Sintered ore strength control device, sintered ore strength control method, sintered ore production method, prediction model generation method, sintered ore strength prediction method, sintering operation guidance system, and terminal device

Abstract

A sintered ore strength control device (10) includes: an acquisition unit (12) that acquires data related to an operation condition of a sintering production facility as input data, wherein the data includes exhaust gas information; a sintered ore strength prediction unit (13) that predicts sintered ore strength after a predetermined time on the basis of the acquired input data; an operation action determination unit (14) that determines an operation action on the basis of the predicted sintered ore strength and a predetermined target range for the sintered ore strength; and an output unit (15) that outputs a change instruction to the sintering production facility or presents a guidance operation amount on the basis of the determined operation action.

Description

Sintered ore strength control device, sintered ore strength control method, sintered ore manufacturing method, prediction model creation method, sintered ore strength prediction method, sintered ore operation guidance system, and terminal device

[0001] The present disclosure relates to a sinter ore strength control device, a sinter ore strength control method, a sinter ore manufacturing method, a prediction model generation method, a sinter ore strength prediction method, a sintering operation guidance system, and a terminal device.

[0002] Sintered ore is used as the iron source for blast furnaces. It is produced by burning small particles of iron ore using the heat from the combustion of a coagulant. Since sintered ore accounts for approximately 80% of the charge to a blast furnace, the properties of sintered ore have a significant impact on blast furnace operation. The strength of sintered ore is particularly important, as low-strength sintered ore is prone to pulverization, which can lead to poor furnace conditions due to poor permeability or uneven gas flow.

[0003] One effective method for increasing strength is to promote combustion to prevent sintered ore from passing through the sinter machine unfired. For example, increasing the ratio of the coagulant (heat source), increasing the ratio of quicklime (which acts as a binder during granulation to improve ventilation), and increasing the proportion of coke fines in the upper layer to facilitate ignition in the ignition furnace are effective methods. Another effective method is to reduce the pallet speed to ensure sufficient firing time. However, there is a trade-off between increasing sintered ore strength and productivity: increasing the ratio of coagulant, quicklime, and coke fines increases production costs, while reducing the pallet speed reduces production volume.

[0004] Representative methods for measuring the strength of sintered ore include the shutter strength test and the tumbler strength test. In the shutter strength test, sintered ore is dropped four times from a height of 2 m, and the strength is defined as the percentage of the measured sample mass of a +10 mm particle size relative to the measured sample mass before the test (shutter strength index (SI), unit [%]). In the tumbler strength test, when a rotation test is performed using a rotating drum, the strength is defined as the percentage of the measured sample mass of a +6.3 mm particle size relative to the measured sample mass before the test (tumbler strength index (TI), unit [%]).

[0005] Patent Document 1 discloses a method for predicting the strength of sintered ore from the calcium ferrite content, slag content, pore size distribution index, and porosity, which are constituent minerals of the sintered ore. Also, a known technique is to measure the change in the FeO ratio in sintered ore in real time by detecting the change in inductance with a coil.

[0006] Japanese Patent Application Publication No. 7-11349

[0007] The method of Patent Document 1 requires sampling tests to identify physical properties and cannot continuously predict sinter strength. Therefore, it is not possible to continuously and early detect strength fluctuations. Furthermore, the method of measuring changes in the FeO content in sinter in real time by detecting changes in inductance with a coil is not widely used due to poor measurement accuracy, and traditional chemical analysis is used for measurement. Here, chemical analysis takes approximately five hours to measure (three hours for cooling in a cooler and sampling, and two hours for analysis). Therefore, an increase in the FeO content in sinter can only be detected five hours later, resulting in a five-hour delay in corrective action. Furthermore, while a higher FeO content tends to increase strength, the two do not necessarily correspond one-to-one.

[0008] In consideration of the above circumstances, the purpose of the present disclosure is to provide a sintered ore strength control device, a sintered ore strength control method, a sintered ore manufacturing method, a prediction model generation method, a sintered ore strength prediction method, a sintering operation guidance system, and a terminal device that can predict sintered ore strength with high accuracy.

[0009] (1) A sintered ore strength control device according to one embodiment of the present disclosure includes: an acquisition unit that acquires data regarding operating conditions in a sintered ore manufacturing facility, including exhaust gas information, as input data; a sintered ore strength prediction unit that predicts the sintered ore strength after a predetermined time based on the acquired input data; an operational action determination unit that determines an operational action based on the predicted sintered ore strength and a predetermined target range for the sintered ore strength; and an output unit that outputs a change instruction to the sintered ore manufacturing facility or presents a guidance operation amount based on the determined operational action.

[0010] (2) As an embodiment of the present disclosure, in (1), the exhaust gas information is an exhaust gas NO. ｘ Concentration, exhaust gas O ２ The information includes at least one of the concentration and the exhaust gas temperature.

[0011] (3) As an embodiment of the present disclosure, in (2), the exhaust gas information includes an exhaust gas CO concentration, an exhaust gas CO ２ Concentration and exhaust gas SO ｘ The concentration includes at least one of:

[0012] (4) As one embodiment of the present disclosure, in any one of (1) to (3), the input data includes at least one of a mixing ratio of a binder and a pallet speed.

[0013] (5) As an embodiment of the present disclosure, in any one of (1) to (4), the input data includes at least one of raw material brand, raw material moisture content, water spray flow rate in a granulation mixer, quicklime blending ratio, iron ore blending ratio, return ore blending ratio, cooler blowing pressure, cooler exhaust gas temperature, layer thickness, and production volume.

[0014] (6) As one embodiment of the present disclosure, in any one of (1) to (5), the sintered ore strength prediction unit predicts the sintered ore strength using, as the input data, data acquired in consideration of a delay time based on a timing at which the sintered ore strength can be identified.

[0015] (7) As one embodiment of the present disclosure, in (6), the time interval for outputting a change instruction to the sintering manufacturing equipment based on the operational action is set to be greater than the delay time going back from the prediction point in time when the manipulated variable is past data.

