Method for predicting cross-sectional dimensions of unequal angle steel, method for manufacturing unequal angle steel, method for generating cross-sectional dimension prediction model, and device for predicting cross-sectional dimensions of unequal angle steel

Abstract

Provided is a method for predicting the cross-sectional dimensions of an unequal angle steel with which it is possible to predict the cross-sectional dimensions of an unequal angle steel manufactured by caliber rolling.　A method for predicting the cross-sectional dimensions of an unequal angle steel in which predictions are made of the cross-sectional dimensions of a second rolled material from among two pieces of unequal angle steel that have been hot-rolled using the same roll set, the method comprising: inputting, into a cross-sectional dimension prediction model, input data including the cross-sectional dimensions of a first rolled material from among the two pieces of unequal angle steel, the parameters for rolling the first rolled material including the inter-roll gap in a rough rolling mill, an intermediate rolling mill, and a finishing rolling mill, and the parameters for rolling the second rolled material including the inter-roll gap in the rough rolling mill, the intermediate rolling mill, and the finish rolling mill; and outputting the cross-sectional dimensions of the second rolled material so as to predict the cross-sectional dimensions of the second rolled material. The category relating to the cross-sectional dimensions of the first rolled material used in the input data includes the category relating to the cross-sectional dimensions of the second rolled material outputted from the cross-sectional dimension prediction model.

Description

Method for predicting cross-sectional dimensions of unequal leg angle steel, method for manufacturing unequal leg angle steel, method for generating cross-sectional dimension prediction model, and device for predicting cross-sectional dimensions of unequal leg angle steel

[0001] The present invention relates to a method for predicting the cross-sectional dimensions of unequal leg angle steel, a manufacturing method for unequal leg angle steel, a method for generating a cross-sectional dimension prediction model, and a cross-sectional dimension prediction device for unequal leg angle steel.

[0002] The background art will be described using an example of an unequal angle steel (hereinafter also referred to as "NAB"), a type of unequal angle steel used as reinforcing steel for structures such as ship hulls, bridges, and tanks. NAB is manufactured by hot-rolling a steel billet. FIG. 1 is a schematic diagram showing the cross-sectional shape of an NAB 10. As shown in FIG. 1, the NAB 10 is composed of a relatively thin, wide long side 12 and a relatively thick, narrow short side 14. There are approximately a dozen combinations of the thickness and width dimensions of the long side 12 and the short side 14. The mechanical properties that NAB 10 must satisfy are determined by the application and the classification society's standard classification, and the chemical composition of NAB 10 is appropriately adjusted accordingly. When NAB 10 is used as a steel material for a structure, the dimensions, thickness, and chemical composition of the long side 12 and the short side 14 are selected depending on the required performance of the structure.

[0003] To manufacture NAB10, a steel billet heated in a heating furnace is subjected to rough rolling, intermediate rolling, and finish rolling using multiple rolling mills to be formed into a cross-sectional shape with the desired dimensions. For example, in the rough rolling and intermediate rolling, the steel billet is rolled through multiple passes using a rough rolling mill and an intermediate rolling mill having a pair of upper and lower rolls each with multiple grooves called calibers, to roughly form the billet into a shape close to the desired cross-sectional shape. Finally, a single pass of finish rolling is performed using a finishing rolling mill having a pair of upper and lower rolls each with one caliber, to form NAB10 with the desired cross-sectional shape.

[0004] The Japanese Industrial Standard JIS G3192:2021 (Shape, Dimensions, Mass, and Tolerances of Hot-Rolled Section Steel) and corresponding foreign and international standards specify tolerances for the cross-sectional dimensions of section steel. For example, NAB10 specifies representative values ​​and allowable values ​​for each dimension, such as long side width A, short side width B, long side thickness t1, short side thickness t2, and short side foot shape C, and requires that all of these dimensions be controlled so that they fall within the allowable range from the representative value. Therefore, rolling conditions such as the roll gap for each rolling mill and each pass are adjusted so that each dimension of NAB10 falls within the allowable range.

[0005] 2 is a schematic diagram illustrating the toe shape. The toe shape is the amount of material shortage (dimension C) at a 45° angle at a corner in the cross-sectional shape as shown in FIG. 2. The short side toe shape of the NAB10 is the amount of material shortage (dimension C) at a 45° angle at a corner on the short side of the NAB10 shown in FIG. 1.

[0006] Patent Document 1 discloses a method for controlling the thickness of H-beams rolled by universal rolling, in which a reduction correction amount determined from the deviation of the web thickness and flange thickness of the previous material from the target values ​​is added to the setup calculation for the next material. Patent Document 2 discloses a method for controlling the thickness of H-beams that specifically defines a method for adjusting the roll positions to eliminate variations in flange thickness at four points on the H-beam.

[0007] Patent Document 3 discloses a method for predicting the cross-sectional dimension of shaped steel using a cross-sectional dimension change prediction model that inputs the difference in rolling operation parameters between a preceding rolled material and a succeeding rolled material of H-section steel and outputs the difference in cross-sectional dimension between the preceding rolled material and the succeeding rolled material. According to Patent Document 3, by acquiring the difference in rolling operation parameters between the preceding rolled material and the succeeding rolled material of H-section steel and the cross-sectional dimension of the preceding rolled material, it is possible to predict the amount of change in the cross-sectional dimension of the H-section steel corresponding to the amount of correction of the rolling operation parameters.

[0008] JP-A-62-188726 JP-A-10-296311 JP-A 2021-194701

[0009] Patent Documents 1 to 3 all disclose methods for controlling the cross-sectional dimensions of H-shaped steel beams rolled by a universal rolling mill, but there was a problem in that the prediction accuracy was low even when applied to unequal leg angle steel beams manufactured by groove rolling rather than by a universal rolling mill.

[0010] The present invention has been made in view of the above-mentioned prior art, and its object is to provide a method for predicting the cross-sectional dimensions of unequal leg angle steels and a device for predicting the cross-sectional dimensions of unequal leg angle steels, which are manufactured by groove rolling. Another object of the present invention is to provide a method for generating a cross-sectional dimension prediction model used in the method for predicting the cross-sectional dimensions of unequal leg angle steels, and a method for manufacturing unequal leg angle steels.

