Width prediction method for rough rolled material, width control method for rough rolled material, method for manufacturing hot-rolled steel sheet, and method for generating width prediction model for rough rolled material

Abstract

This method for predicting the width of a rough rolled product predicts the width of a rough rolled product in a hot rolling line provided with: a heating furnace for heating a slab; a width reduction press device that intermittently performs width reduction on the heated slab; a roughing mill for manufacturing a roughed material by roughing the slab after width reduction; and a finishing mill for producing a hot-rolled steel sheet by finishing rolling the rough rolled material, the method comprising the steps of: predicting the width distribution in the longitudinal direction of the rough rolled material using a width prediction model learned by machine learning, and predicting the width distribution in the longitudinal direction of the rough rolled material using a width prediction model learned by machine learning; the width prediction model includes, as input data, one or more operating parameters selected from among operating parameters of a width press device and one or more operating parameters selected from among operating parameters of a roughing mill, and information on the width distribution in the longitudinal direction of the roughing stock is used as output data.

Description

Technical Field

[0001] The present invention relates to a method for predicting the width of a rough-rolled piece in a hot rolling line, a method for controlling the width of a rough-rolled piece, a method for manufacturing a hot-rolled steel sheet, and a method for generating a width prediction model for a rough-rolled piece. Background Art

[0002] In a hot rolling line, first, a slab as a raw material of a steel sheet is heated by a heating furnace, the width of the slab is adjusted by a width reduction stamping device (finishing press), and a semi-finished steel sheet called a rough-rolled slab (rough-rolled piece) with a thickness of about 30 to 50 mm is manufactured by rough rolling using one or more than two roughing mills. Next, after cutting the front and rear end portions of the rough-rolled slab by a crop shear, the rough-rolled slab is finish-rolled by 5 to 7 finishing mills capable of continuous rolling to manufacture a steel strip with a thickness of about 1.0 to 25.0 mm. Then, finally, the steel strip in a high-temperature state is cooled by a cooling device of an output roller table and wound by a coiler (reeling machine) to become a hot-rolled steel strip. In the hot rolling line, in the width reduction stamping device, the roughing mill, and the finishing mill, plastic deformation is imparted to the steel sheet in the thickness direction and the width direction, so the width of the steel sheet varies complexly in the manufacturing process of the hot-rolled steel strip. On the other hand, the width accuracy of the hot-rolled steel strip directly affects the product yield. Therefore, in the hot rolling line, the width of the steel sheet at the stage before the end of rough rolling and before loading into the finishing mill is controlled (rough width control), and the width of the steel sheet is controlled during the process of passing through the finishing mill (fine width control).

[0003] In the hot rolling line, the width of the steel sheet changes for various reasons, so various techniques for improving the width accuracy of the hot-rolled steel strip have been proposed. For example, Patent Document 1 describes a method for predicting the width change amount under finish rolling at the skid mark portion based on the measurement result of the temperature distribution in the rolling direction of the rolled piece or the change in the rolling load of the roughing mill. In addition, Patent Document 2 describes a method of forming a plurality of stepped portions from the front end portion and the rear end portion of the steel sheet toward the stable portion by using a width reduction stamping device in order to suppress the width variation of the front end portion and the rear end portion of the steel sheet. In addition, Patent Document 3 describes a method of predicting the width variation of the steel sheet after horizontal rolling by learning the relationship between the actual values of the operating conditions of the vertical rolling mill (edging mill) and the horizontal rolling mill constituting the roughing mill and the actual value of the width of the steel sheet after horizontal rolling, and setting the opening degree of the vertical rolling mill so that the difference between the predicted value and the actual value of the width of the steel sheet after horizontal rolling is zero.

[0004] Patent Document 1: Japanese Patent No. 2968647 Gazette

[0005] Patent Document 2: Japanese Patent Laid-Open No. 5-200411 Gazette

[0006] Patent Document 3: Japanese Patent Publication No. 3260616

[0007] The method described in Patent Document 1 predicts the amount of width change in the finishing mill due to the temperature distribution in the rolling direction of the rolled piece generated in the heating furnace. However, the length and width of the rolled piece change due to width reduction, rough rolling, etc. Therefore, it is necessary to determine which position of the rolled piece the temperature distribution generated in the heating furnace corresponds to. However, during the period until the slab becomes a rough-rolled piece, the length of the rolled piece extends about 5 to 10 times, so it is difficult to accurately track the temperature distribution generated in the heating furnace. Therefore, it is difficult to predict the amount of width change with high precision to suppress the occurrence of width defects in the hot-rolled steel sheet.

[0008] On the other hand, the method described in Patent Document 2 uses a width reduction stamping device to suppress the width variation at the front end and the rear end of the steel sheet. However, it is difficult to accurately predict the complex metal flow lines at the front end and the rear end of the steel sheet to form corresponding stepped portions. Therefore, there is a deviation in the width of the steel sheet, and there is a possibility of occurrence of width defects in the hot-rolled steel sheet. In addition, Patent Document 3 describes predicting the width variation of the steel sheet after horizontal rolling by learning the actual values of the operating conditions of the vertical rolling mill and the horizontal rolling mill. However, the width distribution in the length direction of the steel sheet is affected not only by the operating conditions of rough rolling but also by the temperature distribution of the slab generated in the heating furnace and the width variation based on the width reduction stamping device. Therefore, it is difficult to predict the width variation of the steel sheet after horizontal rolling with high precision. Summary of the Invention

[0009] The present invention has been completed to solve the above problems, and an object thereof is to provide a method for predicting the width distribution in the length direction of a rough-rolled piece with high precision. Another object of the present invention is to provide a method for controlling the width of a rough-rolled piece that can accurately control the width in the length direction of the rough-rolled piece. Another object of the present invention is to provide a method for manufacturing a hot-rolled steel sheet capable of manufacturing a hot-rolled steel sheet having excellent width accuracy in the length direction. Another object of the present invention is to provide a method for generating a width prediction model for a rough-rolled piece that can generate a width prediction model for accurately predicting the width distribution in the length direction of the rough-rolled piece.

[0010] The width prediction method for rough-rolled pieces of the present invention predicts the width of the above-mentioned rough-rolled pieces in the hot rolling line. The above-mentioned hot rolling line includes: a heating furnace that heats the slab; a width reduction stamping device that intermittently performs width reduction on the heated slab; a rough rolling mill that rough-rolls the slab after width reduction to manufacture rough-rolled pieces; and a finish rolling mill that finish-rolls the above-mentioned rough-rolled pieces to manufacture hot-rolled steel plates. Among them, it includes the following steps: using a width prediction model learned through machine learning, predicting the width distribution in the length direction of the above-mentioned rough-rolled pieces. The above-mentioned width prediction model includes, as input data, one or more operating parameters selected from the operating parameters of the above-mentioned width reduction stamping device and one or more operating parameters selected from the operating parameters of the above-mentioned rough rolling mill, and uses the information on the width distribution in the length direction of the above-mentioned rough-rolled pieces as output data.