[0016] (8) A sintered ore strength control method according to one embodiment of the present disclosure includes: an acquisition step of acquiring data relating to operating conditions in a sintered ore manufacturing facility, including exhaust gas information, as input data; a sintered ore strength prediction step of predicting the sintered ore strength after a predetermined time based on the acquired input data; an operational action determination step of determining an operational action based on the predicted sintered ore strength and a predetermined target range for the sintered ore strength; and an output step of outputting a change instruction to the sintered ore manufacturing facility or presenting a guidance operation amount based on the determined operational action.

[0017] (9) A method for producing sintered ore according to an embodiment of the present disclosure produces sintered ore in accordance with the change instruction output to the sintering production facility by the sintered ore strength control method according to (8).

[0018] (10) A prediction model generation method according to one embodiment of the present disclosure is a prediction model generation method for generating a prediction model for predicting sintered ore strength, and includes the steps of: acquiring, as explanatory variables, data on operating conditions in a sintering production facility, including exhaust gas information, and, as a dependent variable, the sintered ore strength corresponding to the explanatory variables; and associating the explanatory variables with the dependent variables to use them as learning data, and generating the prediction model by machine learning.

[0019] (11) A sintered ore strength prediction method according to one embodiment of the present disclosure includes: an acquisition step of acquiring data relating to operating conditions in a sintering production facility, including exhaust gas information, as input data; a sintered ore strength prediction step of predicting the sintered ore strength after a predetermined time based on the acquired input data; and an output step of outputting the predicted sintered ore strength.

[0020] (12) A sintering operation guidance system according to one embodiment of the present disclosure includes an acquisition unit that acquires data regarding operating conditions in a sintering manufacturing facility, including exhaust gas information, as input data; a sinter strength prediction unit that predicts the sinter strength after a predetermined time based on the acquired input data; an operation action determination unit that determines an operation action based on the predicted sinter strength and a predetermined target range for the sinter strength; an output unit that presents a guidance operation amount based on the determined operation action; and a display unit that displays the acquired guidance operation amount.

[0021] (13) A terminal device according to one embodiment of the present disclosure is a terminal device that constitutes a sintering operation guidance system together with a sintering operation guidance server, and includes: a terminal communication unit that acquires guidance operation variables for a plurality of operations calculated so that the predicted sinter ore strength after a predetermined time approaches a predetermined target value and operation control indicators related to the production volume of sinter, using a prediction model that predicts sinter strength using data related to operating conditions in a sintering manufacturing facility, including a plurality of operation variables and exhaust gas information, as input data; a display unit that prioritizes and displays the guidance operation variables based on the acquired guidance operation variables and operation control indicators; and an operation action change input unit that allows an operator to select from the displayed guidance operation variables.

[0022] According to the present disclosure, it is possible to provide a sintered ore strength control device, a sintered ore strength control method, a sintered ore manufacturing method, a prediction model generation method, a sintered ore strength prediction method, a sintering operation guidance system, and a terminal device that can predict sintered ore strength with high accuracy.

[0023] FIG. 1 is a diagram showing an overview of a sintering process. FIG. 2 is a diagram showing the prediction error of a prediction model depending on the number of explanatory variables. FIG. 3 is a diagram showing a configuration example of a sinter ore strength control device according to an embodiment of the present disclosure. FIG. 4 is a flowchart showing a sinter ore strength control method according to an embodiment of the present disclosure. FIG. 5 is a diagram showing the results of Example 1. FIG. 6 is a diagram showing the results of Example 2. FIG. 7 is a diagram showing a configuration example of a sintering operation guidance system according to an embodiment of the present disclosure.

[0024] A sinter ore strength control device 10 (see FIG. 3 ), a sinter ore strength control method, a sinter ore manufacturing method, a prediction model generation method, a sinter ore strength prediction method, a sintering operation guidance system, and a terminal device 50 (see FIG. 7 ) according to an embodiment of the present disclosure are described below with reference to the accompanying drawings. The sinter ore strength control method according to this embodiment, in summary, uses data related to operational conditions in a sinter manufacturing facility, including exhaust gas information, as input data, to predict sinter ore strength in real time using a machine learning model. The machine learning model is a trained model generated by machine learning. Based on the predicted value, the system calculates and outputs a manipulated variable for the coagulant blending ratio to bring the sinter ore strength within a target range, thereby controlling the sinter ore strength to an appropriate range and implementing appropriate operational actions. For example, if the sinter ore strength is predicted to exceed a predetermined target range, actions such as reducing the coagulant, increasing the pallet speed, or reducing the quicklime can be implemented. Furthermore, if the sinter ore strength is predicted to fall below the target, actions such as increasing the coagulant, decreasing the pallet speed, or increasing the quicklime can be implemented. Here, the method for measuring the sinter strength is not limited, and for example, a shutter strength test or a tumbler strength test may be used. In addition, in the sinter strength control method, the predicted sinter strength may be output as is. In this case, the sinter strength prediction method is carried out by the sinter strength control device 10 functioning as a sinter strength prediction device.

[0025] Figure 1 shows an overview of the sintering process. Sinter is produced by burning small-sized iron ore using the heat generated by the combustion of a coagulant to produce sintered ore for use as a source of iron in a blast furnace. The raw material for sinter is stored in a hopper and is removed from the hopper. If powdered iron ore were directly charged into the sintering machine, poor ventilation would inhibit the combustion reaction. Therefore, in the granulation process, the iron ore, along with other raw materials such as quicklime and coke, is mixed with water in a granulation mixer to produce granules with larger particle sizes than the original raw materials. The granules are then charged into the sintering machine, ignited in an ignition furnace, and then suctioned from below to gradually promote the combustion reaction in layers from top to bottom. After firing, the sintered ore is discharged from the sintering machine, crushed in a crusher, and then sent to a cooler. After being cooled by air in the cooler duct, it is sorted through a sieve, and the larger particles are sent to the blast furnace as non-defective ("product" in Figure 1). Sintered ore with a small particle size (for example, a particle size of 4 mm or less) is returned to the return ore hopper as return ore, and then cut out from the return ore hopper and fed into the sinter machine again.

