Sintered ore manufacturing apparatus and method for manufacturing sintered ore

Abstract

The sintered ore manufacturing apparatus (10) includes an acquisition unit (12) that acquires data related to operating conditions in a sintered ore manufacturing facility as input data, a sintered ore FeO ratio prediction unit (13) that predicts the sintered ore FeO ratio after a predetermined time based on the acquired input data, an operation variable calculation unit (14) that calculates an operation variable for a mixing ratio of an agglomerating agent based on the predicted sintered ore FeO ratio and a predetermined target range for the sintered ore FeO ratio, and an output unit (15) that outputs the calculated operation variable to other manufacturing facilities or presents it as a guidance operation variable.

Description

[Technical field]

[0001] The present disclosure relates to an apparatus for producing sintered ore and a method for producing sintered ore. [Background technology]

[0002] In the steel industry, the quality of iron ore has declined due to years of mining. As a result, the proportion of fine ore with a high fineness that has been dressed at the mine is increasing, and the sintering process, in which the fine ore is solidified to produce sintered ore before being charged into a blast furnace, is becoming increasingly important.

[0003] Figure 1 shows an overview of the sintering process. Sintered ore is produced by sintering small-grained iron ore with the heat of combustion of the agglomeration agent to be used as the iron source for the blast furnace. The raw material for sintered ore is stored in a hopper and is cut out from the hopper. If powdered iron ore is directly charged into the sintering machine, the combustion reaction is suppressed due to poor ventilation. Therefore, in the granulation process, iron ore is mixed with water in a granulation mixer along with other raw materials such as quicklime and coke, and processed into granules with larger grain sizes than the original raw materials. The granules are charged into the sintering machine, ignited in an ignition furnace, and the combustion reaction progresses gradually in layers from top to bottom due to air suction from below. 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 duct of the cooler, it is sorted by a sieve, and the large grains are sent to the blast furnace as good products ("product" in Figure 1). Sintered ore with small particle size (for example, 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 back into the sinter machine.

[0004] Here, since sintered ore accounts for a high proportion of the charge to the blast furnace (approximately 80% as an example), the properties of sintered ore have a large impact on blast furnace operation. In particular, the reduction disintegration index (RDI), which is an index of the ease with which sintered ore disintegrates when reduced at low temperatures, is important. The higher the RDI, the more easily sintered ore disintegrates, which causes deterioration of the furnace condition due to poor ventilation and non-uniform gas flow. The RDI is closely related to the FeO ratio in sintered ore (hereinafter, sintered ore FeO ratio). For this reason, operations are carried out to reduce the variation in RDI by controlling the sintered ore FeO ratio within a target range using the mixing ratio of agglomeration materials (carbonaceous materials such as fine coke) as an operational variable.

[0005] The FeO ratio of sintered ore is measured by chemical analysis, but the measurement takes time. For example, it may take 3 hours to cool in a cooler and take samples, and 2 hours to analyze, so the total measurement time may be about 5 hours. In this example, if the FeO ratio of sintered ore increases, it can only be detected after 5 hours. In other words, corrective action is delayed by 5 hours. The same is true if the FeO ratio of sintered ore decreases, and it takes about 5 hours to detect the decrease in the FeO ratio of sintered ore, so action is delayed. In order to detect and control fluctuations in the FeO ratio of sintered ore early, it is important to predict the FeO ratio of sintered ore and take action in advance or shorten the measurement time.

[0006] In relation to shortening the measurement time, a method for measuring the FeO content of sintered ore by detecting the change in inductance with a coil is conventionally known. However, the method for detecting the change in inductance with a coil has a large error and generally has low measurement accuracy. Here, it is known that the higher the FeO content of sintered ore, the higher the strength of sintered ore tends to be. For example, 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 the minerals that constitute sintered ore. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Application Publication No. 7-11349 Summary of the Invention [Problem to be solved by the invention]

[0008] However, the method of Patent Document 1 requires random testing to identify physical properties, and cannot continuously predict the strength of sintered ore.