[0011] The operating parameters of the above-mentioned width reduction stamping device and the operating parameters of the above-mentioned rough rolling mill can be operating parameters determined as representative values in the length direction of the above-mentioned rough-rolled pieces.

[0012] The above-mentioned width prediction model can include, as the above-mentioned input data, one or more operating parameters selected from the operating parameters of the above-mentioned heating furnace.

[0013] The above-mentioned width prediction model can include, as the above-mentioned input data, one or more attribute parameters selected from the attribute information of the above-mentioned slab.

[0014] The width control method for rough-rolled pieces of the present invention uses the width prediction method for rough-rolled pieces of the present invention to predict the width distribution in the length direction of the above-mentioned rough-rolled pieces, and sets the operating parameters of the above-mentioned rough rolling mill based on the predicted width distribution.

[0015] The manufacturing method of the hot-rolled steel plate of the present invention performs finish rolling on the rough-rolled piece whose width is controlled by using the width control method for rough-rolled pieces of the present invention, using the above-mentioned finish rolling mill, to manufacture a hot-rolled steel plate.

[0016] A method for generating a width prediction model of a rough rolling workpiece of the present invention generates a width prediction model for predicting the width of the rough rolling workpiece in the hot rolling line. The hot rolling line includes: a heating furnace that heats a slab; a width reduction stamping device that intermittently performs width reduction on the heated slab; a rough rolling mill that rough rolls the slab after width reduction to produce a rough rolling workpiece; and a finishing mill that finishes the rough rolling workpiece to produce a hot rolled steel sheet. The method includes the following steps: obtaining a plurality of learning data, and generating the width prediction model through machine learning using the obtained plurality of learning data. The plurality of learning data includes, as input actual data, one or more operation actual data selected from the operation parameters of the width reduction stamping device and one or more operation actual data selected from the operation parameters of the rough rolling mill, and includes, as output actual data, information on the width distribution in the length direction of the rough rolling workpiece.

[0017] According to the width prediction method of the rough rolling workpiece of the present invention, the width distribution in the length direction of the rough rolling workpiece can be predicted with high accuracy. In addition, according to the width control method of the rough rolling workpiece of the present invention, the width in the length direction of the rough rolling workpiece can be controlled with high accuracy. In addition, according to the manufacturing method of the hot rolled steel sheet of the present invention, a hot rolled steel sheet with excellent width accuracy in the length direction can be manufactured. In addition, according to the method for generating a width prediction model of the rough rolling workpiece of the present invention, a width prediction model that can accurately predict the width distribution in the length direction of the rough rolling workpiece can be generated. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram showing a structural example of a hot rolling line to which the present invention is applied.

[0019] Figure 2 It is shown Figure 1 A schematic diagram showing a structural example of the heating furnace shown.

[0020] Figure 3 Viewed from above Figure 1 A schematic diagram of the interior of the heating furnace shown.

[0021] Figure 4 It is shown Figure 1 A schematic diagram showing a structural example of the width reduction stamping device shown.

[0022] Figure 5 It shows the composition of Figure 1 A diagram showing the structure of the stand of the rough rolling mill shown.

[0023] Figure 6 It is a schematic diagram for explaining an optical width measurement method.

[0024] Figure 7It is a block diagram showing the structure of a width prediction model generation unit as an embodiment of the present invention.

[0025] Figure 8 It is a schematic diagram showing the structure of a width prediction model using a neural network.

[0026] Figure 9 It is a diagram showing an example of the width distribution of a slab after width reduction by a width reduction stamping device.

[0027] Figure 10 It is a diagram showing the width reduction at the front end and the tail end of a rolled product.

[0028] Figure 11 It is a diagram showing the width distribution formed when a rolled product with a constant width in the length direction is horizontally rolled without width reduction using an edger.

[0029] Figure 12 It is a block diagram showing the structure of a width prediction unit as an embodiment of the present invention.

[0030] Figure 13 It is a diagram showing the width distribution of a rough rolled product after width reduction of a slab using a width reduction stamping device and then performing one pass of width rolling using an edger and one pass of horizontal rolling using a horizontal rolling mill.

[0031] Figure 14 It is a diagram showing the prediction result of the width distribution in the length direction of the rough rolled product in the example.

[0032] Figure 15 It is a diagram showing the prediction result of the width distribution in the length direction of the rough rolled product in the example.

[0033] Figure 16 It is a diagram showing the prediction result of the width distribution in the length direction of the rough rolled product in the example. Detailed Embodiments

[0034] Hereinafter, a method for predicting the width of a rough rolled product, a method for controlling the width of a rough rolled product, a method for manufacturing a hot rolled steel sheet, and a method for generating a width prediction model of a rough rolled product as an embodiment of the present invention will be described in detail with reference to the drawings.

[0035] <Hot rolling line>

[0036] First, with reference to Figures 1 to 6 the structure of a hot rolling line to which the present invention is applied will be described.

[0037] Figure 1 It is a schematic diagram showing an example of the structure of a hot rolling line to which the present invention is applied. As Figure 1As shown in the figure, the hot rolling line 1 of the present invention includes: a heating furnace 2, a descaling device 3, a width reduction stamping device 4, a rough rolling mill 5, a finish rolling mill 6, a cooling device 7, and a coiler (reeling machine) 8. The unillustrated cast slab is heated to a specified set temperature after being charged into the heating furnace 2, and is taken out from the heating furnace 2 as a hot slab. The hot slab taken out from the heating furnace 2 is width-reduced to a specified set width by the width reduction stamping device 4 after the primary scale formed on the surface is removed by the descaling device 3. Then, the width-reduced slab becomes a rough rolling billet (rough rolling piece) by being rolled to a specified thickness in the rough rolling mill 5, and is conveyed to the finish rolling mill 6. In the finish rolling mill 6, the rough rolling billet is rolled to the product thickness by 5 to 7 continuous rolling mills. The device called the output roller table on the downstream side of the finish rolling mill 6 is equipped with a cooling device 7, and the steel plate is wound into a coil shape by the coiler 8 after being cooled to a specified temperature. In addition, a plurality of width gauges are provided as width measurement units in the middle of the conveying process of the hot rolling line 1. In Figure 1 In the example shown, a rough rolling exit side width gauge 11 is provided on the exit side of the rough rolling mill 5, and a finish rolling exit side width gauge 12 is provided on the exit side of the finish rolling mill 6. In addition, a coiler front width gauge (coiler inlet side width gauge) 13 for measuring the width of the steel plate before coiling is provided on the exit side of the cooling device 7. Hereinafter, the width of the rough rolling billet measured by the rough rolling exit side width gauge 11 is sometimes referred to as the rough rolling exit side width, the width of the steel plate measured by the finish rolling exit side width gauge 12 is referred to as the finish rolling exit side width, and the width of the steel plate measured by the coiler front width gauge 13 is referred to as the coiler front width.

