with an appropriate and minimum amount of auxiliary material input. . In the technique disclosed in Patent Document 1, a physical model based on a known equilibrium relation (Healy's phosphorus equilibrium formula: phosphorus distribution ratio estimation formula) is used as a constraint condition, and an optimization problem with an auxiliary function cost as an objective function is solved. By solving, economic operation is realized.

  Patent Document 2 discloses a technique related to a method for estimating a phosphorus equilibrium state. In the technique disclosed in Patent Document 2, in order to solve the problem relating to the phosphorus distribution ratio estimation formula disclosed in Patent Document 1, the phosphorus concentration in the molten steel and blow smelting, which are not generally considered in the Hely's phosphorus equilibrium formula A phosphorus distribution ratio estimation formula that reflects the previous charged phosphorus concentration is proposed to improve the accuracy of the distribution formula.

  On the other hand, Non-Patent Document 1 introduces a technique related to a decision tree / regression tree that performs prediction and discrimination by dividing data along a series of procedures without using an explicit function such as regression analysis. Has been. In recent years, a tree-based model is one of the methods often used in the field of data mining, but this tree-based model is a nonlinear regression analysis / discriminant analysis method. In the problem of regression tree (regression tree) and classification, it is called classification tree (decision tree) or decision tree (decision tree). The regression tree model can predict the objective variable (for example, phosphorus distribution ratio, etc.) with explanatory variables (various operating conditions) in a simple branch format, especially when there is interaction or non-linearity between explanatory variables. It is more effective than the multiple regression equation, and is a method used for prediction and discrimination in various fields recently.

Japanese Patent Laid-Open No. 9-256021 JP 2001-152229 A Yamaguchi Kazunori, Takahashi Junichi, Takeuchi Mitsunori: Introduction to Illustrated Basics and Mechanism of Multivariate Analysis, Hidekazu System, 2004

  However, the technique disclosed in Patent Document 1 has a problem in that there is no guarantee that the molten steel temperature assumed under the constraint conditions of the physical model can be realized because the constraint conditions regarding the heat balance are not taken into consideration. In the technique disclosed in Patent Document 1, the influence of other molten steel components, for example, the carbon concentration is a factor of the phosphorus distribution ratio estimation formula. In the actual operation, there is a case where the operation of blowing the carbon concentration below the target value is performed, but there is a problem that the method disclosed in Patent Document 1 cannot cope with the operation. Furthermore, although the technique disclosed in Patent Document 1 is based on Healy's phosphorus equilibrium equation that assumes an equilibrium state for the de-P reaction, it is unlikely that the de-P reaction will reach an equilibrium state in actual operation. There was also a problem that the accuracy was difficult. In addition, in the technique disclosed in Patent Document 1, attention is paid only to the instruction of the auxiliary raw material input amount, and no consideration is given to the molten steel temperature / molten steel carbon concentration, which is the premise thereof. There was a problem in that it was not possible to match the values, and the optimum amount of blown oxygen and auxiliary material input could not be determined from the viewpoint of operation cost.

  Moreover, in the technique currently disclosed by patent document 2, the introduction | transduction form of the phosphorus concentration in the molten steel at the time of blowing in the phosphorus distribution ratio estimation formula and the charging phosphorus concentration before blowing and the phosphorus concentration in the molten steel at the time of blowing And since the calculation method of the parameter with respect to the charged phosphorus concentration before blowing is not disclosed, there is a problem that it is difficult to apply the technique because the truth of the claim cannot be determined. Also, in the technique disclosed in Patent Document 2, as in the technique disclosed in Patent Document 1, attention is paid only to the instruction of the auxiliary raw material input amount, and consideration is given to the molten steel temperature and molten steel carbon concentration. Therefore, there is a problem in that the blowing component / temperature cannot be matched with the target values, and the optimum blowing oxygen amount and auxiliary raw material charging amount cannot be determined from the viewpoint of operation cost. In general, the regression tree model is often applied in social science fields such as marketing. One of the reasons is that it is possible to construct a regression tree model by selecting predictive variables and explanatory variables relatively freely without being bound by scientific principles such as heat balance and mass balance. Therefore, in order to apply the contents disclosed in Non-Patent Document 1 to the converter blowing control involving various in-reactor reactions that are intertwined in a complicated manner, a lot of consideration is given to the principle of scientific principles. There was a technical impediment to carefully selecting effective and appropriate predictors and explanatory variables from among the variables.