[0011] The means for solving the above problems are as follows: [1] A method for predicting the cross-sectional dimension of a succeeding rolled material of two unequal leg angle steels manufactured by hot rolling with the same roll set, comprising: inputting input data including the cross-sectional dimension of a preceding rolled material of the two unequal leg angle steels, rolling parameters including the roll gaps of a roughing mill, an intermediate rolling mill, and a finishing rolling mill for the preceding rolled material, and rolling parameters including the roll gaps of a roughing mill, an intermediate rolling mill, and a finishing rolling mill for the following rolled material, into a cross-sectional dimension prediction model, outputting the cross-sectional dimension of the following rolled material to predict the cross-sectional dimension of the following rolled material, and including an item for the cross-sectional dimension of the preceding rolled material to be output from the cross-sectional dimension prediction model in the item for the cross-sectional dimension of the preceding rolled material used in the input data. [2] The method for predicting cross-sectional dimensions of unequal arm angle steel according to [1], wherein the input data includes guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters of the preceding rolled material, and guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters of the following rolled material. [3] The method for predicting cross-sectional dimensions of unequal arm angle steel according to [1] or [2], wherein the input data includes one or more attribute parameters of the following rolled material. [4] The method for predicting cross-sectional dimensions of unequal arm angle steel according to any of [1] to [3], wherein the cross-sectional dimensions of the following rolled material output from the cross-sectional dimension prediction model include the long side length, short side length, long side thickness, short side thickness, and short side foot shape of the unequal arm angle steel. [5] A method for manufacturing unequal leg angle steel, which specifies rolling parameters that will result in the cross-sectional dimensions of a subsequently rolled material predicted by the cross-sectional dimension prediction method for unequal leg angle steel described in any one of [1] to [4] being within the range of target dimensions, and hot rolling the unequal leg angle steel under rolling conditions that include the specified rolling parameters.[6] A method for generating a cross-sectional dimension prediction model used to predict the cross-sectional dimension of a succeeding rolled material of two unequal leg angle steel bars manufactured by hot rolling with the same roll set, comprising: training a machine learning model using as training data a plurality of data sets each consisting of a set of actual values ​​of input data including the cross-sectional dimension of a preceding rolled material previously manufactured with the same roll set, rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing rolling mill of the preceding rolled material, and rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing rolling mill of the succeeding rolled material, and actual values ​​of the cross-sectional dimension of the succeeding rolled material; generating a cross-sectional dimension prediction model that uses the input data as input and outputs the cross-sectional dimension of the succeeding rolled material; and including an item for the cross-sectional dimension of the preceding rolled material in the item for the cross-sectional dimension of the preceding rolled material to be output from the cross-sectional dimension prediction model. [7] The method for generating a cross-sectional dimension prediction model described in [6], wherein the input data includes guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters of the preceding rolled material, and guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters of the following rolled material. [8] A cross-sectional dimension prediction device for unequal leg angle steel that predicts the cross-sectional dimension of a succeeding rolled material of two unequal leg angle steels manufactured by hot rolling with the same roll set, comprising: a cross-sectional dimension prediction unit that inputs input data including the cross-sectional dimension of a preceding rolled material of the two unequal leg angle steels, rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing rolling mill of the preceding rolled material, and rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing rolling mill of the succeeding rolled material into a cross-sectional dimension prediction model, outputs the cross-sectional dimension of the succeeding rolled material, and predicts the cross-sectional dimension of the succeeding rolled material; and the cross-sectional dimension prediction device for unequal leg angle steel includes an item for the cross-sectional dimension of the preceding rolled material that is output from the cross-sectional dimension prediction model in the item for the cross-sectional dimension of the preceding rolled material used in the input data. [9] The cross-sectional dimension prediction device for unequal leg angle steel as described in [8], wherein the input data includes guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters for the preceding rolled material, and guide positions of a roughing mill, an intermediate rolling mill, and a finishing rolling mill as rolling parameters for the following rolled material.

[10] The cross-sectional dimension prediction device for unequal leg angle steel according to [8] or [9], wherein the input data includes one or more attribute parameters of the subsequently rolled material.

[0012] By implementing the method for predicting the cross-sectional dimensions of unequal leg angle steel according to the present invention, it becomes possible to predict the cross-sectional dimensions of unequal leg angle steel manufactured by groove rolling. In this way, if the cross-sectional dimensions of unequal leg angle steel can be predicted, it becomes possible to specify the rolling conditions for groove rolling that will bring the predicted cross-sectional dimensions into a target range, thereby making it possible to manufacture unequal leg angle steel with cross-sectional dimensions within the target range.

[0013] FIG. 1 is a schematic diagram showing the cross-sectional shape of an NAB. FIG. 2 is a schematic diagram explaining the toe shape. FIG. 3 is a schematic diagram showing the progression of the target cross-sectional shape in each manufacturing process of an NAB. FIG. 4 is a schematic diagram showing the cross-sectional shapes of an unequal-leg equal-thickness angle steel and a spherical flat steel. FIG. 5 is a schematic diagram showing the general configuration of a rolling facility including a cross-sectional dimension prediction device for an unequal-leg angle steel, which can implement the cross-sectional dimension prediction method for an unequal-leg angle steel according to this embodiment. FIG. 6 is a schematic diagram showing an example configuration of a cross-sectional dimension prediction device for an unequal-leg angle steel. FIG. 7 is a flow chart showing the flow of a rolling parameter identification process performed by a rolling parameter identification unit. FIG. 8 is a schematic diagram showing a method for generating a cross-sectional dimension prediction model by a cross-sectional dimension prediction model generation unit.

[0014] An embodiment of the present invention will be described below with reference to the drawings. In the following embodiment, the present invention will be described using an example in which the present invention is applied to NAB10, which is a type of unequal leg angle iron. The embodiment shown below exemplifies an apparatus and method for embodying the technical concept of the present invention, and the technical concept of the present invention is not limited to the following embodiment.

[0015] 3 is a schematic diagram showing the progression of the target cross-sectional shape in each manufacturing process of the NAB 10. The NAB 10 is manufactured by heating a steel billet 11 in a heating furnace, rolling the steel billet 11 in each rolling mill in a rough rolling process, an intermediate rolling process, and a finish rolling process, and then cooling the steel billet 11.