[0038] The hot rolling line 1 includes: a control controller (PLC) 90 that controls each device constituting the hot rolling line 1; a control computer (process computer) 91 that gives control instructions to the control controller 90; and an upper computer 92 that gives manufacturing instructions to the hot rolling line 1. The control computer 91 sets the target value of the width on the rough rolling exit side (rough rolling target width), the target value of the width on the finish rolling exit side (finish rolling target width), and the target value of the width in front of the coiler (coiler front target width) based on the upper computer 92 or the manufacturing instructions from the upper computer 92, and sets the operating conditions of the rough rolling mill 5 and the finish rolling mill 6 to perform width control of the steel plate in the hot rolling line 1. Specifically, the upper computer 92 or the control computer 91 sets the finish rolling target width based on the coiler front target width determined according to the product specifications of the hot rolled steel plate, considering the width change amount of the steel plate generated between the exit side of the finish rolling mill 6 and the coiler front width gauge 13. In addition, the upper computer 92 or the control computer 91 sets the rough rolling target width based on the set finish rolling target width, considering the width change amount of the steel plate in the finish rolling mill 6. Then, the upper computer 92 or the control computer 91 sets the rolling conditions for each pass of rough rolling so that the width of the rough rolled billet after rough rolling is consistent with the rough rolling target width. In addition, the upper computer 92 or the control computer 91 sets the rolling conditions for each stand of finish rolling so that the width of the steel plate after finish rolling is consistent with the finish rolling target width. In addition, the upper computer 92 or the control computer 91 sometimes sets the tension between the finish rolling mill 6 and the coiler 8 and the cooling conditions of the cooling device 7 so that the width of the steel plate before coiling is consistent with the coiler front target width. In this case, there is also a case where dynamic width control is performed in the finish rolling mill 6 while referring to the measured values of the rough rolling exit side width and the finish rolling exit side width. The control controller 90 has the following functions: in addition to the information obtained from the width gauges provided on the hot rolling line 1, it collects the information obtained from various sensors (such as thickness gauges, thermometers, etc.) at a specified sampling period and outputs them to the control computer 91.

[0039] The width prediction method of the rough rolled piece as an embodiment of the present invention is a method for predicting the width distribution in the length direction of the rough rolled billet measured by the rough rolling exit side width gauge 11 provided on the exit side of the rough rolling mill 5. In addition, the width control method of the rough rolled piece as an embodiment of the present invention is a method for controlling the width in the length direction of the rough rolled billet so that the width of the rough rolled billet after rough rolling is consistent with the rough rolling target width.

[0040] 〔Heating furnace〕

[0041] Figure 2 It represents Figure 1 The schematic diagram of the structural example of the heating furnace 2 shown. As Figure 2 shown, in this embodiment, the slab SA is from Figure 2The left side of [it] is loaded into the heating furnace 2. The temperature of the slab SA loaded into the heating furnace 2 may be cooled to around room temperature in the slab yard after casting, or may reach a temperature of around 600 °C during the cooling process. In addition, there is also a case where it is loaded at a temperature of around 600 - 800 °C without passing through the slab yard after casting. The interior of the heating furnace 2 is divided into multiple zones. Generally, on the upstream side, there are heating zones divided into 2 - 8 zones and 1 - 3 soaking zones. In Figure 2 In the example shown, there are 5 heating zones and 1 soaking zone, and the two are collectively referred to as "heating furnace zones" here. Each heating furnace zone is set to a different ambient temperature in such a way that the average temperature of the slab SA loaded into the heating furnace 2 gradually rises to a specified target heating temperature (the target value of the average temperature of the slab SA when it is taken out from the heating furnace 2). In addition, a thermometer 21 for measuring the ambient temperature inside the heating furnace zone is provided above any heating furnace zone.

[0042] Figure 3 is a schematic view of the interior of the heating furnace 2 as viewed from above Figure 1 shown. As Figure 3 shown, the slab SA loaded into the heating furnace 2 passes through each heating furnace zone in turn inside the heating furnace 2 by means of a conveying device called a walking beam 22. In addition, multiple slabs SA are loaded into the heating furnace 2 at the same time and are taken out from the extraction side outlet of the heating furnace 2 in the order in which they are loaded into the heating furnace 2 for hot rolling. The inside of the walking beam 22 is water-cooled, and there are parts that locally hinder the heating of the slab SA through components in direct contact with the slab SA called skids. The part of the slab SA in contact with the skid is called a skid mark, and it becomes a region with a lower temperature compared to the part not in contact with the skid. As described in Patent Document 1, the width of the steel plate varies according to the skid mark, which affects the width accuracy of the hot-rolled steel plate.

[0043] 〔Width reduction stamping device〕

[0044] Figure 4 is a schematic view showing Figure 1 the structural example of the width reduction stamping device 4 shown. The slab SA heated by the heating furnace 2 is width-reduced to a specified set width by the width reduction stamping device 4 after the primary scale formed on the surface is removed by the descaling device 3. As Figure 4 shown, the width reduction stamping device 4 includes a pair of width reduction dies 41. The width reduction dies 41 press down the slab SA in the width direction. The width reduction stamping device 4 drives the width reduction dies 41 using a drive device 42 while conveying the slab SA, and intermittently performs width reduction on the slab SA from both sides in the width direction of the slab SA.

[0045] In the width reduction stamping device 4, pinch rolls 43 or the like are used to convey the slab SA. The width reduction stamping device 4 can change the feed pitch of the slab SA between width reduction passes by changing the driving amount of the pinch rolls 43. The feed pitch means the conveying distance of the slab SA for each width reduction pass in the width reduction stamping device 4. The driving amount of the pinch rolls 43 is controlled by a control controller 90 that controls the width reduction stamping device 4. On the surface of the width reduction die 41 that contacts the slab SA, a parallel portion 41a parallel to the slab conveying direction and an inclined portion 41b that extends in the width direction toward the opposite direction of the slab conveying direction are sequentially formed from the front end side in the slab conveying direction. In the width reduction die 41, in order to suppress slippage relative to the slab SA, there is a case where one or more parallel portions 41a are provided between the inclined portions 41b. The deformation state of the slab SA changes according to the shape of the width reduction die 41, which affects the width accuracy of the steel plate in the roughing mill 5.