[0026] (Sinter Strength Prediction) In this embodiment, the sinter strength is predicted using a prediction model. As described above, the sinter strength can be measured using a shutter strength test or a tumbler strength test, and may be expressed as a shutter strength index or a tumbler strength index in units of [%]. Data related to the operating conditions of the equipment for producing sinter (sinter production equipment), including at least exhaust gas information, is input to the prediction model. The prediction model uses this input data as explanatory variables and the sinter strength as a target variable. The sintering process performed in the sinter production equipment can then be controlled based on the predicted value of the sinter strength. For example, the sinter production equipment can automatically or via an operator's operation take operational actions during the sintering process to keep the sinter strength within a predetermined target range.

[0027] The input data includes exhaust gas information. The exhaust gas information is information on exhaust gas generated during firing, and the exhaust gas NO ｘ Concentration, exhaust gas O ２ The exhaust gas information includes at least one of the exhaust gas CO concentration, the exhaust gas CO ２Concentration and exhaust gas SO ｘ The input data may further include at least one of the following: concentration. In addition to the exhaust gas information, the input data may include data on the operating conditions of the sinter manufacturing equipment. The input data may include at least one of the coagulant blending ratio and pallet speed. The input data may also include at least one of the raw material brand, raw material moisture, spray flow rate in the granulation mixer, quicklime blending ratio, iron ore blending ratio, return ore blending ratio, cooler blast pressure, cooler exhaust gas temperature, layer thickness, and production volume. For example, the input data may include all of these feature quantities, or may further include other operating factors. Here, the raw material brand and raw material moisture refer to the brand of iron ore in the raw material and the moisture content in the raw material, respectively. As described above, iron ore is mixed with water in the granulation mixer along with other raw materials such as quicklime and coke, and the spray flow rate refers to the flow rate at which the water is sprayed to be mixed. The blending ratio of the coagulant, the blending ratio of quicklime, the blending ratio of iron ore, and the blending ratio of return ore are the blending ratios of the coagulant, quicklime, iron ore, and return ore in the raw materials of the sintered ore, respectively. The coagulant is, for example, a carbonaceous material such as coke powder. ｘ Concentration, exhaust gas O ２ concentration, exhaust gas CO concentration, exhaust gas CO ２ Concentration, exhaust gas SO ｘ The concentration and exhaust gas temperature are the concentration and temperature of each component in the exhaust gas generated during firing. The cooler air pressure is the pressure of the air sent by the cooler when cooling the sintered ore. The cooler exhaust gas temperature is the temperature of the exhaust gas sucked in by the cooler. The pallet speed is the transport speed of the pallet that transports the raw materials during firing. The layer thickness is the thickness of the layer of raw materials on the pallet. The production volume is the amount of sintered ore produced.

[0028] Here, among the above-mentioned characteristic quantities, the exhaust gas NO Ｘ The exhaust gas information including the concentration is particularly important. ２ As the firing progresses, the strength of the sintered ore increases, and the O ２ In addition, the partial pressure of CO increases, creating a stronger reducing atmosphere, and NO, an oxide, is consumed. Ｘ Therefore, the generation of exhaust gas NO ＸThe lower the concentration, the higher the strength of the sintered ore. Also, the higher the firing temperature, the higher the exhaust gas temperature. Therefore, the higher the exhaust gas temperature, the higher the strength of the sintered ore.

[0029] The input data preferably includes the mixing ratio of the agglomerating agent and the pallet speed. The higher the mixing ratio of the agglomerating agent, the higher the firing temperature and the higher the strength of the sintered ore. Furthermore, lowering the pallet speed ensures sufficient firing time, preventing a decrease in the strength of the sintered ore.

[0030] The prediction model is not limited to a specific one as long as it is configured to obtain a response variable (sinter ore strength) from explanatory variables. The prediction model may be, for example, a physical model or a machine learning model. In this embodiment, the prediction model is generated by machine learning. As a machine learning method, linear regression, hierarchical model, neural network, decision tree, GBDT (Gradient Boosting Decision Tree), random forest, transformer, etc. can be used, and is not particularly limited. The prediction model is generated, for example, using actual data (past measured values, past set values, etc.) from the sintering process before the sinter ore strength is predicted.

[0031] Figure 2 shows an example of evaluating a prediction model generated by a neural network technique. The number of explanatory variables in machine learning was increased (types of explanatory variables were added), and the error between the predicted value of sinter strength by each machine learning model and the actual sinter strength was measured. For example, when the explanatory variables were exhaust gas NO Ｘ The prediction error of the prediction model based only on the concentration is 1.7% or more. Ｘ Concentration, exhaust gas O ２ The prediction error of the prediction model using the concentration, exhaust gas temperature, pallet speed, coagulant blend ratio, and quicklime blend ratio was 0.66%. As such, the error tended to decrease as the number of explanatory variables increased. Here, Figure 2 shows an example of the order in which explanatory variables were added. For example, if the explanatory variables were exhaust gas NO ＸThe mixing ratio of the coagulant may be added after the concentration, and regardless of the order of addition shown in Figure 2, the error tends to decrease as the number of explanatory variables increases.

