Steel strip temperature prediction method, steel strip temperature control method, steel strip manufacturing method, and steel strip temperature prediction model generation method

Abstract

Proposed are a steel-strip plate temperature prediction method, a steel-strip plate temperature control method, a steel-strip manufacturing method, and a method for generating a steel-strip plate temperature prediction model that make it possible to control with high precision the annealing temperature of a steel-strip so as to be an annealing temperature that is favorable for magnetic properties and thus obtain a steel-strip with stably excellent magnetic properties. This steel-strip plate temperature prediction method includes: an input data acquisition step in which operation data, attribute parameters relating to dimensions of a steel-strip, attribute parameters relating to the composition of the steel-strip, and attribute parameters relating to the shape of the steel-strip are acquired as input data for a steel-strip plate temperature prediction model; and a plate temperature prediction step in which the operation data, the attribute data relating to the dimensions of the steel-strip, the attribute parameters relating to the composition of the steel-strip, and the attribute parameters relating to the shape of the steel-strip that were acquired in the input data acquisition step are input into the plate temperature prediction model so as to output the plate temperature of the steel-strip on the exit side of a soaking area.

Description

[Technical Field]

[0001] The present invention relates to a method for predicting the strip temperature of a steel strip in an annealing process of a steel strip (hot-rolled steel strip), a method for controlling the strip temperature of a steel strip, a method for manufacturing a steel strip, and a method for generating a steel strip temperature prediction model. The present invention relates to a method for annealing a hot-rolled steel strip for an electrical steel sheet containing, in particular, 1.6 to 5.0 mass% Si. [Background technology]

[0002] It is known that heat treatment of hot-rolled electrical steel strip containing 1.6 to 5.0 mass% Si before cold rolling improves the magnetic properties of the product sheet, i.e., it is possible to highly develop the Goss orientation. Hot-rolled sheet annealing equipment generally includes a preheating zone, a heating zone, a soaking zone, and a cooling zone where the annealing process is carried out, and the temperature of the steel strip at the exit of this soaking zone has a significant impact on the magnetic properties of the product sheet.

[0003] In order to control the temperature of the steel strip at the outlet side of the soaking zone in such hot strip annealing equipment, various methods have been proposed. For example, Patent Document 1 discloses a technique for rapidly heating a steel strip in hot-rolled sheet annealing. Furthermore, Patent Document 2 discloses a technique relating to the cooling rate when cooling a steel strip after heating in hot-rolled sheet annealing. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2018-066040 [Patent Document 2] Japanese Patent Application Laid-Open No. 2016-000856 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the method described in Patent Document 1, rapid heating in hot-rolled sheet annealing is used as a means for improving descaling properties after annealing, and there is no suggestion about increasing the precision of the annealing temperature in rapid heating. Also, in the method described in Patent Document 2, although there is a disclosure that hot-rolled sheet structure of grain-oriented electrical steel sheet is improved by annealing the hot-rolled sheet, there is no suggestion about increasing the precision of the annealing temperature in heating.

[0006] As described above, in the production of electrical steel sheets, highly accurate control of the heating temperature in the annealing of hot-rolled sheets is extremely important from the viewpoint of improving magnetic properties. However, hot-rolled steel strips are characterized by their large thickness, poor thickness accuracy, and large thickness variations, which cause fluctuations in the furnace temperature of the annealing furnace, and therefore there are limitations to the ability to control the heating temperature of the steel strip with high precision.

[0007] Therefore, in order to solve the above-mentioned problems of the conventional technology, the present invention aims to propose a method for predicting the sheet temperature of a steel strip, a method for controlling the sheet temperature of a steel strip, a method for manufacturing a steel strip, and a method for generating a sheet temperature prediction model of a steel strip, which enable the annealing temperature of the steel strip to be controlled with high precision to an annealing temperature that is favorable for magnetic properties, thereby enabling the steel strip to obtain stable and excellent magnetic properties. Here, in the annealing equipment, the downstream side of the flow of the steel strip near the outside of the soaking zone is referred to as the "soaking zone outlet side." [Means for solving the problem]

[0008] The inventors have conducted extensive research into a method for annealing a hot-rolled steel strip that can control the temperature of the steel strip on the outlet side of the soaking zone with high precision during the annealing of the hot-rolled steel strip. As a result, when annealing steel strips after hot rolling, a steel strip temperature (strip temperature) prediction model was created using machine learning, and this model was used to predict the strip temperature at the exit of the soaking zone during the annealing process. This heating temperature was then applied to actual annealing as a manufacturing condition. It was found that it was possible to suppress variation in the steel strip temperature (strip temperature) at the exit of the soaking zone and to control the annealing temperature to the target value with high precision.