[0046] 〔Roughing Mill〕

[0047] Return Figure 1 . The roughing mill 5 includes a reversible rolling mill 5a capable of reverse rolling and an irreversible rolling mill 5b capable of performing rolling only in the conveying direction toward the downstream side. In Figure 1 The arrow (solid line) shown below the roughing mill 5 in the figure indicates the reduction pass (the rolling pass that reduces the thickness). In the reversible rolling mill 5a, generally, about 5 to 11 reduction passes are performed in the reversible direction (from the upstream side to the downstream side or from the downstream side to the upstream side). In the final reduction pass, in order to simultaneously perform rolling and convey to the next rolling mill, the number of rolling passes of the reversible rolling mill must be odd, and the roughing billet is conveyed to the rolling mill located on the downstream side while performing rolling. At this time, the time from when the tail end of the roughing billet exits the rolling mill in the current reduction pass to when the rolling direction is reversed and the roughing billet bites into the rolling mill in the next reduction pass is called the inter-pass time of the reversible pass. In addition, the time from when the tail end of the roughing billet exits the reversible rolling mill 5a in the final rolling pass of the reversible pass to when it bites into the irreversible rolling mill 5b is called the inter-pass time of the continuous pass. In addition, the inter-pass time of the reversible pass and the inter-pass time of the continuous pass are combined and called the inter-pass air cooling time. The inter-pass air cooling time represents the time when the roughing billet is air-cooled during conveyance, which affects the temperature change of the roughing billet.

[0048] Figure 5 is a schematic diagram showing the structure of the stand that constitutes Figure 1 the roughing mill 5 shown in the figure. As Figure 5As shown in the figure, the roughing mill 5 includes a horizontal rolling mill 51 that reduces the thickness of the steel plate SB, and an edging mill (vertical rolling mill) 52 that reduces the width of the steel plate SB. The edging mill 52 is a rolling mill in which a pair of rolls are arranged longitudinally and is adjacently provided to the horizontal rolling mill 51. The width reduction (width rolling) of the steel plate SB using the edging mill 52 is usually performed before the horizontal rolling in each rolling pass of the roughing. Therefore, when one edging mill 52 is arranged relative to the reversing rolling mill 5a, the width reduction using the edging mill 52 is only performed in the forward direction, that is, in the odd-numbered passes, and is not performed in the even-numbered passes, that is, in the reverse-direction passes. However, when edging mills 52 are arranged on both sides of the reversing rolling mill 5a, the width reduction can be performed in any rolling pass. The roll opening (roll gap) of the horizontal rolling mill 51 and the opening of the edging mill 52 in each pass of the roughing are set by the control computer 91. A descaling head for spraying descaling water toward the steel plate SB is provided in the roughing mill 5, and descaling is performed on the entrance side of the horizontal rolling mill 51. However, descaling water is not necessarily sprayed in all rolling passes of the roughing. For example, it is sprayed only in the first pass, or only in the odd-numbered passes, and a prescribed descaling pattern is set according to the material of the steel plate SB, etc. The steel plate SB in the state where all the rolling passes preset by the roughing mill 5 are completed is called a roughing billet or a thin slab, and is called a roughing piece in the present invention.

[0049] 〔Width meter〕

[0050] Return Figure 1 . The width of the steel plate in the hot rolling line 1 is measured by the roughing exit side width meter 11, the finishing exit side width meter 12, and the width meter 13 in front of the coiler. Most of these width meters use an optical width measurement method. In the optical width measurement method, a light source is arranged below the rolling line for conveying the steel plate, and an image sensor is arranged above, and the width is measured based on the shadow length of the light emitted from the light source in the width direction of the steel plate during the passage of the steel plate. On the other hand, as a width meter, the width of the steel plate is measured by determining the position of the end portion in the width direction of the steel plate through image processing of the image of the steel plate taken by a camera. The width meter using a camera does not require a light source and an image sensor to be arranged along the width direction of the steel plate, so the device structure becomes simple. However, since the end portion in the width direction of the steel plate is photographed from an inclined direction, if the steel plate floats from the rolling line, it is easy to generate a measurement error in the width. Therefore, it mostly has a function of correcting the measurement error of the width according to the floating amount of the steel plate from the rolling line. Specifically, as Figure 6As shown, when the steel plate SB is conveyed at a height of the floating amount H from the rolling line 14, the measurement error of the width is corrected as follows. First, a set of cameras 15a and 15b arranged in the width direction of the steel plate SB determine the both ends in the width direction of the steel plate SB, and determine the measured value W1 of the width. Next, another set of cameras 15c and 15d arranged in the width direction of the steel plate SB determine the both ends in the width direction of the steel plate SB, and determine the measured value W2 of the width. Then, based on the positional relationship between the cameras 15a and 15b and the cameras 15c and 15d (in the example shown in Figure 6 , the interval D in the width direction between the cameras 15a and 15b and the interval L in the width direction between the camera 15a (15b) and the camera 15c (15d)), the actual width W of the steel plate SB is calculated by the following formula (1) as the measured value of the width of the steel plate SB.

[0051] [Formula 1]

[0052]

[0053] However, as another width measurement method, a method of irradiating a laser in the width direction of the steel plate and receiving the reflected light from the end face of the steel plate and measuring the width based on the distances to the both end faces of the steel plate is sometimes used. Some width gauges have a thermal expansion correction function for converting the width of the steel plate into the width after cooling based on the temperature of the steel plate. The width of the steel plate is measured by a width gauge during the conveyance of the steel plate, and thus the measured value of the width of the steel plate obtained by the width gauge becomes time-series numerical information corresponding to the sampling interval of the width gauge. In addition, the information on the conveyance speed at the position where the steel plate passes through the width gauge is used to convert the relationship between the position in the length direction of the steel plate and the actual value of the width. Then, in the control computer 91, the representative value of the width is calculated based on the measured value of the width obtained. The representative value of the width uses the average value of the widths in the length direction of the steel plate (average width), the measured value of the width of the stable part of the steel plate except for the front and tail ends (stable width), the measured value of the width of the front end part of the steel plate (front end width), the measured value of the width of the tail end part (tail end width), etc. In addition, the minimum value (minimum width), the maximum value (maximum width), etc. of the widths in the length direction of the steel plate are sometimes calculated. The measured value of the width measured by the width gauge is sometimes expressed by the deviation from a preset target width.

[0054] The information on the width distribution in the length direction of the rough-rolled piece in this embodiment refers to the width of the rough-rolled piece measured by the rough-rolling exit side width meter 11 and corresponding to the position information in the length direction. Usually, based on the measurement values of the rough-rolling exit side width meter 11, discrete data for each position divided in the length direction from the front end of the rough-rolled piece are obtained, and thus a plurality of data sets in which the distance from the front end of the rough-rolled piece is associated with the width measurement value are generated. The information on the width distribution in the length direction of the rough-rolled piece is the width measurement value associated with the position data of at least 3 points with respect to the length direction of the rough-rolled piece. However, the information on the width distribution in the length direction of the rough-rolled piece is preferably the width measurement value associated with 100 to 10,000 position data with respect to the length direction of the rough-rolled piece. This is because when the position data is less than 100, it is difficult to determine the width variation in the length direction of the rough-rolled piece, and even if the position data exceeds 10,000, the accuracy of determining the width variation will not be improved. The information on the width distribution can also normalize the position data in the length direction using the length of the rough-rolled piece and be the width measurement value of the rough-rolled piece associated with the normalized position data.

