JP2020002390A - Blowing calculation method, blowing calculation program - Google Patents

Abstract

To provide a blowing calculation technique capable of accurately calculating an input amount of oxygen or the like to be supplied in a blowing step.SOLUTION: In the blowing calculation method according to the present invention, an amount of the oxygen, an amount of a temperature rising material, and an amount of a coolant, required in a first period and a second period until a first moment and a second moment respectively, at which an amount of chemical components contained in a molten steel and a molten steel temperature are measured during a blowing step, are calculated. An amount of the oxygen, an amount of the temperature rising material, and an amount of the coolant required in the second period are calculated using a measured unidentified heat amount and a measured unidentified oxygen amount in addition to the amount of the chemical components, the molten steel temperature and the like measured at the first moment.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a technique for calculating parameters in a blowing process.

  In the steelmaking process, there is a blowing process in which oxygen is blown to hot metal or molten steel in a converter or an AOD. In the subsequent process of the blowing process, a continuous process of solidifying the molten steel to produce a slab or other steel slab is performed. There is a casting process. In order to operate this continuous casting process stably, it is necessary to raise the temperature to the molten steel temperature at which continuous casting is possible in the blowing process, and the target molten steel temperature at the time of the blow stop according to the steel type and casting time is set. . In the blowing process, the molten steel is blown with oxygen to raise the heat by an oxidation reaction of each component. At this time, if the amount of heat rise is insufficient, the target molten steel temperature will not be reached, so a heat-raising material such as coke is charged into the furnace to compensate for the heat rise. In addition, when the amount of heat rise becomes excessive only by the carbon brought in from the hot metal, iron ore or the like is put into the molten steel and cooled. As described above, since the amount of carbon introduced from the hot metal and the amount of coke used for heating are large, the setting of the amount of oxygen to be blown is greatly affected. Therefore, in order to hit the target molten steel temperature and the target carbon concentration at the time of blowing, it is necessary to calculate in advance the amount of oxygen, the amount of heat-raising material, the amount of coolant, etc. in the blowing process. In the blowing process, after the calculated amounts of the heat-up material and the coolant, the molten steel temperature and the molten steel component are measured using a sub-lance at an intermediate stage of the blowing process to confirm the actual values. If the measured values of the molten steel temperature and the carbon concentration deviate greatly from the respective target values, an additional action, specifically, a heating by adding a heating material or a cooling by adding a cooling material is performed. In addition, according to the measured actual carbon concentration and the additional amount of the heating material, the amount of oxygen from the time of measuring the sublance is adjusted again. In these blowing calculations, each input amount can be calculated based on the fact that the balance of the calorific value, the balance of the oxygen amount, and the balance of the mass loss due to the oxidation reaction are balanced throughout the process. However, in the actual blowing process, the balance between the calorific value and the oxygen amount does not match, and there is an unknown calorie and an unknown oxygen amount. Therefore, it is important how to estimate the unknown heat and the unknown oxygen with high accuracy as an issue of the blowing calculation. In the conventional operation, the operator determines unknown heat and unknown oxygen based on his / her own experience, and performs blowing calculation.

  Japanese Patent Application Laid-Open No. H11-163,199 discloses "estimating the unknown heat amount and the unknown oxygen amount included in the static blowing control model formula with high accuracy, and improving the accuracy of the static blowing control. For the purpose, "the operation information t1 to tn in the converter steelmaking is input, and the unknown calorie, which is the difference between the theoretical calorific value and the theoretical heat absorption amount, and the difference between the theoretically generated oxygen amount and the theoretically consumed oxygen amount. A neural network that outputs the unknown oxygen amount is constructed, and the unknown heat amount and the unknown oxygen amount are estimated using the neural network. At this time, learning of the neural network is performed using the operation information at the time of actual operation and the unknown heat amount and the unknown oxygen amount, and the weight coefficient between units constituting the neural network is corrected. ] (See abstract).

JP-A-6-200312

  A conventional blowing method as disclosed in Patent Document 1 estimates an unknown heat amount and an unknown oxygen amount at the time of starting the blowing process, and calculates each parameter in the blowing process based on the estimation result. However, since the unknown calorie and the unknown oxygen amount are different each time the blowing process is performed, the unknown calorie and the unknown oxygen amount estimated at the time before the start of the blowing process are the unknown calorie and the unknown oxygen amount calculated based on the actual measurement values. It may not always be accurate when compared to quantities. Therefore, each input amount obtained in the blowing calculation may not always be highly accurate.

  The present invention has been made in view of the above-described problems, and is intended to reduce the amount of oxygen or the like to be supplied to the blowing step in order to obtain the target molten steel component amount and the molten steel temperature during blowing. An object of the present invention is to provide a blowing calculation technique capable of calculating with high accuracy.

  The blowing calculation method according to the present invention is characterized in that the amount of oxygen required in the first period and the second period until the first time point and the second time point, respectively, when actually measuring the amount of the component contained in the molten steel and the molten steel temperature in the blowing process. , The amount of heating material and the amount of coolant are calculated. The amount of oxygen, the amount of heat-up material, and the amount of coolant required in the second period from the first time to the second time are unknown in addition to the actually measured component amount and the actually measured molten steel temperature at the first time. Calculate using calorific value and unknown oxygen content.

  According to the blowing calculation method according to the present invention, by performing the blowing calculation again using the measured unknown calorific value and the measured unknown oxygen amount, in order to obtain the target molten steel component amount and molten steel temperature during blowing. It is possible to accurately calculate an appropriate amount of input of oxygen or the like.

