Return ore control device, return ore control method, and sintered ore manufacturing method

Abstract

A return ore control device (10), which is used in a sintering manufacturing facility and performs control pertaining to return ore, comprises: an acquisition unit (12) that acquires, as input data, data pertaining to an operation condition in the sintering manufacturing facility; a return-ore hopper level prediction unit (13) that predicts, on the basis of the acquired input data, a return-ore hoper level to be achieved after a prescribed time set in advance; a manipulation amount calculation unit (14) that calculates, on the basis of the predicted return-ore hoper level and a return-ore hopper level target range set in advance, a manipulation amount for manipulation variables selected from the operation condition and a raw-material charging condition; and an output unit (15) that outputs the calculated manipulation amount to the sintering manufacturing facility, or presents the same as a guidance manipulation amount.

Description

[Technical Field]

[0001] The present disclosure relates to a return ore control device, a return ore control method, and a method for producing sintered ore. [Background technology]

[0002] In the steel industry, the quality of iron ore has declined over many years of mining. As a result, the proportion of fine ore with a high fineness that has been dressed at the mine site 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. Sinter is produced by burning small-grain iron ore using the heat from the combustion of a coagulant to be used as the iron source for the 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 grain sizes than the original raw materials. The granules are then charged into the sintering machine, ignited in an ignition furnace, and the combustion reaction proceeds gradually from top to bottom in layers due to air suction from below. After firing, the sinter is discharged from the sintering machine, crushed in a crusher, and sent to a cooler. After being cooled by air in the cooler duct, it is sorted through a sieve, and the larger grains are sent to the blast furnace as non-defective products ("product" in Figure 1). Sintered ore with small particle sizes (for example, particle sizes 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, the return ore ratio is defined as the proportion of return ore in the sintered ore after firing. Effective methods for reducing the return ore ratio and improving yield include promoting combustion to prevent sintered ore from passing through the sinter machine unfired. For example, increasing the blending ratio of the coagulant (heat source), increasing the blending ratio of quicklime (which acts as a binder during granulation) to improve permeability, and increasing the proportion of coke fines in the upper layer to facilitate ignition in the ignition furnace are effective. Reducing the pallet speed is also effective in ensuring sufficient firing time. However, there is a trade-off between reducing the return ore ratio and productivity: increasing the blending ratio of coagulant, quicklime, and coke fines increases production costs, while reducing the pallet speed reduces production volume. Therefore, it is necessary to control the return ore ratio by setting an appropriate target value depending on the operating conditions.

[0005] The return ore rate is measured in a process that follows separation by sieving. Therefore, it takes time from the start of firing until the return ore rate can be measured, approximately two hours for example. It also takes another time (approximately 0.5 hours for example) for the return ore to be fed into the return ore hopper. Therefore, even if the return ore rate increases, it will not be detected until, for example, two hours later, meaning that corrective action will be delayed by two hours. The same applies if the return ore rate decreases, meaning that corrective action will be delayed.

[0006] In order to detect and control fluctuations in the return ore rate at an early stage, it is necessary to predict the future return ore rate and take action in advance. For example, Patent Document 1 discloses a method for predicting the return ore rate from the calcium ferrite content, slag content, pore size distribution index, and porosity, which are constituent minerals of 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] The method of Patent Document 1 requires sampling tests to identify physical properties, and is unable to continuously predict the return ore rate, making it impossible to detect fluctuations in the return ore rate at an early stage.

[0009] Furthermore, the return ore ratio fluctuates in a relatively short cycle of about one hour, for example, due to fluctuations in raw material components or firing conditions. Therefore, if an operation to lower the return ore ratio, such as increasing the blending ratio of quicklime when the return ore ratio temporarily increases, is performed, the blending ratio of quicklime may be excessively increased, resulting in increased production costs. Furthermore, if an operation to increase the return ore ratio, such as decreasing the blending ratio of quicklime when the return ore ratio temporarily decreases, is performed, the blending ratio of quicklime may be excessively decreased, resulting in an increase in the return ore ratio and reduced productivity. Furthermore, attempting to correct the hourly fluctuations in the return ore ratio requires frequent operational actions, which may destabilize the furnace conditions. Hereinafter, operational actions in response to such temporary increases or decreases in the return ore ratio, such as excessively increasing or decreasing the blending ratio of quicklime, are referred to as excessive actions.