  Therefore, the present invention relates to a converter blowing control method that can match the component concentration and temperature at the time of blowing to the target values, and determine the optimum blowing oxygen amount and auxiliary material input amount from the viewpoint of operation cost, and It aims at providing the converter blowing control system used for a converter blowing control method.

  As a result of earnest research to solve the above problems, the present inventors, a mass balance equation based on the reaction in the converter furnace, a distribution ratio estimation equation that defines the distribution ratio of the molten steel components to the molten steel and slag, Constructing an optimization problem using the heat balance equation based on each reaction heat of the converter reactor reaction as a constraint and using the operation amount (injected oxygen amount and various auxiliary material input amounts) cost as an evaluation function The present inventors have found that it is possible to solve the above-described problem of the molten steel temperature by solving the above, and have completed the present invention. The characteristic parts of the present invention are outlined below.

  The blowing component / temperature in the optimization problem is introduced not as a target value (constant) given for each material but as a variable of the optimization problem, and the difference from the target value is evaluated by an objective function. As a result, for example, it becomes possible to cope with a blow-down operation (operation targeting a carbon lower than the target carbon concentration) generally performed when the heat source is insufficient or when de-P is promoted, and an appropriate blow-down instruction is possible.

  Furthermore, regarding the phosphorus distribution ratio estimation formula, a distribution ratio estimation formula based on a regression tree model is introduced instead of a distribution ratio estimation formula based on a multiple regression analysis based on the commonly used Hely's phosphorus equilibrium formula. This time, we found that the distribution ratio can be estimated simply and with high accuracy by expressing the phosphorus distribution ratio estimation formula using a regression tree model. The behavior of the distribution ratio of molten steel components (for example, phosphorus distribution ratio, etc.), which is a limiting condition in the converter blowing control method of the present invention, is highly nonlinear and has a large interaction between explanatory variables. Although the method was insufficient in correspondence, the regression tree model can improve the disadvantages of the conventional method.

  The present invention will be described below. In addition, in order to make an understanding of this invention easy, the reference sign of an accompanying drawing may be appended in parentheses, but this invention is not limited to the form of illustration by it.

  1st this invention determines the quantity of injection | pouring oxygen required in order to make the component concentration and temperature of molten steel at the time of blowing stop, and temperature correspond with target value, and the input amount of auxiliary material, and determined quantity of injection oxygen and auxiliary material This is a static blowing control method for a converter that performs blowing based on the amount of input of the furnace. The balance equation, the heat balance equation of the molten steel, the distribution ratio estimation equation that prescribes the distribution ratio of the components distributed to the molten steel and slag, and the equation related to the operating conditions are included in the above constraints, and the cost of the blown oxygen And a term relating to the cost of the auxiliary material is included in the objective function, and the amount of injected oxygen and the input amount of the auxiliary material are determined by solving the optimization problem. The above problem is solved by the blowing control method.

  Here, specific examples of “component of molten steel” include C, Si, Mn, P, Ti and the like. Furthermore, Mn, P, Fe etc. can be mentioned as a specific example of "the component distributed to molten steel and slag." Furthermore, specific examples of “operation conditions” include conditions relating to the MgO concentration in the slag derived from the viewpoint of protecting the furnace body, conditions relating to basicity derived from the viewpoint of stabilizing de-P, and the like. .

  In the first aspect of the present invention, a term relating to a difference between the component concentration at the time of blowing and the target value of the component concentration, or a term relating to a difference between the temperature at the time of blowing and the target value of the temperature It is preferably included in the function.