[0055] <Generation method of width prediction model>

[0056] Next, with reference to Figures 7 to 11 a generation method of a width prediction model according to an embodiment of the present invention will be described.

[0057] Figure 7 is a block diagram showing the structure of a width prediction model generation unit according to an embodiment of the present invention. As Figure 7 shown, the width prediction model generation unit 100 according to an embodiment of the present invention includes a database unit 101 and a machine learning unit 102. The database unit 101 stores one or more operation performance data selected from the operation performance data of the width reduction stamping device 4, one or more operation performance data selected from the operation performance data of the rough rolling mill 5, and the performance data of the width distribution in the length direction of the rough-rolled piece. The database unit 101 may also store, as needed, one or more operation performance data selected from the operation performance data of the heating furnace and one or more performance data selected from the performance data of the slab attribute information. Specific examples of the operation data stored in the database unit 101 will be described later.

[0058] Actual performance data stored in the database unit 101 is appropriately obtained from the control controller 90, the control computer 91, or the host computer 92. In the present embodiment, the width prediction model generation unit 100 includes a data acquisition unit 103 for collecting such actual performance data. The data acquisition unit 103 temporarily stores the actual performance data and stores it in the database unit 101 after generating a data set in which a plurality of actual performance data are correlated. Since the data stored in the database unit 101 may be obtained at different times, a data set in which the plurality of actual performance data are correlated can be formed by correlating the plurality of actual performance data in the data acquisition unit 103.

[0059] As the actual performance data of the width reduction stamping device 4 and the actual performance data of the rough rolling mill 5 included in the data set stored in the database unit 101, actual performance data determined as representative values in the length direction of a single rough rolling piece can be used. The representative value in the length direction of the rough rolling piece refers to a single data representing the operating conditions in the length direction of the rough rolling piece. As the representative value in the length direction of the rough rolling piece, data that sets a representative position determining the position in the length direction of the rough rolling piece and is related to the operating conditions at the set representative position can be used. In addition, as the representative value in the length direction of the rough rolling piece, the average value or standard deviation of the operating data in the length direction of the rough rolling piece can be used. In addition, as the representative value in the length direction of the rough rolling piece, data related to operating conditions fixed (not changing) with respect to the length direction of the rough rolling piece can also be used. For example, when performing width reduction on a slab, a representative position can be set in the stable part of the slab, and the actual performance data set with respect to the representative position can be used. When performing rough rolling, a representative position can also be set in the stable part of the rough rolling piece, and the actual performance data set with respect to the representative position can be used. However, the set representative position is not limited to the stable part of the rough rolling piece, and a representative position can also be set at the 1 / 4 position and 3 / 4 position of the total length of the slab with respect to the end on the downstream side in the conveying direction of the hot rolling line, or a representative position can be set starting from the front end or the tail end, and the actual performance data can be obtained. In this case, the operating parameters of the width reduction stamping device and the operating parameters of the rough rolling mill are determined as representative values as operating parameters for a single rough rolling piece, and they are correlated with the width distribution of the rough rolling piece. That is, the width prediction model of the present embodiment has the following characteristics: as a prediction result, the width at positions different in the length direction from the position in the length direction of the rough rolling piece correlated with the input operating parameters is output.

[0060] The width prediction model generation unit 100 can be provided in the control computer 91 for controlling the hot rolling line 1 to manufacture steel plates. Additionally, the width prediction model generation unit 100 can also be provided in the upper computer 92 that issues manufacturing instructions to the control computer 91, or in an independent computer capable of communicating with other devices. Further, the width prediction model generation unit 100 can also be configured as a device separate from the database unit 101 using a device capable of receiving the data sets stored in the database unit 101. The database unit 101 stores more than 100 data sets. Preferably, more than 10,000 data sets, and more preferably more than 100,000 data sets are stored in the database unit 101. There are cases where the data stored in the database unit 101 is screened as needed.

[0061] The machine learning unit 102 uses the data sets stored in the database unit 101 to generate a width prediction model M that predicts the width distribution in the length direction of the rough rolled piece through machine learning using multiple learning data. The above-mentioned multiple learning data includes, as input actual data, one or more operation actual data selected from the operation parameters of the width reduction stamping device 4 and one or more operation actual data selected from the operation parameters of the rough rolling mill 5, and the width distribution information in the length direction of the rough rolled piece is used as output actual data. The machine learning model for generating the width prediction model M can be any machine learning model as long as it achieves a practically sufficient width prediction accuracy. For example, it is sufficient to use commonly used neural networks (including deep learning, convolutional neural networks, etc.), decision tree learning, random forest, support vector regression, etc. Additionally, an ensemble model combining multiple models can also be used. For example, the width prediction model M can be generated through machine learning using the general neural network shown. In particular, if deep learning is used, the problem of multicollinearity is not considered, and other operation parameters related to the width distribution of the rough rolled piece can be freely selected as inputs, so the prediction accuracy of the width distribution of the rough rolled piece can be improved. For example, the middle layer of the neural network is set to 2 layers, and the number of nodes is set to 3 each. As the activation function, an activation function using the sigmoid function can be used. Figure 8 The width prediction model M can be generated through machine learning using the general neural network shown. In particular, if deep learning is used, the problem of multicollinearity is not considered, and other operation parameters related to the width distribution of the rough rolled piece can be freely selected as inputs, so the prediction accuracy of the width distribution of the rough rolled piece can be improved. For example, the middle layer of the neural network is set to 2 layers, and the number of nodes is set to 3 each. As the activation function, an activation function using the sigmoid function can be used.