[0016] 4A and 4B are schematic diagrams showing the cross-sectional shapes of an unequal leg equal thickness angle iron 16 and a spherical flat iron 22. Fig. 4A is a schematic diagram showing the cross-sectional shape of an unequal leg equal thickness angle iron (hereinafter referred to as "ABS") 16, in which the thickness of the short side 18 is equal to the thickness of the long side 20. Fig. 4B is a schematic diagram showing the cross-sectional shape of a spherical flat iron (hereinafter referred to as "BP (valve plate)") 22, in which the short side 24 is tapered. ABS 16 and BP 22 are each a type of unequal leg angle iron.

[0017] ABS16 and BP22 are manufactured by the same groove rolling as NAB10. Therefore, the method for predicting the cross-sectional dimensions of unequal leg angle steel, the method for manufacturing unequal leg angle steel, and the rolling equipment including the device for predicting the cross-sectional dimensions of unequal leg angle steel according to this embodiment can also be applied to predicting the cross-sectional dimensions of ABS16 and BP22 and manufacturing ABS16 and BP22. Therefore, the unequal leg angle steels targeted by the method for predicting the cross-sectional dimensions of unequal leg angle steel, the method for manufacturing unequal leg angle steel, and the rolling equipment including the device for predicting the cross-sectional dimensions of unequal leg angle steel according to this embodiment include NAB10, ABS16, and BP22.

[0018] 5 is a schematic diagram showing the overall configuration of a rolling facility 100 including an unequal leg angle steel cross-sectional dimension prediction device 44 on which the method for predicting cross-sectional dimensions of unequal leg angle steel according to this embodiment can be implemented. As shown in FIG. 5, the rolling facility 100 includes a heating furnace 30 that heats a steel slab 11 to a predetermined temperature, a roughing mill 32 that rolls the steel slab 11 heated in the heating furnace 30, intermediate rolling mills 34 and 36, and a finishing rolling mill 38. The rolling facility 100 further includes a hot dimension gauge 40 that measures the hot dimensions of the unequal leg angle steel after finish rolling, a process computer 42, and an unequal leg angle steel cross-sectional dimension prediction device 44.

[0019] The billet 11 is charged into a heating furnace 30 and heated to a temperature above the austenite temperature range (for example, 1100 to 1300°C). The billet 11 extracted from the heating furnace 30 is transported to the rolling mills by a plurality of table rollers installed on the outlet side of the heating furnace 30, and is hot rolled using grooved rolls in a roughing mill 32, intermediate rolling mills 34 and 36, and a finishing rolling mill 38.

[0020] The hot dimensions (long side length A, short side length B, long side thickness t1, short side thickness t2, and short side toe shape C) of the finish-rolled NAB 10 are measured using a hot dimension gauge 40. Next, the NAB 10 is cut to a predetermined product length using a hot saw (not shown). The NAB 10 cut to the predetermined product length is cooled using a cooling device such as a cooling bed (not shown). In this manner, the steel billet 11 is hot-rolled to produce the NAB 10.

[0021] In the following description, the long side length of the NAB10 will be referred to as "A," the short side length as "B," the long side thickness as "t1," the short side thickness as "t2," and the short side foot shape as "C." In this embodiment, the cross-sectional dimension items refer to various dimensions of the NAB10, such as the long side length "A," the short side length "B," the long side thickness "t1," the short side thickness "t2," and the short side foot shape "C."

[0022] In rough rolling, a roughing mill 32 applies multiple passes of groove rolling to a rectangular cross-section steel billet 11 to adjust the size of the billet 11 and form it into a rough billet. In intermediate rolling, an intermediate rolling mill 34 applies multiple passes of groove rolling to the rough billet to form it into a shape with an upward convex shape near the joint between the long and short sides. Furthermore, in finish rolling, a finish rolling mill 38 having a caliber for the final product shape forms the rolled material into the product shape. The number of calibers through which the rolled material passes and the number of passes vary depending on the size of the product.

[0023] The hot dimension meter 40 measures the cross-sectional dimensions of the NAB 10 after finish rolling. The hot dimension meter 40 measures A, B, t1, t2, and C of the NAB 10 shown in FIG. 1 in the hot state as the cross-sectional dimensions of the NAB 10. The hot dimension meter 40 preferably measures the above cross-sectional dimensions over the entire length, and uses the average value of the measured cross-sectional dimensions.

[0024] The hot dimension gauge 40 measures the hot temperature used to convert the hot cross-sectional dimension into the cold cross-sectional dimension. The hot dimension gauge 40 outputs the measured hot dimensions and hot temperature of the NAB 10 to the process computer 42. In addition to the hot dimension gauge 40, a temperature sensor for measuring the hot temperature may be provided near the hot dimension gauge 40. In this case, the hot dimension gauge 40 does not need to have a temperature measurement function.

[0025] The process computer 42 can be a general-purpose computer such as a workstation or a personal computer. The process computer 42 is connected to the roughing mill 32, the intermediate rolling mills 34 and 36, the finishing mill 38, and the hot dimension gauge 40 by wire or wirelessly, and controls the manufacturing process of the NAB 10. The process computer 42 also acquires attribute parameters of the NAB 10 from a higher-level computer. The attribute parameters include information regarding the target dimensions of the cross-sectional dimensions of the NAB 10 and the steel type classification. The target dimensions of the NAB 10 are the target dimensions (typical values ​​and allowable values) of A, B, t1, t2, and C.

[0026] The process computer 42 sets rolling parameters such as the number of passes and roll spacing of each rolling mill in accordance with the attribute parameters of the NAB 10. The set values ​​of normal rolling parameters are set as table values ​​associated with the attribute parameters of the NAB 10 based on past rolling performance.

[0027] The process computer 42 outputs the set rolling parameters to the roughing mill 32, the intermediate mills 34 and 36, and the finishing mill 38. Each rolling mill controls the roll gap, roll position, rolling load, etc. of each rolling mill in accordance with the acquired rolling parameters. The process computer 42 acquires the rolling parameters, hot roll dimensions, hot roll temperature, identification numbers for identifying the roll sets of each rolling mill, etc.

