NOX concentration prediction method, sintered ore manufacturing method, and control device

Abstract

Provided is a NOx concentration prediction method capable of predicting NOx concentration of sintering exhaust gas in consideration of raw material segregation within a raw material charging layer.　This NOx concentration prediction method is for predicting the NOx concentration of a sintering exhaust gas and comprises: a raw material information acquisition step; a granulation condition acquisition step; a granulation result estimation step; a charging condition acquisition step; a charging result estimation step; a firing result estimation step; and a NOx concentration prediction step in which the sintering temperature in a raw material charging layer and the air volume of exhaust gas directly below the raw material charging layer are estimated using the previous steps, and the NOx concentration of the sintering exhaust gas is predicted using the estimated values.

Description

NOx concentration prediction method, sintered ore manufacturing method and control device

[0001] The present invention relates to a NOx concentration prediction method for predicting the NOx concentration of sintering exhaust gas discharged from a sintering machine, a control device, and a method for producing sintered ore using the NOx concentration prediction method.

[0002] Sinter, one of the raw materials used in the blast furnace pig iron production process, is made up of iron ore as the main raw material, and calcium oxide (CaO) such as limestone and silica stone. ２ It is produced by adding a few percent of water to auxiliary materials containing the above, return fines, and solid fuel, mixing them, granulating them, and sintering the resulting pseudo-particles using the heat of combustion of the solid fuel. The surface of the raw material packed bed filled with granulated pseudo-particles is ignited, and air is drawn in from below, burning the solid fuel in the packed bed from top to bottom, sintering the raw material.

[0003] At this time, the sintering exhaust gas sucked from below the sinter machine contains combustion products from the filling tank, and therefore contains nitrogen oxides (NOx). Most of the NOx in sintering exhaust gas is fuel NOx, produced by the oxidation of nitrogen contained in the fuel. NOx emissions are regulated to prevent air pollution. In the production of sintered ore, it is necessary to reduce and manage the amount of NOx emissions contained in sintering exhaust gas. The NOx content of sintering exhaust gas varies depending on various factors in sintering operations, such as the type of ore used, the amount of fuel used, and the suction conditions. For this reason, the NOx concentration in sintering exhaust gas is monitored and managed to stay within a control value (target value) lower than the regional NOx emission regulation value.

[0004] As described above, it is necessary to keep the emission amount of NOx, which is a type of environmental pollutant, within a controlled value. For this reason, efforts have been made to reduce the NOx in the sintering exhaust gas from the sintering process.

[0005] Patent Document 1 discloses a modified carbonaceous material for sintered ore production that is coated with a coating containing 36 mass % or more of Ca derived from a lime-based raw material. According to Patent Document 1, modified carbonaceous material of 1 mm or more has a coating density of 4 to 42 g / m ２ , 5 to 20 g / m for modified carbonaceous materials of 0.25 to less than 1 mm ２It is said that by incorporating Ca into the coating and introducing a coating layer that melts at high temperatures onto the surface of the carbonaceous material, it is possible to suppress combustion of the carbonaceous material at low temperatures and reduce NOx emissions.

[0006] Patent Document 2 discloses a method for suppressing the generation of NOx in the sintering exhaust gas by setting the particle size of the carbonaceous material used during sintering to a predetermined value or more. According to Patent Document 2, by using a coarse-grained carbonaceous material with an average particle size of 2 mm or more, or an average particle size of 1 mm or more as fuel without high-temperature heat treatment, it is possible to suppress the amount of NOx emissions.

[0007] An operating method has also been proposed for predicting an increase in NOx emissions when operating conditions fluctuate in the production of sintered ore and preventing the increase in NOx emissions. Patent Document 3 discloses a technology for increasing the blending ratio of low-nitrogen anthracite in accordance with the increase in the crystallization water content when changing the composition ratio of multiple brands of iron ore and if the changed iron ore has an increased crystallization water content. Patent Document 3 states that by increasing the blending ratio of low-nitrogen anthracite, it is possible to prevent an increase in NOx emissions due to an increase in the amount of crystallization water.

[0008] Attempts have also been made to suppress the increase in NOx emissions by using mathematical models. Patent Document 4 describes a combustion model in a sintered bed that takes into account the combustion reaction of coke breeze particles and the material and heat balances, and calculates the CO concentration and O concentration in the gas film of coke breeze particles during the combustion process. ２ According to Patent Document 4, a method for evaluating the amount of NOx generated from the ratio of the temperature of the sintered layer, the CO concentration of the combustion exhaust gas, and the O concentration is used to calculate the amount of NOx generated. ２ It is said that the amount of NOx generated can be evaluated by calculating the concentration. Patent Document 4 also describes that by evaluating the amount of fuel NOx generated using a mathematical model based on the principle of combustion, it is possible to consider various factors that can suppress the amount of NOx generated.

[0009] Japanese Patent No. 5621653 Japanese Patent Laid-Open No. 53-37103 Japanese Patent Laid-Open No. 2022-129696 Japanese Patent No. 5447192

[0010] Conventionally, many techniques have been developed to reduce the NOx content in sintering exhaust gas and suppress the amount of NOx emissions, as disclosed in Patent Document 1 and Patent Document 2. As shown in Patent Document 3, the effects of several operational factors related to the generation of NOx have been elucidated, and a method for preventing an increase in the amount of NOx generated has also been developed.

[0011] However, the amount of NOx generated is not determined only by the combustion temperature and particle size of the carbonaceous material or the amount of water of iron ore crystallization, as shown in Patent Documents 1 to 3, but is affected by various factors such as the amount of carbonaceous material used as fuel, the amount of water in granulation, the permeability of the raw material layer, and the thickness of the raw material layer. For this reason, these technologies predict future increases and decreases in NOx in advance, and carry out operational procedures to suppress the increase before the amount of NOx emissions increases, and actually perform NOx control. Ｘ There was a limit to what could be done before emissions increased.

[0012] Patent Document 4 discloses a model that describes the amount of NOx generated using a mathematical model based on combustion reactions. However, the model disclosed in Patent Document 4 does not take into account the segregation of raw materials in the height direction of the raw material charging layer. For this reason, there are limitations to the prediction of the amount of NOx generated. Although the model disclosed in Patent Document 4 can estimate the amount of NOx generated in a ladle test, it is difficult to estimate the amount of NOx emitted from an actual sintering machine.

[0013] In a practical sintering machine, segregation of various raw materials, including fuel, occurs in the raw material charging layer depending on the type and properties of the raw materials charged in the raw material charging layer and the granulation conditions. The segregation state in the raw material charging layer in such a practical sintering machine has a significant impact on the temperature distribution in the sintering layer, which is related to the NOx content in the sintering exhaust gas. Therefore, in a practical sintering machine, unless the raw material segregation in the raw material charging layer is taken into consideration, it is difficult to predict the NOx concentration in the sintering exhaust gas with high accuracy.

[0014] The present invention has been made in view of the above problems of the prior art, and its object is to provide a method and a control device for predicting the NOx concentration of sintering exhaust gas taking into account the segregation of raw materials in the raw material charging layer. Another object of the present invention is to provide a method for producing sintered ore using the NOx concentration prediction method.