1 is a configuration diagram of a blowing calculation device 100 according to Embodiment 1. FIG. It is a schematic diagram of a blowing process in the case of using a converter. It is a conceptual diagram explaining the heat balance and oxygen balance in a blowing process. FIG. 4 is a diagram illustrating a calculation example for explaining a blowing calculation method according to the first embodiment. 9 is a flowchart illustrating an operation of a yield estimation unit 113. It is an example of a screen presented by blowing machine 100.

<Embodiment 1>
FIG. 1 is a configuration diagram of the blowing computer 100 according to the first embodiment of the present invention. The blowing calculation device 100 is a device that calculates blowing parameters such as the amount of oxygen to be supplied to the molten steel blowing process. The blowing calculation device 100 includes a CPU (Central Processing Unit) 110 and a storage device 120.

  The CPU 110 executes an input amount calculation unit 111, an unknown amount estimation unit 112, a yield estimation unit 113, and a residual amount estimation unit 114. Details of these functional units will be described later. These functional units can be configured using hardware such as a circuit device that implements the function, or can be configured by the CPU 110 executing software that implements the function. In the following description, it is assumed that these functional units are configured as software. For convenience of description, these functional units may be mainly described as operation units, but it is the CPU 110 that actually executes these functional units.

  The input amount calculation unit 111 calculates an oxygen amount, a heating material amount, and a cooling material amount required in the blowing process. The calculation procedure will be described later. The unknown amount estimating unit 112 estimates the unknown heat amount and the unknown oxygen amount in the blowing process to be performed based on the learning result data 121. The yield estimating unit 113 estimates the amount of the heating material actually input to the converter among the heating materials input to the blowing process (details of the second embodiment). The residual amount estimating unit 114 estimates the amount of a specific element component remaining in the molten steel in the converter (for details, Embodiment 3).

  The learning result data 121 is data in which a result of learning a parameter in a blowing process performed in the past by machine learning is recorded. Specific examples of the parameters will be described later. Most simply, the parameters themselves in the blowing process performed in the past can be recorded as the learning result data 121. The machine learning device may be provided in the blowing calculation device 100 itself, or may use a machine learning device outside the blowing calculation device 100 and use the learning result as learning result data 121. You may.

  FIG. 2 is a schematic diagram of a blowing process in the case of using a converter. Scrap and hot metal are charged into the converter, and after the start of blowing, ferroalloys (such as FeCr), a heating material, a cooling material, and the like are charged. The steelmaking process of oxidizing carbon and decarburizing by blowing oxygen or a mixed gas containing oxygen and nitrogen or argon onto the molten iron in which these are melted is generally called a blowing process. In addition, it is common that smelting for reducing other impurity components in the molten iron is performed simultaneously by using the slag formed by the added auxiliary raw material and the generated oxide through the blowing process.

  FIG. 3 is a conceptual diagram illustrating a heat balance and an oxygen balance in the blowing process. In the blowing process, molten iron (hot metal in a steel process using a blast furnace and a converter), scrap, ferroalloys (alloys containing chromium, etc.), and auxiliary materials (burnt lime, etc.) are supplied, and oxygen is supplied. Thus, the target amounts of the respective element components at the target time and the target temperature at the target time are realized.

When scrap or alloyed iron is charged into a furnace containing molten iron, the temperature of the molten iron decreases. To raise the temperature of the molten iron to the target temperature, a heating material such as coke is introduced. The amount of heat input to the blowing process is the heat of oxidation reaction of each component such as carbon, silicon, chromium, etc. contained in molten iron, scrap, ferromagnetic iron, etc., and the heat of oxidation reaction of carbon etc. contained in heating materials such as coke. is there. Here, each heat of oxidation reaction is a value excluding the sensible heat of the generated oxide. The heat of oxidation reaction of carbon includes the primary combustion heat when CO gas is generated from the molten iron and the furnace internal space. in including a portion of the heat generated by the secondary combustion CO is further combusted to CO ₂ (secondary combustion heat), since the remainder of the secondary combustion heat is dissipated to the outside of the system as such sensible heat of the exhaust gas, A product obtained by multiplying the secondary combustion heat by a predetermined heating efficiency is a heat input component. Here, the ratio (secondary combustion rate) at which the secondary combustion of the CO gas occurs and the heating efficiency may be, for example, a standard value obtained from past operation results, or may be an operating condition based on past operation results. It may be determined accordingly. On the other hand, the heat output in the blowing step is the sensible heat increase and heat of fusion of each raw material such as molten iron and scrap. It is a principle that these heats are balanced, but in practice, unknown heat components such as radiant heat from the furnace side wall exist. Also, if the actual secondary combustion rate or heat transfer efficiency fluctuates from the estimated values, or if there is an error in the evaluation of the oxidation amount of chromium or the like, the fluctuations in the heat balance due to these factors should be included in the unknown heat. become. Therefore, the actual heat balance has a relationship as shown in FIG.

  In the steel making process, components such as carbon are burned by blowing oxygen to the molten iron, whereby steel can be refined. In the blowing calculation, the amount of oxygen necessary to achieve the target component amount at the target time is calculated. Since oxygen is generated by reduction of oxides such as manganese ore in addition to the oxygen to be blown, the sum of these becomes the oxygen input amount in the blowing step. The amount of oxygen consumed in the oxidation reaction of each component such as silicon and chromium in addition to carbon and the amount of oxygen consumed in the secondary combustion are the amount of oxygen output in the blowing step. The amount of oxygen consumed for the oxidation of each component in the molten iron can be calculated by measuring the change in the concentration of each component, but it is difficult to accurately evaluate the amount of oxygen consumed for the oxidation of iron and the secondary combustion, It is thought that it may fluctuate depending on the difference of blowing conditions. Therefore, there is an unknown amount of oxygen in the same manner as the unknown amount of heat, so that the actual oxygen amount balance has a relationship as shown in FIG.