[0062] The Machine Learning Department 102 can also improve the inference accuracy of the width distribution of the rough-rolled product by dividing the data set stored in the Database Department 101 into training data and test data for learning. For example, the Machine Learning Department 102 can also use the training data to learn the weight coefficients of the neural network, appropriately change the structure of the neural network (the number of intermediate layers, the number of nodes) in such a way that the correct solution rate of the width distribution of the rough-rolled product in the test data becomes higher, and generate the width prediction model M. The update of the weight coefficients can use the error backpropagation method. The width prediction model M can also be updated to a new model by relearning every six months or annually, for example. This is because the more data stored in the Database Department 101 increases, the more accurately the width distribution of the rough-rolled product can be predicted. By updating the width prediction model M based on the latest data, a width prediction model M that reflects the changes in the manufacturing conditions of the hot-rolled steel sheet manufactured using the hot rolling line can be generated. In the present embodiment, when manufacturing a hot-rolled steel sheet using the hot rolling line, information on the width distribution in the length direction of the rough-rolled product can be predicted based on the operating conditions determined as the representative values in the length direction of the rolled product. The width prediction of the rough-rolled product in the prior art obtains the actual performance data of the operating conditions for each position in the length direction of the rolled product, and establishes a correspondence relationship with the actual performance data of the width obtained for each position in the length direction of the rough-rolled product. Thus, in the case where the slab drawn out from the heating furnace 2 is plastically deformed by the width reduction stamping device 4 and the rough rolling mill 5 and is greatly extended as a rough-rolled product, it is necessary to track the position in the length direction of the rolled product to establish a correspondence relationship between the operating conditions of the width reduction stamping device 4 and the rough rolling mill 5 and the position information of the rough-rolled product. In contrast, in the present embodiment, since the representative value of the operating conditions in the length direction of the rolled product is directly correlated with the width distribution of the rough-rolled product, it is not necessary to track a specific position of the rolled product during rough rolling.

[0063] 〔Slab Attribute Information〕

[0064] The slab attribute information that can be used as the input for the width prediction model M refers to information related to the slab dimensions that affect the width change of the rolled product in the width reduction stamping device 4 and the rough rolling mill 5 of the hot rolling line 1. The information related to the slab dimensions refers to information related to the thickness, width, length, and weight of the slab. The information related to the slab dimensions affects the temperature change of the steel sheet in the hot rolling line 1 and affects the deformation resistance of the rolled product during width reduction and rough rolling. In particular, since the behavior of temperature drop is different at the front end, the tail end, and the stable part of the rolled product, as a result, it is correlated with the width distribution in the length direction of the rough-rolled product.

[0065] 〔Operating Parameters of the Heating Furnace〕

[0066] The operating parameters of the reheating furnace that can be used as the input for the width prediction model M are parameters representing the operating conditions when heating the slab in the reheating furnace 2 of the hot rolling line 1, and refer to information that affects the width change of the rolled piece in the width reduction stamping device 4 and the rough rolling mill of the hot rolling line 1. The operating parameters of the reheating furnace 2 can use the temperature of the slab when it is charged into the reheating furnace 2, the residence time of a specific furnace zone in the reheating furnace 2, the ambient temperature of the final furnace zone of the reheating furnace 2, and the temperature of the slab withdrawn from the reheating furnace 2. According to these parameters, the behavior of temperature drop changes at the front end, tail end, and stable part of the rolled piece, so there is a correlation with the width distribution in the length direction of the rough rolled piece.

[0067] In addition, information such as the charging position of the slab in the reheating furnace 2 and the positional relationship between the slab in the reheating furnace 2 and other slabs can also be used. As Figure 3 shown, the charging position of the slab in the reheating furnace 2 can use the information about the in-furnace charging position P representing the distance between the walking beam 22 at one end of the reheating furnace 2 and the lengthwise end of the slab SA. This is because the position of the skid mark on the slab SA in the length direction changes due to the in-furnace charging position P, which affects the position where width variation occurs in the length direction of the steel plate. The charging position of the slab SA can also use parameters represented by the distance D1 between the lengthwise end of the slab SA and the furnace wall of the reheating furnace 2, the interval between the walking beam 22 (fixed skid) and the moving skid 23 of the reheating furnace 2 in which the slab SA is charged.

[0068] In addition, as information about the positional relationship between the slab in the reheating furnace 2 and other slabs, as Figure 3 shown, the distance (charging interval) D2 between the slab and other adjacent slabs in the reheating furnace 2 can be used. If the charging interval D2 changes in the reheating furnace 2, the temperature of the end face of the slab changes. This is because it affects the temperature distribution of the rolled piece in the hot rolling line 1, thereby affecting the width distribution of the rough rolled piece. As information about the positional relationship between the slab in the reheating furnace 2 and other slabs, the length and thickness of other slabs, the difference in the distance from the front end of the slab to be predicted and the front end of other slabs to the furnace wall of the reheating furnace, etc. can also be used.

[0069] 〔Operating parameters of the width reduction stamping device〕

[0070] The operating parameters of the width reduction stamping device 4 that can be used as the input for the width prediction model M refer to the operating conditions when performing width reduction on the heated slab. The operating parameters of the width reduction stamping device 4 can use the operating parameters related to the amount of width reduction of the slab. The operating parameters related to the amount of width reduction of the slab include the amount of width reduction at the representative position in the length direction of the slab and the feed pitch of the slab between width reduction passes. According to the operating parameters related to the amount of width reduction of the slab, a difference is generated between the shape of the dogbone (thickness distribution in the width direction) in the stable part of the slab after width reduction and the dogbone shape of the front and tail ends, which affects the width distribution in the length direction of the rough rolled piece. That is, even if the amount of width reduction, which is the representative value in the length direction of the slab, is constant with respect to the length direction of the slab, the dogbone shape is different in the stable part, the front end part, and the tail end part of the slab, so it affects the width distribution in the length direction of the rough rolled piece. In addition, even when the feed pitch of the slab between width reduction passes is constant, a difference in the dogbone shape is generated between the stable part and the front and tail ends, so it affects the width distribution in the length direction of the rough rolled piece. Regarding the amount of width reduction of the slab, sometimes different amounts of width reduction are set for the front end part, the stable part, and the tail end part of the slab respectively. In this case, as the representative value of the operating parameters of the width reduction stamping device 4, any of the amounts of width reduction set for the front end part, the stable part, and the tail end part of the slab can also be used. In addition, regarding the feed pitch of the slab between width reduction passes, sometimes different feed pitches are set for the front end part, the stable part, and the tail end part of the slab respectively. In this case, as the representative value of the operating parameters of the width reduction stamping device 4, any of the feed pitches set for the front end part, the stable part, and the tail end part of the slab can also be used.

[0071] In addition, the operating parameters of the width reduction stamping device 4 can use the operating parameters related to the die shape applicable to the width reduction stamping device 4. The operating parameters related to the die shape are representative values fixed with respect to the length direction of the slab. For example, the length of the parallel part 41a and the angle of the inclined part 41b formed on the width reduction die 41 can be used. This is because, according to the shape of the width reduction die 41, a difference in the deformation state is generated between the stable part and the front and tail ends of the slab, which affects the width distribution in the length direction of the rough rolled piece. Figure 9 An example of the width distribution after width reduction by the width reduction stamping device 4 with the amount of width reduction and the feed pitch constant in the length direction of the slab is shown. As Figure 9 shown, it can be seen that periodic width variations (also called stamping marks) occur from the front end part to the tail end part of the slab after width reduction. Since the width reduction using the width reduction stamping device 4 is an intermittent process, parts with a wide width and parts with a narrow width appear at regular intervals. Thereby, it affects the width distribution in the length direction of the rough rolled piece.