[0015] The means for solving the above problems are as follows: [1] A method for predicting the NOx concentration of sintering exhaust gas discharged from a sintering machine that sinters granulated particles obtained by adding water to a sintering raw material containing an iron-containing raw material and a carbon-containing raw material, the method comprising: a raw material information acquisition step of acquiring raw material information including the particle size, component composition, and blending ratio of each raw material contained in the sintering raw material; a granulation condition acquisition step of acquiring granulation conditions including the moisture content of the sintering raw material when granulating the granulated particles from the sintering raw material; and a step of inputting the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content. and a component estimation model that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and outputs the content of a specific component of the granulated particles in one of the particle size classifications obtained by classifying the particle sizes of the granulated particles into a plurality of particle size classifications, thereby making it possible to estimate the particle sizes of the granulated particles and the composition of the granulated particles for each particle size by inputting the raw material information and the granulation conditions into the granulation estimation model. a granulation result estimation step of estimating a granulation result including a chemical composition; a charging condition acquisition step of acquiring charging conditions when charging the granulated particles into the sinter machine to form a raw material charging layer; a charging result estimation step of estimating particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer based on the granulation result and the charging conditions using a charging estimation model that indicates a content ratio of the granulated particles of each particle size in each charging section obtained by dividing the height direction of the raw material charging layer into a plurality of charging sections; a sintering result estimation step of estimating the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer using a sintering estimation model including a heat transfer model that can calculate the sintering temperature at a predetermined position in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on at least one of the above; and a NOx concentration prediction step of predicting the NOx concentration of the sintering exhaust gas using the sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer and the amount of carbon-containing raw material used per unit time included in the sintering raw material.Ｘ Concentration prediction method. [2] The NOx concentration prediction method according to [1], wherein the NOx concentration prediction step predicts the NOx concentration of the sintering exhaust gas using the following formula (1): NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas, T is the highest temperature (K) among the sintering temperatures in the raw material charging layer, M is the amount (ton / h) of the carbon-containing raw material contained in the sintering raw material used per unit time, N is the nitrogen content (mass%) of the carbon-containing raw material contained in the sintering raw material, U is the exhaust gas flow rate (mol / h) immediately below the raw material charging layer, and A (K －１), and B(-) is a constant. [3] A method for producing sintered ore, comprising: a production condition specifying step for specifying production conditions under which the NOx concentration of the sintered ore exhaust gas predicted by the NOx concentration prediction method according to [1] or [2] satisfies a predetermined target value; and a sintered ore production step for producing sintered ore under the production conditions specified in the production condition specifying step.[4] A control device for predicting the NOx concentration of sintering exhaust gas discharged from a sintering machine that sinters granulated particles obtained by adding water to sintering raw materials containing an iron-containing raw material and a carbon-containing raw material, the control device comprising: a raw material information acquisition unit that acquires raw material information including the particle size, component composition, and blending ratio of each raw material contained in the sintering raw material; a granulation condition acquisition unit that acquires granulation conditions including the moisture content of the sintering raw material when granulating the granulated particles from the sintering raw material; and a control device that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and acquires the NOx concentration of the sintering exhaust gas discharged from the sintering machine. By inputting the raw material information and the granulation conditions into a granulation estimation model including: a particle size estimation model that outputs the content of the granulated particles in one particle size division among a plurality of particle size divisions obtained by dividing the particle sizes of the granulated particles; and a component estimation model that inputs the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sinter raw material, and the moisture content, and outputs the content of a specific component of the granulated particles in one particle size division among a plurality of particle size divisions obtained by dividing the particle sizes of the granulated particles, the particle size of the granulated particles and the composition of the granulated particles for each particle size can be estimated. a charging condition acquisition unit that acquires charging conditions when the granulated particles are charged into the sinter machine to form a raw material charging layer; a charging result estimation unit that estimates particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer based on the granulation result and the charging conditions using a charging estimation model that indicates the content ratio of the granulated particles of each particle size in each charging section obtained by dividing the height direction of the raw material charging layer into a plurality of charging sections; a sintering result estimation unit that estimates the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer using a sintering estimation model including a heat transfer model that can calculate the sintering temperature at a predetermined position in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on one of the above, and a NOx concentration prediction unit that predicts the NOx concentration of the sintering exhaust gas using the sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer and the amount of carbon-containing raw material contained in the sintering raw material used per unit time. [5] The control device according to [4], wherein the NOx concentration prediction unit predicts the NOx concentration of the sintering exhaust gas using the following formula (1):NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the above formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas, T is the highest temperature (K) among the sintering temperatures in the raw material charging layer, M is the amount of carbon-containing raw material contained in the sintering raw material used per unit time (ton / h), N is the nitrogen content ratio (mass%) of the carbon-containing raw material contained in the sintering raw material, U is the exhaust gas flow rate (mol / h) directly below the raw material charging layer, and A (K. －１ ), B(-) is a constant. [6] The control device according to [4] or [5], further comprising a production condition specifying unit that specifies production conditions under which the NOx concentration of the sintering exhaust gas predicted by the NOx concentration prediction unit satisfies a predetermined target value.

[0016] In the NOx concentration prediction method and control device according to the present invention, the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer, which are used to predict the NOx concentration of the sintering exhaust gas, are estimated using the raw material segregation in the raw material charging layer. Therefore, the NOx concentration prediction method and control device according to the present invention are capable of predicting the NOx concentration of the sintering exhaust gas taking into account the raw material segregation in the raw material charging layer.

[0017] FIG. 1 is a schematic diagram showing a configuration example of a sintering facility including a control device according to this embodiment. FIG. 2 is a schematic diagram showing a configuration example of a control device. FIG. 3 is a schematic diagram showing a NO according to this embodiment. Ｘ Fig. 4 is a flow diagram showing an example of a concentration prediction method and a sinter ore manufacturing method using the prediction method. Fig. 4 is a graph showing an example of a charging result when granulated particles having a predetermined particle size distribution are charged into a pallet under predetermined charging conditions. Fig. 5 is a graph showing an example of an estimation result of the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer by the sintering result estimation unit. Fig. 6 is a graph showing the time progression of the actual increase / decrease and the predicted increase / decrease of the NOx concentration of the sintering exhaust gas. Fig. 7 is a graph showing the correlation between the actual increase / decrease and the predicted increase / decrease of the NOx concentration shown in Fig. 6.

[0018] The present invention will be described below through embodiments of the present invention. The following embodiments are preferred examples of the present invention, and the present invention is not limited to these embodiments.

[0019] 1 is a schematic diagram showing an example of the configuration of a sintering facility 10 including a control device 12 according to this embodiment. The control device 12 is a device capable of implementing the NOx concentration prediction method according to this embodiment.

[0020] The sintering equipment 10 is equipment capable of producing sintered ore from sinter raw materials containing an iron-containing raw material and a carbon-containing raw material. The sintering equipment 10 includes a control device 12, a granulator 14, a sintering machine 16, a crusher 18, a cooler 20, and a sieving device 22. The granulator 14 granulates granulated particles from the sinter raw materials containing the iron-containing raw material and the carbon-containing raw material. When the granulator 14 granulates the granulated particles, granulation water is added to the sinter raw materials. The sinter raw materials may further include a calcium oxide (CaO)-containing raw material as an auxiliary raw material.

[0021] The granulated particles produced by the granulator 14 are transported to a sintering machine 16. The granulator 14 may be any granulator capable of producing granulated particles. The granulator 14 may be, for example, a drum mixer.

[0022] The sintering machine 16 may be any sintering machine that sinters granulated particles. The sintering machine 16 is, for example, a Dwight Lloyd type sintering machine. The sintering machine 16 has a sintering raw material supply device 24, a pallet 26, an ignition furnace 28, and a wind box 29. The sintering raw material supply device 24 charges the granulated particles supplied from the granulator 14 into the pallet 26.

[0023] The pallet 26 is an endless moving pallet. Granulated particles are charged from the sintering raw material supply device 24, forming a raw material charging layer on the pallet 26. The ignition furnace 28 ignites the carbon-containing raw material contained in the surface layer of the raw material charging layer formed on the pallet 26.

[0024] The wind box 29 sucks air downward from the raw material charging layer formed on the pallet 26. When the wind box 29 sucks air downward from the raw material charging layer, the combustion and molten materials in the raw material charging layer move downward from the raw material charging layer. In this way, the combustion and molten materials move downward from the raw material charging layer, causing the raw material charging layer to be sintered. By sintering the raw material charging layer, a sintered cake is obtained from the raw material charging layer.