[0028] When the process computer 42 acquires the hot dimensions and hot temperature of the NAB 10, it converts the hot cross-sectional dimensions into cold cross-sectional dimensions. The process computer 42 converts the hot cross-sectional dimensions of the NAB 10 into cold cross-sectional dimensions by predicting thermal contraction when the NAB 10 is cooled to room temperature in the rolling mill using the hot temperature and linear expansion coefficient. Here, the cold cross-sectional dimensions are the cross-sectional dimensions (A, B, t1, t2, C) of the NAB 10 after it has been cooled to room temperature in the rolling mill. The process computer 42 stores the attribute parameters, rolling parameters, and cross-sectional dimensions in a database of the process computer 42 in association with the production lot number of the NAB 10.

[0029] The unequal leg angle steel cross-sectional dimension prediction device 44 predicts the cross-sectional dimension of the subsequently rolled material of two unequal leg angle steels produced by hot rolling with the same roll set. The unequal leg angle steel cross-sectional dimension prediction device 44 predicts the cross-sectional dimension of the subsequently rolled material before the billet 11 of the subsequently rolled material is rolled in the roughing mill 32. Considering that the predicted cross-sectional dimension of the subsequently rolled material is used to identify the rolling parameters, which are the rolling conditions of each rolling mill, it is preferable that the unequal leg angle steel cross-sectional dimension prediction device 44 predicts the cross-sectional dimension of the billet 11 of the subsequently rolled material before it is extracted from the heating furnace 30.

[0030] The cross-sectional dimensions of the subsequently rolled material predicted by the unequal leg angle steel cross-sectional dimension prediction device 44 may be one or more of A, B, t1, t2, and C of NAB10. However, since all of the above cross-sectional dimensions are subject to management in NAB10, it is preferable that the unequal leg angle steel cross-sectional dimension prediction device 44 predicts all of the cross-sectional dimensions A, B, t1, t2, and C of NAB10. In the following embodiment, an example will be described in which the unequal leg angle steel cross-sectional dimension prediction device 44 predicts all of the cross-sectional dimensions of NAB10.

[0031] The unequal leg angle steel cross-sectional dimension prediction device 44 obtains the cross-sectional dimensions of the preceding rolled material, the rolling parameters of the preceding rolled material, and the rolling parameters of the succeeding rolled material from the process computer 42. The unequal leg angle steel cross-sectional dimension prediction device 44 inputs input data including these into a cross-sectional dimension prediction model and outputs the cross-sectional dimensions of the succeeding rolled material, thereby predicting the cold cross-sectional dimensions of the succeeding rolled material. In this embodiment, an example will be described in which the unequal leg angle steel cross-sectional dimension prediction device 44 predicts the cold cross-sectional dimensions of the succeeding rolled material, but this is not limiting, and the hot cross-sectional dimensions may also be predicted. When the unequal leg angle steel cross-sectional dimension prediction device 44 predicts the hot cross-sectional dimensions, the process computer 42 stores the hot cross-sectional dimensions in its database.

[0032] The unequal leg angle steel cross-sectional dimension prediction device 44 may specify rolling parameters that will cause the predicted cross-sectional dimensions of the subsequently rolled material to fall within the target dimension range. The unequal leg angle steel cross-sectional dimension prediction device 44 may output the specified rolling parameters to the process computer 42, thereby reflecting the rolling parameters in the rolling conditions for hot rolling.

[0033] Next, an unequal-leg angle steel cross-sectional dimension prediction device 44 that predicts the cross-sectional dimensions of the NAB 10 will be described. FIG. 6 is a schematic diagram showing an example of the configuration of the unequal-leg angle steel cross-sectional dimension prediction device 44. The unequal-leg angle steel cross-sectional dimension prediction device 44 can be a general-purpose computer such as a workstation or a personal computer. The unequal-leg angle steel cross-sectional dimension prediction device 44 has a control unit 46, an input unit 48, an output unit 50, and a storage unit 52. The control unit 46 is, for example, a CPU, and functions as a data acquisition unit 54, a cross-sectional dimension prediction unit 56, a rolling parameter identification unit 58, and a cross-sectional dimension prediction model generation unit 60 by executing a program stored in the storage unit 52.

[0034] The input unit 48 may be, for example, a keyboard, a touch panel integrated with a display, or the like. The output unit 50 may be, for example, an LCD or CRT display. The storage unit 52 may be, for example, an updatable flash memory, a hard disk built-in or connected via a data communication terminal, a memory card, or other information recording medium and its read / write device. The storage unit 52 stores programs and data for implementing each function of the unequal leg angle steel cross-sectional dimension prediction device 44. The storage unit 52 also stores a database 62 and a cross-sectional dimension prediction model 64. The database 62 stores a set of actual values ​​for a preceding rolled material and a following rolled material of two unequal leg angle steels previously manufactured using the same rolling equipment 100 (each rolling mill having the same roll set). The actual values ​​for the preceding rolled material and the following rolled material include the actual values ​​of the respective cross-sectional dimensions and rolling parameters. At least 1,000, preferably at least 3,000, data sets are stored. This data set is used as training data for generating a cross-sectional dimension prediction model 64. The preceding rolled material is the NAB 10 that has been rolled first by the same roll set, and the following rolled material is the NAB 10 that is rolled after the preceding rolled material. The preceding rolled material and the following rolled material may be rolled materials that are rolled consecutively, or may be rolled materials that are not rolled consecutively.

[0035] It is preferable that the database 62 sequentially acquires the above-mentioned data sets for the unequal leg angle steel manufactured by the rolling equipment 100 from the process computer 42 and stores them. In this case, an upper limit may be set on the number of data sets stored in the database 62, and the oldest data sets may be deleted so as not to exceed the upper limit.

[0036] The cross-section dimension prediction model 64 is a trained machine learning model that has been trained using multiple data sets stored in the database 62 as training data. The cross-section dimension prediction model 64 is a trained machine learning model that receives the rolling parameters of the preceding rolled material, the cross-section dimensions (A, B, t1, t2, C) of the preceding rolled material, and the rolling parameters of the succeeding rolled material as input, and outputs the cross-section dimensions (A, B, t1, t2, C) of the succeeding rolled material.