[0032] In addition, various operational factors (operational variables for the sintering process) can be used as input data for the prediction model, but the time it takes for the operational variables to affect the strength of sintered ore varies depending on the operational variables. For example, the mixing ratio of the coagulant, which is data related to raw materials, is located upstream of the sintering process, so it can affect the strength of sintered ore after 2.5 hours. For example, the NOx content of exhaust gas, which is data related to firing, can affect the strength of sintered ore after 2.5 hours. ｘ Concentration, exhaust gas O ２ The concentration and exhaust gas temperature are located downstream of the sintering process, and therefore may affect the sinter strength after 1.5 hours. For example, the cooler exhaust gas temperature, which is data related to the cooler, is located downstream of the sintering process, and therefore may affect the sinter strength after 1.5 hours. Furthermore, the time until the sinter strength is affected may vary depending on the pallet speed. Therefore, for accurate prediction, it is preferable that the input data for the prediction model be acquired taking into account a delay time based on the timing at which the sinter strength can be determined. Here, the timing at which the sinter strength can be determined may be, for example, when the sinter strength is determined by sieving after cooling in the cooler and whether the particle size meets the standard (the time of product determination). The sinter strength may then be determined at the time of product determination. The operating variables that affected the sinter at this determination time may be acquired taking into account the delay time. Similarly, the prediction model is preferably generated by machine learning using training data in which data corresponding to explanatory variables (input data) extracted from actual data in the sintering process and the sinter strength extracted from the actual data are matched, taking the delay time into account. For example, a prediction formula for predicting the sinter strength (y) at time t is shown in Equation (1) below.

[0033] y(t) = f(x １ (t-1.5), x ２ (t-1.5), x ３ (t-1.5), x ４ (t-1.5), x ５ (t-2.5)) Formula (1)

[0034] Here, the unit of time t is hours. １ (t) is the exhaust gas NO at time t ｘ is the concentration. ２ (t) is the exhaust gas O at time t ２ is the concentration. ３ (t) is the exhaust gas temperature at time t. x ４ (t) is the pallet speed at time t. x ５ (t) is the mixing ratio of the coagulant at time t. Also, "t-1.5" indicates 1.5 hours before time t. Each explanatory variable is acquired taking into account the time (delay time) until it affects the strength of the sintered ore.

[0035] A process for predicting sinter strength is performed using the input data and prediction model described above. The sinter strength after a predetermined time (e.g., 1.5 hours) is predicted so that corrective action is not delayed. The predetermined time is preferably, for example, the shortest delay time among the variables used in the prediction formula. By setting the predetermined time in this manner, future sinter strength can be predicted using current or past performance data. Here, when a prediction calculation is required far into the future, future values ​​may be required for some variables. In this case, for some variables for which future values ​​are required, the current measured values ​​may be assumed to continue as they are, and the current measured values ​​may be used as future values.

[0036] (Determining Operational Actions) Furthermore, manipulated variables for the sintering process are calculated so that the predicted value of sinter strength falls within the target range. In this embodiment, if the sinter strength exceeds a preset target range in the future, an action is taken to reduce the blending ratio of the agglomerating agent to lower the sinter strength. Furthermore, if the sinter strength falls below the target range in the future, an action is taken to increase the blending ratio of the agglomerating agent to increase the sinter strength. In calculating the manipulated variables, a machine learning model using the manipulated variables as explanatory variables may be constructed, and the manipulated variables that make the predicted value of sinter strength match the target value (e.g., the median of the target range) may be calculated by inverse analysis. Inverse analysis may include a method of selecting one or more manipulated variables to be changed and using optimization calculation, or a method of using the influence coefficient of the predicted value on the manipulated variable of the machine learning model. As another example, a physical model and a machine learning model may be used in combination to calculate the manipulated variables. In this case, a simulation of changing the blending ratio of the agglomerating agent is performed using the physical model. Then, the future operating state (e.g., flue gas NOx) predicted by the simulation is calculated. ｘ Concentration, exhaust gas O ２ The strength of the sintered ore is predicted using a machine learning model with inputs such as the concentration of the sintered ore and the like. The blending ratio of the coagulant that matches the predicted value of the sintered ore strength and the target value may be calculated by inverse analysis.

[0037] As described above, increasing the blending ratio of the agglomerating agent, which is a heat source, to promote reduction is an effective measure for increasing sinter strength. Other effective measures for increasing sinter strength include promoting combustion and improving ventilation to prevent sinter from passing through the sinter machine unfired. For example, increasing the blending ratio of quicklime, which acts as a binder during granulation, is effective. Increasing the proportion of coke fines in the upper layer is also effective for facilitating ignition in the ignition furnace. Reducing the pallet speed is also effective for increasing sinter strength, as it secures firing time. Conversely, reducing the blending ratio of the agglomerating agent to suppress reduction is effective for decreasing sinter strength. Reducing the blending ratio of quicklime and increasing the pallet speed are also effective measures for decreasing sinter strength.

[0038] Here, if the mixing ratio of the coagulation agent is changed significantly at once, the furnace conditions may change significantly and operation may become unstable. Therefore, in order to prevent excessive operation (excessive operation amount), a range of operation amount per operation may be determined, and operation within the range may be performed multiple times.

[0039] The calculated manipulated variable (the blending ratio of the coagulant in this embodiment) may be output as a change instruction to, for example, a process computer that manages the sintering process. The output of the manipulated variable includes output as operation guidance to the operator operating the sintering machine. That is, the manipulated variable may be output so that the appropriate manipulated variable is reflected in the sintering process through the operator's judgment. The information output as operation guidance includes at least the calculated manipulated variable and may be displayed on a display visible to the operator. The selection of one of the multiple manipulated variables is determined appropriately depending on the operational status. For example, if production volume is prioritized, adjustment of the blending ratio of the coagulant or quicklime may be prioritized. Furthermore, if production volume is sufficient and sufficient, adjustment of the pallet speed may be prioritized. Furthermore, when the sinter ore strength control device 10 functions as an operation guidance device, all manipulated variables may be presented to the operator, and the operator may determine the manipulated variable.

[0040] Here, the process of determining and using the manipulated variables involves, in summary, (Step 1) predicting the strength of sintered ore, (Step 2) determining the manipulated variables based on the prediction, and (Step 3) using the manipulated variables.

[0041] In the above (Step 1), prediction may be performed according to the data observation time of the operation conditions with a short delay time. At this time, for the data of the operation conditions with a long delay time, the saved past data is used (x in Equation (2)). ５ In the example of the above formula (1), if the time t when the exhaust gas information is acquired is used as the reference time, the prediction formula is expressed as the following formula (2).