  In addition to heat balance and oxygen balance, there is also a mass balance. The mass of the material and the like charged into the blowing process is (mass of hot metal) + (mass of scrap) + (mass of ferroalloy) + (mass of heating material) + (mass of coolant) . The mass of the molten steel after the completion of the blowing process is a value obtained by subtracting the reduced amount of those element components whose mass is reduced by the oxidation reaction.

  The heat balance, the oxygen balance, and the mass balance described above are described by three equations describing respective calculation formulas. These three equations are defined as three variables of a heating material amount, an oxygen amount, and a cooling material amount (hereinafter, abbreviated as heating material amount / oxygen amount / cooling material amount) to be supplied to the blowing process. By solving the simultaneous inequalities consisting of, these three variables can be obtained. This is generally called a blowing calculation. However, since unknown heat and unknown oxygen are included in each equation, these values must be filled in to solve the simultaneous inequalities.

  FIG. 4 is a diagram illustrating a calculation example for explaining the blowing calculation method according to the first embodiment. In the blowing calculation, generally, first, the amount of ferroalloys to be introduced, the amount of auxiliary materials, and the amount of heating material / oxygen / cooling material are determined so that molten iron having a target composition and temperature can be obtained throughout blowing. However, it is common that most of the ferroalloys and auxiliary raw materials are introduced at the beginning of blowing. Next, the target molten steel temperature and the target component amount at a certain point in the blowing process are set, and the above-mentioned iron alloy and auxiliary raw material input amounts are assumed so that these target values are realized. The amount of heat-raising material / oxygen / cooling material to be charged up to is calculated. In the conventional blowing process, for example, at a certain oxygen amount at which oxygen is blown, the amount of each elemental component and the molten steel temperature are actually measured using a device called a sublance (SL), and based on the result, the oxygen amount blown thereafter Etc. are being adjusted. However, the blowing calculation is performed only once at the start of the blowing process, and the adjustment of the heating material amount / oxygen amount / cooling material amount after the measurement of the sublance is performed by the decarburization amount and the temperature increase required from the measurement of the sublance. In order to obtain the amount, a simple calculation using an appropriately assumed decarboxylation efficiency and a heating coefficient was performed. Then, based on the unknown heat and the unknown oxygen estimated at the time of starting the blowing process, the heating material amount / oxygen amount / cooling material amount to be charged is obtained, and the blowing is performed using the result. Therefore, there is considerable variation in the carbon content and temperature of the molten steel at the time of measurement with the sublance, and even if the adjustment after the measurement of the sublance is performed as described above, the target of the finally obtained molten steel components and the molten steel temperature Medium accuracy was not always high. In the first embodiment, in view of the above problem, the blowing calculation is performed twice.

  Specifically, before starting the blowing process, the target temperature (SL1 target temperature) and the target component (SL1 target concentration, in this case, the residual carbon concentration here) at the first sublance point (SL1) are set as the target component values. ) And the target temperature (SL2 target temperature) and the target component (SL2 target concentration) at the second sublance time point (SL2). Before the start of blowing, the first blowing calculation is performed to obtain the molten iron of the target composition and temperature throughout the blowing. The primary calculation value of the amount / cooling material amount is obtained, and (a) the heating material amount / oxygen amount / cooling material amount to be supplied in the first period from the start of blowing to SL1 is obtained. (B) Heating material amount / oxygen amount / cooling material amount to be supplied in the second period from SL1 to SL2, or heating material amount to be supplied in the combined period of the first period and the second period / The amount of oxygen / the amount of cold material may be determined preliminary. The blowing calculation is performed by the input amount calculation unit 111. At this time, the unknown amount estimating unit 112 estimates the unknown heat and the unknown oxygen in each period used in each blowing calculation. The heat-raising material amount / oxygen amount / cooling material amount obtained in the above (a) is based on the temperature and carbon concentration of the molten steel (carbon concentration and chromium concentration in the blowing of chromium-containing molten steel) at the time of SL1 injection. If the hit accuracy is good, the hit accuracy of SL2 can be improved. As a result, stable continuous casting in the next step is enabled.

  At the time of SL1, the molten steel temperature is actually measured (actually measured temperature), and the concentration of each component remaining in the molten steel in the furnace is actually measured (measured concentration). The measurement of the component concentration is performed by a device analyzer capable of performing a rapid analysis such as a spark discharge emission analysis or a fluorescent X-ray analysis on an analysis sample collected from molten iron by SL1. The carbon concentration can be measured in a shorter time by measuring the solidification temperature of the molten iron sample. Further, for example, the weight of carbon may be calculated from the material balance in the furnace by a method described in a third embodiment described later. The unknown quantity estimating unit 112 calculates the actual values of the unknown heat and the unknown oxygen (the actually measured unknown heat and the actually measured unknown oxygen) in the first period from the start of blowing to SL1 based on the actually measured temperature and the actually measured concentration. Also at SL2, the actual values of unknown heat and unknown oxygen in the second period from SL1 to SL2 are calculated based on the actually measured temperature of the molten steel and the respective component concentrations. The actual values of the unknown heat and the unknown oxygen in the first period and the actual values of the unknown heat and the unknown oxygen in the second period calculated in this manner will be described later so as to be usable in the blowing calculation of the subsequent charge. It is desirable to record in the storage device 120 such that a part of the learning result data 121 is added or updated together with other blowing parameters.