[0037] Next, we will explain the processing executed by the data acquisition unit 54 and the cross-sectional dimension prediction unit 56. The data acquisition unit 54 acquires the rolling parameters of the preceding rolled material, the cross-sectional dimensions (A, B, t1, t2, C) of the preceding rolled material, and the rolling parameters of the succeeding rolled material as input data from the process computer 42. The data acquisition unit 54 outputs the acquired input data to the cross-sectional dimension prediction unit 56.

[0038] When the cross-section dimension prediction unit 56 acquires the input data, it reads out the cross-section dimension prediction model 64 from the storage unit 52. The cross-section dimension prediction unit 56 inputs the input data into the cross-section dimension prediction model 64, and causes it to output the cross-section dimensions of the subsequently rolled material. In this way, the cross-section dimension prediction unit 56 predicts the cross-section dimensions of the subsequently rolled material. The cross-section dimension prediction unit 56 may also display the predicted cross-section dimensions of the subsequently rolled material on the output unit 50. This allows the operator to confirm the predicted value of the cross-section dimension of the subsequently rolled material by visually checking the output unit 50.

[0039] The input data for the cross-sectional dimension prediction model 64 includes the cross-sectional dimensions (A, B, t1, t2, C) of the preceding rolled material, the rolling parameters of the preceding rolled material, and the rolling parameters of the following rolled material. The preceding rolled material is NAB10 manufactured in a rolling facility 100 in which the roll sets of each rolling mill are the same, so there is a high correlation between the cross-sectional dimensions of the preceding rolled material and the cross-sectional dimensions of the following rolled material. Therefore, by including the cross-sectional dimensions of the preceding rolled material in the input data for the cross-sectional dimension prediction model 64, the prediction accuracy of the cross-sectional dimensions of the following rolled material by the cross-sectional dimension prediction model 64 is improved.

[0040] The rolling parameters of the preceding rolled material affect the cross-sectional dimensions of the preceding rolled material. As described above, the cross-sectional dimensions of the preceding rolled material are highly correlated with the cross-sectional dimensions of the following rolled material, and therefore the rolling parameters of the preceding rolled material are also correlated with the cross-sectional dimensions of the following rolled material. The rolling parameters of the following rolled material are highly correlated with the cross-sectional dimensions of the following rolled material. Therefore, by including the rolling parameters of the preceding rolled material and the following rolled material in the input data of the cross-sectional dimension prediction model 64, the prediction accuracy of the cross-sectional dimensions of the following rolled material by the cross-sectional dimension prediction model 64 is improved. The rolling parameters are, for example, one or more of the set values ​​or measured values ​​of the parameters that are rolling conditions for hot rolling shown below.

[0041] <Rolling parameters> - Time from when the billet 11 is tapped until it is charged into the heating furnace 30 - Temperature of the billet 11 when it is charged into the heating furnace 30 - Time the billet 11 is in the heating furnace 30 - Roll spacing for each pass of the roughing mill 32, the intermediate rolling mills 34, 36, and the finishing mill 38 - Total number of rotations of each roll of the roughing mill 32, the intermediate rolling mills 34, 36, and the finishing mill 38 - Guide positions before and after the roughing mill 32, the intermediate rolling mills 34, 36, and the finishing mill 38 - Month when rolling was performed - Thickness of the billet 11, width of the billet 11, length of the billet 11, weight of the billet 11

[0042] The guides are provided before and after the roughing mill 32, the intermediate mills 34 and 36, and the finishing mill 38 to adjust the pass line of the rolled material. This allows the width position of the rolled material to be aligned with the appropriate position of the groove, thereby controlling the thickness of the toe properly. As such, the guide position affects the rolling of the rolled material by the groove, and therefore correlates with the cross-sectional dimension of the subsequently rolled material. The temperature and humidity at the manufacturing site of NAB10 change depending on the month in which rolling is performed. Since the temperature and humidity at the manufacturing site affect the cross-sectional dimension of the subsequently rolled material, the month in which rolling is performed correlates with the cross-sectional dimension of the subsequently rolled material.

[0043] It is preferable that the input data of the cross-sectional dimension prediction model 64 further includes one or more attribute parameters of the NAB 10. The attribute parameters of the NAB 10 are, for example, the parameters described below.

[0044] <Attribute parameters> ・Reference dimensions of target dimensions ・Steel type classification

[0045] Of these attribute parameters, the reference dimension of the target dimension affects the cross-sectional dimension of the NAB 10. The steel type classification affects the composition and structure of oxide scale in the heating furnace 30, and affects the deformation resistance during hot rolling. Therefore, it can be said that these attribute parameters also affect the cross-sectional dimension of the NAB 10. Therefore, by including these attribute parameters in the input data of the cross-sectional dimension prediction model 64, the prediction accuracy of the cross-sectional dimension prediction model 64 is improved.

[0046] Next, a description will be given of the rolling parameter specification process performed by the rolling parameter specification unit 58. Fig. 7 is a flow chart showing the flow of the rolling parameter specification process performed by the rolling parameter specification unit 58. In the following description, an example will be given in which the rolling parameter specification unit 58 specifies the roll spacing in each pass of the roughing mill 32, the intermediate mills 34 and 36, and the finishing mill 38 as a rolling parameter.

[0047] 7 is started at the same time as the prediction of the cross-sectional dimension of the following rolled material is started. First, the data acquisition unit 54 acquires the cross-sectional dimension of the preceding rolled material, the roll spacing of the preceding rolled material, and the roll spacing of the following rolled material as input data from the process computer 42 (step S101). The data acquisition unit 54 outputs the acquired input data to the cross-sectional dimension prediction unit 56.

[0048] When the cross-sectional dimension prediction unit 56 acquires the input data, it reads out the cross-sectional dimension prediction model 64 from the storage unit 52, inputs the input data into the cross-sectional dimension prediction model 64, and outputs the cross-sectional dimensions of the NAB 10. As a result, the cross-sectional dimension prediction unit 56 predicts the cross-sectional dimensions of the NAB 10 cooled to room temperature (step S102). The cross-sectional dimension prediction unit 56 outputs the predicted cross-sectional dimensions to the rolling parameter identification unit 58.

[0049] The rolling parameter specifying unit 58 determines whether or not all the predicted cross-sectional dimensions are within the range of the target dimensions of NAB 10 (step S103). The range of the target dimensions of NAB 10 may be stored in advance in the storage unit 52, or may be input by an operator via the input unit 48.