[0009] The method for predicting the strip temperature of a steel strip according to the present invention, which advantageously solves the above problems, is configured as follows. [1] A method for predicting the sheet temperature of a steel strip at the outlet side of the soaking zone in an annealing process of a steel strip using hot-rolled sheet annealing equipment, comprising: an input data acquisition step of acquiring, as input data to be used in a pre-generated steel strip sheet temperature prediction model, one or more parameters selected from operational parameters of the hot-rolled sheet annealing equipment, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the chemical composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip, which are set as operational conditions of the steel strip manufacturing process; and a sheet temperature prediction step of inputting the operational data, attribute parameters related to the dimensions of the steel strip, attribute parameters related to the chemical composition of the steel strip, and attribute parameters related to the shape of the steel strip acquired in the input data acquisition step into the sheet temperature prediction model, thereby outputting the sheet temperature of the steel strip at the outlet side of the soaking zone. [2] In the above [1], the strip temperature prediction model is a method for predicting the strip temperature of a steel strip, the method comprising: a learning data acquisition step of acquiring a plurality of learning data using as input data one or more operation history data selected from operation history data of the hot-rolled strip annealing equipment, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the component composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip; and output data being information related to the strip temperature of the steel strip at the outlet of the soaking zone in the annealing process based on the input data; and a strip temperature prediction model generation step of generating the strip temperature prediction model by machine learning using the plurality of learning data acquired in the learning data acquisition step. [3] In the above [2], the method for predicting board temperature uses, as the machine learning, any one of a neural network, a decision tree learning, a random forest, and a support vector regression.

[0010] The method for controlling the temperature of a steel strip according to the present invention, which advantageously solves the above-mentioned problems, and the method for manufacturing a steel strip using the method for controlling the temperature of the steel strip, are configured as follows. [4] A method for controlling the temperature of a steel strip, comprising the steps of predicting the temperature of the steel strip at the outlet side of the soaking zone in the annealing process using the method for predicting the temperature of a steel strip described in any one of [1] to [3] above, comparing the predicted temperature of the steel strip with a predetermined target temperature, and resetting one or more operational parameters selected from the operational parameters of the hot-rolled strip annealing equipment so that the difference falls within a predetermined temperature range. [5] A method for manufacturing a steel strip, comprising a step of annealing a hot-rolled strip using the method for controlling the temperature of a steel strip described in [4] above.

[0011] The method for generating a steel strip temperature prediction model according to the present invention, which advantageously solves the above-mentioned problems, is configured as follows. [6] A method for generating a strip temperature prediction model for a steel strip in a hot-rolled strip annealing facility, the method generating a strip temperature prediction model for predicting the strip temperature of the steel strip at the soaking zone outlet side of an annealing process, the method including: a learning data acquisition step of acquiring a plurality of learning data using as input data one or more operation history data selected from operation history data of the hot-rolled strip annealing facility, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the chemical composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip, and information related to the strip temperature of the steel strip at the soaking zone outlet side in the annealing process based on the input data; and a strip temperature prediction model generation step of generating, by machine learning, a strip temperature prediction model using the plurality of learning data generated by the learning data acquisition step, the operation data, the attribute parameters related to the dimensions of the steel strip, the attribute parameters related to the chemical composition of the steel strip, and the attribute parameters related to the shape of the steel strip as input data, and the strip temperature of the steel strip at the soaking zone outlet side as output data. [7] In the above [6], the method for generating a steel strip temperature prediction model uses any one of a neural network, a decision tree learning, a random forest, and a support vector regression as the machine learning. [Effects of the Invention]