[0009] In view of the above circumstances, an object of the present disclosure is to provide a sintered ore manufacturing apparatus and a sintered ore manufacturing method that can predict the sintered ore FeO ratio with high accuracy and control it to be within a target range. [Means for solving the problem]

[0010] (1) An apparatus for producing sintered ore according to an embodiment of the present disclosure, An acquisition unit that acquires data on operating conditions in a sintering manufacturing facility as input data; A sintered ore FeO ratio prediction unit that predicts a sintered ore FeO ratio after a predetermined time based on the acquired input data; An operation amount calculation unit that calculates an operation amount of a mixture ratio of a coagulant based on the predicted sintered ore FeO ratio and a predetermined target range of the sintered ore FeO ratio; The control unit includes an output unit that outputs the calculated manipulated variable to another manufacturing facility or presents the calculated manipulated variable as a guidance manipulated variable.

[0011] (2) As an embodiment of the present disclosure, in (1), The input data are the raw material brand, raw material moisture content, water spray flow rate in the granulation mixer, coagulant blend ratio, quicklime blend ratio, iron ore blend ratio, return ore blend ratio, exhaust gas NO x concentration, exhaust gas O2 concentration, exhaust gas CO concentration, exhaust gas CO2 concentration, exhaust gas SO x At least one of the concentration, exhaust gas temperature, cooler blower pressure, cooler exhaust gas temperature, pallet speed, layer thickness, and production volume is included.

[0012] (3) As an embodiment of the present disclosure, in (1) or (2), The sintered ore FeO percentage prediction unit predicts the sintered ore FeO percentage using, as the input data, data acquired in consideration of a delay time based on a timing at which the sintered ore FeO percentage can be identified.

[0013] (4) As an embodiment of the present disclosure, in any one of (1) to (3), The present invention further includes a learning unit that evaluates a prediction error of the predicted sintered ore FeO ratio and performs additional learning or re-learning of a prediction model used in the sintered ore FeO ratio prediction unit according to a predetermined criterion.

[0014] (5) A method for producing sintered ore according to one embodiment of the present disclosure includes: An acquisition step of acquiring data on operating conditions in a sintering manufacturing facility as input data; A sintered ore FeO ratio prediction step of predicting a sintered ore FeO ratio after a predetermined time based on the acquired input data; A manipulation amount calculation step of calculating a manipulation amount of a mixture ratio of a coagulant based on the predicted sintered ore FeO ratio and a predetermined target range of the sintered ore FeO ratio; The method further includes an output step of outputting the calculated manipulated variable to another manufacturing facility or presenting the calculated manipulated variable as a guidance manipulated variable. Effect of the Invention

[0015] According to the present disclosure, it is possible to provide a sintered ore manufacturing apparatus and a sintered ore manufacturing method that can predict the sintered ore FeO ratio with high accuracy and control it to be within a target range. [Brief description of the drawings]

[0016] [Figure 1] FIG. 1 is a diagram showing an outline of the sintering process. [Diagram 2] FIG. 2 is a diagram showing the error of the predicted value of the prediction model according to the number of explanatory variables. [Diagram 3]FIG. 3 is a diagram illustrating a configuration example of a sintered ore manufacturing apparatus according to an embodiment of the present disclosure. [Figure 4] FIG. 4 is a flow chart illustrating a method for producing sintered ore according to one embodiment of the present disclosure. [Diagram 5] FIG. 5 shows the results of Example 1. [Figure 6] FIG. 6 shows the results of Example 2. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, a sintered ore manufacturing apparatus 10 (see FIG. 3) and a sintered ore manufacturing method according to an embodiment of the present disclosure will be described with reference to the drawings. In summary, the sintered ore manufacturing method according to this embodiment predicts the sintered ore FeO ratio in real time using a machine learning model. The machine learning model is a trained model generated by machine learning. Then, based on the predicted value, an operation amount of the agglomeration agent blend ratio is calculated and output so that the sintered ore FeO ratio falls within a target range, thereby controlling the sintered ore FeO ratio to an appropriate range and performing appropriate operational actions.