[0025] The crusher 18 crushes the sintered cake supplied from the sintering machine 16. The crusher 18 supplies the crushed sintered cake to a cooler 20. The cooler 20 cools the crushed sintered cake supplied from the crusher 18. The crushed sintered cake cooled by the cooler 20 is supplied to a sieving device 22.

[0026] The sieving device 22 sieves the crushed sinter cake cooled by the cooler 20 according to the particle size of the crushed material. The sieving device 22 sieves the crushed sinter cake, for example, into sinter ore having a particle size of 5 mm or more and return ore having a particle size of less than 5 mm. In this way, sieving by the sieving device 22 is finally performed to produce sintered ore. The return ore sieved by the sieving device 22 may be blended with the sinter raw material and reused as a raw material for sintered ore.

[0027] The control device 12 acquires raw material information, which is information about the raw materials contained in the sintering raw material supplied to the granulator 14. The control device 12 acquires granulation conditions, which include information about the moisture content of the sintering raw material when the granulator 14 granulates granulated particles from the sintering raw material.

[0028] The control device 12 estimates the granulation result of the granulated particles granulated by the granulator 14 based on the acquired raw material information and granulation conditions. The granulation result estimated by the control device 12 includes the particle size of the granulated particles and the component composition of the granulated particles for each particle size. The control device 12 acquires the charging conditions when the sintering raw material supply device 24 charges the granulated particles onto the pallet 26 to form a raw material charging layer.

[0029] The control device 12 estimates the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer formed on the pallet 26 based on the estimated granulation result and the acquired charging conditions. The control device 12 estimates the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on the estimated particle size segregation and component segregation. In this case, the control device 12 may estimate the sintering temperature in the raw material charging layer and the exhaust gas flow rate based on either the particle size segregation or the component segregation, rather than both.

[0030] The control device 12 uses the estimated sintering temperature and exhaust gas flow rate in the raw material charging layer to predict the NOx concentration of the sintering exhaust gas discharged from the sintering machine 16. As a result, the control device 12 can predict the NOx concentration of the sintering exhaust gas by taking into account the segregation of raw materials in the raw material charging layer. Ｘ Furthermore, the control device 12 identifies the sinter ore production conditions under which the predicted NOx concentration in the sintering exhaust gas satisfies a predetermined target value, and sets the production conditions in each device of the sintering equipment 10. This makes it possible to produce sinter ore while satisfying the target value of the NOx concentration in the sintering exhaust gas.

[0031] Next, the control device 12 will be described. Fig. 2 is a schematic diagram showing an example of the configuration of the control device 12. The control device 12 may be a general-purpose computer such as a workstation or a personal computer, or may be a dedicated computer configured to function as the control device 12 of the sintering equipment 10. The control device 12 has a control unit 30, an input unit 32, an output unit 34, a memory unit 36, and a communication unit 38.

[0032] The control unit 30 reads programs, data, etc. stored in the memory unit 36 ​​and executes various functions. The control unit 30 controls the granulator 14, the sintering machine 16, the crusher 18, the cooler 20, and the sieving device 22. The control unit 30 executes the programs read from the memory unit 36, causing the control unit 30 to function as a raw material information acquisition unit 40, a granulation condition acquisition unit 42, a granulation result estimation unit 44, a charging condition acquisition unit 46, a charging result estimation unit 48, a firing result estimation unit 50, a NOx concentration prediction unit 52, and a production condition identification unit 54.

[0033] The input unit 32 includes one or more input interfaces that detect user input and acquire input information based on the user's operation. The input unit 32 includes, for example, physical keys, capacitive keys, a display, etc. The output unit 34 includes one or more output interfaces that output information to notify the user. The output unit 34 includes, for example, a display that outputs information as an image, a speaker that outputs information as audio, etc.

[0034] The storage unit 36 ​​is, for example, a flash memory, a hard disk, an optical memory, etc. A part of the storage unit 36 ​​may be located outside the control device 12. In this case, a part of the storage unit 36 ​​may be a hard disk, a memory card, etc. connected to the control device 12 via an arbitrary interface. The storage unit 36 ​​stores programs for the control unit 30 to execute each function, data used by the programs, etc.

[0035] The communication unit 38 includes at least one of a communication module compatible with wired communication and a communication module compatible with wireless communication. The control device 12 can communicate with other terminal devices and the like via the communication unit 38.

[0036] Next, the processes executed by the raw material information acquisition unit 40, the granulation condition acquisition unit 42, the granulation result estimation unit 44, the charging condition acquisition unit 46, the charging result estimation unit 48, the firing result estimation unit 50, and the NOx concentration prediction unit 52 will be described.

[0037] The raw material information acquisition unit 40 acquires raw material information, which is information about raw materials contained in the sintering raw material supplied to the granulator 14. The raw material information includes information about the particle size, component composition, and blending ratio of each raw material contained in the sintering raw material. For example, if the sintering raw material includes an iron-containing raw material, a CaO-containing raw material, and a carbon-containing raw material, the raw material information includes information about the particle size, component composition, and blending ratio of each of the iron-containing raw material, the CaO-containing raw material, and the carbon-containing raw material.

[0038] The iron-containing raw material is, for example, iron ore. The CaO-containing raw material is, for example, limestone. The carbon-containing raw material is, for example, coke powder. The particle size information of each raw material includes information on the content ratio for each predetermined particle size range. The information on the component composition of each raw material includes, for example, carbon (C) concentration, moisture concentration, calcium oxide (CaO) concentration, and aluminum oxide (Al ２ O ３ ) concentration.

[0039] The raw material information acquisition unit 40 may acquire the raw material information by an operator inputting the raw material information into the input unit 32. The operator can acquire the raw material information in advance by sieving, chemically analyzing, or the like, each raw material.

[0040] The raw material information acquisition unit 40 may acquire the raw material information by receiving the raw material information input by an operator to another terminal device via the communication unit 38. The raw material information acquisition unit 40 outputs the acquired raw material information to the granulation condition acquisition unit 42.

[0041] The granulation condition acquisition unit 42 acquires granulation conditions including information on the moisture content of the sintering raw material when the granulator 14 granulates the sintering raw material into granulated particles. The granulation condition acquisition unit 42 may acquire the granulation conditions from the granulator 14, or may acquire the granulation conditions through an input operation by an operator to the input unit 32.

[0042] The moisture content of the sintering raw material when the granulator 14 granulates the sintering raw material into granulated particles is determined from the moisture content of each raw material contained in the sintering raw material and the amount of granulation water added to the granulator 14. The moisture content of each raw material can be measured using, for example, an infrared moisture meter.

[0043] The granulation conditions acquired by the granulation condition acquisition unit 42 may include information on the moisture content of the sintering raw material when the granulator 14 granulates the sintering raw material into granulated particles, but may also include information on other conditions set when the granulator 14 granulates the sintering raw material into granulated particles. For example, the granulation conditions may also include information on the space factor, rotation speed, and residence time of the granulator 14. The granulation condition acquisition unit 42 outputs the granulation conditions and the raw material information acquired from the raw material information acquisition unit 40 to the granulation result estimation unit 44.

[0044] The granulation result estimation unit 44 estimates the granulation results of the granulated particles granulated by the granulator 14 based on the raw material information and granulation conditions acquired from the granulation condition acquisition unit 42. The granulation results estimated by the control device 12 include the particle sizes of the granulated particles and the component compositions of the granulated particles for each particle size.