  In the first aspect of the present invention (including modifications, the same applies hereinafter), the distribution ratio estimation formula is constructed by a regression tree model having the operation conditions as explanatory variables and the component distribution ratio as an objective variable, The target value of the component concentration is preferably included.

  The second aspect of the present invention is a converter blowing control system (10) used in the converter blowing control method according to the first aspect of the present invention, and at least information relating to charging conditions for the converter, and An input unit (1) for inputting information on target values of component concentration and temperature at the time of blowing, and a constraint condition setting means (2) for setting a constraint condition based on information input from the input unit (1) Objective function setting means (3) for setting an objective function based on information inputted from the input unit (1), constraint conditions set by the constraint condition setting means (2), and objective function setting means (3 ) An optimization calculation means (4) for solving the optimization problem constructed by the objective function set by (1) and calculating an optimal solution; a correction input of information input from the input unit (1); and a constraint condition , Objective function and calculation Outputting information optimal solution is available, input and output unit (5), characterized in that it comprises a a converter blowing control system from (10), to solve the above problems.

  In the second aspect of the present invention, specific examples of “information input from the input unit (1)” include charging weight for each charge, charging components (C, Si, Mn, P, etc.), charging temperature, Also, information related to charging conditions represented by the hot metal ratio, etc., target components (C, Si, Mn, P, etc.) for each charge, target temperature, target basicity organized by material code, etc. Information on the target value can be given.

  According to the first aspect of the present invention, since the value of the objective function can be minimized in consideration of the molten steel temperature, molten steel carbon concentration, etc., the component concentration and temperature at the time of blowing are matched with the target values, and the operation is performed. It is possible to provide a converter blowing control method capable of determining the optimum blown oxygen amount and auxiliary raw material input amount from the viewpoint of cost. Furthermore, it is easy to match the component concentration and temperature at the time of blowing to the target value by considering the difference between the component concentration and the target concentration at the time of blowing and the difference between the temperature and the target temperature at the time of blowing. Become. In addition, since the distribution ratio estimation formula is constructed with a regression tree model, it is possible to accurately estimate a wide range of P distribution ratios, so it is applicable to hot metal after hot metal pretreatment. A blowing control method can be provided.

  According to the second aspect of the present invention, there is provided a converter blowing control system used for implementing the converter blowing control method according to the first aspect of the present invention. Therefore, it is possible to provide a converter blowing control system that can implement a converter blowing control method that can determine the optimum amount of injected oxygen and auxiliary raw material input from the viewpoint of operating costs. In the second aspect of the present invention, considering the difference between the component concentration and the target concentration at the time of blowing and the difference between the temperature and the target temperature at the time of blowing, the component concentration and the temperature at the time of blowing are matched with the target values. It becomes easy to make. In addition, since the distribution ratio estimation formula is constructed using a regression tree model, a wide range of P distribution ratios can be accurately estimated, so a converter blowing control system applicable to hot metal after hot metal pretreatment is provided. can do.

  Embodiments of the present invention will be described below.

  First, Table 1 shows the definition of variables in the optimization problem constructed in the present invention, and Table 2 shows the definition of constants.

  Next, a mass balance equation and a heat balance equation based on the reaction in the converter shown in FIG. 1 are shown below.

式１は、Ｍｎの収支式である。式１において、５５はＭｎの原子量、７１はＭｎＯの原子量を意味する。

Equation 1 is a Mn balance equation. In Formula 1, 55 means the atomic weight of Mn, and 71 means the atomic weight of MnO.

式２は、Ｐの収支式である。式２において、６２は２Ｐの原子量、１４２はＰ_２Ｏ_５の原子量を意味する。

Equation 2 is a balance equation of P. In Formula 2, 62 means the atomic weight of 2P, and 142 means the atomic weight of P ₂ O ₅ .