[0042] y(t+1.5)=f(x １ (t), x ２ (t), x ３ (t), x ４ (t), x ５ (t-1.0)) Formula (2)

[0043] In the above (Step 2), the difference between the predicted value of the sintered ore strength obtained using the prediction formula and the target value is calculated, and the manipulated variable is calculated by inverse analysis so as to approach the target value.

[0044] In the above (Step 3), the manipulated variables (such as the coagulant blending ratio, the quicklime blending ratio, and the pallet speed) may be directly given as change instructions to the sinter machine control device (an example of sinter manufacturing equipment). The manipulated variables and the predicted value of the sinter ore strength may be displayed on a display device such as a monitor as guidance manipulated variables for the operator. The operator may then input the presented value or a modified value using an input device such as a keyboard, thereby transmitting a change instruction to the sinter machine control device. Alternatively, multiple ways of giving manipulated variables may be displayed, and the selected manipulated variable may be determined as the final manipulated variable using an input device such as a pointer or a keyboard, and an instruction may be transmitted to the sinter machine control device. The content displayed on the display as the guidance manipulated variables is not limited, but the time series change in the actual value of the sinter ore strength may be displayed together with the predicted value of the sinter ore strength. Furthermore, the flue gas NOx, which is the operating condition (input data) used to predict the sinter ore strength, may be displayed as a change instruction to the sinter machine control device. ｘ Concentration, exhaust gas O ２ In addition to time-series changes in concentration and exhaust gas temperature, operational control indicators such as actual values ​​of each manipulated variable, upper and lower limit control values, and target sintering production rate may be displayed.

[0045] Here, the control period (period of changing the manipulated variable) needs to be set to be longer than the delay time going back from the time of prediction when the manipulated variable used as input data at the time of predicting the sinter strength is past data. This is to ensure that the results of the previous change in the manipulated variable are reflected in the predicted value when the next prediction for changing the manipulated variable is made. In the above formula (2), the mixing ratio of the coagulating agent (x ５ (t-1.0)) uses data from one hour before the prediction time point (that is, the delay time going back from the prediction time point is one hour), so a control period longer than one hour is adopted.

[0046] FIG. 3 is a diagram showing the configuration of the sinter ore strength control device 10 according to this embodiment. The sinter ore strength control device 10 is part of a sinter production facility and controls the strength of sinter ore. Among the sinter production facilities, the facilities other than the sinter ore strength control device 10 may be referred to as other production facilities. As shown in FIG. 3, the sinter ore strength control device 10 includes a memory unit 11, an acquisition unit 12, a sinter ore strength prediction unit 13, an operation action determination unit 14, and an output unit 15. The sinter ore strength control device 10 acquires data on the operating conditions of the sinter production facility, i.e., the above-mentioned input data, from an operation data server 60. The input data includes exhaust gas NO Ｘ Concentration, exhaust gas O ２The actual values ​​include actual values ​​of characteristic quantities such as concentration, exhaust gas temperature, and coagulant blending ratio. The actual values ​​include measured values ​​and set values ​​of manipulated variables. The sinter strength control device 10 may also acquire a target range for sinter strength from an operation data server 60. The operation data server 60 can communicate with the sinter strength control device 10 via a network and may be implemented, for example, by a computer that collects data related to sinter ore production. The network may be, for example, the Internet. The sinter strength control device 10 performs the above-described process, i.e., predicts sinter strength using a prediction model and calculates manipulated variables for manipulated variables such as the coagulant blending ratio so that future sinter strength is maintained within the target range. In this embodiment, the sinter strength control device 10 also has a function to output manipulated variables of manipulated variables to the sinter production equipment or a function to present guidance manipulated variables via the output unit 15. When the output unit 15 presents the guidance manipulated variables, the sinter strength control device 10 functions as an operation guidance device. The display unit 30 displays the guidance manipulated variables output from the sinter strength control device 10 (operation guidance device). The sinter ore strength control device 10 may be configured as a computer (for example, a process computer that manages the operation of the sintering machine or a sintering operation guidance server 40 described later) separate from the operation data server 60. The display unit 30 may be a display device such as a liquid crystal display (LCD) or an organic electroluminescence panel (OLED). The display unit 30 may also be realized by a display of a terminal device 50 (see FIG. 7) such as a smartphone or a tablet.

[0047] Here, the prediction model may be generated by the sinter ore strength control device 10 and stored in the memory unit 11, or may be generated by another computer and stored in the memory unit 11. When the sinter ore strength control device 10 generates a prediction model, it may further include a model generation unit that generates the prediction model and stores it in the memory unit 11. The model generation unit and the acquisition unit 12 may execute a prediction model generation method for generating a prediction model for predicting sinter ore strength. The prediction model generation method may include a data acquisition step and a prediction model generation step. In the data acquisition step, the acquisition unit 12 acquires, as explanatory variables, data related to operating conditions in the sinter production equipment, including exhaust gas information, and, as a response variable, the sinter ore strength corresponding to the explanatory variables. These data may be extracted from actual data in the sintering process. Furthermore, in the prediction model generation step, the model generation unit may associate the explanatory variables with the response variable to generate learning data and generate the prediction model by machine learning.

[0048] The components of the sinter ore strength control device 10 are described below. The memory unit 11 stores a prediction model. The memory unit 11 also stores programs and data related to the control of sinter ore strength. The memory unit 11 may store acquired input data. The memory unit 11 may store a target range for sinter ore strength. The memory unit 11 may store various information obtained in processing for controlling sinter ore strength. The memory unit 11 may be configured to include any memory device, such as a semiconductor memory device, an optical memory device, or a magnetic memory device. The semiconductor memory device may include, for example, a semiconductor memory. The memory unit 11 may include multiple types of memory devices.

[0049] The acquisition unit 12 acquires input data. The acquisition unit 12 acquires the feature amount of the input data in consideration of a delay time based on the timing at which the strength of sintered ore can be identified.

[0050] The sintered ore strength prediction unit 13 predicts the strength of the sintered ore after a predetermined time based on the acquired input data. A prediction model is used to predict the strength of the sintered ore.