[0045] When estimating the granulation result, the granulation result estimation unit 44 estimates the granulation result using a granulation estimation model including a particle size estimation model and a component estimation model. The particle size estimation model and the component estimation model may be stored in the storage unit 36. The particle size estimation model is a trained machine learning model that, when the blending ratio of each raw material contained in the sintering raw material and the moisture content of the sintering raw material at the time of granulation are input, outputs the content of granulated particles in one of a plurality of particle size classifications obtained by classifying the particle sizes of the granulated particles.

[0046] The granulation result estimation unit 44 estimates the content of granulated particles in one particle size range by inputting the blending ratio of each raw material and the granulation conditions, including the moisture content of the sintered raw material during granulation, into the particle size estimation model read from the storage unit 36. The content of granulated particles in one particle size range is an example of the particle size of the granulated particles. The particle size of the granulated particles may be expressed by other indicators.

[0047] The particle size classification may be, for example, three levels. When there are three levels of particle size classification, the classification may be, for example, over 8 mm, 2.8 mm or more and 8.0 mm or less, and less than 2.8 mm. The particle size classification is not limited to three levels, and may be a particle size classification divided into more levels. The granulation result estimation unit 44 repeats the same process for the other particle size classifications and estimates the content of granulated particles in each particle size classification. The granulation result estimation unit 44 can determine the particle size distribution of the granulated particles by estimating the content of granulated particles in all particle size classifications.

[0048] The component estimation model is a trained machine learning model that, when inputted with the blending ratio of each raw material contained in the sintering raw material and the moisture content of the sintering raw material during granulation, outputs the content of a specific component of the granulated particles in one of multiple particle size categories into which the particle sizes of the granulated particles are divided.

[0049] The granulation result estimation unit 44 estimates the content of a specific component in granulated particles in one particle size division by inputting the blending ratio of each raw material and the granulation conditions, including the moisture content of the sintered raw material during granulation, into a component estimation model read from the storage unit 36. The granulation result estimation unit 44 repeats the same process for other particle size divisions to estimate the content of the specific component in granulated particles in each particle size division. The granulation result estimation unit 44 can determine the content of the specific component in granulated particles by estimating the content of the specific component in granulated particles in all particle size divisions. The granulation result estimation unit 44 outputs the estimated content of granulated particles in each particle size division and the content of the specific component in granulated particles in each particle size division to the charging result estimation unit 48.

[0050] The particle size estimation model and the component estimation model are generated in advance by machine learning using a large number of data sets, each set consisting of a pair of actual input values ​​and actual output values, and are stored in the storage unit 36. The inputs to the particle size estimation model and the component estimation model may include particle sizes of other raw materials. The inputs to the particle size estimation model and the component estimation model may include granulation conditions such as the rotation speed and residence time of the granulator 14.

[0051] In the above description, a case has been described in which machine learning models are used as the particle size estimation model and the component estimation model, but the particle size estimation model and the component estimation model may also be multiple regression models. In this case, the inputs of the machine learning model become explanatory variables of the multiple regression model, and the outputs of the machine learning model become target variables of the multiple regression model. In the multiple regression model, each parameter of the multiple regression model may be calculated in advance using a large number of data sets each consisting of a pair of actual input values ​​and actual output values, and may be stored in the storage unit 36.

[0052] If the particle size estimation model and the component estimation model are machine learning models, the particle size estimation model and the component estimation model may be updated by performing machine learning using a later acquired dataset.If the particle size estimation model and the component estimation model are multiple regression models, each parameter of the multiple regression model may be updated using a later acquired dataset.

[0053] The charging condition acquisition unit 46 acquires the charging conditions when the sintering raw material supply device 24 charges granulated particles onto the pallet 26 to form a raw material charging layer. The charging condition acquisition unit 46 may acquire the charging conditions from the sintering machine 16, or may acquire the charging conditions through an input operation by an operator to the input unit 32.

[0054] The charging conditions acquired by the charging condition acquisition unit 46 are, for example, information on the speed of the pallet 26, the opening degree of the sub-gate provided in the sintering raw material supply device 24, and the angle of the chute provided in the sintering raw material supply device 24. The charging condition acquisition unit 46 outputs the acquired charging conditions to the charging result estimation unit 48.

[0055] The memory unit 36 ​​stores a table showing the content ratio of granulated particles of each particle size for each charging section, which is obtained by dividing the height direction of the raw material charging bed into a plurality of sections, for each particle size distribution of the granulated particles and each charging condition. The charging result estimation unit 48 acquires the granulation results of the granulated particles from the granulation result estimation unit 44 and acquires the charging conditions from the charging condition acquisition unit 46, and then reads out the table corresponding to the particle size distribution and the charging conditions from the memory unit 36. Here, the granulation results of the granulated particles estimated by the granulation result estimation unit 44 mean the content of granulated particles in each particle size section of the granulated particles and the content of a specific component of the granulated particles in each particle size section of the granulated particles.

[0056] The content ratio of granulated particles of each particle size for each charging section, which divides the raw material charging bed into a plurality of sections in the height direction, can be calculated in advance for each charging section by performing a simulation simulating the sintering raw material supply device 24 of the sinter machine 16 using DEM (Discrete Element Method). For this reason, the above simulation is performed for each particle size distribution of granulated particles and charging conditions, and a table showing the content ratio of granulated particles of each particle size for each charging section is obtained. This makes it possible to estimate the content ratio of granulated particles of each particle size for each charging section under the particle size distribution of granulated particles and charging conditions. The table showing the content ratio of granulated particles of each particle size for each charging section may be obtained in advance and stored in the memory unit 36.

[0057] The charging result estimation unit 48 estimates the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer formed on the pallet 26 based on the granulation result acquired from the granulation result estimation unit 44 and the charging conditions acquired from the charging condition acquisition unit 46. The charging result estimation unit 48 estimates the particle size segregation of the granulated particles in the height direction of the raw material charging layer using a table showing the content ratio of granulated particles of each particle size for each charging section.

[0058] The component composition of the granulated particles by particle size is estimated by the granulation result estimation unit 44. Therefore, the charging result estimation unit 48 can also estimate the component segregation of the granulated particles in the height direction of the raw material charging layer using the particle size segregation of the granulated particles in the height direction of the raw material charging layer and the component composition by particle size for each charging section.

[0059] The table showing the content ratio of granulated particles of each particle size for each charging division is an example of a charging estimation model. As the charging estimation model, a trained machine learning model or a multiple regression model that outputs the content ratio of granulated particles of each particle size for each charging division when the content, component composition, and charging conditions of granulated particles for each particle size division are input may be used.

[0060] The charging result estimation unit 48 outputs the estimated particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer to the sintering result estimation unit 50. The sintering result estimation unit 50 estimates the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer acquired from the charging result estimation unit 48. In this case, the sintering result estimation unit 50 may estimate the sintering temperature in the raw material charging layer and the exhaust gas flow rate based on either the particle size segregation or the component segregation, rather than both.

[0061] The firing result estimation unit 50 estimates the sintering temperature and exhaust gas flow rate in the raw material charging layer, for example, using a heat transfer model disclosed in Reference 1 shown below. The exhaust gas flow rate is estimated as the exhaust gas flow rate (amount of gas molecules) immediately below the raw material charging layer. The exhaust gas flow rate can be obtained by calculating the pressure loss of the fluid in each control volume, as disclosed in Reference 1.

[0062] The information input to the heat transfer model may be the thickness of the raw material charging layer, the speed of the pallets 26, the temperature in the ignition furnace 28, the component concentrations in the raw material charging layer, the void fraction in the raw material charging layer, and the exhaust gas flow rate. The set values ​​of the sinter machine 16 may be used for the thickness of the raw material charging layer, the speed of the pallets 26, and the temperature in the ignition furnace 28. The component concentrations in the raw material charging layer and the void fraction in the raw material charging layer are values ​​calculated using the particle size segregation and component segregation estimated by the charging result estimation unit 48. The exhaust gas flow rate may be a value measured in the sinter machine 16. Since the temperature in the ignition furnace 28 is determined by the gas flow rate in the ignition furnace 28, the gas flow rate in the ignition furnace 28 may be input to the heat transfer model instead of the temperature in the ignition furnace 28.