  The unknown amount estimating unit 112 estimates unknown heat and unknown oxygen in the first period from the known blowing parameters using the learning result data 121 at the start of the blowing process. At this time, as will be described later, it is desirable to estimate unknown heat and unknown oxygen based on a result of searching a past blowing process close to the blowing process of the charge from each blowing parameter by machine learning, as described later. Further, unknown heat and unknown oxygen in the first period are obtained as function expressions of blowing parameters by machine learning or multiple regression analysis performed based on the learning result data 121 in advance, and are calculated from known blowing parameters. You can also. In these cases, the known blowing parameters include, for example, the amount of scrap charged before the start of the blowing process, the amount of iron alloy charged before the start of the blowing process, and the The amount of molten iron charged, the first target component amount at the first time point (SL1), the second target component amount at the second time point (SL2), the first target molten steel temperature at the first time point, and the second at the second time point. The target molten steel temperature is included, and in addition to these, the temperature of the molten iron charged before the start of the blowing process, the concentration of each component, and the amount of ferrochrome and other ferrochrome to be charged in the first period are also included. Good.

  The unknown quantity estimating unit 112 further obtains a learning result from the calculated actual values of the unknown heat and the unknown oxygen in the first period and other known blowing parameters including the molten steel temperature and the respective molten steel component concentrations actually measured in SL1. Using the data 121, the unknown heat and the unknown oxygen during the second period of the blowing of the charge are estimated. In this case as well, it is desirable to estimate unknown heat and unknown oxygen based on a result of searching a past blowing process close to the blowing process of the charge from each blowing parameter by machine learning by machine learning. Further, unknown heat and unknown oxygen in the second period are obtained as a function formula of blowing parameters by machine learning or multiple regression analysis performed based on the learning result data 121 in advance, and are calculated from known blowing parameters. You can also. In these cases, the known blowing parameters include, for example, the amount of scrap charged before the start of the blowing process, the amount of iron alloy charged before the start of the blowing process, and the The amount of molten iron charged, the amount of ferro-alloy charged during the blowing process up to the first time, the first measured component amount and the first measured molten steel temperature at the first time, the second target component amount at the second time, and The second target molten steel temperature and the actual values of unknown heat and unknown oxygen in the first period are included. Here, the second target component amount and the second target molten steel temperature may be the same as those set earlier, or may be set again.

  The input amount calculation unit 111 uses the estimated values of the unknown heat and the unknown oxygen in the first period estimated by the unknown amount estimation unit 112, in addition to the blowing parameters determined at the start of the blowing, to start the blowing. A first blowing calculation is performed to determine the amount of heat-raising material / the amount of oxygen / the amount of cold material to be charged in the first period from to SL1. The input amount calculation unit 111 estimates the unknown heat and the unknown oxygen in the second period estimated by the unknown amount estimation unit 112 in addition to the actually measured values such as the molten steel temperature and the respective molten steel component concentrations at the time of SL1. Using the measured values and the estimated values of the unknown heat and the unknown oxygen in the second period, the heating material amount / oxygen amount / cooling material amount to be supplied in the second period from SL1 to SL2 is obtained. Is performed for the second time. Thereby, the accuracy of the amount of heat-raising material / the amount of oxygen / the amount of cold material to be supplied in the second period in order to obtain the target amount of carbon in the molten steel and the target temperature of the molten steel at the second time point can be improved as compared with the conventional case. it can. As a result, it is possible to reduce the error between the target value of the amount of carbon in the molten steel and the temperature of the molten steel obtained at the second time point.

  When performing blasting of chromium-containing molten iron, when performing each of the above-mentioned blasting calculations, separate the estimated values of unknown heat and unknown oxygen and estimate the chromium oxidation amount in each period. In addition, it is necessary to estimate the amount of oxygen consumed for chromium oxidation and the heat of reaction due to chromium oxidation and use them in the blowing calculation. The estimation of the amount of chromium oxidation in each of these periods includes the actual value of the amount of chromium oxidation obtained by the unknown amount estimating unit 112 from the change in the component concentration of the steel in each period, similarly to the estimation of the unknown heat and the unknown oxygen. Using the learning result data 121, it can be estimated by machine learning based on a result of searching for a past blowing process close to the blowing process of the charge from the known blowing parameters at the time of each blowing calculation by machine learning. At this time, the actual value of the chromium oxidation amount in each period may be calculated by the unknown amount estimating unit 112 at each blowing and accumulated in the learning result data 121. Further, the chromium oxidation amount in each period is previously obtained as a function formula of the blowing parameter by machine learning or multiple regression analysis performed based on the learning result data 121 including the actual value of the chromium oxidation amount in each period. , Can also be calculated from known blowing parameters. Furthermore, as described later, the amount of chromium oxidation can also be calculated sequentially during blowing from the calculation of the material balance using the information of exhaust gas analysis, so by checking the transition tendency of the amount of chromium oxidation during blowing. Alternatively, it can be used as a material for determining whether the estimated value of the chromium oxidation amount in each period is within an appropriate range.

<Embodiment 1: Summary>
Before starting the blowing process, the blowing calculation device 100 according to the first embodiment performs the blowing calculation in accordance with the target molten steel temperature and the target component amount in each of SL1 and SL2, and measures the measured molten steel temperature at the time of SL1. The blowing calculation in the second period is performed according to the measured component amount. Thereby, the accuracy of the amount of heat-raising material, the amount of oxygen, and the amount of cold material to be charged in SL1 to SL2 in order to obtain the target amount of carbon in the molten steel and the target temperature of the molten steel in SL2 can be increased as compared with the related art.