[0051] The operational action determination unit 14 determines an operational action based on the predicted sintered ore strength and a predetermined target range of the sintered ore strength.

[0052] The output unit 15 outputs a change instruction to the sintering production equipment (or the other production equipment described above) or presents a guidance operation amount based on the determined operation action. When the sinter ore strength prediction method is executed, the output unit 15 may output the predicted sinter ore strength to the display unit 30 or the like.

[0053] When the manipulated variable is output from the output unit 15 to another manufacturing facility, the facility that receives the manipulated variable may automatically update the manipulated variable to manufacture sintered ore. In addition, the operator may change the operating conditions of the sintering machine based on the guidance manipulated variable displayed on the display unit 30.

[0054] As described above, the sinter ore strength control device 10 can be realized, for example, by a computer. The computer includes, for example, a memory, a hard disk drive (storage device), a CPU (processing device), and the like. The program can be stored in the hard disk drive, and when executed by the CPU, it is read from the hard disk drive to the memory. Data during processing is stored in the memory, and if necessary, stored in the hard disk drive. The memory unit 11 can be realized, for example, by a storage device. The acquisition unit 12, the sinter ore strength prediction unit 13, the operational action determination unit 14, and the output unit 15 can be realized, for example, by a CPU executing the program.

[0055] FIG. 4 is a flowchart showing the sintered ore strength control method (sintered ore manufacturing method) according to this embodiment.

[0056] The acquisition unit 12 acquires input data (step S1, acquisition step). The sinter strength prediction unit 13 predicts the sinter strength after a predetermined time using the input data and a prediction model (step S2, sinter strength prediction step). The operational action determination unit 14 determines an operational action based on the predicted sinter strength and a predetermined target range of sinter strength (step S3, operational action determination step). The output unit 15 outputs a change instruction based on the determined operational action (step S4, output step). In the sinter ore manufacturing method, sinter is manufactured in accordance with the change instruction. Here, in the output step, a guidance operation amount may be presented instead of outputting the change instruction. Furthermore, when the sinter ore strength prediction method is executed, the operational action determination step may be omitted, and the predicted sinter ore strength may be output in the output step.

[0057] (Example 1) A specific example (Example) in which the optimal manipulated variable manipulated value is determined by the above method will be described below. In Example 1, in a sintering line, the strength of sintered ore 1.5 hours ahead was predicted using a machine learning model with a neural network. The input data included the raw material brand, raw material moisture, water spray flow rate in the granulation mixer, coagulant blending ratio, quicklime blending ratio, iron ore blending ratio, return ore blending ratio, exhaust gas NOx, and the like. ｘ Concentration, exhaust gas O ２ concentration, exhaust gas CO concentration, exhaust gas CO ２ The timing at which the strength of sintered ore can be determined was based on the time when the exhaust gas analysis value was acquired, and the data acquired taking into account the delay time was used as input data to predict the strength of sintered ore. ｘThe delay time for exhaust gas information such as concentration was 1.5 hours, and the prediction time was determined based on this delay time. The predicted sinter strength value 1.5 hours ahead was displayed on a display visible to the operator. The sinter strength was measured using a shutter strength test. The target range for sinter strength was set to 91% to 92%. If the predicted sinter strength value fell below 91%, an action was taken to increase the coagulant blend ratio. If the predicted sinter strength value exceeded 92%, an action was taken to decrease the coagulant blend ratio. The coagulant blend ratio manipulation amount per operation was fixed at 0.04%, and operations (operational actions) were performed every two hours. The operator, upon viewing the displayed guidance manipulation amount, judged the appropriateness and decided to execute the operation. As shown in Figure 5, in conventional operation (operation in which the operator makes adjustments based on experience), the sinter strength variation was 2.0% after 1,000 hours of operation. The sinter strength variation is the standard deviation. By applying the method of this embodiment, the variation in sinter ore strength after 1,000 hours of operation was 1.1%. By applying the method of this embodiment, it became possible to reduce the variation in sinter ore strength. Here, the time interval (2 hours in this embodiment) for outputting a change instruction to the sinter manufacturing equipment based on an operational action is set to be longer than the delay time of the manipulated variable used to predict the sinter ore strength. Because the manipulated variable used for the prediction is past data, the time interval is set so that the next operational action is performed after the change in the manipulated variable is reflected in the predicted value. Here, the time interval for outputting a change instruction to the sinter manufacturing equipment corresponds to the control period or the manipulated variable change period.

[0058] In Example 2, the strength of sintered ore 1.5 hours later was predicted using a machine learning model based on a neural network in a sintering line. The input data included the raw material brand, raw material moisture content, water spray flow rate in the granulation mixer, coagulant blending ratio, quicklime blending ratio, iron ore blending ratio, return ore blending ratio, exhaust gas NOx, and the like. ｘ Concentration, exhaust gas O ２ concentration, exhaust gas CO concentration, exhaust gas CO ２The timing at which the strength of sintered ore can be determined was based on the time when the exhaust gas analysis value was acquired, and the data acquired taking into account the delay time was used as input data to predict the strength of sintered ore. ｘ The delay time for exhaust gas information such as concentration was 1.5 hours, and the prediction time was determined based on this delay time. The predicted sinter strength value 1.5 hours ahead was displayed on a display visible to the operator. The sinter strength was measured using a shutter strength test. The target range for sinter strength was set to 91% to 92%. If the predicted sinter strength value fell below 91%, an action to increase the coagulant blend ratio was implemented. If the predicted sinter strength value exceeded 92%, an action to decrease the coagulant blend ratio was implemented. The coagulant blend ratio manipulation amount was determined by inverse analysis based on the calculation results of the machine learning model, and an operation (operational action) was performed every two hours. The operator, upon viewing the presented guidance manipulation amount, judged whether it was acceptable and decided to execute it. As shown in Figure 6, in conventional operation (operation in which the operator makes adjustments based on experience), the sinter strength variation was 2.0% after 1,000 hours of operation. The sinter strength variation is the standard deviation. By applying the method of this embodiment, the variation in sinter strength was 0.9% after 1000 hours of operation. By applying the method of this embodiment, it became possible to reduce the variation in sinter strength. Here, the method of back analysis is not particularly limited, but the bisection method was used in this example.