[0072] As described above, by determining the operating parameters of the width reduction stamping device 4 as representative values in the length direction of a slab, the mode of width variation in the length direction is characterized for the slab after width reduction. That is, even without determining the operating parameters at each position along the length direction of the slab, it is possible to characterize the distribution related to the width variation in the length direction for the slab after width reduction.

[0073] 〔Operating parameters of roughing mill〕

[0074] The operating parameters of the roughing mill 5 input to the width prediction model M mean the rolling operation conditions that affect the width distribution in the length direction of the rolled piece in any rolling pass of the roughing performed on the roughing mill 5. The operating parameters of the roughing mill 5 preferably include the rolling conditions of the horizontal rolling mill 51 and the edging mill 52 that make up the roughing mill 5. As the rolling conditions of the horizontal rolling mill 51, the roll opening, work roll diameter, entrance side plate thickness, exit side plate thickness, reduction ratio, roughing target width, rolling load, and steel plate temperature in any rolling pass can be used. Even if the operating conditions in horizontal rolling are constant with respect to the length direction of the rolled piece, the width expansion behavior is different in the stable part and the front and tail end parts of the rolled piece, so it affects the width distribution in the length direction of the rough rolled piece.

[0075] As the rolling conditions of the edging mill 52, the edging mill opening, edging mill roll diameter, plate thickness, width reduction ratio, and width reduction load in any rolling pass can be used. This is because these rolling conditions affect the width expansion behavior in horizontal rolling of the rolled piece and also affect the difference in width expansion between the stable part and the front and tail end parts of the rolled piece. In addition, even if the operating conditions of width rolling are constant with respect to the length direction of the rolled piece, the dogbone formation behavior is different in the stable part and the front and tail end parts of the rolled piece, so it affects the width distribution in the length direction of the rough rolled piece. The rolling conditions as the operating parameters of the roughing mill have a great influence on the width distribution of the rough rolled piece, so the rolling conditions corresponding to all rolling passes of roughing can also be included in the operating parameters of the roughing mill. For the edging mill opening, sometimes different edging mill openings are set at the front end part, stable part, and tail end part of the rolled piece. In this case, as the representative value of the operating parameters of the roughing mill 5, any edging mill opening set for the front end part, stable part, and tail end part of the rolled piece can also be used.

[0076] In addition, preferably, the operating parameters of the roughing mill 5 include the set value or actual value of the outlet side plate thickness from the first pass to the final pass of roughing. This is the so-called pass schedule of roughing. If the pass schedule is different, the thickness of the rolled piece changes during transportation between rolling passes, so the temperature distribution during air cooling changes, resulting in differences in the deformation behavior in the width direction based on the position in the length direction of the rolled piece. Also, the operating parameters of the roughing mill 5 can include information related to the air cooling time between rolling passes and the presence or absence of descaling water spraying in any rolling pass. This is because, although they are operating conditions that do not change in the length direction of the rolled piece, they affect the difference in the amount of temperature drop between the stable part and the front and tail ends of the rolled piece, resulting in differences in the deformation behavior in the width direction based on the position in the length direction of the rolled piece. For the roll opening of the horizontal rolling mill 51, sometimes different roll openings are set for the front end, stable part, and tail end of the rolled piece respectively. In this case, as a representative value of the operating parameters of the roughing mill 5, any of the roll openings set for the front end, stable part, and tail end of the rolled piece can also be used.

[0077] Figure 10 Schematically shows the width reduction at the front end and tail end of the rolled piece produced by roughing. Width reduction refers to the behavior where the width at the front end or tail end is smaller than that at the stable part due to roughing. This is because, when the width of the rolled piece is rolled using an edger, in the stable part, when the width direction reduction is performed using the edger roll, the materials before and after in the length direction become the resistance to the width direction reduction, forming a width direction thickness distribution called dogbone where the deformation is concentrated at the width direction ends of the rolled piece. Then, when the rolled piece with dogbone is horizontally rolled, the plastic flow in the width direction is promoted by the thickened part formed at the width direction ends, so the width recovery becomes larger. In contrast, when the width of the front and tail ends is rolled using an edger, since there is no material in front or behind in the length direction, the resistance to the width direction reduction becomes weaker, so it is easier for the rolled piece to thicken toward the center of the width direction of the rolled piece. Thus, during horizontal rolling, the plastic flow in the width direction is suppressed, and the width recovery is smaller compared to the stable part. In this way, since the thickening behavior of the rolled piece using the edger is different between the stable part and the front and tail ends of the rolled piece, the width expansion produced by horizontal rolling is different according to the position in the length direction of the rolled piece, resulting in width reduction.

[0078] On the other hand, Figure 11Schematically shows the width distribution formed in the case where no width reduction using an edger is performed and the rolled material with a constant width in the length direction is horizontally rolled. In the case of horizontal rolling, in the stable part, due to the resistance (constraint) of the materials present before and after in the length direction to the plastic flow in the width direction, the width expansion of the rolled material is small. In contrast, at the front and tail ends, there is no material that constrains the plastic flow in the width direction at either the front or the rear in the length direction, so the width expansion of the rolled material becomes larger. Therefore, if a rolled material with a constant width in the length direction is horizontally rolled, as Figure 11 shown, the width expansion at the front end and the tail end of the rolled material becomes larger.

[0079] As described above, even if the operating conditions of the roughing mill 5 are constant in the length direction of the rolled material, the width of the rolled material in the edger and the horizontal rolling mill that make up the roughing mill 5 has a constant pattern in the length direction. Therefore, in the present embodiment, as the operating parameters of the roughing mill, it is possible to use the operating parameters determined as the representative values in the length direction of the rolled material to establish a correspondence with the information on the width distribution in the length direction of the rough rolled material.

[0080] <Width prediction method for rough rolled material>

[0081] Next, with reference to Figure 12 , Figure 13 a width prediction method for a rough rolled material according to an embodiment of the present invention will be described.

[0082] A width prediction method for a rough rolled material according to an embodiment of the present invention includes the following steps: By inputting the operating parameters of the width reduction stamping device 4 and the operating parameters of the roughing mill 5 into the width prediction model M and outputting the information on the width distribution in the length direction of the rough rolled material, the width distribution in the length direction of the rough rolled material is predicted. The width prediction unit that executes the step of predicting the width distribution of the rough rolled material can be provided in the control computer 91 for controlling the hot rolling line 1. In addition, the width prediction unit can also be provided in the upper computer 92 that gives a manufacturing instruction to the control computer 91, or can also be provided in an independent computer capable of communicating with other devices. In addition, the width prediction unit can also use a device that can receive the width prediction model M generated by the width prediction model generation unit 100 and is configured as a device separately provided from the width prediction model generation unit 100. Hereinafter, with reference to Figure 12 the operation of the width prediction unit according to an embodiment of the present invention will be described.