[0018] In the method for producing sintered ore according to the present embodiment, the sintered ore FeO ratio is predicted using a prediction model. The sintered ore FeO ratio is the ratio of FeO to all components in the sintered ore, and may be indicated by "%" as in the present embodiment. In the following description, "%" means mass %. In the prediction model, data on the operating conditions in a facility (sintering production facility) that produces sintered ore, including a sintering machine and a return ore hopper, is input as input data. The prediction model uses the input data as explanatory variables and the sintered ore FeO ratio as a response variable. Then, the sintering process performed in the sintering production facility can be controlled based on the predicted value of the sintered ore FeO ratio. For example, the sintering production facility can perform an operation action automatically or through the operation of an operator so that the sintered ore FeO ratio is within a predetermined target range in the sintering process.

[0019] Input data are raw material brand, raw material moisture, water flow rate in the granulation mixer, coagulant blend ratio, quicklime blend ratio, iron ore blend ratio, return ore blend ratio, exhaust gas NO x concentration, exhaust gas O2 concentration, exhaust gas CO concentration, exhaust gas CO2 concentration, exhaust gas SO x The input data includes at least one of the following: concentration, exhaust gas temperature, cooler air pressure, cooler exhaust gas temperature, pallet speed, layer thickness, and production amount. For example, the input data may include all of these feature quantities, and other operation factors may be used in combination. Here, the raw material brand and raw material moisture are 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 together with other raw materials such as quicklime and coke, and the spray flow rate is the flow rate when the mixed water is sprayed. The blending ratio of the coagulating agent, the blending ratio of quicklime, the blending ratio of iron ore, and the blending ratio of return ore are the blending ratios of the coagulating agent, quicklime, iron ore, and return ore in the raw material for sintered ore, respectively. The coagulating agent is, for example, a carbonaceous material such as fine coke. Exhaust gas NO x concentration, exhaust gas O2 concentration, exhaust gas CO concentration, exhaust gas CO2 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 blowing 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. Additionally, the production volume is the amount of sintered ore produced.

[0020] Here, among the above features, exhaust gas NO X The following are particularly important: exhaust gas NO concentration, exhaust gas O2 concentration, exhaust gas temperature, and the mixing ratio of the coagulant. X It is preferable that the composition includes the concentration, exhaust gas O2 concentration, exhaust gas temperature, and the mixing ratio of the coagulant. As the firing proceeds, O2 is consumed. Then, the CO partial pressure increases and NO Xis consumed, reduction proceeds and the FeO ratio of sintered ore increases. Also, the higher the firing temperature, the higher the exhaust gas temperature, so it is possible to predict the FeO ratio of sintered ore based on the exhaust gas temperature. If the mixing ratio of the agglomeration agent decreases, the FeO ratio of sintered ore decreases. Also, if the mixing ratio of the agglomeration agent increases, the FeO ratio of sintered ore increases.

[0021] The prediction model is not limited to a specific one as long as it is configured to obtain the objective variable (sintered ore FeO ratio) from the 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 are not particularly limited. The prediction model is generated, for example, using actual data (past measured values ​​and past set values, etc.) in the sintering process before the sintered ore FeO ratio is predicted.

[0022] Figure 2 shows an example of evaluating a prediction model generated by the neural network technique. The number (type) of explanatory variables in machine learning was added, and the error between the predicted value of the sinter FeO ratio by each machine learning model and the actual sinter FeO ratio was measured. For example, when the explanatory variables are exhaust gas NO X The prediction error of the prediction model based only on concentration is more than 0.6%. In contrast, for example, if the explanatory variable is exhaust gas NO X The prediction error of the prediction model, which is the concentration, exhaust gas O2 concentration, exhaust gas temperature, coagulant mixing ratio, and cooler exhaust gas temperature, is 0.45%. As shown above, the error tends to decrease as the number of explanatory variables is increased. Here, Figure 2 shows an example of the order in which explanatory variables are added. For example, if the explanatory variables are exhaust gas NO X The mixing ratio of the coagulant can be added after the concentration, and there is a tendency for the error to decrease as the number of explanatory variables is increased, regardless of the order of addition in Figure 2.