[0010] In view of the above circumstances, an object of the present disclosure is to provide a return ore control device, a return ore control method, and a sintered ore manufacturing method that can reduce excessive action and suppress an increase in production costs. [Means for solving the problem]

[0011] The present inventors have conducted extensive research into ways to solve the above-mentioned problems and have found that controlling the return ore hopper level, instead of directly controlling the return ore rate, is effective in preventing excessive action. Since the amount of return ore corresponding to fluctuations in the return ore rate is integrated and reflected in the return ore hopper level, it is possible to indirectly manage the return ore rate by controlling the return ore hopper level to fall within a target range. Furthermore, by evaluating fluctuations in the amount of return ore as an integrated value, the effects of temporary increases or decreases in the return ore rate can be reduced.

[0012] (1) A return ore control device according to an embodiment of the present disclosure includes: A return ore control device used in a sintering production facility for controlling return ore, an acquisition unit that acquires data regarding operating conditions of the sintering manufacturing facility as input data; a return ore hopper level prediction unit that predicts a return ore hopper level after a predetermined time based on the acquired input data; an operation amount calculation unit that calculates an operation amount of an operation variable selected from the operation conditions and the raw material charging conditions based on the predicted return ore hopper level and a predetermined target range of the return ore hopper level; and an output unit that outputs the calculated manipulated variable to the sintering manufacturing equipment or presents it as a guidance manipulated variable.

[0013] (2) As one embodiment of the present disclosure, in (1), The input data includes the actual measurement value of the return ore hopper level and the amount of return ore cut out, and further includes 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 NO x The parameters include at least one of concentration, exhaust gas CO concentration, exhaust gas CO2 concentration, exhaust gas O2 concentration, exhaust gas temperature, cooler blowing pressure, pallet speed, layer thickness, and production amount.

[0014] (3) As an embodiment of the present disclosure, in (1) or (2), the return ore hopper level prediction unit predicts the return ore hopper level using a prediction model, The prediction model is generated by machine learning using learning data in which data corresponding to the input data extracted from performance data in a sintering process corresponds to the return ore hopper level extracted from the performance data, taking a delay time into consideration.

[0015] (4) A return ore control method according to an embodiment of the present disclosure includes: A return ore control method used in a sintering production facility, which controls return ore, an acquisition step of acquiring data on operating conditions of the sintering manufacturing facility as input data; a return ore hopper level prediction step of predicting a return ore hopper level after a predetermined time based on the acquired input data; an operation amount calculation step of calculating an operation amount of an operation variable selected from the operation conditions and the raw material charging conditions based on the predicted return ore hopper level and a predetermined target range of the return ore hopper level; and an output step of outputting the calculated manipulated variable to the sintering manufacturing equipment or presenting it as a guidance manipulated variable.

[0016] (5) A method for producing sintered ore according to an embodiment of the present disclosure includes: (4) Sintered ore is produced using the operation amount output to the sinter production facility by the return ore control method. [Effects of the Invention]

[0017] According to the present disclosure, it is possible to provide a return ore control device, a return ore control method, and a sintered ore manufacturing method that can reduce excessive action and suppress an increase in production costs. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 1 is a diagram showing an outline of the sintering process. [Figure 2] FIG. 2 is a diagram showing the error of the predicted value of the prediction model depending on the number of explanatory variables. [Figure 3] FIG. 3 is a diagram illustrating a configuration example of a return ore control device according to an embodiment of the present disclosure. [Figure 4] FIG. 4 is a flowchart showing a return ore control method according to one embodiment of the present disclosure. [Figure 5] FIG. 5 shows the results of Example 1. [Figure 6] FIG. 6 shows the results of Example 2. DETAILED DESCRIPTION OF THE INVENTION

[0019] Hereinafter, a return ore control device, a return ore control method, and a sinter ore manufacturing method according to an embodiment of the present disclosure will be described with reference to the drawings. The return ore control method according to this embodiment generally predicts a return ore hopper level 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 variable is calculated and output so as to bring the return ore hopper level within a target range, thereby controlling the return ore hopper level to an appropriate range and executing an appropriate operational action.