式３は、ＭｇＯの収支式である。

Equation 3 is a balance equation of MgO.

式４は、スラグの収支式である。式４において、６０はＳｉＯ_２の原子量、２８はＳｉの原子量、８０はＴｉＯ_２の原子量、４８はＴｉの原子量、７２はＦｅＯの原子量、５６はＦｅの原子量を意味する。

Equation 4 is a slag balance equation. In Formula 4, 60 means the atomic weight of SiO ₂ , 28 means the atomic weight of Si, 80 means the atomic weight of TiO ₂ , 48 means the atomic weight of Ti, 72 means the atomic weight of FeO, and 56 means the atomic weight of Fe.

式５は、Ｆｅの収支式である。

Equation 5 is a balance equation of Fe.

式６は、酸素の収支式である。式６において、para_C［(Nm³/ton)/%］、para_Si［(Nm³/ton)/%］、para_Mn［(Nm³/ton)/%］、para_P［(Nm³/ton)/%］、及び、para_Ti［(Nm³/ton)/%］は、各成分（Ｃ、Ｓｉ、Ｍｎ、Ｐ、Ｔｉ）の酸化反応式から理論的に得られる単位成分当たりの必要酸素量を意味する。

Equation 6 is an oxygen balance equation. In Equation 6, para_C [(Nm ³ / ton) /%], para_Si [(Nm ³ / ton) /%], para_Mn [(Nm ³ / ton) /%], para_P [(Nm ³ / ton) /% ] And para_Ti [(Nm ³ / ton) /%] mean the required oxygen amount per unit component theoretically obtained from the oxidation reaction formula of each component (C, Si, Mn, P, Ti). .

式７は、熱収支式である。式７において、seibun.jは、各成分（Ｃ、Ｓｉ、Ｍｎ、Ｐ、Ｔｉ）を意味し、α_seibun.jは、各成分（Ｃ、Ｓｉ、Ｍｎ、Ｐ、Ｔｉ）の酸化反応による単位成分当たりの温度上昇量を意味する。

Equation 7 is a heat balance equation. In Formula 7, seibun.j means each component (C, Si, Mn, P, Ti), and α _seibun.j is a unit obtained by oxidation reaction of each component (C, Si, Mn, P, Ti). It means the amount of temperature rise per component.

  Next, the distribution ratio estimation formula is shown below. The coefficients (para_mn1 to 4, para_p1 to 4, para_fe1 to 4) in the following equations 8 to 11 are obtained by multiple regression. However, for the P distribution ratio, the P distribution ratio based on the regression tree model is used. FIG. 2 shows an example of a method for specifying the P distribution ratio by the regression tree model. Further, FIG. 3 shows the estimation result of the P distribution ratio by the regression tree model shown in FIG. 2 and the estimation result of the P distribution ratio using the Healy distribution formula. FIG. 3A shows the estimation result of the P distribution ratio using the Healy's distribution formula, and FIG. 3B shows the estimation result of the P distribution ratio using the regression tree model. 3A and 3B, the vertical axis represents the estimated value of the P distribution ratio, and the horizontal axis represents the actual value of the P distribution ratio. 2 and 3, Lp means P distribution ratio. As shown in FIG. 3, when the P distribution ratio is estimated by the regression tree model, the standard deviation σ is 7.68. However, when the P distribution ratio is estimated using the Healy distribution formula, the standard deviation σ is 15. 26. That is, by using the regression tree model, it is possible to estimate the P distribution ratio with higher accuracy than when using the Healy distribution formula. Therefore, if the P distribution ratio is estimated using a regression tree model, the present invention can be applied to hot metal after hot metal pretreatment which is often performed in recent years.

式８は、Ｍｎの分配比推定式である。

Formula 8 is an Mn distribution ratio estimation formula.

式９は、Ｐの分配比推定式である。

Equation 9 is a P distribution ratio estimation equation.