[0023] Here, when the prediction model is a neural network, it is desirable to use a model (prediction model using hyperparameters) optimized by adjusting hyperparameters so as to minimize the prediction error. The hyperparameters include the number of neurons, number of layers, learning rate, number of epochs, dropout rate, or type of activation function of the neural network.

[0024] 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 FeO ratio in sintered ore varies. For example, the mixing ratio of the agglomerate, which is data related to raw materials, is located upstream of the sintering process, so it can affect the FeO ratio in sintered ore after 7 hours. For example, the NO2 content of exhaust gas, which is data related to firing, x The concentration, exhaust gas O2 concentration, and exhaust gas temperature are located in the midstream of the sintering process, so they can affect the sinter FeO ratio after 5 hours. For example, the cooler air pressure and cooler exhaust gas temperature, which are data related to the cooler, are located downstream of the sintering process, so they can affect the sinter FeO ratio after 4.5 hours. The time until the sinter FeO ratio is affected can also vary depending on the pallet speed. Therefore, for accurate prediction, it is preferable that the input data of the prediction model is acquired taking into account a delay time based on the timing at which the sinter FeO ratio can be identified. Here, the timing at which the sinter FeO ratio can be identified may be, for example, the time when the process of sorting the sinter FeO ratio by a sieve after cooling in the cooler and determining whether the particle size meets the standard is executed. The operation variables that affected the sinter at this identification time may be acquired taking into account the above-mentioned delay time. Similarly, it is preferable that the prediction model is generated by machine learning using learning data in which data corresponding to the explanatory variables (input data) extracted from the actual data in the sintering process and the sinter FeO ratio extracted from the actual data are associated with each other taking into account the delay time.

[0025] Using the input data and the prediction model as described above, a process of predicting the sinter FeO ratio is executed. The sinter FeO ratio after a predetermined time (for example, 5 hours) is predicted so that corrective action is not delayed. For accurate prediction, it is preferable that the input data considers a delay time based on the timing at which the sinter FeO ratio can be specified. In addition, the manipulated variable of the sintering process is calculated so that the predicted value of the sinter FeO ratio falls within the target range. In this embodiment, if the sinter FeO ratio exceeds the target range set in advance in the future, an action is performed to reduce the mixing ratio of the coagulation agent to reduce the sinter FeO ratio. In addition, if the sinter FeO ratio falls below the target range in the future, an action is performed to increase the mixing ratio of the coagulation agent to increase the sinter FeO ratio. In the calculation of the manipulated variable, a machine learning model with the manipulated variable as an explanatory variable is constructed, and the manipulated variable that makes the predicted value of the sinter FeO ratio match the target value (for example, the median of the target range) may be calculated by inverse analysis. As another example, in the calculation of the manipulated variable, a physical model and a machine learning model may be used in combination. In this case, a simulation of changing the mixing ratio of the coagulant is performed using a physical model. Then, the future operating state (e.g., flue gas NO x The FeO ratio of sintered ore is predicted using a machine learning model with inputs such as the concentration of FeO in the sintered ore, the exhaust gas O2 concentration, etc. The mixing ratio of the agglomeration agent that matches the predicted value of the FeO ratio of sintered ore predicted in this way with the target value may be calculated by inverse analysis.

[0026] As described above, an effective action for increasing the sinter FeO ratio is to increase the mixing ratio of the agglomeration material, which is a heat source, to promote reduction. Conversely, an effective action for decreasing the sinter FeO ratio is to reduce the mixing ratio of the agglomeration material to suppress reduction. Here, if the mixing ratio of the agglomeration material is changed significantly at once, the furnace conditions may change significantly and operation may become unstable. Therefore, in order to prevent excessive operation (excessive amount of operation), a range of the amount of operation per time may be determined, and the operation may be performed multiple times within the range.