  The reason is described below. In SL1, when the actual molten steel temperature and the actual carbon amount deviate from the respective target ranges with respect to the target molten steel temperature and the target carbon weight due to some influence, the second period calculated based on the target molten steel temperature and the target component of SL1. Cannot be blown by setting the heat-up material / oxygen amount / coolant. Therefore, the blowing calculation is performed using the actual molten steel temperature and the actual molten steel component at the point of time SL1. Here, it is considered that the cause of the error in the actual molten steel temperature and the actual carbon amount of SL1 is due to an error in the estimated value of unknown heat and unknown oxygen in the first period from the start of blowing to SL1. As described above, some of these unknown quantities are considered to be based on the error in the estimated values of the heat dissipated in the furnace, the secondary combustion rate and the heating efficiency, and the errors in the estimated values of the oxidation amounts of iron and chromium. It is considered that the unknown amount in the first period and the unknown amount in the second period of SL1 and SL2 are related to each other. Therefore, the relationship between the two is evaluated by machine learning, etc., and the unknown heat and unknown oxygen in the second period are estimated using the actual values of the unknown heat and unknown oxygen in the first period. The unknown heat and unknown oxygen used for the blowing calculation in the second period can be set based on the values, so that the target amount of carbon in the molten steel and the molten steel temperature to be obtained in SL1 to SL2 are obtained in SL2. Heat material / oxygen amount / cooling material amount can be obtained with higher accuracy.

<Embodiment 2>
As a result of the calculation of the blowing performance described in the first embodiment, the actual values of the unknown heat and the unknown oxygen in the first period and the second period may be negative values. According to the findings of the present inventors, this phenomenon occurs when the amount of the heat-up material supplied is large. As a cause of this phenomenon, the present inventors have found that a part of the heating material charged into the blowing process is not actually charged into the furnace but spilled out of the furnace or blown before entering the molten steel. Despite the fact that the oxygen and the coke react directly with each other, the blowing calculation assumes that all the heating material reacts after being dissolved in the molten steel. . In the second embodiment of the present invention, a procedure for correcting unknown oxygen based on this concept will be described. The configuration of the blowing computer 100 is the same as that of the first embodiment.

  FIG. 5 is a flowchart illustrating the operation of the yield estimation unit 113. Hereinafter, each step of FIG. 5 will be described.

(FIG. 5: Step S501)
The yield estimating unit 113 initializes a ratio (hereinafter referred to as a yield) of the heat-up material supplied to the blowing process, which actually enters the furnace and reacts. Here, the yield when all the heat raising materials are actually charged into the furnace is set to 1.0, and this is set as the initial value. When coke is used as the heating material, the ratio of the fixed carbon amount in the entire coke may be analyzed in advance, and may be used as the initial value of the yield.

(FIG. 5: Step S502)
The yield estimating unit 113 calculates the amount (mass) of each element component at the time of SL1 and SL2 by using the actually measured value of each element component (concentration) of the molten iron at the time of SL1 and SL2. In addition, the yield estimating unit 113 calculates the charged amounts and the respective component concentrations of the raw materials such as the scrap, the amount of the iron alloy, and the amount of the molten iron charged before the start of the blowing process, and the amount of the alloy charged during the blowing process. The total amount of each component charged or charged into the furnace at the time of SL1 and SL2 is calculated from the calculation of the material balance based on the input amounts of the respective additive materials such as iron and a heating material and the respective component concentrations. Here, when calculating the total amount of each component, only the input amount of the heating material is multiplied by the yield value.

(FIG. 5: Step S503)
The yield estimation unit 113 calculates unknown oxygen (SL1 unknown oxygen) in the first period and calculates unknown oxygen (SL2 unknown oxygen) in the second period using the result of step S502. Specifically, it corresponds to the total input amount of each element component in each period calculated in step S502 and the decrease amount of each element component in each period calculated from the actual value of each element component amount at the end of each period. Oxygen consumption can be determined, and unknown oxygen can be determined according to the oxygen balance equation.

(FIG. 5: Step S504)
The yield estimation unit 113 checks whether both the SL1 unknown oxygen and the SL2 unknown oxygen are positive values. If both are positive values, this flowchart ends. If any of them is a negative value, a predetermined amount (for example, 0.01) is subtracted from the current yield value, and the process returns to step S502 to repeat the same process. This makes it possible to indirectly estimate the amount of heat-raising material actually charged into the furnace.

(FIG. 5: Step S504: Supplement No. 1)
In this step, if the unknown oxygen is a positive value, it is assumed to be normal. However, instead of this, the unknown oxygen is within a certain positive value range (for example, if the unknown oxygen is It may be determined whether or not the calculation result of the unknown oxygen is normal based on whether or not the calculation result is within the range.

(FIG. 5: Step S504: Supplement No. 2)
In this step, the yield of the heating material is determined based on whether or not the unknown oxygen is within a predetermined range, because the oxygen balance can be relatively accurately grasped and the calculated ratio of the unknown oxygen is relatively high. Because there are few. That is, at the time of SL1, each component amount can be calculated to some extent accurately based on the actually measured values, and the ratio of oxygen consumed for secondary combustion and iron oxidation is relatively small, so that the unknown oxygen is also relatively small. This is because it can be calculated accurately. The second embodiment makes use of this fact and assumes that if the calculated unknown oxygen is negative, some of the heating materials charged to the blowing step were not charged to the furnace. Things.

<Embodiment 3: Residual carbon weight>
In the above embodiment, it has been described that the blowing calculation is performed at the start of the blowing process and at the time of SL1. In the actual blowing process, the operator may reset the timing of SL2 according to the actual measurement value at the time of SL1. In a third embodiment of the present invention, a function for assisting this work will be described. The configuration of the blowing computer 100 is the same as that of the first embodiment.

Regarding the exhaust gas in the blowing step, the flow rate of the exhaust gas, the concentration of CO in the exhaust gas, and the concentration of CO ₂ in the exhaust gas can be measured by sensors installed near the furnace. This measured value can be obtained at an arbitrary timing. The residual amount estimating unit 114 can obtain the carbon weight remaining in the furnace by the following procedure using the acquired actual measurement values.