[0063] Reference 1: Koichiro Ohno and four others, "Effect of Coke Combustion Rate Equation on Numerical Simulation of Estimating Temperature Distribution in Sintering Process Layer," Iron and Steel, Vol. 101, 2015, No. 1, pp. 19-24

[0064] The sintering result estimation unit 50 can calculate the sintering temperature at a predetermined position in the raw material charging layer and the exhaust gas temperature immediately below the raw material charging layer at set time intervals by using the above-mentioned heat transfer model. The time interval may be, for example, one second. The sintering result estimation unit 50 outputs the estimated sintering temperature and exhaust gas flow rate to the NOx concentration prediction unit 52. The NOx concentration prediction unit 52 predicts the NOx concentration of the sintering exhaust gas using the sintering temperature and exhaust gas flow rate immediately below the raw material charging layer estimated by the sintering result estimation unit 50.

[0065] The NOx concentration prediction unit 52 predicts the NOx concentration of the sintering exhaust gas using, for example, the following formula (1): The following formula (1) including constants A and B may be stored in advance in the storage unit 36 ​​by an operator's input operation to the input unit 32.

[0066] NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the above formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas. T is the highest sintering temperature (K) in the raw material charging layer. M is the amount of carbon-containing raw material contained in the sintering raw material used per unit time (ton / h). N is the nitrogen content (mass%) of the carbon-containing raw material contained in the sintering raw material. U is the exhaust gas flow rate (amount of gas molecules mol / h) directly below the raw material charging layer. A(K －１ ), B(-) is a constant.

[0067] The constant A is preferably a constant of 0.033 or more and 0.036 or less, and the constant B is preferably a constant of 52 or more and 54 or less. The nitrogen content of the carbon-containing raw material contained in the sintering raw material may be measured in advance and the measured value may be used as N. Generally, the nitrogen content of the carbon-containing raw material is in the range of 0.5 mass% or more and 2.5 mass% or less. Therefore, any value within this range may be used instead of the above measured value.

[0068] When predicting the NOx concentration using the above formula (1), the NOx concentration prediction unit 52 uses the estimated sintering temperature and exhaust gas flow rate in the raw material charging layer to identify the highest sintering temperature and the exhaust gas flow rate immediately below the raw material charging layer. The NOx concentration prediction unit 52 calculates the amount of carbon-containing raw material used per unit time using the content of carbon-containing raw material contained in the sintering raw material acquired by the raw material information acquisition unit 40. In this way, the highest sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer, and the amount of carbon-containing raw material used per unit time in the above formula (1) are determined, so the NOx concentration prediction unit 52 can predict the NOx concentration in the sintering exhaust gas using the above formula (1).

[0069] Instead of the above formula (1), the NOx concentration prediction unit 52 may predict the NOx concentration of the sintering exhaust gas using a trained machine learning model or a multiple regression model. In this case, the trained machine learning model or the multiple regression model is a model that receives the highest sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer, and the amount of carbon-containing raw material used per unit time as inputs, and outputs the NOx concentration of the sintering exhaust gas.

[0070] In this way, the control device 12 estimates the sintering temperature in the raw material charging layer and the exhaust gas flow rate directly below the raw material charging layer, which are used to predict the NOx concentration of the sintering exhaust gas, using the raw material segregation in the raw material charging layer. Therefore, the control device 12 and the NOx concentration prediction method implemented by this device according to this embodiment are control device 12 and NOx concentration prediction method that predict the NOx concentration of the sintering exhaust gas taking into account the raw material segregation in the raw material charging layer. And, by predicting the NOx concentration of the sintering exhaust gas taking into account the raw material segregation in the raw material charging layer, the NOx concentration of the sintering exhaust gas can be predicted with high accuracy.

[0071] Next, the processing executed by the production condition specification unit 54 will be described. The production condition specification unit 54 specifies production conditions under which the NOx concentration in the sintering exhaust gas predicted by the NOx concentration prediction unit 52 satisfies a predetermined target value for the NOx concentration in the sintering exhaust gas. The production condition specification unit 54 sets the specified production conditions as the production conditions for each device constituting the sintering equipment 10, and produces sintered ore under the specified production conditions. This makes it possible to produce sintered ore while the NOx concentration in the sintering exhaust gas satisfies the target concentration value.

[0072] 3 is a flow diagram showing an example of a NOx concentration prediction method according to the present embodiment and a sintered ore production method using the prediction method. The processing executed by the production condition specification unit 54 will be described in detail with reference to FIG. 3. The flow shown in FIG. 3 is started, for example, when the input unit 32 receives a start instruction from an operator.

[0073] The raw material information acquisition unit 40 acquires raw material information, which is information about the raw materials contained in the sintering raw material supplied to the granulator 14 (step S101). The processing of step S101 is a raw material information acquisition step. The raw material information acquisition unit 40 outputs the acquired raw material information to the granulation condition acquisition unit 42.

[0074] The granulation condition acquisition unit 42 acquires granulation conditions including information on the moisture content of the sintering raw material when the granulator 14 granulates granulated particles from the sintering raw material (step S102). The processing of step S102 is a granulation condition acquisition step. The granulation condition acquisition unit 42 outputs the raw material information and the granulation conditions to the granulation result estimation unit 44.

[0075] The granulation result estimation unit 44 estimates the granulation result of the granulated particles granulated by the granulator 14 based on the acquired raw material information and granulation conditions (step S103). The granulation result of the granulated particles estimated by the granulation result estimation unit 44 includes the particle size of the granulated particles and the component composition of the granulated particles for each particle size. The processing of step S103 is a granulation result estimation step. The granulation result estimation unit 44 outputs the estimated granulation result of the granulated particles to the charging result estimation unit 48.

[0076] The charging condition acquisition unit 46 acquires the charging conditions when the sintering raw material supply device 24 charges the granulated particles onto the pallet 26 to form a raw material charging layer (step S104). The processing of step S104 is a charging condition acquisition step. The charging condition acquisition unit 46 outputs the acquired charging conditions to the charging result estimation unit 48.

[0077] The charging result estimation unit 48 estimates the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer formed on the pallet 26 using the granulation results of the granulated particles acquired from the granulation result estimation unit 44 and the charging conditions acquired from the charging condition acquisition unit 46 (step S105). The processing of step S105 is a charging result estimation step. The charging result estimation unit 48 outputs the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer to the sintering result estimation unit 50.

[0078] The sintering result estimation unit 50 estimates the sintering temperature and the exhaust gas flow rate in the raw material charging layer using at least one of the particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer (step S106), and uses this to identify the maximum sintering temperature and the exhaust gas flow rate immediately below the raw material charging layer. The processing of step S106 is the sintering result estimation step. The sintering result estimation unit 50 outputs the identified maximum sintering temperature and the exhaust gas flow rate immediately below the raw material charging layer to the NOx concentration prediction unit 52.

[0079] The NOx concentration prediction unit 52 predicts the NOx concentration of the sintering exhaust gas using the maximum sintering temperature and the exhaust gas flow rate just below the raw material charging layer specified by the sintering result estimation unit 50 (step S107). The processing of step S107 is a NOx concentration prediction step of the sintering exhaust gas. As described above, the NOx concentration prediction method according to this embodiment includes a raw material information acquisition step, a granulation condition acquisition step, a granulation result estimation step, a charging condition acquisition step, a charging result estimation step, a sintering result estimation step, and a NOx concentration prediction step.