[0012] According to the present invention, the temperature of a steel strip at the soaking zone exit of a hot-rolled strip annealing facility, which performs heat treatment on a hot-rolled steel strip before cold rolling, is predicted by a strip temperature prediction method using a strip temperature prediction model, making it possible to control the steel strip temperature at the soaking zone exit to a predetermined target temperature with high accuracy. As a result, the variation in the steel strip temperature at the soaking zone exit of the hot-rolled strip annealing facility is reduced, the magnetic properties of the product strip are stably improved, and the occurrence of magnetic defects can be suppressed. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a schematic diagram showing an example of hot-rolled sheet annealing equipment used for annealing a hot-rolled steel strip according to the present invention. FIG. [Figure 2] 1 is a graph showing an example of the thermal history of a steel strip in the annealing of a hot-rolled steel strip according to the present invention. [Figure 3] FIG. 10 is an explanatory diagram showing a method for generating a strip temperature prediction model on the outlet side of the equalizing zone. [Figure 4] FIG. 2 is an explanatory diagram showing a strip temperature prediction method and a strip temperature control method on the outlet side of the equalizing zone. DETAILED DESCRIPTION OF THE INVENTION

[0014] A method for predicting the temperature of a steel strip in an annealing process of a hot-rolled steel strip, a method for controlling the temperature of a steel strip, and a method for manufacturing a steel strip according to embodiments of the present invention will be described. The method for predicting the strip temperature of a steel strip according to this embodiment predicts the strip temperature of a steel strip on the outlet side of a soaking zone in hot-rolled strip annealing equipment that applies heat treatment to a hot-rolled steel strip before cold rolling. At least after the hot rolling process, the thin steel sheet is wound into a coil and then subjected to heat treatment and the like, and therefore in this embodiment, the thin steel sheet is referred to as a "steel strip."

[0015] Hereinafter, this embodiment will be specifically described with reference to the drawings. Figure 1 shows a hot-rolled sheet annealing facility according to this embodiment, which applies heat treatment to a hot-rolled steel strip for electrical steel sheet before cold rolling. Arrow F in Figure 1 indicates the line traveling direction. The hot-rolled sheet annealing facility is broadly divided into a pre-heating zone 2, a heating zone 3, a soaking zone 4, and a cooling zone 5. The annealing process is a process in which the temperature of a steel strip 1 is raised from near room temperature, maintained at a predetermined temperature, and then lowered to near room temperature. In the hot-rolled sheet annealing equipment shown in Figure 1, the annealing process is carried out in a preheating zone 2, a heating zone 3, a soaking zone 4, and a cooling zone 5.

[0016] The preheating zone 2 and heating zone 3 are equipment for raising the temperature of the steel strip 1 to a preset temperature. In the preheating zone 2, an induction heating device or a direct-fire or radiant combustion burner is used. In the heating zone 3, a direct-fire or radiant combustion burner is used. These heating devices have a large heating capacity and a fast response, making it easy to change the temperature rise history when changing the heat cycle. The soaking zone 4 is equipment for maintaining the steel strip 1 at a predetermined temperature, and has a heating capacity sufficient to compensate for the heat dissipated by the furnace body.

[0017] The cooling zone 5 is a facility for cooling the steel strip 1 to a predetermined temperature, and uses gas jet cooling as a cooling means. Gas jet cooling is a cooling means in which gas is sprayed onto the surface of the steel strip 1 from a nozzle.

[0018] Additionally, thermometers for measuring the surface temperature of the steel strip 1 are installed at multiple locations in the preheating zone 2, heating zone 3, soaking zone 4, and cooling zone 5. In particular, in the preheating zone 2 and heating zone 3, where the temperature change of the steel strip 1 is large, thermometers are installed at the inlet and outlet of the preheating zone 2 and heating zone 3, and the actual heating rate of the preheating zone 2 and heating zone 3 is calculated by measuring the surface temperature of the steel strip 1 at those locations.

[0019] Furthermore, a radiation thermometer is used as the thermometer, which continuously measures the surface temperature of the center of the steel strip 1 in the width direction. However, the thermometer is not limited to a radiation thermometer, and a profile radiation thermometer, which measures the temperature distribution in the width direction, may also be used. Furnace thermometers are also installed to measure not only the surface temperature of the steel strip 1 but also the atmospheric temperature inside the furnace in each zone of the annealing process. The measured surface temperature and atmospheric temperature of the steel strip 1 are output to a process computer that controls the hot-rolled sheet annealing equipment and supervises its operation.