  Here, when the target P concentration is 0.01% or more and the blow-off temperature is lower than 1650 ° C., the P distribution ratio (Lp) is 48.P when the P distribution ratio is specified using the regression tree model shown in FIG. 5 and the above equation 9 is expressed as follows.

式１１は、Ｆｅの分配比推定式である。式８、式９、及び、式１１におけるＣＳは塩基度を意味し、下記式１２によって表される。

Equation 11 is an Fe distribution ratio estimation equation. CS in Formula 8, Formula 9, and Formula 11 means basicity and is represented by Formula 12 below.

  Next, an example of constraint conditions due to operation constraints is shown. The following equation 13 is a furnace body protection, and the following equation 14 is an operational constraint for the purpose of de-P stabilization, etc., and is defined using constants (para_MgO, para_CS) set for each charge. If there are other necessary operation constraints, they may be added as appropriate.

上記式１〜式１４が、最適化問題の制約条件である。

The above formulas 1 to 14 are the constraint conditions for the optimization problem.

  Next, the objective function of the optimization problem is shown in Equation 15.

式１５の第一項が副原料コストに関する項、第二項が送酸量コストに関する項、第三項が歩留コストに関する項である。本発明の転炉吹錬制御方法では、式１〜式１５によって構築される最適化問題を、遂次二次計画法等に代表される数理計画法を用いて解くことにより、コストミニマムな操作量（吹込酸素の量及び副原料の投入量）を算出することができる。

The first term of Equation 15 is a term related to the auxiliary raw material cost, the second term is a term related to the acid feed amount cost, and the third term is a term related to the yield cost. In the converter blowing control method of the present invention, the optimization problem constructed by the equations 1 to 15 is solved by using a mathematical programming method represented by a sequential quadratic programming method, etc. The amount (the amount of blown oxygen and the input amount of auxiliary materials) can be calculated.

  Another example of the objective function of the optimization problem is shown in Equation 16.

式１６は、式１５の目的関数に、吹止め成分濃度とその目標値との差に対するペナルティ項（第四項〜第六項）、及び、吹止め温度とその目標値との差に対するペナルティ項（第七項）を追加している。目的関数として、式１５に代えて式１６を使うことによって、例えば、熱源不足時や脱Ｐ促進のために行われる目標炭素濃度よりもさらに小さな吹止め炭素濃度を狙う「吹下げ操業」に対応可能な、操作量（吹込酸素の量及び副原料の投入量）を算出することができる。

Equation 16 includes, in the objective function of Equation 15, a penalty term (fourth to sixth terms) for the difference between the blowing component concentration and its target value, and a penalty term for the difference between the blowing temperature and its target value. (Section 7) is added. By using Equation 16 instead of Equation 15 as the objective function, for example, it corresponds to “Blow Down Operation” aiming at a blown carbon concentration that is smaller than the target carbon concentration performed for lack of heat source or for promoting de-P It is possible to calculate a possible operation amount (amount of blown oxygen and an input amount of auxiliary materials).

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 4 is a conceptual diagram showing an example of a converter blowing control system of the present invention. As shown in FIG. 4, the converter blowing control system 10 of the present invention includes an input unit 1, a constraint condition setting unit 2, an objective function setting unit 3, an optimization calculation unit 4, and an input / output unit 5. . Hereinafter, the function of each part with which the converter blowing control system 10 is provided is demonstrated.

  The input unit 1 includes information regarding charging conditions such as charging weight for each charge, charging component concentration (C, Si, Mn, P, etc.), charging temperature, hot metal ratio, etc., and target components for each charge. This is an interface used when inputting information on target values represented by concentration (C, Si, Mn, P, etc.), target temperature, target basicity organized for each material code, and the like.

  The constraint condition setting means 2 sets the constraint conditions represented by, for example, the above formulas 1 to 14, for each charge, based on the information regarding the charging conditions and the information regarding the target value input via the input unit 1. It is means to do.