[0080] The NOx concentration prediction unit 52 outputs the predicted NOx concentration in the sintering exhaust gas to the production condition specification unit 54. The production condition specification unit 54 determines whether the NOx concentration in the sintering exhaust gas predicted by the NOx concentration prediction unit 52 satisfies a predetermined target value (step S108). The target value of the NOx concentration in the sintering exhaust gas is stored in advance in the storage unit 36 ​​by the operator via the input unit 32.

[0081] The manufacturing condition specifying unit 54 reads out the target value of the NOx concentration from the memory unit 36 ​​and compares the NOx concentration of the sintering exhaust gas predicted by the NOx concentration prediction unit 52 with the target value of the NOx concentration. When the predicted NOx concentration of the sintering exhaust gas is equal to or lower than the target value of the NOx concentration, the manufacturing condition specifying unit 54 determines that the predicted NOx concentration of the sintering exhaust gas satisfies the target value (step S108: Yes). On the other hand, when the predicted NOx concentration of the sintering exhaust gas is higher than the target value of the NOx concentration, the manufacturing condition specifying unit 54 determines that the predicted NOx concentration of the sintering exhaust gas satisfies the target value (step S108: Yes). Ｘ It is determined that the concentration does not satisfy the target value (step S108: No).

[0082] When it is determined that the predicted value of the NOx concentration in the sintering exhaust gas does not satisfy the target value (step S108: No), the production condition specifying unit 54 changes the sinter ore production conditions used to predict the NOx concentration in the sintering exhaust gas (step S109). The production condition specifying unit 54 changes, for example, the thickness of the raw material charging layer, the speed of the pallet 26, the temperature in the ignition furnace 28, or the gas flow rate of the ignition furnace 28, which are used by the firing result estimating unit 50 to estimate the sintering temperature in the raw material charging layer and the exhaust gas flow rate.

[0083] The manufacturing condition determination unit 54 returns the process to step S106 and performs steps S106 to S108 again using the changed manufacturing conditions. In this way, steps S106 to S109 are repeatedly performed until it is determined in step S108 that the NOx concentration in the sintering exhaust gas predicted satisfies the target value. This makes it possible to determine manufacturing conditions (sintering conditions) under which the predicted NOx concentration in the sintering exhaust gas satisfies the target value. One unit of change in the firing conditions is predetermined for each item of the firing conditions to be changed.

[0084] The production condition specifying unit 54 may change the charging conditions acquired by the charging condition acquiring unit 46 instead of the above production conditions. In this case, the production condition specifying unit 54 performs the processing of steps S105 to S108 again using the changed charging conditions (dashed arrow in FIG. 3). In this way, the processing of steps S105 to S109 is repeatedly performed until it is determined that the NOx concentration of the sintering exhaust gas predicted in step S108 satisfies the target value. This makes it possible to identify production conditions (charging conditions) that are determined to cause the predicted NOx concentration of the sintering exhaust gas to satisfy the target value. One unit of change of the charging conditions is predetermined for each item of the charging conditions to be changed.

[0085] The manufacturing condition determination unit 54 may change the moisture content of the sinter raw material acquired by the granulation condition acquisition unit 42 instead of the above manufacturing conditions. If the granulation conditions acquired by the granulation condition acquisition unit 42 include granulation conditions such as the rotation speed and residence time of the granulator 14, these granulation conditions may be changed in the process of step S109. In this case, the manufacturing condition determination unit 54 performs the processes of steps S103 to S108 again using the changed granulation conditions (dotted arrows in FIG. 3). In this manner, the processes of steps S103 to S109 are repeatedly performed until it is determined in step S108 that the NOx concentration of the sintering exhaust gas predicted satisfies the target value. This makes it possible to identify manufacturing conditions (granulation conditions) that are determined to result in the predicted NOx concentration of the sintering exhaust gas satisfying the target value. One unit of change in the granulation conditions is predetermined for each item of the granulation conditions to be changed. If it is not determined that the NOx concentration in the sintering exhaust gas satisfies the target value for all combinations of manufacturing conditions, the output unit 34 may display a message to that effect or an error, and the flow may be terminated.

[0086] On the other hand, if it is determined that the predicted value of the NOx concentration in the sintering exhaust gas satisfies the target value (step S108: Yes), the production condition specification unit 54 specifies the production conditions used to predict the NOx concentration as production conditions that satisfy the target value of the NOx concentration (step S110). The processing of this step S110 is the production condition specification step. The production condition specification unit 54 sets the specified production conditions as production conditions for each device of the sintering equipment 10. Thereafter, sintered ore is produced in the sintering equipment 10 for which the production conditions have been set (step S111). The processing of this step S111 is the sintered ore production step.

[0087] In this way, by predicting the NOx concentration in the sintering exhaust gas while changing various production conditions, it is possible to identify production conditions under which the NOx concentration in the sintering exhaust gas satisfies the target value. Then, by producing sintered ore in the sintering equipment 10 that reflects the identified production conditions, it becomes possible to produce sintered ore while satisfying the target value of the NOx concentration in the sintering exhaust gas.

[0088] Next, the test results for confirming the estimation of the granulation results will be explained. Table 1 below shows the particle size distribution of granulated particles granulated with various sinter raw materials and moisture contents, and the actual values ​​of the carbon concentration for each particle size. In the example shown in Table 1, coke powder is used as the carbon-containing raw material.

[0089]

[0090] In the granulation test, the sinter raw material was granulated under 24 granulation conditions, T1 to T24, to produce granulated particles. Table 1 shows the blending ratio, coke fine particle size, and moisture content as the granulation conditions for T1 to T24, and also shows the particle size distribution of the granulated particles and the carbon concentration for each particle size as actual values.

[0091] In the blending ratios shown in Table 1, raw material A is iron ore produced in South America. Raw material B is iron ore produced in South America of a different brand from raw material A. Raw material C is iron ore produced in Australia. In the coke breeze particle size, "-2 mm" indicates the proportion of coke breeze with a particle size of less than 2 mm. "-1 mm" indicates the proportion of coke breeze with a particle size of less than 1 mm. Coke breeze with a particle size of less than 2 mm is coke breeze that can be sieved under a sieve with 2 mm openings, and coke breeze with a particle size of less than 1 mm is coke breeze that can be sieved under a sieve with 1 mm openings.

[0092] In the particle size distribution of the granulated particles, "+8.0" indicates the proportion of granulated particles having a particle size of more than 8.0 mm. "2.8-8.0" indicates the proportion of granulated particles having a particle size of 2.8 mm or more and 8.0 mm or less. "-2.8" indicates the proportion of granulated particles having a particle size of less than 2.8 mm.

[0093] Granulated particles having a particle size of more than 8.0 mm are granulated particles that can be sieved through a sieve with an opening of 8.0 mm. Granulated particles having a particle size of 2.8 mm or more and 8.0 mm or less are granulated particles that can be sieved through a sieve with an opening of 8.0 mm and can be sieved through a sieve with an opening of 2.8 mm. Granulated particles having a particle size of less than 2.8 mm are granulated particles that can be sieved through a sieve with an opening of 2.8 mm.

[0094] In the carbon concentration for each particle size, "+8.0" indicates the concentration of carbon contained in granulated particles having a particle size of more than 8.0 mm. "2.8-8.0" indicates the concentration of carbon contained in granulated particles having a particle size of 2.8 mm or more and 8.0 mm or less. "-2.8" indicates the concentration of carbon contained in granulated particles having a particle size of less than 2.8 mm.

[0095] A particle size estimation model and a component estimation model were created using the actual values ​​shown in Table 1, and these were used to estimate the content of granulated particles in each particle size division and the carbon concentration for each particle size. This confirmation was performed using two types of estimation: estimation using a machine learning model and estimation using a multiple regression model.