[0020] In the return ore control method according to this embodiment, a prediction model is used to predict the return ore hopper level. As shown in FIG. 1 , the return ore hopper level is the amount of return ore returned and charged into the return ore hopper, and may be expressed, for example, as a percentage (%) of the maximum amount of the return ore hopper, as in this embodiment. The return ore hopper level is not limited to a percentage, and may be expressed, for example, as a height from the bottom of the return ore hopper (height level) or as a volume of the return ore.

[0021] The return ore hopper level is estimated by predicting the return ore hopper level increase / decrease amount at a predetermined future time based on the current return ore hopper level for each control cycle of the return ore hopper level. The return ore hopper level increase / decrease amount is calculated based on the predicted value Y p and the amount of return ore cut from the hopper, u, can be predicted using the following equation (1).

[0022] H p (t+k)=H(t)+Y p (t+k)-u(t+k) … (1)

[0023] Here, H(t) is the actual measurement value of the return ore hopper level at time t. If time t is the current time, H(t) corresponds to the current return ore hopper level. H p(t+k) is the predicted value of the return ore hopper level at time t+k, which is a predetermined time k ahead of time t. p (t+k) is a predicted value of the amount of return ore generated from time t to time t+k, a predetermined time k from time t. u(t+k) is the amount of return ore cut from the return ore hopper from time t to time t+k, a predetermined time k from time t. Here, u(t+k), the amount of return ore cut from the return ore hopper, may be calculated assuming that the actual value of the return ore cut amount at time t is maintained. The amount of return ore cut is not frequently corrected or changed. Therefore, assuming that the current actual value is maintained often does not cause a significant problem in prediction accuracy. Furthermore, even if the amount of return ore cut is corrected or changed, the correction is reflected in the timing of the next control cycle after the correction or change, so it does not significantly affect prediction accuracy.

[0024] Y is the predicted value of the amount of return ore generated from time t to future time t+k p A predictive model may be used to calculate (t+k), where Y p Sometimes, the term "predicted amount of return ore generated" is used without specifying (t+k). Time k is determined depending on the equipment structure or the transport time of raw materials. For example, time k can be determined taking into account the time from the start of firing until the return ore rate is measured (e.g., 2 hours) or the time until the return ore is fed into the return ore hopper (e.g., 0.5 hours).

[0025] The prediction model receives as input data data relating to the operating conditions of a facility for producing sinter (sinter production facility), which includes a sinter machine, a return ore hopper, and the like. The prediction model uses this input data as explanatory variables and the amount of return ore generated from time t to time t+k as a response variable. A predicted value for the return ore hopper level is obtained from the predicted value for the amount of return ore generated, the return ore hopper level at the time of prediction, and the amount of return ore discharged from the return ore hopper. Then, the sintering process performed in the sinter production facility can be controlled based on the predicted value for the return ore hopper level. For example, the sinter production facility can automatically or via an operator's operation take operational action during the sintering process to keep the return ore hopper level within a predetermined target range.