[0020] Figure 2 is a graph showing the thermal history of a steel strip 1 in a hot-rolled sheet annealing facility, which performs heat treatment on hot-rolled steel strips for electrical steel sheets before cold rolling. The horizontal axis represents time, and the vertical axis represents the steel strip temperature. The steel strip temperature is, for example, the surface temperature of the steel strip 1. The annealing process is carried out using a preheating zone 2, a heating zone 3, a soaking zone 4, and a cooling zone 5. To prevent variations in the strip temperature depending on the longitudinal position of the steel strip 1, the transport speed of the steel strip 1 during the annealing process is maintained constant. However, when steel strips 1 with different thicknesses, widths, steel types, etc. are welded together, the line speed may change before and after the weld. For this reason, the shape of the thermal history graph may vary depending on the measurement position on the steel strip 1.

[0021] <Method for generating a steel strip temperature prediction model> FIG. 3 shows a method for generating a prediction model for the strip temperature on the outlet side of the equalizing zone according to this embodiment. The database stores operational performance data of the hot-rolled sheet annealing equipment, performance data of attribute parameters relating to the dimensions of the steel strip, performance data of attribute parameters relating to the chemical composition of the steel strip, performance data of attribute parameters relating to the shape of the steel strip, and performance data regarding the sheet temperature at the exit side of the steel strip equalizing zone.

[0022] Details of the operational performance data of the hot-rolled strip annealing facility will be described later. However, performance data selected from the operational performance data held by the process computer that manages the operation of the hot-rolled strip annealing facility is sent to a database of a strip temperature prediction model generation unit at the outlet of the soaking zone. The operational performance data of the hot-rolled strip annealing facility is then associated with each other by coil number, etc., and stored in the database as a set of data sets. At this time, one data set is acquired for each steel strip, and stored in the database.

[0023] The database also contains one or more parameters selected from attribute parameters related to the dimensions of the steel strip, such as the thickness, width, and length of the steel strip. Actual data on the attribute parameters related to the dimensions of the steel strip is stored in a process computer or a host computer together with the coil number as actual values ​​in the hot rolling process, and a data set can be constructed by sending this data to the database as appropriate. This is because, by adding the attribute parameters related to the dimensions of the steel strip to the input, the strip temperature prediction model at the outlet of the soaking zone can be widely applied to steel strips of different dimensions.

[0024] The database also contains one or more parameters selected from attribute parameters related to the chemical composition of the steel strip. Actual data on the attribute parameters related to the chemical composition of the steel strip is stored in a process computer or a host computer together with the coil number as actual values ​​in the steelmaking process, and a data set can be constructed by sending this data to the database as appropriate. This is because, by adding the attribute parameters related to the chemical composition of the steel strip to the input, the strip temperature prediction model at the outlet of the soaking zone can be widely applied to steel strips with different chemical compositions.

[0025] The database also contains one or more parameters selected from attribute parameters related to the shape of the steel strip. Actual data on the attribute parameters related to the shape of the steel strip is stored in a process computer or a host computer together with the coil number as actual values ​​in the hot rolling process, and a data set can be constructed by sending the data to the database as appropriate. This is because, by adding attribute parameters related to the shape of the steel strip to the input, the strip temperature prediction model at the outlet of the soaking zone can be widely applied to steel strips with different shapes.

[0026] The number of data sets in the database used to generate the strip temperature prediction model at the strip temperature outlet side of the equalizing zone in this embodiment is preferably 200, and more preferably 1000 or more.

[0027] In this embodiment, using the database created in this manner, one or more types of input data are selected from one or more operational performance data selected from the operational performance data of the hot-rolled sheet annealing equipment, one or more parameters selected from the attribute parameter information regarding the dimensions of the steel strip, one or more parameters selected from the attribute parameter information regarding the component composition of the steel strip, and one or more parameters selected from the attribute parameter information regarding the shape of the steel strip, and a sheet temperature prediction model at the outlet side of the steel strip equalizing zone is generated by machine learning using the input data.

[0028] Any known learning method may be applied as the machine learning method. The type of machine learning model is not limited as long as it provides a practically sufficient prediction accuracy for the strip temperature at the strip equalizing zone outlet. For example, a known machine learning method such as a neural network may be used. Other examples of the method include decision tree learning, random forest, support vector regression, and Gaussian process. An ensemble model combining a plurality of models may also be used. The strip temperature prediction model at the outlet of the soaking zone may be updated as needed using the latest learning data. This is because it can respond to long-term changes in the operating conditions of the hot-rolled strip annealing facility.