[0096] When using a machine learning model, estimation was performed using a particle size estimation model and a component estimation model.The particle size estimation model was used, which outputs the content of granulated particles for each particle size class when the particle size of the coke fines, the blending ratio of the sintering raw materials, and the moisture content at the time of granulation are input.The component estimation model was used, which outputs the carbon concentration of granulated particles for each particle size when the particle size of the coke fines, the blending ratio of the sintering raw materials, and the moisture content at the time of granulation are input.

[0097] On the other hand, when estimating using a multiple regression model, the particle size estimation model used a model in which the particle size of the coke fines, the blending ratio of the sintering raw materials, and the moisture content during granulation were used as explanatory variables, and the content of granulated particles in each particle size class was used as the objective variable.The component estimation model used a model in which the particle size of the coke fines, the blending ratio of the sintering raw materials, and the moisture content during granulation were used as explanatory variables, and the carbon concentration for each particle size was used as the objective variable.

[0098] The correlation coefficients between the estimated values ​​and the actual values ​​when estimation is performed using a machine learning model and when estimation is performed using a multiple regression model are shown in Table 2. For the case when a machine learning model is used, the correlation coefficients when a neural network is used and when a C&R Tree (Classification and Regression Tree) is used are shown.

[0099]

[0100] As shown in Table 2, a high correlation coefficient was obtained when either model was used. These results confirmed that the particle size of granulated particles and the concentration of specific components for each particle size can be estimated with high accuracy by using a trained machine learning model or multiple regression model.

[0101] Next, the estimation of the charging result will be described. Fig. 4 is a graph showing an example of the charging result when granulated particles having a predetermined particle size distribution are charged into the pallet 26 under predetermined charging conditions. As shown in Fig. 4, in this embodiment, the content ratio of granulated particles having particle sizes of +8.0 mm and -2.8 mm and the content ratio of carbon are estimated for each division in the height direction of the raw material charging layer.

[0102] The content ratio of granulated particles of each particle size in each division in the height direction of the raw material charging bed shown in Figures 4(a) and (b) can be obtained by performing a simulation using DEM for each particle size distribution of the granulated particles and charging conditions. The carbon content ratio of each division in the height direction of the raw material charging bed as shown in Figure 4(c) can also be calculated from the carbon concentration of granulated particles of each particle size estimated by the component estimation model and the content ratio of particles of each particle size in each division in the height direction of the raw material charging bed.

[0103] In this way, by performing a simulation using DEM, the content ratio of granulated particles of each particle size in the height direction of the raw material charging layer can be obtained. Therefore, if a DEM simulation is performed in advance for each particle size distribution of the granulated particles and each charging condition, and a table showing the content ratio of granulated particles of each particle size in the height direction of the raw material charging layer is created, it can be seen that the charging result can be predicted using the table.

[0104] Next, the estimation of the sintering results will be described. Figure 5 is a graph showing an example of the sintering temperature in the raw material charging bed and the flue gas flow rate directly below the raw material charging bed estimated by the sintering result estimation unit 50. As shown in Figure 5, in the heat transfer model according to this embodiment, the raw material charging bed is divided into multiple microelements, and the temperature distribution in the sintering bed and the flue gas flow rate directly below the sintering bed are calculated using a sintering estimation model including a heat transfer model in each element region. The flue gas flow rate directly below the sintering bed is obtained from the pressure drop calculation in each control volume as disclosed in Reference 1 above. The calculation is performed assuming S'-type and P-type coke distributions in the pseudo-particles. This allows the flue gas flow rate and temperature in the sintering bed for each longitudinal position to be estimated. By setting the carbon content for each microelement using the carbon content distribution shown in Figure 4, it becomes possible to calculate the temperature distribution in the sintering bed and the flue gas flow rate directly below the sintering bed, taking into account segregation in the sintering bed.

[0105] Next, an example will be described in which the correlation coefficient R between the actual value of the NOx concentration in the sintering exhaust gas when sintered ore was produced in an actual sintering machine using the sintering raw materials of raw material blends 1 to 3 and the predicted value of the NOx concentration in the sintering exhaust gas was confirmed. The blends of the sintering raw materials used in the example and the amounts of the raw materials used are shown in Table 3 below.

[0106]

[0107] The sintering raw materials of raw material blends 1 to 3 shown in Table 3 were granulated in the granulator 14, charged onto the pallet 26 of the sintering machine 16 to form a raw material charging layer, and fired in the sintering machine 16 to produce sintered ore. For the production of this sintered ore, the temperature in the raw material charging layer and the exhaust gas flow rate immediately below the charging layer were estimated using the granulation result estimation unit 44, the charging result estimation unit 48, and the firing result estimation unit 50, and the NOx concentration of the sintering exhaust gas was predicted using the estimated temperature of the raw material charging layer and the exhaust gas flow rate immediately below the charging layer. The NOx concentration of the sintering exhaust gas was predicted using the following equation (1).

[0108] NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the above formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas. T is the highest sintering temperature (K) in the raw material charging layer. M is the amount of carbon-containing raw material contained in the sintering raw material used per unit time (ton / h). N is the nitrogen content (mass%) of the carbon-containing raw material contained in the sintering raw material. U is the exhaust gas flow rate (amount of gas molecules mol / h) directly below the raw material charging layer. A(K －１ ), B(-) is a constant.

[0109] In this example, T was used as the highest temperature among the estimated temperatures in the raw material charging layer. M was used as the amount of carbon-containing raw material used per unit time calculated from the content of carbon-containing raw material contained in the sintering raw material. N was used as 1.75. U was used as the total estimated exhaust gas flow rate directly below the charging layer in each wind box. A was used as a constant between -0.020 and -0.036, and B was used as a constant between 51 and 55. The correlation coefficient R between the actual and estimated values ​​of the NOx concentration in the sintering exhaust gas and the difference between the average of the actual values ​​and the average of the estimated values ​​are shown in Tables 4 and 5 below.

[0110]

[0111]

[0112] As shown in Tables 4 and 5, by setting A to -0.036 or more and -0.020 or less and B to 51 or more and 55 or less, the correlation coefficient between the estimated value of NOx concentration and the actual value of NOx concentration became 0.70 or more for any of raw material blends 1 to 3. From these results, it was confirmed that by using the above formula (1) in which A is -0.036 or more and -0.020 or less and B is 51 or more and 55 or less, the NOx concentration of the sintering exhaust gas can be predicted with high accuracy.

[0113] On the other hand, when looking at the difference between the actual value and the predicted value of the NOx concentration in the sintering exhaust gas, in Invention Examples 1 to 5, where A was set to -0.020, the difference exceeded 100 ppm. As A became smaller, the difference between the actual value and the predicted value of the NOx concentration in the sintering exhaust gas became smaller, but in Invention Examples 6 to 10, where A was set to -0.030, the difference exceeded 35 ppm for many of the blended raw materials.

[0114] In contrast, in Examples 12 to 14, 17 to 19, and 22 to 24, in which A was set to -0.036 or more and -0.033 or less and B was set to 52 or more and 54 or less, the difference between the actual and predicted NOx concentrations in the sintering exhaust gas was 35 ppm or less. From these results, it was confirmed that when using the above formula (1), A is preferably -0.036 or more and -0.020 or less, more preferably -0.036 or more and -0.030 or less, and even more preferably -0.036 or more and -0.033 or less. It was confirmed that B is preferably 51 or more and 55 or less, and more preferably 52 or more and 54 or less.

[0115] In particular, in the above formula (1), it is preferable that A is set to be not less than −0.036 and not more than −0.033, and B is set to be not less than 52 and not more than 54. It has been confirmed that by setting A and B within these ranges, the correlation coefficient between the actual value and the predicted value of the NOx concentration of the sintering exhaust gas can be made 0.7 or more, and the difference between the actual value and the predicted value can be made 35 ppm or less.