[0026] The input data for the prediction model of the return ore generation amount are the raw material brand [id], raw material moisture [%], water spray flow rate in the granulation mixer [ton / hr], coagulant blending ratio [%], quicklime blending ratio [%], iron ore blending ratio [%], return ore blending ratio [%], and exhaust gas NO x The input data includes at least one of the following: concentration [%], flue gas CO concentration [%], flue gas CO2 concentration [%], flue gas O2 concentration [%], flue gas temperature [°C], cooler air pressure [kPa], pallet speed [m / min], bed thickness [mm], and production volume [ton / hr]. For example, the input data may include all of these feature quantities, or other operational factors may be used in combination. Here, the raw material brand and raw material moisture refer to the brand of iron ore in the raw material and the moisture content of the raw material, respectively. Raw material brands are classified into multiple groups based on their raw material composition, and are labeled according to which group they belong to. Each label is treated as weight data corresponding to that label. The raw material moisture indicates the moisture content of each raw material by weight ratio. As mentioned above, iron ore is mixed with water in a granulation mixer along with other raw materials such as quicklime and coke. The water spray flow rate refers to the flow rate at which the water is sprayed during mixing. The mixing ratio of the coagulant, the mixing ratio of quicklime, the mixing ratio of iron ore, and the mixing ratio of return ore are the mixing ratios of the coagulant, quicklime, iron ore, and return ore in the raw materials for sintered ore, respectively. xThe concentration, exhaust gas CO concentration, exhaust gas CO2 concentration, exhaust gas O2 concentration, and exhaust gas temperature are the concentrations and temperatures of each component of the exhaust gas generated during sintering. The cooler air pressure is the pressure of the air sent by the cooler when cooling the sintered ore. The pallet speed is the transport speed of the pallet that transports the raw materials during sintering. 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. Specifically, the production volume may be determined based on the weight and transport speed measured by a Merrick-type conveyor scale or the like installed on the outlet side of the sintering machine.

[0027] Here, among the above feature quantities, the cooler air pressure and exhaust gas NO X The concentration, blending ratio of return ore and production volume are particularly important. Input data include at least cooler blast pressure, exhaust gas NO X It is preferable that the return ore ratio is configured to include the concentration, the blending ratio of return ore, and the production volume. The higher the return ore ratio, the finer the particle size of the sintered ore becomes, which worsens ventilation and reduces the air suction pressure when cooling the sintered ore in the cooler (increases negative pressure). Therefore, the higher the cooler blast pressure, the higher the return ore hopper level. Here, the sensor for measuring the cooler blast pressure may be any sensor that can measure pressure, and for example, a strain gauge type, metal gauge type, semiconductor gauge type, or semiconductor diaphragm type pressure gauge can be used. Furthermore, when there is insufficient heat, the CO partial pressure drops, causing the exhaust gas NO x Therefore, the concentration of NO in the exhaust gas increases. x As the concentration increases, the return ore rate increases, and the return ore hopper level also increases. As production increases, the amount of return ore increases, and the return ore hopper level also increases. As the blending ratio of return ore increases, the amount of return ore cut out of the return ore hopper increases, and the return ore hopper level also decreases.

[0028] The prediction model is not limited to a specific one as long as it is configured to obtain the objective variable (the amount of return ore generation) from the explanatory variables. The prediction model may be, for example, a physical model or a machine learning model. In the present embodiment, the prediction model is generated by machine learning. Examples of machine learning techniques that can be used include linear regression, neural networks, decision trees, GBDT (Gradient Boosting Decision Tree), and random forests, and are not particularly limited. Here, the prediction model may be a model that directly predicts the return ore hopper level. In this case, a learning model may be constructed using the actual measured value of the return ore hopper level and the return ore cut amount as essential input data in addition to the explanatory variables of the prediction model for the return ore generation amount. Furthermore, a model that estimates the future return ore hopper level may be constructed using the explanatory variables of the prediction model for the cut amount and the return ore generation amount based on the actual measured value of the return ore hopper level using a hierarchical Bayesian estimation technique. The prediction model is generated using, for example, actual data (such as past measured values ​​and past set values) from the sintering process before the return ore hopper level is predicted.