(Calculation procedure 1 of carbon residue)
The residual amount estimation unit 114 acquires the exhaust gas flow rate, the CO concentration in the exhaust gas, and the CO ₂ concentration in the exhaust gas. The residual amount estimation unit 114 calculates the CO flow rate by multiplying the exhaust gas flow rate by the CO concentration, and calculates the CO ₂ flow rate by multiplying the exhaust gas flow rate by the CO ₂ concentration.

(Calculation procedure 2 of carbon residue)
The weight of carbon in the exhaust gas is the sum of the weight of carbon contained in the CO in the exhaust gas and the weight of carbon contained in the CO ₂ in the exhaust gas. The weight of carbon contained in the CO in the exhaust gas can be determined by (CO flow rate / gas volume of 1 mol (22.4 L) × carbon atomic weight (12)). The weight of carbon contained in CO ₂ in the exhaust gas can be determined by (CO ₂ flow rate / 22.4 × 12). The residual amount estimating unit 114 obtains the carbon weight in the exhaust gas according to the above formula.

(Calculation procedure 3 of carbon residue)
The weight of carbon in the exhaust gas can be considered as the amount of oxidized carbon. The residual amount estimating unit 114 can use this to calculate the weight of carbon remaining in the furnace. By displaying the calculated residual carbon weight (and its change over time) on, for example, an operation screen used by the operator, the operator's work can be assisted.

<Embodiment 3: Weight of residual chromium>
For example, when refining stainless steel or the like, a chromium alloy (FeCr: ferrochrome) or the like may be used as a material. The change in the weight of chromium remaining in the furnace can be used as reference information when the operator determines whether or not the estimated value of the chromium oxidation amount used in the blowing calculation substantially matches the actual value. The residual amount estimating unit 114 calculates the residual chromium weight in the furnace according to the following procedure.

(Calculation procedure 1 of residual amount of chromium)
In the blowing of chromium-containing molten iron having a chromium concentration of 3% by mass or more, it can be assumed that all of the injected oxygen is used for oxidizing chromium except for the one used for oxidizing carbon. The amount of oxygen used for the oxidation can be determined by the following formula: (the amount of oxygen used for chromium oxidation) = (the amount of oxygen injected)-(the amount of oxygen used for the oxidation of CO and CO ₂ in the exhaust gas) Amount)-(Amount of oxygen entrained from air)

(Calculation procedure of chromium residual amount 1-1)
The amount of oxygen blown can be obtained according to the control value.

(Calculation procedure 1-2 of chromium residual amount)
The amount of oxygen used for the oxidation of CO and CO _{2 in} the exhaust gas can be determined by the same procedure as that for calculating the weight of carbon in the exhaust gas. That is, the CO flow rate and the CO ₂ flow rate in the exhaust gas are obtained, and can be calculated based on the oxygen atomic weight and the composition formula.

(Calculation procedure 1-3 of residual amount of chromium)
The residual amount estimation unit 114 acquires the exhaust gas flow rate, the CO concentration in the exhaust gas, the CO ₂ concentration in the exhaust gas, the O ₂ concentration in the exhaust gas, and the H ₂ concentration in the exhaust gas. When the other components in the exhaust gas is assumed to be the only N _2, by subtracting the respective concentration of 100% can be obtained N ₂ concentration in the exhaust gas. Since those from N ₂ N ₂ is mixed with air and oxygen engulfed in the exhaust gas, according to the composition ratio of the air from the N ₂ amount excluding the latter N ₂ amount to determine the amount of oxygen involving the air be able to.

(Calculation procedure 2 of residual amount of chromium)
The oxidation reaction of chromium is represented by the following equation: Cr + 0.75 × O ₂ = 0.5 × Cr ₂ O ₃ . Therefore, the residual amount estimating unit 114 can obtain the oxidized chromium amount (molar number) by multiplying the obtained oxygen amount (molar number) used for chromium oxidation by 0.75. The residual amount estimation unit 114 can use this to calculate the weight of chromium remaining in the furnace. By displaying the calculated residual chromium weight (and its temporal change) on, for example, an operation screen used by the operator, the operator's work can be assisted.

  FIG. 6 is an example of a screen presented by the blowing calculation device 100. After the point of time SL1, the transition of the residual carbon concentration and the residual chromium concentration can be similarly calculated and displayed, starting from the actually measured component concentration at the point of time SL1. By displaying the calculation result and the target value on the screen for each of the carbon weight% and the chromium weight%, the operator's work can be assisted. For example, the operator looks at the transition of the calculation results of the carbon residual amount and the chromium residual amount during the period of SL1 to SL2, resets the timing of SL2, and determines whether the estimated value of the chromium oxidation amount in the blowing calculation is appropriate. can do. The blowing calculation apparatus 100 may re-perform the second blowing calculation based on, for example, the reset SL2. Thus, even when the timing of SL2 is changed in the middle of the process, the blowing calculation can be performed with high accuracy.

<Embodiment 4>
In the above embodiment, it has been described that the unknown amount estimating unit 112 estimates the unknown heat and the unknown oxygen in the first period and the second period, respectively, at the start of the blowing process using the learning result data 121. In a fourth embodiment of the present invention, a specific example will be described.