[0059] (Sintering Operation Guidance System) FIG. 7 is a diagram showing the configuration of the sintering operation guidance system according to this embodiment. The sintering operation guidance system may be configured, for example, as shown by the dashed line in FIG. 7 , with a sintering operation guidance server 40 and a terminal device 50. The sintering operation guidance server 40 is a computer that further includes a communication unit 16 in addition to the functions of the sinter ore strength control device 10. The terminal device 50 functions as the display unit 30. The terminal device 50 also functions as an operation action change input unit 51 that acquires action change inputs from an operator according to the displayed content, and a terminal communication unit 52. The terminal communication unit 52 receives guidance operation variables, operation performance values, and operation management indexes calculated by the sintering operation guidance server 40, and transmits operation variable change inputs determined based on the inputs from the operation action change input unit 51. The terminal device 50 may be realized by, for example, a mobile terminal device such as a tablet or a computer. The sintering operation guidance server 40 and the terminal device 50 can mutually transmit and receive data via a network such as the Internet. The sintering operation guidance server 40 and the terminal device 50 may be located in the same place (for example, in the same factory) or may be located physically apart. Furthermore, the sintering operation guidance system is not limited to the above configuration, and may be configured to further include, for example, an operation data server 60. The operation data server 60 is capable of communicating with the sintering operation guidance server 40 and the terminal device 50 via a network. The operation data server 60 may be located in the same place as the sintering operation guidance server 40 or the terminal device 50 or may be located physically apart.

[0060] The configuration of the sintering operation guidance server 40 is almost the same as the configuration of the sinter ore strength control device 10 in Fig. 3, but a communication unit 16 is added for transmitting and receiving data to and from the terminal device 50. That is, the sintering operation guidance server 40 includes a storage unit 11, an acquisition unit 12, a sinter ore strength prediction unit 13, an operation action determination unit 14, an output unit 15, and a communication unit 16. The output unit 15 of the sintering operation guidance server 40 may output a change instruction to a control device of the sintering machine or a process computer that controls each facility device, or may present a guidance operation amount, based on the operation action determined by the operation action determination unit 14.

[0061] The terminal device 50, together with the sintering operation guidance server 40, constitutes a sintering operation guidance system and displays guidance operation variables. The terminal device 50 includes at least a display unit 30. As described above, the display unit 30 may be a liquid crystal display or the like. The terminal device 50 may also include a terminal communication unit 52 that acquires guidance operation variables from the sintering operation guidance server 40. The terminal communication unit 52 receives and acquires, for example, operation control index data for determining the selection of operation variables in addition to the guidance operation variables from the sintering operation guidance server 40. The guidance operation variables may include the mixing ratio of the coagulant, the mixing ratio of quicklime, and the pallet speed, and may be guidance operation variables for a plurality of different operation patterns. The operation control indexes may also include the actual value of the sinter ore production rate, the target value of the sinter ore production rate, upper and lower limits of the sinter ore production rate, the actual value of each operation variable, and the upper and lower limits of each operation variable. The operation control indexes may also include time-series data such as exhaust gas temperature and exhaust gas component concentration, and time-series data of the sinter strength actual value. The display unit 30 may display the acquired guidance operation variables and operation control indicators and prompt the operator to select an operation variable. In this case, the guidance operation variables may be displayed in order of priority. For example, using the control range (upper and lower limits of the target value) of the actual sinter production rate, if the actual sinter production rate is below the target, the operation variables may be presented with priority given to the blending ratio of the coagulant and quicklime. Furthermore, if the sinter production rate exceeds the target value, operation variables including the operation variable for reducing the pallet speed may be displayed with priority given. Furthermore, the cost of the coagulant and quicklime may be added to the evaluation and reflected in the display priority order for the coagulant and quicklime. The operation action change input unit 51 inputs the operation variable selected by the operator using an input device such as a pointer device or a keyboard from the multiple displayed guidance operation variables into the terminal device 50. If no input is received within a predetermined time after displaying a new operation variable, the operation action change input unit 51 may execute a process of inputting the operation variable with the highest priority as input information. The terminal communication unit 52 further transmits the final operation amount, which is input to the terminal device 50 by the operation action change input unit 51, to the sintering operation guidance server 40.The operation action determination unit 14 of the sintering operation guidance server 40 may output the final operation amount to the control device of each facility. As another example, the terminal communication unit 52 of the terminal device 50 may directly transmit the final operation amount to the control device of each facility.

[0062] Here, the configuration of the sintering operation guidance system in FIG. 7 is one example. For example, the sintering operation guidance system may be configured to include at least one of the sintering operation guidance server 40 and the terminal device 50, or may be configured to include another information processing device. Also, some of the functional units included in the sintering operation guidance server 40 and the functional units included in the terminal device 50 may be omitted. For example, the sintering operation guidance system may be configured only with the sintering operation guidance server 40 having the display unit 30. In this configuration, the guidance operation variable acquisition unit may be omitted. That is, the sintering operation guidance system may include the acquisition unit 12, the sinter ore strength prediction unit 13, the operation action determination unit 14, the output unit 15, and the display unit 30. In this configuration, the communication unit 16 and the terminal communication unit 52 may be omitted. Also, some of the functional units included in the sintering operation guidance server 40 in the example of FIG. 7 may be included in the terminal device 50. 7, some of the functional units included in the terminal device 50 may be included in the sintering operation guidance server 40. For example, the sintering operation guidance system may be configured so that the operation action determination unit 14 and the output unit 15 are included in the terminal device 50 instead of the sintering operation guidance server 40.