[0029] Figure 2 shows an example of an evaluation of a prediction model generated by a neural network technique. The number (type) of explanatory variables in machine learning was added, and the error between the predicted value of the return ore generation amount by each machine learning model and the actual return ore generation amount was measured. For example, the prediction error of a prediction model whose only explanatory variable is cooler blast pressure is 7.3%. Here, the error was evaluated using the mean absolute error (MAE). In contrast, when the explanatory variables are, for example, cooler blast pressure and exhaust gas NO x The prediction error using the prediction model for concentration, production volume, and return fine blending ratio was 3.2%. As such, a tendency was observed for the error to decrease as the number of explanatory variables increased. Figure 2 shows an example of the order in which explanatory variables can be added; for example, the return fine blending ratio can be added after the cooler blast pressure as an explanatory variable. Regardless of the addition order in Figure 2, a tendency for the error to decrease as the number of explanatory variables is increased.

[0030] 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 return ore hopper level, i.e., the time it takes to affect the amount of return ore to be fed into the return ore hopper, varies. For example, the blending ratio of quicklime, which is data related to raw materials, is located upstream of the sintering process, so it can affect the return ore hopper level after 3.5 hours. For example, the NOx content of exhaust gas, which is data related to firing, can affect the return ore hopper level after 3.5 hours. x The concentration of sintered ore is located midstream in the sintering process, so it can affect the return ore hopper level after 2.5 hours. For example, the cooler blast pressure, which is data related to the cooler, is located downstream in the sintering process, so it can affect the return ore hopper level after 1.5 hours. Furthermore, the time until the return ore generation amount is affected can 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 return ore hopper level can be determined. Here, the timing at which the return ore hopper level can be determined is, for example, the time when the return ore hopper level is transported to and charged into the return ore hopper after being cooled in the cooler and sorted through a sieve to determine whether the particle size is below the standard. The operation variables that affected the sintered ore at this time are acquired taking the above-mentioned delay time into account. 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 are matched to the return ore hopper level extracted from the actual data, taking the delay time into account.

[0031] A process for predicting the return ore hopper level is performed using input data including the actual measured return ore hopper level and the cut-out amount of return ore, as described above, and a prediction model. The return ore hopper level after a predetermined time (e.g., 2.5 hours) is predicted to prevent delays in corrective action. Furthermore, manipulated variables for the sintering process are calculated every separately determined control period (e.g., every hour) so that the predicted return ore hopper level falls within the target range. If the return ore hopper level is higher than the upper limit of the target range, an action is taken to lower the return ore rate. If it is lower than the lower limit of the target range, an action is taken to increase the return ore rate. Otherwise, no action is taken. This prevents excessive action. To calculate the manipulated variables, a machine learning model using the manipulated variables as explanatory variables may be constructed, and the manipulated variables may be calculated by inverse analysis so that the predicted return ore hopper level matches the target value (e.g., the median of the target range).

[0032] To lower the return ore hopper level, promoting combustion to prevent sintered ore from passing through the sinter machine unburned is an effective measure. For example, increasing the blending ratio of the coagulating agent, which is a heat source, is considered. Other effective measures include increasing the blending ratio of quicklime, which acts as a binder during granulation to improve ventilation, increasing the proportion of coke fines in the upper layer to facilitate ignition in the ignition furnace, and ensuring sufficient firing time by reducing the pallet speed. Effective measures to raise the return ore hopper level include decreasing the blending ratio of the coagulating agent, decreasing the blending ratio of quicklime, reducing the proportion of coke fines in the upper layer, and increasing the pallet speed. As mentioned above, there is a trade-off between reducing the return ore hopper level and productivity. Increasing the blending ratios of coagulating agent, quicklime, and coke fines increases production costs, while reducing the pallet speed reduces production volume. Therefore, the target range of the return ore hopper level is preferably determined from the perspective of production cost and production volume, and it is preferable to maintain the return ore hopper level within the target range. Here, the corrective action may be to determine a range of the manipulated variable's manipulated amount per operation and perform the operation within that range so as to prevent excessive operation (excessive manipulated amount).