[0050] The rolling parameter specifying unit 58 compares all the predicted cross-sectional dimensions with the target dimension range for each dimension. If any one of the predicted cross-sectional dimensions is not included in the target dimension range, the rolling parameter specifying unit 58 determines that the predicted cross-sectional dimension is not within the target dimension range (step S103: No).

[0051] If the rolling parameter specifying unit 58 determines that the predicted cross-sectional dimensions are not within the range of the target dimensions, it changes the roll spacing in each pass of each rolling mill, which is the rolling parameter of the subsequently rolled material used as input data (step S104). The roll spacing is changed, for example, by changing the roll spacing in increments of 20 levels within the allowable range of the roll spacing, and the processing from step S102 is executed again using input data including the changed roll spacing. In step S103, the rolling parameter specifying unit 58 repeatedly executes the processing of steps S104, S102, and S103 until all predicted cross-sectional dimensions are included in the range of the target dimensions.

[0052] On the other hand, if all the predicted cross-sectional dimensions are included in the range of the target dimensions, the rolling parameter specifying unit 58 determines that the predicted cross-sectional dimensions are within the range of the target dimensions (step S103: Yes). The rolling parameter specifying unit 58 specifies the roll gap used in the input data as a roll gap that can bring the cross-sectional dimensions within the range of the target dimensions (step S105). If it is determined in the first processing of step S103 that the cross-sectional dimensions are within the range of the target dimensions, there is no need to change the roll gap, and therefore the processing of steps S105 to S107 may be skipped.

[0053] The rolling parameter specifying unit 58 outputs the specified roll gap to the process computer 42, which reflects the specified roll gap in the rolling conditions of each rolling mill (step S106), and this process ends. Thereafter, in the rolling facility 100, the NAB 10 is produced by hot rolling under rolling conditions including the specified roll gap. This makes it possible to produce an NAB 10 that falls within the range of the target cross-sectional dimensions.

[0054] In the processing of step S104, the rolling parameter specifying unit 58 may set all rolling parameters to about 20 levels within the tolerance range centered on the reference dimension, and create input data by combining these. For example, if the rolling parameter is the roll spacing for each pass of each rolling mill, the rolling parameter specifying unit 58 may set the roll spacing for each pass to 10 levels in 0.1 mm increments centered on the reference rolling condition for both the widening and narrowing sides of the roll spacing for each pass of each rolling mill, and input the combined input data into the cross-sectional dimension prediction model.

[0055] The rolling parameter specifying unit 58 specifies a combination of roll gaps for the subsequently rolled material included in the input data used to output the predicted value of the cross-sectional dimension that is within the range of the allowable value and is the same as or closest to the representative value, among all the output predicted values ​​of the cross-sectional dimension. The rolling parameter specifying unit 58 may specify the specified combination of roll gaps as a roll gap that can roll NAB10 having the target cross-sectional dimension, and output the combination of roll gaps to the process computer 42.

[0056] Next, a method for generating the cross-sectional dimension prediction model 64 used to predict the cross-sectional dimensions of the NAB 10 will be described. The data acquisition unit 54 acquires, from the process computer 42, actual values ​​of the cross-sectional dimensions of the preceding and succeeding rolled materials of the NAB 10 previously manufactured by the rolling equipment 100, as well as actual values ​​of the rolling parameters of the preceding and succeeding rolled materials. The data acquisition unit 54 stores a set of data sets, each consisting of the acquired actual values, in the database 62 of the storage unit 52. The preceding and succeeding rolled materials are not limited to unequal leg angle steels hot-rolled consecutively, and the succeeding rolled material may be any unequal leg angle steel that has been hot-rolled after the preceding rolled material. The number of data sets stored in the database 62 is preferably 1,000 or more, and more preferably 3,000 or more.

[0057] Fig. 8 is a schematic diagram showing a method for generating a cross-sectional dimension prediction model 64 by the cross-sectional dimension prediction model generation unit 60. As shown in Fig. 8, the cross-sectional dimension prediction model generation unit 60 reads out a machine learning model stored in advance in the storage unit 52, and performs machine learning on the machine learning model using a data set stored in the database 62 as training data to generate a trained machine learning model. The data set serving as training data is a data set that includes, as a set, the actual value of the cross-sectional dimension of the preceding rolled material, the actual value of the cross-sectional dimension of the succeeding rolled material, the actual value of the rolling parameters of the preceding rolled material, and the actual value of the rolling parameters of the succeeding rolled material.

[0058] The machine learning model trained in this manner becomes the cross-sectional dimension prediction model 64. The cross-sectional dimension prediction model generation unit 60 stores the generated cross-sectional dimension prediction model 64 in the storage unit 52. The machine learning model used in this embodiment may be any of the commonly used neural networks, decision tree learning, random forests, support vector regression, Gaussian processes, and nearest neighbor methods.

[0059] The cross section dimension prediction model generating unit 60 may update the cross section dimension prediction model 64 to a new cross section dimension prediction model by, for example, retraining the machine learning model every month or every year. The data acquiring unit 54 acquires a data set every time an NAB 10 is manufactured in the rolling equipment 100 and stores the data set in the database 62. Therefore, the database 62 stores performance data of newly manufactured NAB 10, and the number of data sets stored in the database 62 increases.

[0060] As the number of data sets and the number of training data increase, the prediction accuracy of the cross-sectional dimension prediction model 64 improves, making it possible to predict the cross-sectional dimensions of the subsequently rolled material with higher accuracy. Furthermore, by using a new data set, the latest state of the rolling equipment 100 is reflected in the cross-sectional dimension prediction model 64, so by updating the cross-sectional dimension prediction model 64, it becomes possible to predict the cross-sectional dimensions of the subsequently rolled material with even higher accuracy.

[0061] As described above, the cross-sectional dimension prediction device 44 and cross-sectional dimension prediction method for unequal leg angle steel according to this embodiment make it possible to predict the cross-sectional dimensions of unequal leg angle steel manufactured by groove rolling. If the cross-sectional dimensions of unequal leg angle steel can be predicted, it becomes possible to identify the rolling conditions for groove rolling under which the predicted cross-sectional dimensions fall within the target dimension range, thereby realizing the manufacture of unequal leg angle steel whose cross-sectional dimensions fall within the target dimension range.