[0029] [Operational parameters for hot strip annealing equipment] As the operation parameters in the hot strip annealing equipment, any operation parameters that affect the strip temperature of the steel strip on the outlet side of the soaking zone can be used. Using the example of the thermal history of the steel strip in the annealing process shown in Figure 2, the following operational parameters for the annealing process can be used. For example, the operational parameters for preheating zone 2 may be the time it takes for the steel strip 1 to pass through preheating zone 2 and the amount of temperature rise, or an average heating rate calculated from these values ​​may be used. The operational parameters for heating zone 3 may be the time it takes for the steel strip 1 to pass through heating zone 3 and the amount of temperature rise, or an average heating rate calculated from these values ​​may be used. The operational parameters for soaking zone 4 may be the time it takes for the steel strip 1 to pass through soaking zone 4 and the amount of temperature rise, or an average heating rate calculated from these values ​​may be used. After measurement, these values ​​are output to the process computer from the preheating zone 2, heating zone 3 and soaking zone 4.

[0030] At least one of the operational parameters of the annealing process described above is used as training data (learning data). However, the operational parameters are not limited to these, and the control output values ​​of the heating devices in the preheating zone 2, heating zone 3, and soaking zone 4, as well as the flow rates of gas and air fed into the preheating zone 2, heating zone 3, and soaking zone 4 may also be used as operational parameters. This is because these operational parameters are used to control the temperature history of the steel strip 1 in the annealing process.

[0031] [Steel strip dimension attribute parameters] In this embodiment, the input data for the strip temperature prediction model at the soaking zone outlet further includes one or more parameters selected from attribute parameters related to the dimensions of the steel strip, such as the thickness, width, and length of the steel strip. These parameters affect the heat transfer behavior in hot-rolled strip annealing equipment, and therefore affect the strip temperature of the steel strip at the soaking zone outlet, even if the furnace atmosphere temperature is the same. This makes it possible to generate a strip temperature prediction model that predicts the strip temperatures of steel strips with various dimensions at the soaking zone outlet, for use as steel strips that undergo heat treatment before cold rolling after hot rolling for electrical steel sheets, thereby expanding the scope of application of the strip temperature prediction model at the soaking zone outlet.

[0032] [Attribute parameters related to the composition of steel strips] In this embodiment, the input data for the strip temperature prediction model at the outlet of the equalizing zone further includes one or more parameters selected from attribute parameters relating to the chemical composition of the steel strip. These factors affect the heat transfer behavior in hot strip annealing equipment and the target range of strip temperature at the outlet of the soaking zone. Therefore, even if the atmosphere temperature inside the furnace is the same, they affect the strip temperature at the outlet of the soaking zone, and further affect the strip temperature prediction model at the outlet of the soaking zone. Therefore, a sheet temperature prediction model can be generated to predict the sheet temperature of steel strips having various chemical compositions at the outlet of the equalizing zone, which are used as steel strips that are heat-treated before cold rolling after hot rolling for electrical steel sheets, thereby expanding the range of application of sheet temperature prediction models at the outlet of the equalizing zone.

[0033] The attribute parameters relating to the chemical composition of the steel strip may include the contents of C, Si, Mn, P, and S as chemical components contained in the steel strip. The attribute parameters relating to the chemical composition of the steel strip may also include the contents of Cu, Ni, Cr, Al, Nb, Ti, V, N, and B. However, it is not necessary to use all of these compositional compositions as attribute parameters relating to the compositional composition of the steel strip. It is preferable to select the compositional composition according to the type of steel strip to be manufactured in the hot-rolled sheet annealing equipment.

[0034] [Steel strip shape attribute parameters] In this embodiment, the input data for the strip temperature prediction model at the soaking zone outlet further includes one or more parameters selected from attribute parameters related to the shape of the steel strip, such as the steepness and warpage of the steel strip. These parameters affect the heat transfer behavior in the hot-rolled strip annealing equipment, and therefore affect the strip temperature at the soaking zone outlet even at the same furnace ambient temperature. In particular, the shape of the steel strip has a significant effect on the heat transfer behavior in the strip width direction. Therefore, it is possible to generate a sheet temperature prediction model that predicts the sheet temperature at the outlet side of the equalizer zone for steel strips having various shapes, such as steel strips after hot rolling for electrical steel sheets that are heat-treated before cold rolling, thereby expanding the range of application of the sheet temperature prediction model at the outlet side of the equalizer zone.