[0033] Here, the calculated manipulated variables may be output so that they can be reflected by a process computer that manages the sintering process. Here, the output of the manipulated variables includes output as operation guidance for an operator who operates the sintering machine. In other words, the manipulated variables may be output so that appropriate manipulated variables for the manipulated variables are reflected in the sintering process through the judgment of the operator. The information output as operation guidance includes at least the calculated manipulated variables for the manipulated variables, and may be displayed on a display that the operator can see.

[0034] FIG. 3 is a diagram showing the configuration of a return ore control device 10 according to this embodiment. The return ore control device 10 is used in a sinter manufacturing facility and performs control related to the return ore, including the above-described return ore control method. As shown in FIG. 3, the return ore control device 10 includes a storage unit 11, an acquisition unit 12, a return ore hopper level prediction unit 13, an operation amount calculation unit 14, and an output unit 15. The return ore control device 10 acquires data related to the operation conditions in the sinter manufacturing facility, i.e., the above-described input data, from an operation data server 60. The input data includes the cooler blowing air pressure, exhaust gas NOx, in addition to the actual measurement value of the return ore hopper level and the return ore cut amount from the return ore hopper. XThe return ore control device 10 may acquire a target range for the return ore hopper level from the operation data server 60. The operation data server 60 can communicate with the return ore control device 10 via a network and may be implemented, for example, by a computer that manages the production of sintered ore. The network may be, for example, the Internet. The return ore control device 10 performs the above-described process, i.e., predicts the return ore hopper level using a prediction model and calculates manipulated variables for the cooler air pressure and other manipulated variables so that the future return ore hopper level is maintained within the target range. In this embodiment, the return ore control device 10 has a function to output manipulated variables for the manipulated variables to the sinter production equipment or a function to present them as guidance manipulated variables via the output unit 15. When the output unit 15 presents the guidance manipulated variables, the return ore control device 10 functions as an operation guidance device. The display unit 30 displays the guidance manipulated variables output from the return ore control device 10 (operation guidance device). The return ore control device 10 may be configured as a computer (e.g., 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 also be realized by a display of a terminal device such as a smartphone or a tablet. The terminal device can communicate with the return ore control device 10 via a network. A sintering operation guidance server having the functions of the return ore control device 10 and a terminal device having the functions 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 (e.g., in the same plant) or may be physically separated from each other. The sintering operation guidance system may further include an operation data server 60.

[0035] Here, the prediction model may be generated by the return fine control device 10 and stored in the storage unit 11, or may be generated by another computer and stored in the storage unit 11. When the return fine control device 10 generates the prediction model, it may further include a model generation unit that generates the prediction model by the above-mentioned method and stores it in the storage unit 11.

[0036] The components of the return ore control device 10 will be described below. The memory unit 11 stores a prediction model. The memory unit 11 also stores programs and data related to return ore hopper level control. The memory unit 11 may store acquired input data. The memory unit 11 may store a target range for the return ore hopper level. The memory unit 11 may store various information obtained in the process for controlling the return ore hopper level. 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.

[0037] 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 return ore hopper level can be identified.

[0038] The return ore hopper level prediction unit 13 predicts the return ore hopper level after a predetermined time based on the acquired input data. A prediction model is used to predict the return ore hopper level.

[0039] The manipulated variable calculation unit 14 calculates manipulated variables selected from the operating conditions and the raw material charging conditions based on the predicted return ore hopper level and a predetermined target range for the return ore hopper level. The raw material charging conditions include, for example, the blending ratios of the coagulant, quicklime, iron ore, or return ore.

[0040] The output unit 15 outputs the calculated manipulated variable to the sintering production equipment, or displays it on the display unit 30 as a guidance manipulated variable.

[0041] When the manipulated variables of the manipulated variables are output from the output unit 15 to the sinter production equipment, the sinter production equipment may automatically update the manipulated variables to the output manipulated variables to produce sinter ore. That is, the return ore control method according to this embodiment may be executed as part of a production method for producing sinter ore. Furthermore, the operator may change the operating conditions of the sinter machine based on the guidance manipulated variables displayed on the display unit 30. When the return ore hopper level exceeds a predetermined target range in the future, the operator may take an action to lower the return ore rate and lower the return ore hopper level. Furthermore, when the return ore hopper level falls below the target range, the operator may take an action to increase the return ore rate and raise the return ore hopper level. Such operational guidance for the sinter machine may be executed as part of a production method for producing sinter ore.