[0027] Here, the calculated manipulated variable (the mixing ratio of the coagulant in this embodiment) may be output to, for example, a process computer that manages the sintering process. Here, the output of the manipulated variable includes an output as operation guidance for an operator who operates the sintering machine. In other words, the manipulated variable may be output so that an appropriate manipulated variable is reflected in the sintering process through the judgment of the operator. The information output as operation guidance includes at least the calculated manipulated variable, and may be displayed on a display that the operator can see.

[0028] FIG. 3 is a diagram showing the configuration of a sintered ore manufacturing apparatus 10 according to this embodiment. The sintered ore manufacturing apparatus 10 is a part of a sintering production facility, and controls the sintered ore FeO ratio, including the above-mentioned sintered ore manufacturing method. Among the sintering production facilities, equipment other than the sintered ore manufacturing apparatus 10 may be referred to as other production facilities. As shown in FIG. 3, the sintered ore manufacturing apparatus 10 includes a memory unit 11, an acquisition unit 12, a sintered ore FeO ratio prediction unit 13, an operation amount calculation unit 14, and an output unit 15. The sintered ore manufacturing apparatus 10 may further include a learning unit 16. The sintered ore manufacturing apparatus 10 acquires data on the operating conditions in the sintering production facility, i.e., the above-mentioned input data, from an operation data server 60. The input data includes exhaust gas NO XThe actual values ​​of the characteristic quantities, such as the concentration, the exhaust gas O2 concentration, the exhaust gas temperature, and the mixing ratio of the coagulating agent, are included. The actual values ​​include the measured values ​​and the set values ​​of the operation variables. The sintered ore manufacturing apparatus 10 may acquire the target range of the sintered ore FeO ratio from the operation data server 60. The operation data server 60 may be realized by a computer that can communicate with the sintered ore manufacturing apparatus 10 via a network and collects data on the production of sintered ore, for example. The network is, for example, the Internet. The sintered ore manufacturing apparatus 10 executes the above process, that is, a process of predicting the sintered ore FeO ratio using the prediction model, and obtaining the operation amount of the operation variable, such as the mixing ratio of the coagulating agent, so that the future sintered ore FeO ratio is maintained within the target range. In this embodiment, the sintered ore manufacturing apparatus 10 has a function of outputting the operation amount of the operation variable to the sintered ore manufacturing equipment by the output unit 15 or a function of presenting it as a guidance operation amount. When the output unit 15 presents the guidance operation amount, the sintered ore manufacturing apparatus 10 functions as an operation guidance device. The display unit 30 displays the guidance operation amount output from the sintered ore manufacturing apparatus 10 (operation guidance device). The sintered ore manufacturing apparatus 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) 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 electro-luminescence panel (OLED). The display unit 30 may be realized by a display of a terminal device such as a smartphone or tablet. The terminal device can communicate with the sintered ore manufacturing apparatus 10 via a network. A sintering operation guidance server having the function of the sintered ore manufacturing apparatus 10 and a terminal device having the function of the display unit 30 may configure a sintering operation guidance system. The sintering operation guidance server and the terminal device may be located in the same place (for example, in the same factory) or may be located physically apart. The sintering operation guidance system may further include an operation data server 60.

[0029] Here, the prediction model may be generated by the sintered ore manufacturing apparatus 10 and stored in the memory unit 11, or may be generated by another computer and stored in the memory unit 11. When the sintered ore manufacturing apparatus 10 generates the prediction model, the apparatus may further include a model generation unit that generates the prediction model by the above-mentioned method and stores it in the memory unit 11. Here, the learning unit 16 may also function as the model generation unit.

[0030] The components of the sintered ore manufacturing apparatus 10 are described below. The memory unit 11 stores a prediction model. The memory unit 11 also stores a program and data related to the control of the sintered ore FeO ratio. The memory unit 11 may store acquired input data. The memory unit 11 may store a target range of the sintered ore FeO ratio. The memory unit 11 may store various information obtained in the process for controlling the sintered ore FeO ratio. The memory unit 11 may include any memory device such as a semiconductor memory device, an optical memory device, and 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.