[0035] <Method for predicting steel strip temperature> The following steps are used to predict the strip temperature at the outlet side of the equalizing zone of a hot strip annealing facility. [Input data acquisition step] First, one or more parameters selected from the operating parameters of the hot-rolled sheet annealing equipment, which are set as operating conditions for the steel strip manufacturing process, one or more parameters selected from the attribute parameters related to the dimensions of the steel strip, one or more parameters selected from the attribute parameters related to the component composition of the steel strip, and one or more parameters selected from the attribute parameters related to the shape of the steel strip are obtained as input data. [Board temperature prediction step] Next, the operational data acquired in the input data acquisition step, attribute parameters relating to the dimensions of the steel strip, attribute parameters relating to the chemical composition of the steel strip, and attribute parameters relating to the shape of the steel strip are input into a strip temperature prediction model, and the strip temperature of the steel strip at the outlet of the equalizing zone is output.

[0036] <Steel strip temperature control method> Fig. 4 shows a method for controlling the strip temperature at the outlet of the soaking zone and a method for manufacturing a steel strip using the above-mentioned method for predicting the strip temperature at the outlet of the soaking zone.The control flow shown in Fig. 4 is started when the leading end of the steel strip, the strip temperature of which is to be predicted at the outlet of the soaking zone, reaches the outlet of the soaking zone. At this point, the input data for the strip temperature prediction model at the soaking zone outlet are the operational performance data of the hot-rolled strip annealing equipment obtained in the hot-rolled strip annealing equipment, attribute parameter information relating to the dimensions of the steel strip, attribute parameter information relating to the chemical composition of the steel strip, and attribute parameter information relating to the shape of the steel strip. The step of acquiring this input data corresponds to the input data acquisition step described above. Furthermore, the operational performance data of the hot-rolled strip annealing facility or the set values ​​of the operational conditions of the hot-rolled strip annealing facility at that time may be used as input data for the strip temperature prediction model at the outlet of the soaking zone. Using the data acquired in this way as input, the strip temperature prediction model is used to predict the strip temperature of the steel strip at the outlet of the soaking zone.

[0037] In this embodiment, a target range for the strip temperature at the soaking zone outlet is also set in the host computer, and the predicted strip temperature at the soaking zone outlet is compared with this target range. The target range for the strip temperature at the soaking zone outlet is set based on past operational results for the strip temperature at the soaking zone outlet so as to reduce variation compared to conventional methods. For example, for a hot-rolled steel strip for electrical steel sheet containing 3.4 mass% Si, the target range for the strip temperature at the soaking zone outlet can be set to 1015 to 1025°C. At this time, the operational condition setting unit of the hot-rolled sheet annealing facility compares the target range of the strip temperature at the soaking zone outlet, which is set in advance as described above, with the predicted strip temperature at the soaking zone outlet. If the predicted strip temperature at the soaking zone outlet satisfies the conditions of the target range, the operational conditions of the hot-rolled sheet annealing facility are determined as initially set and sent to the control unit of the hot-rolled sheet annealing facility. On the other hand, if the predicted strip temperature at the soaking zone outlet is outside the target range, the operational conditions of the hot-rolled sheet annealing facility are reset.

[0038] Furthermore, the strip temperature prediction model generation unit at the outlet of the soaking zone shown in FIG. 3, the strip temperature prediction unit at the outlet of the soaking zone shown in FIG. 4, and the operating condition setting unit of the hot-rolled strip annealing equipment may be realized by a computer capable of communicating with a host computer and various devices constituting the hot-rolled strip annealing equipment. As an example, this computer may be a process computer. The configuration of the computer is not particularly limited and may include, for example, a memory (storage device), a CPU (processing device), a hard disk drive (HDD), a communication control unit for connecting to a network, a display device, and an input device. Here, various processes of the unit for generating a strip temperature prediction model at the soaking zone outlet, the unit for predicting the strip temperature at the soaking zone outlet, and the unit for setting operational conditions for hot-rolled strip annealing equipment may be executed by the CPU. The database in FIG. 3 may be implemented by a hard disk drive. Furthermore, the strip temperature prediction model at the soaking zone outlet may be stored in the memory. Furthermore, collection of learning data may be implemented by the communication control unit. Furthermore, it is possible to carry out a method for manufacturing a steel strip including an annealing step of a hot-rolled strip using the above-mentioned method for controlling the temperature of the steel strip. [Example]