[0042] As described above, the return ore control device 10 may be realized, for example, by a computer. The computer includes, for example, a memory, a hard disk drive (storage device), and a CPU (processing device). 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 a hard disk drive (HDD). The memory unit 11 may be realized, for example, by a storage device. The acquisition unit 12, the return ore hopper level prediction unit 13, the operation amount calculation unit 14, and the output unit 15 may be realized, for example, by a CPU that reads and executes the program.

[0043] Fig. 4 is a flowchart showing a return ore control method according to this embodiment. The return ore control device 10 may calculate manipulated variables and output them as guidance manipulated variables according to the flowchart shown in Fig. 4. The return ore control method shown in Fig. 4 is also an operation guidance method, and may be executed as part of a sinter ore production method.

[0044] The acquisition unit 12 acquires input data (step S1, acquisition step). The return ore hopper level prediction unit 13 predicts the return ore hopper level after a predetermined time using the input data and a prediction model (step S2, return ore hopper level prediction step). The manipulated variable calculation unit 14 calculates the manipulated variable of the manipulated variable so that the predicted value of the return ore hopper level falls within a target range (step S3, manipulated variable calculation step). The output unit 15 outputs the calculated manipulated variable of the manipulated variable (step S4, output step).

[0045] Example 1 Hereinafter, specific examples (Examples) in which the optimal manipulated variable manipulated values ​​are determined by the above-mentioned control method will be described. In Example 1, in a sintering line, the amount of return ore generated 2.5 hours in advance was predicted using a machine learning model with a neural network, and the return ore hopper level was predicted based on the actual measurement value of the return ore hopper level and the amount of return ore cut out. 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 The parameters were concentration, flue gas CO concentration, flue gas CO2 concentration, flue gas O2 concentration, flue gas temperature, cooler air pressure, pallet speed, layer thickness, and production volume. The target range of the return ore hopper level was set to 40% to 60%. When the return ore hopper level fell below 40%, an action to reduce quicklime was taken. Furthermore, when the return ore hopper level rose above 60%, an action to increase quicklime was taken. The amount of quicklime manipulated per operation was fixed at 0.1%, and an operation (action) was performed every hour. As shown in Figure 5, in operation using conventional control (direct control of the return ore rate), the quicklime usage rate (quicklime blending ratio) averaged 1.5% over 100 hours of operation. By applying the method of this embodiment, the quicklime usage rate averaged 1.3% over 100 hours of operation. In this example, the method of this embodiment did not result in an excessively high quicklime blending ratio. In other words, the proportion of quicklime used could be reduced without excessive action, and production costs could be kept down.

[0046] Example 2 In Example 2, the amount of return ore generated in the sintering line 2.5 hours in advance was predicted using a machine learning model with a neural network, and the return ore hopper level was predicted based on the actual measurement value of the return ore hopper level and the cut-out amount of the return ore. 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, and exhaust gas NO. x The parameters were concentration, flue gas CO concentration, flue gas CO2 concentration, flue gas O2 concentration, flue gas temperature, cooler air pressure, pallet speed, layer thickness, and production volume. The target range of the return ore hopper level was set to 40% to 60%. When the return ore hopper level fell below 40%, an action to reduce quicklime was implemented. Furthermore, when the return ore hopper level rose above 60%, an action to increase quicklime was implemented. The quicklime manipulation amount was determined by back analysis based on the calculation results of the machine learning model. While the back analysis is not limited to a specific method, in this example, multiple calculations were performed by changing the input data of the machine learning model, and the optimal quicklime manipulation amount was searched for using a dichotomy method. Operations (actions) were performed every hour. As shown in Figure 6, in operation using conventional control (direct control of the return ore rate), the quicklime usage rate (quicklime blending ratio) averaged 1.5% over 100 hours of operation. By applying the method of this embodiment, the quicklime usage rate averaged 1.2% over 100 hours of operation. In this example, the method of this embodiment did not result in an excessively high blending ratio of quicklime. In other words, the blending ratio of quicklime was suppressed without excessive action, and an increase in production costs was suppressed.