[0063] As described above, the sinter ore strength control device 10, sinter ore strength control method, sinter ore manufacturing method, prediction model generation method, sinter ore strength prediction method, sintering operation guidance system, and terminal device 50 according to the present embodiment can predict sinter ore strength with high accuracy, as is clear from the above examples. Furthermore, based on the highly accurate prediction of sinter ore strength, the sinter ore strength can be controlled to be within a target range. Furthermore, by suppressing the variation in sinter ore strength, the variation in the reduction degradation index (RDI) during blast furnace operation can also be reduced, making it possible to avoid deterioration of furnace conditions.

[0064] Although embodiments of the present disclosure have been described based on the drawings and examples, it should be noted that those skilled in the art would easily be able to make various modifications or alterations based on the present disclosure. Therefore, it should be noted that these modifications and alterations are included within the scope of the present disclosure. For example, functions included in each component or step can be rearranged so as not to cause logical inconsistencies, and multiple components or steps can be combined or divided into one. The embodiments of the present disclosure can also be realized as a program executed by a processor included in an apparatus or a storage medium on which a program is recorded. It should be understood that these are also included within the scope of the present disclosure.

[0065] The configuration of the sintered ore strength control device 10 shown in Figure 3 is one example. The sintered ore strength control device 10 does not need to include all of the components shown in Figure 3. Furthermore, the sintered ore strength control device 10 may include components other than those shown in Figure 3. For example, the sintered ore strength control device 10 may be configured to further include a display unit 30.

[0066] REFERENCE SIGNS LIST 10 Sintered ore strength control device 11 Memory unit 12 Acquisition unit 13 Sintered ore strength prediction unit 14 Operation action determination unit 15 Output unit 16 Communication unit 30 Display unit 40 Sintering operation guidance server 50 Terminal device 51 Operation action change input unit 52 Terminal communication unit 60 Operation data server

Claims

1. An acquisition unit that acquires, as input data, data related to operating conditions in a sintering production facility including exhaust gas information; a sintered ore strength prediction unit that predicts the strength of sintered ore after a predetermined time based on the acquired input data; an operating action determination unit that determines an operating action based on the predicted sintered ore strength and a target range of the predetermined sintered ore strength; and an output unit that outputs a change instruction to the sintering production facility or presents a guidance operation amount based on the determined operating action. A sintered ore strength control device comprising:

2. The exhaust gas information is exhaust gas NO ｘ concentration, exhaust gas O ２ The sintered ore strength control device according to claim 1, including at least one of the concentration and the exhaust gas temperature.

3. The exhaust gas information includes at least one of an exhaust gas CO concentration, an exhaust gas CO ２ concentration, and an exhaust gas SO ｘ concentration. The sintered ore strength control device according to claim 2.

4. The sintered ore strength control device according to any one of claims 1 to 3, wherein the input data includes at least one of a blending ratio of a binder and a pallet speed.

5. The sintered ore strength control device according to any one of claims 1 to 4, wherein the input data includes at least one of a raw material brand, raw material moisture, watering flow rate in a granulating mixer, blending ratio of quicklime, blending ratio of iron ore, blending ratio of return ore, cooler blowing pressure, cooler exhaust gas temperature, layer thickness, and production amount.

6. The sintered ore strength control device according to any one of claims 1 to 5, wherein the sintered ore strength prediction unit predicts the sintered ore strength using, as the input data, data obtained in consideration of a delay time based on a timing at which the sintered ore strength can be specified.

7. The sintered ore strength control device according to claim 6, wherein a time interval for outputting a change instruction to the sintering production facility based on the operating action is set to be greater than a delay time traced back from a prediction time when the operating variable is past data.

8. An acquisition step of acquiring, as input data, data related to operating conditions in a sintering production facility including exhaust gas information; a sintered ore strength prediction step of predicting the strength of sintered ore after a predetermined time based on the acquired input data; an operating action determination step of determining an operating action based on the predicted sintered ore strength and a target range of the predetermined sintered ore strength; and an output step of outputting a change instruction to the sintering production facility or presenting a guidance operation amount based on the determined operating action. A sintered ore strength control method including:

9. A method for manufacturing sintered ore, comprising manufacturing sintered ore according to the change instruction output to the sintering production facility in the sintered ore strength control method according to claim 8.

10. A method for generating a prediction model for predicting the strength of sintered ore, the method comprising: obtaining, as explanatory variables, data related to operating conditions in a sintering production facility including exhaust gas information, and, as a target variable, the strength of sintered ore corresponding to the explanatory variables; and associating the explanatory variables with the target variable to obtain learning data, and generating the prediction model by machine learning.

11. A method for predicting the strength of sintered ore, the method comprising: an acquisition step of obtaining, as input data, data related to operating conditions in a sintering production facility including exhaust gas information; a step of predicting the strength of sintered ore after a predetermined time based on the obtained input data; and an output step of outputting the predicted strength of sintered ore.

12. A sintering operation guidance system comprising: an acquisition unit that obtains, as input data, data related to operating conditions in a sintering production facility including exhaust gas information; a sintered ore strength prediction unit that predicts the strength of sintered ore after a predetermined time based on the obtained input data; an operation action determination unit that determines an operation action based on the predicted strength of sintered ore and a target range of the strength of sintered ore determined in advance; an output unit that presents a guidance operation amount based on the determined operation action; and a display unit that displays the obtained guidance operation amount.

13. A terminal device that constitutes a sintering operation guidance system together with a sintering operation guidance server, the terminal device comprising: a terminal communication unit that obtains, as input data, data related to operating conditions in a sintering production facility including a plurality of operation amounts and exhaust gas information, and uses a prediction model for predicting the strength of sintered ore to obtain guidance operation amounts for a plurality of operations calculated such that the predicted strength of sintered ore after a predetermined time approaches a predetermined target value, and obtains operation management indices related to the production amount of sintered ore as a plurality of operation patterns; a display unit that assigns priorities to the obtained guidance operation amounts and displays them based on the guidance operation amounts and the operation management indices; and an operation action change input unit that enables an operator to select the displayed guidance operation amounts.

Images