[0031] The acquiring unit 12 acquires input data. It is preferable that the acquiring unit 12 acquires the feature amount of the input data in consideration of a delay time based on the timing at which the sintered ore FeO ratio can be identified.

[0032] The sintered ore FeO ratio prediction unit 13 predicts the sintered ore FeO ratio after a predetermined time based on the acquired input data. A prediction model is used to predict the sintered ore FeO ratio.

[0033] The manipulated variable calculation unit 14 calculates manipulated variables based on the predicted sintered ore FeO ratio and a predetermined target range of the sintered ore FeO ratio. The manipulated variables include the mixing ratio of the agglomeration agent.

[0034] The output unit 15 outputs the calculated operation amount to other production equipment (equipment other than the sinter ore production apparatus 10), or displays it on the display unit 30 as a guidance operation amount.

[0035] 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. The operator may change the operating conditions of the sintering machine based on the guidance manipulated variable shown on the display unit 30. When the sintered ore FeO ratio exceeds a preset target range in the future, the operator may take an action to lower the mixing ratio of the agglomeration agent to lower the sintered ore FeO ratio. When the sintered ore FeO ratio falls below the target range, the operator may take an action to raise the mixing ratio of the agglomeration agent to raise the sintered ore FeO ratio.

[0036] The learning unit 16 evaluates the error of the prediction model (i.e., the prediction error of the predicted sintered ore FeO ratio) and performs additional learning or re-learning of the prediction model according to a predetermined standard to maintain the accuracy of the prediction model.

[0037] The prediction error is the difference between the actual value and the predicted value, and can be evaluated at the time when the actual value of the sinter FeO ratio is obtained. The learning unit 16 evaluates the performance of the prediction model by monitoring the prediction error. The learning unit 16 can maintain the accuracy of the prediction model by additionally learning or re-learning the prediction model using recent data that has generated an error larger than the reference. Additional learning is additional machine learning performed on the prediction model using new learning data that is added. Re-learning is re-machine learning performed on the prediction model using the learning data used in the previous machine learning (i.e., learning data used in the past).

[0038] For example, when an error obtained by comparing an actual value with a corresponding predicted value at the time of obtaining the actual value exceeds several times (e.g., 3σ) the standard deviation (σ) of the prediction error at the time of creating the prediction model, the learning unit 16 may perform additional learning using several recent actual values ​​including the actual value as learning data. In addition, the learning unit 16 may monitor the average error and standard deviation of the prediction error based on data for a longer period, and determine thresholds for each of the average error and standard deviation. For example, when the past average error or standard deviation in a predetermined recent period exceeds the average error threshold or standard deviation threshold, the learning unit 16 may perform additional learning of the prediction model using data for that period to update the prediction model. Here, as described above, the analysis of the sintered ore FeO ratio takes about 5 hours. This time includes the time for one cycle from cutting the raw material to firing it and turning it into sintered ore. It is desirable to evaluate the error of the prediction model using data for a time longer than the time required for the analysis of the sintered ore FeO ratio.

[0039] When performing re-learning instead of additional learning, the learning unit 16 may use learning data obtained by adding data from a predetermined most recent period described in the above additional learning to the learning data used to learn the prediction model in use. In the case of re-learning, it is preferable to use data from a longer period than in the case of additional learning, and for example, data from several days (data accumulated on a daily basis) may be used. Also, for example, data from about one month may be used.

[0040] The sintered ore manufacturing apparatus 10 may be realized by, for example, a computer as described above. The computer includes, for example, a memory, a hard disk drive (storage device), and a CPU (processing device). The program may be stored in the hard disk drive, and when executed by the CPU, the program is read from the hard disk drive to the memory. Data during processing is stored in the memory, and if necessary, is stored in the hard disk drive. The storage unit 11 may be realized by, for example, a storage device. The acquisition unit 12, the sintered ore FeO ratio prediction unit 13, the operation amount calculation unit 14, the output unit 15, and the learning unit 16 may be realized by, for example, a CPU that executes a program.