  The objective function setting means 3 sets the objective function expressed by, for example, the above formula 15 or 16 for each charge based on the information about the charging conditions and the information about the target value input via the input unit 1. It is means to do.

  The optimization calculation means 4 calculates the optimization problem constructed by the constraint conditions set by the constraint condition setting means 2 and the objective function set by the objective function setting means 3 by mathematical techniques such as successive quadratic programming. It is a means for solving using a planning method and calculating the optimum solution of the manipulated variables (the amount of blown oxygen and the input amount of auxiliary materials). Information about the optimal solution calculated by the optimization calculation means 4 is sent to the input / output unit 5.

  The input / output unit 5 has interface functions such as displaying the calculated optimum solution (the amount of injected oxygen and the input amount of the auxiliary raw material) and information on the constraint condition / objective function, and the correction input of the objective value. Yes.

  FIG. 5 is a flowchart showing a flow of actual processing when the converter blowing control method of the present invention is implemented using the converter blowing control system 10 shown in FIG. Hereinafter, the converter blowing control method of this invention is demonstrated, referring FIG.4 and FIG.5.

  When implementing the converter blowing control method of this invention, first, the information regarding charging conditions and the information regarding target value are input via the input part 1 (process S1). Next, based on the information input in step S1, a constraint condition is set by the constraint condition setting unit 2, and an objective function is set by the objective function setting unit 3 (step S2). When constraint conditions and objective functions are set in step S2, an optimization problem is constructed based on these. Then, the optimization calculation means 4 solves the optimization problem using a mathematical programming method such as a sequential quadratic programming method, and an optimal solution of the manipulated variables (the amount of blown oxygen and the input amount of the auxiliary material) is obtained. Calculated (step S3). Information on the optimal solution calculated in step S3 is sent to the input / output unit 5 and is confirmed by the operator (step S4). In step S4, for example, when the operator determines that there is a possibility that the amount of quicklime calculated in light of past experience is small and P removal is insufficient, the operator corrects target values and parameters shown in Table 2. Data is input to the input / output unit 5 (step S5), and the process returns to step S1. On the other hand, when it is determined in step S4 that the optimum solution has been obtained, the optimum solution is used as an instruction value, and actual blowing is performed based on the instruction value.

  According to the above procedure, according to the converter blowing control method and the converter blowing control system 10 of the present invention, the blowing component / temperature is matched with the target value, and the operation amount (the amount of blown oxygen and The input amount of the auxiliary material) can be determined. This makes it possible to reduce manufacturing costs.

  Hereinafter, the present invention will be further described with reference to FIGS. 4 and 5 and examples.

  Table 3 shows a part of the information input to the input unit 1.

  From Table 2 and Table 3, HM [C%] = 4.00, HM [Si%] = 0.10, HM [Mn%] = 0.30, HM [P%] = 0.05, HMT = 1400 AIM [C%] = 0.14, AIM [Mn%] = 0.50, AIM [P%] = 0.03, and AIMT = 1650. In Formula 15, since no penalty term relating to the difference between the Si concentration at the time of blowing stop and the target value of Si concentration is provided, for the sake of convenience, the target value column of Si [%] in Table 3 is set to 0.00. It was described.

  After inputting the information shown in Table 3 and other information to the input unit 1 (step S1), the constraint condition and the objective function are set in step S2, and the optimization problem is solved based on the set constraint condition and the objective function. It was constructed. Thereafter, in step S <b> 3, an optimum solution of the constructed optimization problem is calculated, and the calculated optimum solution is output to the input / output unit 5. The transition of the optimal solution output to the input / output unit 5 is shown in FIGS. 6 to 10 are diagrams showing the transition of the optimum solution of quick lime, which is a representative auxiliary material (FIG. 6), the diagram showing the transition of the optimum solution of the carbon concentration at the time of blowing (FIG. 7), and the optimum of the scale FIG. 8 shows the transition of the solution (FIG. 8), FIG. 9 shows the transition of the optimal solution for the amount of oxygen blown in (oxygen basic unit), and FIG. 10 shows the transition of the optimal solution for the objective function (FIG. 10). . The vertical axis of FIGS. 6 to 10 represents the calculation result of the optimal solution of quick lime (FIG. 6), the calculation result of the optimal solution of carbon concentration (FIG. 7), the calculation result of the optimal solution of scale (FIG. 8), and the amount of oxygen blown in FIG. 9 shows the calculation result of the optimal solution of (oxygen basic unit) (FIG. 9) and the calculation result of the optimal solution of the objective function (FIG. 10). The horizontal axis in FIGS. is there.