[0062] The present invention is not limited to the above embodiment and various modifications may be made. In the present embodiment, the cross-sectional dimension prediction method has been described as an example in which all cross-sectional dimensions (A, B, t1, t2, C) of the NAB 10 are predicted. However, the present invention is not limited to this. The cross-sectional dimension prediction method according to the present embodiment is only required to predict one or more of the cross-sectional dimensions A, B, t1, t2, and C. In this case, the input data of the cross-sectional dimension prediction model 64 used to predict the cross-sectional dimensions only needs to include one or more cross-sectional dimensions to be predicted. For example, when predicting A of a subsequently rolled material using the cross-sectional dimension prediction method for unequal leg angle steel according to the present embodiment, the input data of the cross-sectional dimension prediction model 64 only needs to include at least A of the preceding rolled material, and may also include other cross-sectional dimensions (B, t1, t2, C). In other words, the cross-sectional dimension item of the preceding rolled material used in the input data of the cross-sectional dimension prediction model 64 only needs to include the cross-sectional dimension item of the subsequently rolled material to be output from the cross-sectional dimension prediction model.

[0063] In the present embodiment, an example has been shown in which the rolling equipment 100 shown in Fig. 5 has the process computer 42 and the unequal leg angle steel cross-sectional dimension prediction device 44, but this is not limiting. For example, the process computer 42 may have the function of the unequal leg angle steel cross-sectional dimension prediction device 44, and these may be configured as a single device.

[0064] In the present embodiment, an example has been shown in which the control unit 46 in the unequal leg angle steel cross-sectional dimension prediction device 44 shown in Fig. 6 has the data acquisition unit 54, the cross-sectional dimension prediction unit 56, the rolling parameter identification unit 58, and the cross-sectional dimension prediction model generation unit 60, but this is not limited to this. If the unequal leg angle steel cross-sectional dimension prediction method is implemented in the unequal leg angle steel cross-sectional dimension prediction device 44, the control unit 46 does not need to have the rolling parameter identification unit 58. If the cross-sectional dimension prediction model 64 is generated externally and stored in the storage unit 52, the cross-sectional dimension prediction model generation unit 60 does not need to be included.

[0065] In this embodiment, an example has been described in which the cross-sectional dimensions of NAB10 are predicted using the unequal leg angle steel cross-sectional dimension prediction device 44, but the unequal leg angle steel for which cross-sectional dimensions are predicted is not limited to NAB10. The unequal leg angle steel cross-sectional dimension prediction device 44 can similarly predict the cross-sectional dimensions of ABS16 and BP22, which are unequal leg angle steels manufactured by groove rolling.

[0066] Next, an example will be described in which a steel billet was groove-rolled using the rolling equipment 100 shown in Figure 5 to produce an NAB having a long side length of 250 mm and a short side length of 90 mm. In this example, the cross-sectional dimensions (A, B, t1, t2, C) of the subsequently rolled material were predicted using the rolling parameters and cross-sectional dimensions of the preceding rolled material and the rolling parameters of the subsequently rolled material. The output data and input data of the cross-sectional dimension prediction model used for the prediction are shown in Table 1 below.

[0067]

[0068] The cross-sectional dimensions of the preceding and following rolled materials were measured using a hot dimension gauge installed downstream of the finishing mill (A, B, t1, t2, C), and the cold (room temperature) cross-sectional dimensions were calculated by converting the hot dimensions to room temperature dimensions using the hot temperature and linear expansion coefficient. To generate the cross-sectional dimension prediction model, actual manufacturing data for 1,000 NABs was used as training data, and a neural network with six intermediate layers and 300 nodes each was trained to generate the cross-sectional dimension prediction model. A sigmoid function was used as the activation function.

[0069] Using this cross-sectional dimension prediction model, the roll spacing of each pass of the roughing mill, intermediate mill, and finishing mill for the following rolled material was specified so that all of the cross-sectional dimensions of the following rolled material would be within the target dimension range. The roll spacing of each of the roughing mill, intermediate mill, and finishing mill was specified based on the reference rolling conditions of each rolling mill, and 10 levels were set in 0.1 mm increments for each of the sides where the roll spacing was uniformly widened and narrowed for each pass of each rolling mill. Then, input data combining these roll spacings was input into the cross-sectional dimension prediction model, and predicted values ​​of the cross-sectional dimensions for all combinations were output.

[0070] Among the output cross-sectional dimensions, the roll spacing of the subsequently rolled material included in the input data used to output the predicted value of the cross-sectional dimension that was the same as or closest to the reference dimension was identified as the roll spacing that could roll an NAB with the target cross-sectional dimension. One hundred NABs were manufactured by hot rolling under rolling conditions that included the roll spacing of each pass identified in this way. When the cross-sectional dimensions of the manufactured NABs were measured, all of the cross-sectional dimensions of the 100 NABs were found to be within the target dimension range.

[0071] In contrast, the roll gaps of the hot rolling mill, intermediate rolling mill, and finishing rolling mill were set manually by an operator as in the previous example, and 100 NABs were manufactured under rolling conditions including the roll gaps set in the same manner as in the above example. When the cross-sectional dimensions of the manufactured NABs were measured, it was found that some of the cross-sectional dimensions of 17 of the 100 NABs were outside the tolerances, requiring additional cross-sectional dimension correction or remanufacturing. This resulted in an increase in the manufacturing cost of the NABs.

[0072] Thus, all of the cross-sectional dimensions of all 100 NABs manufactured using the roll gap specified using the cross-sectional dimension prediction model were within the target dimension range. Such results would not be obtained unless the cross-sectional dimensions of the NAB of the subsequently rolled material could be predicted with high accuracy. Therefore, the above results confirm that the cross-sectional dimension prediction method according to this embodiment can predict with high accuracy the cross-sectional dimensions of the unequal leg angle steel that will be the subsequently rolled material. Furthermore, it was confirmed that the prediction method can be used to identify rolling parameters for unequal leg angle steel that fall within the target dimension range, and that unequal leg angle steel with the target dimension can be manufactured by rolling under rolling conditions that include these rolling parameters.