  Each time the blowing process is performed, the blowing calculation device 100 records the following actual values as the learning result data 121: (a) the scrap amount input before the start of the blowing process, and (b) the blowing process. (C) Amount of molten iron charged before the start of the blowing process, (d) Amount of scrap charged during the blowing process, (e) Amount charged during the blowing process (F) measured component concentration and measured molten steel temperature at time SL1, (g) measured component concentration and measured temperature of molten steel at time SL2, (h) unknown heat and unknown oxygen during the first period, (i) Unknown heat and unknown oxygen in two periods, (j) target component concentration at SL1, (k) target molten steel temperature at SL1, (l) target component concentration at SL2, (m) target molten steel temperature at SL2. When the blowing of the chromium-containing molten iron is targeted, as described above, in addition to these, the actual values of the chromium oxidation amount in the first period and the chromium oxidation amount in the second period are calculated, and the learning result data 121 is obtained. And it is desirable to use them for these estimations by machine learning.

  The unknown quantity estimating unit 112 acquires a feature quantity vector having the above (a) to (m) as elements at the time of blowing calculation, and obtains the above (a) to (f) in the past blowing step described by the learning result data 121. A vector-to-vector distance is calculated between the vector and (m). The unknown quantity estimating unit 112 assumes that, of the past blowing processes, the one with the obtained inter-vector distance is smaller than the current blowing process, and determines the unknown heat and unknown oxygen in the current blowing process. Used as an estimate of heat and unknown oxygen.

  The unknown quantity estimation unit 112 may narrow down the search target when searching for a past blowing process close to the current blowing process. For example, (i) the difference between the measured component amount at the point in time SL1 and the target component amount at the point in time SL1, (ii) the difference between the measured component amount at the point in time SL2 and the target component amount at the point in time SL2, (iii) SL1 At least one of a difference between the measured molten steel temperature at the time and the target molten steel temperature at the time SL1 and (iv) a difference between the measured molten steel temperature at the time SL2 and the target molten steel temperature at the time SL2 is within a predetermined range. The past blowing process that has been settled may be searched.

  There may be a plurality of past blowing processes in which the distance between vectors is small to some extent. In this case, weighting is applied to each of (a) to (m), and the distance between the vectors is scored to rank the past blowing steps, and the past blowing step having the highest rank is selected. You may. Further, for example, the unknown quantity estimating unit 112 presents the past blowing steps arranged in the rank order together with the main blowing parameters to the operator, and the operator selects a plurality of appropriate charges from among them and averages them. And so on. For example, it is conceivable to place importance on (j) to (m) over other parameters and rank them in ascending order of their difference. Other parameters may give smaller weights. Weighting of these parameters may be determined in advance by machine learning or the like so that the estimated values of unknown heat and unknown oxygen are close to the actual values from past learning result data.

The unknown quantity estimation unit 112 may use a machine learning device to search for unknown heat and unknown oxygen in a past blowing process closest to the blowing process to be performed, instead of the above-described method using the inter-vector distance. In this case, for example, in the first blowing calculation, (a) to (e) and (j) to (m) are input to the machine learning device, and the machine learning device inputs the closest past blowing data. The machine learning device may be configured in advance to search for unknown heat and unknown oxygen in the process.
When searching for unknown heat and unknown oxygen in the second period from SL1 to SL2 using the machine learning device, the blowing parameters (a) to (m) are input to the machine learning device and machine learning is performed. The machine may be configured in advance with a machine learning device so as to search for the unknown heat and oxygen in the closest blowing process in the past.

  Other parameters that are useful to learn using machine learning include: (N) Number of furnace uses: Each time the furnace is used, the refractory on the inner wall of the furnace wears out, so the unknown heat and the unknown oxygen gradually change. Therefore, it is useful to learn the correspondence between the number of furnace uses and unknown heat / oxygen. (O) Elapsed time from the previous time when molten steel was discharged from the furnace to the next charging of hot metal or molten steel: Since the refractory temperature of the inner wall of the furnace varies depending on this elapsed time, unknown heat and unknown oxygen Affect. Therefore, it is useful to learn the correspondence between this elapsed time and unknown heat and unknown oxygen. (P) Elapsed time from the previous time when molten steel was discharged from the furnace until the next start of blowing: For the same reason as in (o).

<Regarding Modification of the Present Invention>
The present invention is not limited to the embodiments described above, and includes various modifications. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described above. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of one embodiment can be added to the configuration of another embodiment. In addition, a part of the configuration of each embodiment can be added, deleted, or replaced with another configuration.

  In the above embodiment, the blowing process using the converter has been described, but the blowing calculation method according to the present invention can be used in other steelmaking processes (for example, those using AOD). In this case, the hot metal components in the above description may be replaced with those used in the steelmaking process.

100: blowing machine 110: CPU
111: Input amount calculation unit 112: Unknown amount estimation unit 113: Yield estimation unit 114: Residual amount estimation unit 120: Storage device 121: Learning result data

Claims (10)