[0039] The present embodiment will be specifically described below using examples, but the present invention is not limited to these examples. In the hot-rolled sheet annealing facility shown in Figure 1, 200 coils of steel strip were produced by heat treating hot-rolled steel strip for electrical steel sheet before cold rolling. The input data were actual data of attribute information related to the dimensions of the steel strip to be charged into the hot-rolled sheet annealing facility, actual data of attribute information related to the chemical composition of the steel strip to be charged into the hot-rolled sheet annealing facility, actual data of attribute information related to the shape of the steel strip to be charged into the hot-rolled sheet annealing facility, and actual operational data of the operational parameters of the hot-rolled sheet annealing facility.

[0040] A plurality of learning data were obtained using the strip temperature at the soaking zone outlet of the hot strip annealing equipment based on the input actual data as output data. Here, the output data based on the input data refers to actual data (obtained strip temperature at the soaking zone outlet of the steel strip) output in response to the input actual data being used as the operational setting conditions, etc. A prediction model for the strip temperature at the soaking zone outlet, trained by machine learning using the acquired plurality of learning data, was generated by the method shown in Figure 3.

[0041] In generating the strip temperature prediction model at the soaking zone outlet, the strip thickness and width were used as input attribute parameters related to the dimensions of the steel strip. The Si, Mn, and Al contents were also input as attribute parameters related to the steel strip's chemical composition. The steepness was also input as an attribute parameter related to the shape of the steel strip. Furthermore, the strip temperatures at the inlet and outlet of heating zone 3 and the conveying speed at which the leading edge of the steel strip passed through heating zone 3 were also input as operational data for the hot-rolled strip annealing equipment. Here, the strip temperature at the outlet side of the soaking zone of the steel strip obtained as learning data is the strip temperature measured using a radiation thermometer at the outlet side of the soaking zone of the hot-rolled strip annealing equipment. The strip temperature prediction model at the outlet of the soaking zone generated in this way was applied to the strip temperature prediction section at the outlet of the soaking zone in the strip temperature control at the outlet of the soaking zone shown in Fig. 4. 100 coils of steel strip were produced from hot-rolled steel strip for electrical steel sheet, which had been heat-treated before cold rolling.

[0042] At this time, using the above-mentioned sheet temperature prediction model at the soaking zone outlet, the sheet temperature of the steel strip at the soaking zone outlet of the hot-rolled sheet annealing equipment was predicted as output data, and the operating parameters of the hot-rolled sheet annealing equipment were reset so that the predicted sheet temperature at the soaking zone outlet was within a predetermined allowable range (1015 to 1025°C). The flow shown in Figure 4 begins after the leading edge of the steel strip reaches the soaking zone exit side. Strip temperature data was then collected using a radiation thermometer at the soaking zone exit side of these steel strips. The results showed that 95% of the steel strips were within the allowable strip temperature range (1015-1025°C) at the soaking zone exit side. As a result, the magnetic properties of the electrical steel sheets were stable with little variation.

[0043] On the other hand, as a comparative example, a similar experiment was conducted using the method described in Patent Document 1. As a result, 65% of the steel strips had a strip temperature within the allowable range on the outlet side of the soaking zone. As a comparative example, a similar experiment was conducted using the method described in Patent Document 2. As a result, 50% of the steel strips had a strip temperature within the allowable range on the outlet side of the soaking zone. As a result, the magnetic properties of the electromagnetic steel sheets in the comparative example varied more than in the inventive example.