[0073] DESCRIPTION OF SYMBOLS 10 NAB 11 Steel billet 12 Long side 14 Short side 16 Unequal leg equal thickness angle steel (ABS) 18 Short side 20 Long side 22 Spherical flat steel (BP) 24 Short side 30 Heating furnace 32 Roughing rolling mill 34 Intermediate rolling mill 36 Intermediate rolling mill 38 Finishing rolling mill 40 Hot dimension gauge 42 Process computer 44 Unequal leg angle steel cross-sectional dimension prediction device 46 Control unit 48 Input unit 50 Output unit 52 Storage unit 54 Data acquisition unit 56 Cross-sectional dimension prediction unit 58 Rolling parameter identification unit 60 Cross-sectional dimension prediction model generation unit 62 Database 64 Cross-sectional dimension prediction model 100 Rolling equipment

Claims

1. A method for predicting the cross-sectional dimensions of an unequal-sided angle steel, which predicts the cross-sectional dimensions of the later-rolled material among two unequal-sided angle steels manufactured by hot rolling in the same roll set, Input data including the cross-sectional dimensions of the first-rolled material among the two unequal-sided angle steels, rolling parameters including the roll spacing of the roughing mill, intermediate mill, and finish mill of the first-rolled material, and rolling parameters including the roll spacing of the roughing mill, intermediate mill, and finish mill of the second-rolled material are input to the cross-sectional dimension prediction model, and the cross-sectional dimensions of the second-rolled material are output to predict the cross-sectional dimensions of the second-rolled material. A method for predicting the cross-sectional dimensions of an unequal-sided angle steel, wherein the item for the cross-sectional dimensions of the preceding rolled material used in the input data includes an item for the cross-sectional dimensions of the succeeding rolled material output from the cross-sectional dimension prediction model.

2. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 1, wherein the input data includes the guide positions of the roughing mill, intermediate mill, and finish mill as rolling parameters for the preceding rolled material, and the guide positions of the roughing mill, intermediate mill, and finish mill as rolling parameters for the succeeding rolled material.

3. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 1, wherein the input data includes one or more attribute parameters of the subsequent rolled material.

4. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 2, wherein the input data includes one or more of the attribute parameters of the subsequent rolled material.

5. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 1, wherein the cross-sectional dimensions of the subsequent rolled material output from the cross-sectional dimension prediction model include the length of the long side, the length of the short side, the thickness of the long side, the thickness of the short side, and the shape of the short side toe of the unequal-sided angle steel.

6. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 2, wherein the cross-sectional dimensions of the subsequent rolled material output from the cross-sectional dimension prediction model include the length of the long side, the length of the short side, the thickness of the long side, the thickness of the short side, and the shape of the short side toe of the unequal-sided angle steel.

7. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 3, wherein the cross-sectional dimensions of the subsequent rolled material output from the cross-sectional dimension prediction model include the length of the long side, the length of the short side, the thickness of the long side, the thickness of the short side, and the shape of the short side toe of the unequal-sided angle steel.

8. The method for predicting the cross-sectional dimensions of an unequal-sided angle steel according to claim 4, wherein the cross-sectional dimensions of the subsequent rolled material output from the cross-sectional dimension prediction model include the length of the long side, the length of the short side, the thickness of the long side, the thickness of the short side, and the shape of the short side toe of the unequal-sided angle steel.

9. The rolling parameters are identified such that the cross-sectional dimensions of the subsequent rolled material predicted by the cross-sectional dimension prediction method for unequal-sided angle steel according to any one of claims 1 to 8 are within the range of the target dimensions. A method for producing unequal-leg angle steel, comprising hot rolling under rolling conditions including specified rolling parameters.

10. A method for generating a cross-sectional dimension prediction model used to predict the cross-sectional dimensions of the later-rolled material among two unequal-sided angle steels manufactured by hot rolling with the same roll set, A machine learning model is trained using multiple datasets as training data, each dataset consisting of actual values ​​of input data including the cross-sectional dimensions of a previously manufactured rolled material using the same roll set, rolling parameters including the roll spacing of the roughing, intermediate, and finishing mills of the previously manufactured rolled material, and rolling parameters including the roll spacing of the roughing, intermediate, and finishing mills of the subsequent rolled material, along with actual values ​​of the cross-sectional dimensions of the subsequent rolled material. This generates a cross-sectional dimension prediction model that takes the input data as input and outputs the cross-sectional dimensions of the subsequent rolled material. A method for generating a cross-sectional dimension prediction model, wherein the item for the cross-sectional dimension of the preceding rolled material used in the input data includes an item for the cross-sectional dimension of the succeeding rolled material output from the cross-sectional dimension prediction model.

11. The method for generating a cross-sectional dimension prediction model according to claim 10, wherein the input data includes the guide positions of the roughing mill, intermediate mill, and finish mill as rolling parameters for the preceding rolled material, and the guide positions of the roughing mill, intermediate mill, and finish mill as rolling parameters for the succeeding rolled material.

12. An unequal-leg angle steel cross-sectional dimension prediction device that predicts the cross-sectional dimensions of the later-rolled material among two unequal-leg angle steels manufactured by hot rolling with the same roll set, The system includes a cross-sectional dimension prediction unit that inputs input data, including the cross-sectional dimensions of the first-rolled material among the two unequal-sided angle steels, rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing mill of the first-rolled material, and rolling parameters including the roll spacing of the roughing mill, intermediate rolling mill, and finishing mill of the second-rolled material, into a cross-sectional dimension prediction model, and outputs the cross-sectional dimensions of the second-rolled material to predict the cross-sectional dimensions of the second-rolled material. A cross-sectional dimension prediction device for unequal-sided angle steel, wherein the input data includes an item for the cross-sectional dimensions of the preceding rolled material, which is output from the cross-sectional dimension prediction model, as an item for the cross-sectional dimensions of the succeeding rolled material.

13. The cross-sectional dimension prediction device for unequal-sided angle steel according to claim 12, wherein the input data includes the guide positions of the roughing mill, intermediate rolling mill, and finishing mill as rolling parameters for the preceding rolled material, and the guide positions of the roughing mill, intermediate rolling mill, and finishing mill as rolling parameters for the succeeding rolled material.

14. The cross-sectional dimension prediction device for unequal-sided angle steel according to claim 12 or claim 13, wherein the input data includes one or more attribute parameters of the subsequent rolled material.