[0116] Fig. 6 is a graph showing the time transition of the actual increase / decrease amount and the predicted increase / decrease amount of the NOx concentration of the sintering exhaust gas. In Fig. 6, the horizontal axis is the elapsed time (h) and the vertical axis is the increase / decrease amount (ppm) of the NOx concentration of the sintering exhaust gas. Fig. 6 shows the increase / decrease amount of the NOx concentration of the sintering exhaust gas when raw material blend 1 is used for invention example 17 where A is -0.035 and B is 53.

[0117] In Fig. 6, the solid line shows the time transition of the actual increase / decrease amount of NOx concentration, and the dashed line shows the time transition of the predicted increase / decrease amount of NOx concentration. As shown in Fig. 6, the time transition of the actual increase / decrease amount of NOx concentration in the sintering exhaust gas and the time transition of the predicted increase / decrease amount were in good agreement.

[0118] Figure 7 is a graph showing the correlation between the actual increase / decrease and the predicted increase / decrease of the NOx concentration shown in Figure 6. In Figure 7, the horizontal axis represents the predicted increase / decrease of the NOx concentration in the sintering exhaust gas, and the vertical axis represents the actual increase / decrease of the NOx concentration in the sintering exhaust gas. As shown in Figure 7, the correlation coefficient R between the predicted increase / decrease and the actual increase / decrease of the NOx concentration in the sintering exhaust gas was high at 0.71. From these results, it was confirmed that the NOx concentration in the sintering exhaust gas can be predicted with high accuracy by using the NOx concentration prediction method according to this embodiment.

[0119] REFERENCE SIGNS LIST 10 Sintering equipment 12 Control device 14 Granulator 16 Sintering machine 18 Crusher 20 Cooler 22 Sieving device 24 Sintering raw material supply device 26 Pallet 28 Ignition furnace 29 Wind box 30 Control unit 32 Input unit 34 Output unit 36 ​​Memory unit 38 Communication unit 40 Raw material information acquisition unit 42 Granulation condition acquisition unit 44 Granulation result estimation unit 46 Charging condition acquisition unit 48 Charging result estimation unit 50 Sintering result estimation unit 52 NOx concentration prediction unit 54 Manufacturing condition identification unit

Claims

1. A method for predicting the NOx concentration of sintering exhaust gas emitted from a sintering machine that sinters granulated particles obtained by adding water to sintering raw materials containing iron-containing raw materials and carbon-containing raw materials, comprising: a raw material information acquisition step of acquiring raw material information including the particle size, component composition, and blending ratio of each raw material contained in the sintering raw materials; and a granulation condition acquisition step of acquiring granulation conditions including the moisture content of the sintering raw materials when granulating the granulated particles from the sintering raw materials. a granulation result estimation step of estimating a granulation result including the particle size of the carbon-containing raw material and a component composition of the granulated particles for each particle size by inputting the raw material information and the granulation conditions into a granulation estimation model including: a particle size estimation model that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and outputs the content of the granulated particles in one of a plurality of particle size classifications into which the particle sizes of the granulated particles are classified; and a component estimation model that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and outputs the content of a specific component of the granulated particles in one of a plurality of particle size classifications into which the particle sizes of the granulated particles are classified; and a charging condition acquisition step of acquiring charging conditions for charging the granulated particles into the sinter machine to form a raw material charging layer. a charging result estimating step of estimating particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer based on the granulation result and the charging conditions, using a charging estimation model that indicates the content ratio of the granulated particles of each particle size in each of a plurality of charging sections obtained by dividing the height direction of the raw material charging layer into a plurality of sections; a sintering result estimating step of estimating the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on at least one of the particle size segregation and the component segregation, using a sintering estimation model that includes a heat transfer model that can calculate the sintering temperature at a predetermined position in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer; and a NOx concentration estimating step of predicting the NOx concentration of the sintering exhaust gas using the sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer, and the amount of carbon-containing raw material used per unit time included in the sintering raw material.

2. The NOx concentration prediction method according to claim 1, wherein the NOx concentration prediction step predicts the NOx concentration of the sintering exhaust gas using the following formula (1): NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas, T is the highest temperature (K) among the sintering temperatures in the raw material charging layer, M is the amount of carbon-containing raw material used per unit time (ton / h) contained in the sintering raw material, N is the nitrogen content ratio (mass%) of the carbon-containing raw material contained in the sintering raw material, U is the exhaust gas flow rate (mol / h) directly below the raw material charging layer, and A (K －１ ), B(-) is a constant.

3. A method for producing sintered ore, comprising: a production condition specification step for specifying production conditions under which the NOx concentration of sintered ore exhaust gas predicted by the NOx concentration prediction method according to claim 1 or 2 satisfies a predetermined target value; and a sintered ore production step for producing sintered ore under the production conditions specified in the production condition specification step.

4. A control device for predicting the NOx concentration of sintering exhaust gas emitted from a sintering machine that sinters granulated particles obtained by adding water to sintering raw materials containing iron-containing raw materials and carbon-containing raw materials, comprising: a raw material information acquisition unit that acquires raw material information including the particle size, component composition, and blending ratio of each raw material contained in the sintering raw materials; and a granulation condition acquisition unit that acquires granulation conditions including the moisture content of the sintering raw materials when granulating the granulated particles from the sintering raw materials. a granulation result estimation unit that estimates a granulation result including the particle size of the carbon-containing raw material and a component composition of the granulated particles for each particle size by inputting the raw material information and the granulation conditions into a granulation estimation model including: a particle size estimation model that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and outputs the content of the granulated particles in one of a plurality of particle size classifications into which the particle sizes of the granulated particles are classified; and a component estimation model that receives as input the particle size of the carbon-containing raw material, the blending ratio of each raw material contained in the sintering raw material, and the moisture content, and outputs the content of a specific component of the granulated particles in one of a plurality of particle size classifications into which the particle sizes of the granulated particles are classified; and a charging condition acquisition unit that acquires charging conditions when the granulated particles are charged into the sinter machine to form a raw material charging layer. a charging result estimation unit that estimates particle size segregation and component segregation of the granulated particles in the height direction of the raw material charging layer based on the granulation result and the charging conditions, using a charging estimation model that indicates the content ratio of the granulated particles of each particle size in each of a plurality of charging sections obtained by dividing the height direction of the raw material charging layer into a plurality of sections; a sintering result estimation unit that estimates the sintering temperature in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer based on at least one of the particle size segregation and the component segregation, using a sintering estimation model that includes a heat transfer model that can calculate the sintering temperature at a predetermined position in the raw material charging layer and the exhaust gas flow rate immediately below the raw material charging layer; and a NOx concentration prediction unit that predicts the NOx concentration of the sintering exhaust gas using the sintering temperature in the raw material charging layer, the exhaust gas flow rate immediately below the raw material charging layer, and the amount of carbon-containing raw material used per unit time included in the sintering raw materials.

5. The control device according to claim 4, wherein the NOx concentration prediction unit predicts the NOx concentration of the sintering exhaust gas using the following formula (1): NOx = (A x (T - 273) + B) x M x N / (100 x 14 x U) ... (1) In the formula (1), NOx is the NOx concentration (ppm) of the sintering exhaust gas, T is the highest temperature (K) among the sintering temperatures in the raw material charging layer, M is the amount of carbon-containing raw material used per unit time (ton / h) contained in the sintering raw material, N is the nitrogen content ratio (mass%) of the carbon-containing raw material contained in the sintering raw material, U is the exhaust gas flow rate (mol / h) directly below the raw material charging layer, and A(K －１ ), B(-) is a constant.

6. A control device according to claim 4 or claim 5, further comprising a manufacturing condition specifying unit that specifies manufacturing conditions under which the NOx concentration of the sintering exhaust gas predicted by the NOx concentration prediction unit satisfies a predetermined target value.

Images