[0044] As described above, by applying the method for predicting the strip temperature at the outlet of the equalizer zone according to the present invention, it is possible to predict the strip temperature at the outlet of the equalizer zone in hot-rolled strip annealing equipment with high accuracy, and also to reduce the variation in the strip temperature at the outlet of the steel strip equalizer zone, thereby suppressing the variation in the magnetic properties of the steel sheet. [Industrial Applicability]

[0045] The technology of the present invention is a technology that aims to minimize the quality variation of product strips by controlling the temperature of the steel strip at the outlet of the soaking zone to a predetermined temperature with high accuracy using a strip temperature prediction model. This technology can be applied not only to hot-rolled steel strips for electrical steel sheets, but also to all metal strips that require annealing temperature control. [Explanation of symbols]

[0046] F: Direction of steel strip travel 1: Steel strip 2: Pre-tropical zone 3: Heating zone 4:Solid temperature 5: Cooling zone

Claims

1. A method for predicting the sheet temperature of a steel strip at an outlet side of an equalizing zone in an annealing process of a steel strip after hot rolling for an electrical steel sheet containing 1.6 to 5.0 mass% Si using hot-rolled sheet annealing equipment, an input data acquisition step of acquiring, as input data to be used in a pre-generated steel strip temperature prediction model, one or more parameters selected from the operation parameters of the hot-rolled sheet annealing equipment, which are set as operation conditions of the steel strip manufacturing process, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the component composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip; a strip temperature prediction step of inputting the operational data, attribute parameters relating to the dimensions of the steel strip, attribute parameters relating to the chemical composition of the steel strip, and attribute parameters relating to the shape of the steel strip acquired in the input data acquisition step into the strip temperature prediction model, thereby outputting the strip temperature of the steel strip at the outlet side of the soaking zone; A method for predicting the strip temperature of a steel strip, including:

2. The strip temperature prediction model uses as input data one or more operation performance data selected from operation performance data of the hot-rolled strip annealing equipment, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the component composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip, Information regarding the sheet temperature of the steel strip at the outlet side of the soaking zone in the annealing process based on the input data is output as output data, a learning data acquisition step of acquiring a plurality of learning data; a strip temperature prediction model generation step of generating the strip temperature prediction model by machine learning using the plurality of learning data acquired in the learning data acquisition step; The method for predicting the strip temperature of a steel strip according to claim 1, wherein the temperature is generated by:

3. The method for predicting the strip temperature of a steel strip according to claim 2, wherein the machine learning is performed using any one of a neural network, a decision tree learning, a random forest, and a support vector regression.

4. A method for controlling the sheet temperature of a steel strip, comprising the steps of predicting the sheet temperature of the steel strip at the outlet side of the soaking zone in the annealing process using the steel strip sheet temperature prediction method according to any one of claims 1 to 3, comparing the predicted sheet temperature of the steel strip with a predetermined target temperature, and resetting one or more operation parameters selected from the operation parameters of the hot-rolled sheet annealing equipment so that the difference falls within a predetermined temperature range.

5. A method for producing a steel strip, comprising a step of annealing a hot-rolled sheet for an electrical steel sheet containing 1.6 to 5.0 mass% Si using the method for controlling the sheet temperature of a steel strip according to claim 4.

6. A method for generating a steel strip temperature prediction model for generating a steel strip temperature prediction model for predicting a steel strip temperature at an outlet side of an equalizing zone in an annealing process of a steel strip after hot rolling for an electrical steel sheet containing 1.6 to 5.0 mass% Si using hot-rolled sheet annealing equipment, One or more operation performance data selected from the operation performance data of the hot-rolled sheet annealing equipment, one or more parameters selected from attribute parameters related to the dimensions of the steel strip, one or more parameters selected from attribute parameters related to the component composition of the steel strip, and one or more parameters selected from attribute parameters related to the shape of the steel strip are used as input data, Information regarding the sheet temperature of the steel strip at the outlet side of the soaking zone in the annealing process based on the input data is output as output data, a learning data acquisition step of acquiring a plurality of learning data; a strip temperature prediction model generation step of generating a strip temperature prediction model using a plurality of learning data generated by the learning data acquisition step, by machine learning, to generate a strip temperature prediction model in which operational data, attribute parameters related to the dimensions of the steel strip, attribute parameters related to the component composition of the steel strip, and attribute parameters related to the shape of the steel strip are used as input data, and the strip temperature of the steel strip on the outlet side of the equalizing zone is used as output data.

7. The method for generating a steel strip sheet temperature prediction model according to claim 6, wherein the machine learning is performed using any one of a neural network, a decision tree learning, a random forest, and a support vector regression.

Images