[0041] 4 is a flowchart showing the method for producing sintered ore according to the present embodiment. The sintered ore production apparatus 10 calculates the manipulated variables according to the flowchart shown in FIG.

[0042] The acquisition unit 12 acquires input data (step S1, acquisition step). The sintered ore FeO ratio prediction unit 13 predicts the sintered ore FeO ratio after a predetermined time using the input data and a prediction model (step S2, sintered ore FeO ratio prediction step). The manipulated variable calculation unit 14 calculates the manipulated variable of the manipulated variable so that the predicted value of the sintered ore FeO ratio falls within a target range (step S3, manipulated variable calculation step). The output unit 15 outputs the calculated manipulated variable (step S4, output step).

[0043] Example 1 Hereinafter, specific examples (embodiments) in which the optimal manipulated variable operation amount is determined by the above method will be described. In Example 1, in a sintering line, the FeO ratio of sintered ore 5 hours ahead was predicted using a machine learning model with a neural network. The input data were the raw material brand, raw material moisture, water spray flow rate in the granulation mixer, coagulant blend ratio, quicklime blend ratio, iron ore blend ratio, return ore blend ratio, exhaust gas NOx, and the like. xThe parameters were concentration, exhaust gas O2 concentration, exhaust gas CO2 concentration, exhaust gas CO2 concentration, exhaust gas temperature, cooler air pressure, pallet speed, layer thickness, and production volume. The target range of sintered ore FeO ratio was set to 6% to 7%. When the sintered ore FeO ratio was below 6%, an action was taken to increase the mixing ratio of the coagulating agent. Also, when the sintered ore FeO ratio was higher than 7%, an action was taken to decrease the mixing ratio of the coagulating agent. The operation amount of the mixing ratio of the coagulating agent per time was fixed at 0.02%, and an operation (action) was performed every 2 hours. As shown in FIG. 5, in the conventional operation (operation in which the operator adjusts based on experience), the variation of the sintered ore FeO ratio was 0.90% after 100 hours of operation. The variation of the sintered ore FeO ratio is the standard deviation. By applying the method of this embodiment, the variation of the sintered ore FeO ratio was 0.50% after 100 hours of operation. By applying the method of this embodiment, it is possible to reduce the variation in the FeO ratio of sintered ore.

[0044] Example 2 In Example 2, the FeO ratio of sintered ore 5 hours ahead was predicted using a machine learning model based on a neural network in a sintering line. The input data were the raw material brand, raw material moisture, water spray flow rate in the granulation mixer, coagulant blend ratio, quicklime blend ratio, iron ore blend ratio, return ore blend ratio, exhaust gas NO xThe parameters were concentration, exhaust gas O2 concentration, exhaust gas CO2 concentration, exhaust gas CO2 concentration, exhaust gas temperature, cooler blowing pressure, pallet speed, layer thickness and production volume. The target range of the sintered ore FeO ratio was set to 6% to 7%. When the sintered ore FeO ratio was below 6%, an action was taken to increase the mixing ratio of the coagulating agent. Also, when the sintered ore FeO ratio was higher than 7%, an action was taken to decrease the mixing ratio of the coagulating agent. The operation amount of the coagulating agent mixing ratio was determined by inverse analysis based on the calculation results of the machine learning model. Although the inverse analysis is not limited to a specific method, in this embodiment, the input data of the machine learning model was changed and calculations were performed multiple times, and the optimal operation amount of the coagulating agent mixing ratio was searched for by the dichotomy method. An operation (action) was performed every two hours. As shown in FIG. 6, in the conventional operation (adjustment based on the operator's experience), the variation of the sintered ore FeO ratio was 0.90% after 100 hours of operation. The variation of the sintered ore FeO ratio is the standard deviation. By applying the method of this embodiment, the variation in the sintered ore FeO ratio was 0.45% after 100 hours of operation. By applying the method of this embodiment, it became possible to reduce the variation in the sintered ore FeO ratio.