  From FIG. 6 to FIG. 10, since the change of the objective function value is almost eliminated, the optimal solution that becomes the cost minimum is reached in the number of iterations of 10 times or less. And in this calculation, it was an instruction that “the blowing carbon concentration should be blown below the target value”. From the above, according to the converter blowing control method of the present invention, the blowing component concentration and temperature are made to coincide with the target values, and the minimum operation amount (the amount of blown oxygen and the amount of auxiliary material input) is determined. be able to. Therefore, according to the present invention, a great effect of reducing the manufacturing cost can be achieved.

It is a figure which shows the outline | summary of the reaction in a converter furnace. It is a figure which shows the identification method of P distribution ratio by a regression tree model. It is a figure which shows the estimation result of P distribution ratio. It is a conceptual diagram which shows the example of a form of the converter blowing control system of this invention. It is a flowchart which shows the form example of the converter blowing control method of this invention. It is a figure which shows transition of the optimal solution of quicklime. It is a figure which shows transition of the optimal solution of the carbon concentration at the time of a blowing stop. It is a figure which shows transition of the optimal solution of a scale. It is a figure which shows transition of the optimal solution of the blowing oxygen amount (oxygen basic unit). It is a figure which shows transition of the optimal solution of an objective function.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Input part 2 ... Restriction condition setting means 3 ... Objective function setting means 4 ... Optimization calculation means 5 ... Input / output part 10 ... Converter blowing control system

Claims (4)

Determine the amount of blown oxygen and the amount of auxiliary material required to make the concentration and temperature of molten steel coincide with the target values at the time of blowing, and based on the determined amount of blown oxygen and the amount of charged auxiliary material A static blowing control method for a converter,
Build an optimization problem that optimizes the value of the objective function under the constraints,
The material balance equation of the molten steel component, the heat balance equation of the molten steel, the distribution ratio estimation equation that defines the distribution ratio of the component distributed to the molten steel and slag, and the equation relating to the operating conditions are the constraints. Included
A term relating to the cost of the blown oxygen and a term relating to the cost of the auxiliary material are included in the objective function,
The converter blowing control method, wherein the amount of the blown oxygen and the input amount of the auxiliary material are determined by solving the optimization problem. Further, the term relating to the difference between the component concentration at the time of blowing and the target value of the component concentration, or the term relating to the difference between the temperature and the target value of the temperature at the time of blowing is included in the objective function. The converter blowing control method of Claim 1 characterized by these. The distribution ratio estimation formula is constructed by a regression tree model having an operation condition as an explanatory variable and a distribution ratio of the component as an objective variable, and the operation condition includes a target value of the component concentration. 3. A converter blowing control method according to claim 1 or 2. It is a converter blowing control system used for the converter blowing control method of any one of Claims 1-3,
At least an input unit for inputting information on charging conditions to the converter, and information on target values of the component concentration and temperature at the time of blowing,
A constraint condition setting means for setting the constraint condition based on the information input from the input unit;
An objective function setting means for setting the objective function based on the information input from the input unit;
Optimization calculation means for solving the optimization problem constructed by the constraint condition set by the constraint condition setting means and the objective function set by the objective function setting means and calculating an optimal solution;
An input / output unit capable of outputting the correction input of the information input from the input unit, the constraint condition, the objective function, and the information of the calculated optimal solution;
A converter blowing control system, comprising:

Images