[0047] As described above, as is clear from the above examples, the return ore control device 10, the return ore control method, and the sintered ore manufacturing method according to the present embodiment can reduce excessive action by controlling the return ore hopper level, thereby suppressing an increase in production costs.

[0048] 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.

[0049] The configuration of the return ore control device 10 shown in Fig. 3 is one example. The return ore control device 10 does not need to include all of the components shown in Fig. 3. Furthermore, the return ore control device 10 may include components other than those shown in Fig. 3. For example, the return ore control device 10 may be configured to further include a display unit 30.

[0050] In the above embodiment, the prediction model may be either a "prediction model for the amount of return ore generated (first prediction model)" or a "model for directly predicting the return ore hopper level (second prediction model)." Here, it can be considered that there is no substantial difference between the two models in terms of obtaining a predicted value for the return ore hopper level. However, as described above, the second prediction model is constructed using the actual measured value of the return ore hopper level and the return ore cut-out amount as input data, and can impart dynamic characteristics, including delays, to the operation of the return ore hopper level. For example, if a temporary return ore buffer is present in the actual return ore transport path, constructing a model including delays can be expected to further improve prediction accuracy. [Explanation of symbols]

[0051] 10 Return ore control device 11 Storage section 12 Acquisition Department 13 Return ore hopper level prediction section 14 Operation amount calculation section 15 Output section 30 Display section 60 Operational Data Server

Claims

1. A return ore control device used in a sintering production facility for controlling return ore, an acquisition unit that acquires data regarding operating conditions of the sintering manufacturing facility as input data; a return ore hopper level prediction unit that predicts a return ore hopper level after a predetermined time based on the acquired input data; an operation amount calculation unit that calculates an operation amount of an operation variable selected from the operation conditions and the raw material charging conditions based on the predicted return ore hopper level and a predetermined target range of the return ore hopper level; an output unit that outputs the calculated manipulated variable to the sintering manufacturing equipment or presents it as a guidance manipulated variable, the input data includes an actual measurement value of the return ore hopper level and the amount of return ore cut out, and further includes at least one of 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, flue gas NO x concentration, flue gas CO concentration, flue gas CO 2 concentration, flue gas O 2 concentration, flue gas temperature, cooler blowing pressure, pallet speed, layer thickness, and production amount.

2. the return ore hopper level prediction unit predicts the return ore hopper level using a prediction model, 2. The return ore control device according to claim 1, wherein the prediction model is generated by machine learning using learning data in which data corresponding to the input data extracted from performance data in a sintering process and the return ore hopper level extracted from the performance data are associated with each other in consideration of a delay time.

3. A return ore control method used in a sintering production facility, which controls return ore, an acquisition step of acquiring data on operating conditions of the sintering manufacturing facility as input data; a return ore hopper level prediction step of predicting a return ore hopper level after a predetermined time based on the acquired input data; an operation amount calculation step of calculating an operation amount of an operation variable selected from the operation conditions and the raw material charging conditions based on the predicted return ore hopper level and a predetermined target range of the return ore hopper level; an output step of outputting the calculated manipulated variable to the sintering manufacturing equipment or presenting it as a guidance manipulated variable, the input data includes an actual measurement value of the return ore hopper level and the cut-out amount of the return ore, and further includes at least one of 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, flue gas NO x concentration, flue gas CO concentration, flue gas CO 2 concentration, flue gas O 2 concentration, flue gas temperature, cooler blowing pressure, pallet speed, layer thickness, and production amount.

4. A method for producing sintered ore, comprising the steps of: producing sintered ore using the operation amount output to the sintering production facility by the return ore control method according to claim 3 ;

Images