[0045] As described above, the sintered ore manufacturing apparatus 10 and the sintered ore manufacturing method according to the present embodiment can predict the sintered ore FeO ratio with high accuracy and control it to be within a target range, as is clear from the above examples. By suppressing the variation in the sintered ore FeO ratio, the variation in the reduction degradation index (RDI) is also reduced in the blast furnace operation, and deterioration of the furnace condition can be avoided.

[0046] Although the embodiments of the present disclosure have been described based on the drawings and examples, it should be noted that those skilled in the art can easily make various modifications or corrections based on the present disclosure. Therefore, it should be noted that these modifications or corrections are included in the scope of the present disclosure. For example, the functions included in each component or each step can be rearranged so as not to be logically inconsistent, and multiple components or steps can be combined into one or divided. The embodiments of the present disclosure can also be realized as a program executed by a processor included in the device or a storage medium on which a program is recorded. It should be understood that these are also included in the scope of the present disclosure.

[0047] The configuration of the sintered ore manufacturing apparatus 10 shown in Fig. 3 is one example. The sintered ore manufacturing apparatus 10 does not need to include all of the components shown in Fig. 3. Furthermore, the sintered ore manufacturing apparatus 10 may include components other than those shown in Fig. 3. For example, the sintered ore manufacturing apparatus 10 may be configured to further include a display unit 30. [Explanation of symbols]

[0048] 10. Sintered ore manufacturing equipment 11 Storage section 12 Acquisition Department 13 Sinter FeO ratio prediction section 14 Operation amount calculation section 15 Output section 16 Learning Department 30 Display section 60 Operational Data Server

Claims

1. An acquisition unit that acquires data on operating conditions in a sintering manufacturing facility as input data; A sintered ore FeO ratio prediction unit that predicts a sintered ore FeO ratio after a predetermined time based on the acquired input data; An operation amount calculation unit that calculates an operation amount of a mixing ratio of a coagulant based on the predicted sintered ore FeO ratio and a predetermined target range of the sintered ore FeO ratio; An output unit that outputs the calculated operation amount to another manufacturing facility or presents it as a guidance operation amount.

2. The input data includes the raw material brand, raw material moisture content, water spray flow rate in the granulation mixer, coagulant blend ratio, quicklime blend ratio, iron ore blend ratio, return ore blend ratio, exhaust gas NO ｘ Concentration, exhaust gas O ２ concentration, exhaust gas CO concentration, exhaust gas CO ２ Concentration, exhaust gas SO ｘ The sintered ore manufacturing apparatus according to claim 1 , wherein the at least one of the concentration, the exhaust gas temperature, the cooler blowing pressure, the cooler exhaust gas temperature, the pallet speed, the layer thickness, and the production amount is included.

3. The sintered ore FeO ratio prediction unit predicts the sintered ore FeO ratio using data acquired taking into account a delay time based on a timing at which the sintered ore FeO ratio can be identified as the input data. The sintered ore manufacturing apparatus according to claim 1 or 2.

4. The sintered ore manufacturing apparatus according to claim 1 or 2, further comprising a learning unit that evaluates a prediction error of the predicted sintered ore FeO ratio and performs additional learning or re-learning of a prediction model used in the sintered ore FeO ratio prediction unit according to a predetermined criterion.

5. An acquisition step of acquiring data on operating conditions in a sintering manufacturing facility as input data; A sintered ore FeO ratio prediction step of predicting a sintered ore FeO ratio after a predetermined time based on the acquired input data; A manipulation amount calculation step of calculating a manipulation amount of a mixing ratio of a coagulant based on the predicted sintered ore FeO ratio and a predetermined target range of the sintered ore FeO ratio; and an output step of outputting the calculated manipulated variable to another manufacturing facility or presenting the calculated manipulated variable as a guidance manipulated variable.

Images