A blowing calculation method for calculating the amount of oxygen required in the blowing process, the amount of heat-up material, and the amount of coolant,
A first calculation step of calculating the amount of oxygen, the amount of heat-raising material, and the amount of coolant required in a first period from a start time of the blowing process to a first time,
Obtaining a first measured component amount obtained by actually measuring an element component amount contained in the molten steel at the first time point;
Obtaining a first actually measured molten steel temperature obtained by actually measuring the molten steel temperature at the first time point;
Obtaining an actually measured unknown heat obtained by calculating the unknown heat at the first time point based on an actual measurement result;
A step of acquiring the measured unknown oxygen obtained by calculating the unknown oxygen at the first time point based on the measured result;
A second calculation step of calculating the amount of oxygen, the amount of heat-raising material, and the amount of coolant required in a second period from the first time to a second time of the blowing process;
Has,
In the first calculation step, the amount of scrap charged before the start of the blowing process, the amount of ferro-alloy input before the start of the blowing process, the amount of molten iron charged before the start of the blowing process, Using the first target component amount at the first time point, the second target component amount at the second time point, the first target molten steel temperature at the first time point, and the second target molten steel temperature at the second time point, Calculate the amount of oxygen required in the period, the amount of heat-up material, and the amount of coolant,
In the second calculation step, the amount of scrap charged before the start of the blowing process, the amount of ferro-alloy input before the start of the blowing process, the amount of molten iron charged before the start of the blowing process, The amount of ferromagnetic iron charged during the blowing process, the first measured component amount, the first measured molten steel temperature, the measured unknown heat, the measured unknown oxygen, the second target component amount, and the second target molten steel temperature And calculating the amount of oxygen, the amount of heat-raising material, and the amount of coolant required in the second period. The blowing calculation method further includes a prediction step of predicting unknown heat and unknown oxygen in the first period and the second period, respectively.
In the prediction step, the first learning is performed by using a result of learning by machine learning a relationship among blowing parameters in a past blowing process, unknown heat in a past blowing process, and unknown oxygen in a past blowing process. The blowing calculation method according to claim 1, wherein unknown heat and unknown oxygen are predicted in a period and the second period, respectively. A blowing calculation method for calculating the amount of oxygen required in the blowing process, the amount of heat-up material, and the amount of coolant,
A first calculation step of calculating the amount of oxygen, the amount of heat-raising material, and the amount of coolant required in a first period from a start time of the blowing process to a first time,
Obtaining an actually measured component amount obtained by actually measuring an element component amount contained in the molten steel at the first time point;
Obtaining an actual measured molten steel temperature obtained by actually measuring the molten steel temperature at the first time point;
Obtaining an actually measured unknown heat obtained by calculating the unknown heat at the first time point based on an actual measurement result;
A step of acquiring the measured unknown oxygen obtained by calculating the unknown oxygen at the first time point based on the measured result;
A second calculation step of calculating the amount of oxygen, the amount of heat-raising material, and the amount of coolant required in a second period from the first time to a second time of the blowing process;
A prediction step of predicting unknown heat and unknown oxygen in the first period and the second period,
Has,
In the prediction step, the first learning is performed by using a result of learning by machine learning a relationship among blowing parameters in a past blowing process, unknown heat in a past blowing process, and unknown oxygen in a past blowing process. A blowing calculation method characterized by predicting unknown heat and unknown oxygen in a period and the second period, respectively. In the first calculation step, a first unknown oxygen in the first period and a second unknown oxygen in the second period are calculated.
The blowing calculation method further includes a heating material yield calculation step of calculating a ratio of the heating material amount actually input to the furnace,
In the heating material yield calculation step, while decreasing the heating material amount by a predetermined amount from an initial value, until the first unknown oxygen and the second unknown oxygen both fall within a predetermined positive value range, The blowing calculation method according to claim 1 or 3, wherein the ratio is calculated by repeating a first calculation step and the second calculation step. The blowing calculation method further includes:
Obtaining the exhaust gas flow rate in the blowing process, the CO concentration in the exhaust gas, and the CO ₂ concentration in the exhaust gas,
Calculating the CO flow rate in the exhaust gas and the CO ₂ flow rate in the exhaust gas using the exhaust gas flow rate, the CO concentration, and the CO ₂ concentration;
Calculating a carbon weight remaining after the blowing step using the CO flow rate and the CO ₂ flow rate;
Outputting the carbon weight;
The blowing calculation method according to claim 1 or 3, wherein: The blowing calculation method further includes a step of resetting the second time point after the step of outputting the carbon weight,
The method according to claim 5, further comprising: performing the second calculation step after resetting the second time point. The blowing calculation method further includes:
Calculating the intake oxygen amount introduced for the blowing step,
Calculating the amount of exhaust gas oxygen contained in CO and CO ₂ in the exhaust gas in the blowing process,
Calculating the amount of entrained oxygen contained in the air entrained in the blowing step,
Calculating the amount of Cr oxygen used to oxidize chromium in the blowing process, using the amount of oxygen absorbed, the amount of exhaust gas oxygen, and the amount of entrained oxygen;
Using the Cr oxygen amount to calculate the weight of chromium remaining after the blowing step;
Outputting the chrome weight;
The blowing calculation method according to claim 1 or 3, wherein: The blowing calculation method further includes:
Obtaining a second actually measured component amount obtained by actually measuring an element component amount contained in the molten steel at the second time point;
A similar blowing step searching step of searching for a similar blowing step similar to the above blowing step among the completed blowing steps performed in the past,
Outputting unknown heat and unknown oxygen in the similar blowing process,
Has,
In the similar blowing process searching step, a difference between the first actually measured component amount and the first target component amount, a difference between the first actually measured molten steel temperature and the first target molten steel temperature, (2) at least one of a difference between the actually measured component amount and the second target component amount and a difference between the second actually measured steel temperature and the second target molten steel temperature falls within a predetermined range. The blowing process is the search target,
In the similar blowing step searching step, the scrap amount charged before the start of the blowing step, the amount of alloy iron charged before the start of the blowing step, and the amount of molten iron charged before the start of the blowing step The amount of ferro-alloy charged during the blowing step, the measured unknown heat, at least one of the measured unknown oxygen, and the corresponding blowing parameters in the completed blowing step are compared with each other. 4. The blowing calculation method according to claim 1, wherein the completed blowing step in which a difference falls within a predetermined range is searched for as the similar blowing step. 5. In the similar blowing process search step, the first target component amount, the first target molten steel temperature, the second target component amount, and the second target molten steel temperature, and the corresponding blowing in the completed blowing process. 9. The blowing calculation method according to claim 8, wherein the blowing parameters are compared with the smelting parameters, and are output as search results in ascending order of the difference between the two.   A blowing calculation program for causing a computer to execute the blowing calculation method according to any one of claims 1 to 9.