[0083] Figure 12 is a block diagram showing the structure of the width prediction unit according to an embodiment of the present invention. Figure 12The operation of the width prediction unit 110 shown is performed before the rough rolling of the hot-rolled steel sheet manufactured in the hot rolling line 1 is completed. The operation of the width prediction unit 110 can be performed, for example, at the stage when the slab, which is the object of width prediction, is withdrawn from the heating furnace 2. At the stage when the slab is withdrawn from the heating furnace 2, the control computer 91 calculates at least the set values of the operating conditions of the width reduction press device 4 and the rough rolling mill 5, and these set values can be used as the input to the width prediction model M. In this case, the information related to the slab size loaded into the heating furnace 2 is stored in the host computer 92, and the actual performance data of the operating parameters of the heating furnace 2 can be obtained. In addition, the operation of the width prediction unit 110 can be performed, for example, at the stage after the width of the steel sheet, which is the object of width prediction, is reduced by the width reduction press device 4. For the slab SA after width reduction, the actual performance data of the operating parameters of the width reduction press device 4 are obtained by the control controller 90 or the control computer 91, and these actual performance data are used as the input to the width prediction model M, and the set values of the operating parameters regarding the rough rolling mill 5 are set by the control computer 91.

[0084] The operation of the width prediction unit 110 can be performed, for example, even in the middle pass of the rough rolling performed by the rough rolling mill 5. It is only necessary to determine the actual performance data of the operating parameters of the width reduction press device 4 and use the actual performance values of the operating parameters up to the middle pass of the rough rolling and the set values of the next pass and subsequent passes as the input to the width prediction model M. Figure 12 The input data acquisition unit 111 shown acquires the actual performance values or set values of the operating parameters in the hot rolling line 1 held by any one of the control controller 90, the control computer 91, and the host computer 92 as described above. Then, the width prediction unit 110 obtains the information on the width distribution of the rough rolled piece by inputting the input data acquired by the input data acquisition unit 111 into the width prediction model M. As described above, the operation of the width prediction unit 110 can be performed at each stage of each process when the steel sheet passes through the hot rolling line 1, so it can be performed multiple times during the manufacturing process of one steel sheet.

[0085] The information on the width distribution of the rough rolled piece output by the width prediction unit 110 can also be displayed on a monitor or the like connected to the width prediction unit 110. Based on the output display of the width distribution output by the width prediction unit 110, at least one of the rough rolling target width, the finish rolling target width, and the target width in front of the coiler can be reset, thereby being able to suppress the width defect of the hot-rolled steel sheet. Explain the significance of using one or more operating parameters selected from the operating parameters of the width reduction press device 4 and one or more operating parameters selected from the operating parameters of the rough rolling mill 5 as the input to the width prediction model M. Figure 13This is a diagram showing the width distribution of the rough rolled product after width reduction of the slab using the width reduction stamping device 4, followed by one pass of width rolling using the edger and one pass of horizontal rolling using the horizontal rolling mill. Figure 13 The vertical axis of Figure 13 represents the difference between the width of the stable part and the width of the front end part of the rough rolled product, which is an index representing the width distribution of the rough rolled product. Additionally, the horizontal axis represents the width reduction amount of the stable part based on the width reduction stamping device 4. In this case, it shows the results when the width reduction amount of the stable part of the width reduction stamping, which is an operating parameter of the width reduction stamping device 4, is changed within the range of 0 to 300 mm, and the width reduction amount of the stable part by the edger, which is an operating parameter of the roughing mill 5, is changed within the range of 5 to 20 mm. According to Figure 13 it can be seen that the width difference between the stable part and the front end part changes in such a way as to produce a minimum value with respect to the width reduction amount of the width reduction stamping device 4. Additionally, it can be seen that the width reduction amount of the width reduction stamping device 4 at which the width difference between the stable part and the front end part takes the minimum value varies depending on the width reduction amount of the edger of the roughing mill 5. That is, the width reduction amount of the width reduction stamping device 4 for minimizing the width difference between the stable part and the front end part, which is an index of the width distribution in the length direction of the rough rolled product, varies depending on the width reduction amount of the edger of the roughing mill 5. Therefore, as the input to the width prediction model M, by combining one or more operating parameters selected from the operating parameters of the width reduction stamping device 4 and one or more operating parameters selected from the operating parameters of the roughing mill 5, the prediction accuracy of the width distribution in the length direction of the rough rolled product can be improved.

[0086] <Width control method for rough rolled product>

[0087] Next, a width control method for the rough rolled product as an embodiment of the present invention will be described.

[0088] The width control method for the rough-rolled piece according to an embodiment of the present invention is based on the information of the width distribution in the length direction of the rough-rolled piece predicted as described above, and the operation parameters of the roughing mill 5 are reset. In the above width prediction method for the rough-rolled piece, the information of the width distribution in the length direction of the rough-rolled piece, which is the output of the width prediction model M, represents the width distribution in the length direction of the rough-rolled piece from the front end to the tail end. The width control in the hot rolling line is usually performed in such a way that the minimum plate width in the length direction of the hot-rolled steel sheet is not lower than the lower limit value of the product width of the hot-rolled steel sheet. This is because if a width-insufficient area occurs in a part of the length direction of the hot-rolled steel sheet, that part needs to be cut off. However, if a part of the length direction is cut off, the single weight of the coil does not meet the product specifications, so the whole becomes a defective product, resulting in a significant reduction in the yield. Therefore, a lower limit value is also set for the width of the rough-rolled piece, and it is controlled so that a part lower than the lower limit value of the roughing target width does not occur in a part of the length direction of the rough-rolled piece.

[0089] According to the present embodiment, it is possible to predict the width distribution in the length direction of the rough-rolled piece from the front end to the tail end before the end of all passes of roughing. Therefore, in the case where it is predicted that a part lower than the lower limit value of the roughing target width occurs in a part of the length direction of the rough-rolled piece, the operation parameters of the roughing mill are set, and it is controlled so that a part lower than the lower limit value of the roughing target width does not occur over the entire length in the length direction of the rough-rolled piece. Specifically, in the final pass of roughing, it is only necessary to set the edger opening of the edger 52 to be increased. In this case, the edger opening of the edger 52 may also be larger than the set value only for the part predicted to be lower than the lower limit value of the roughing target width in the length direction of the rough-rolled piece. The front end and the tail end of the rough-rolled piece are sometimes cut off by the crop shear located upstream of the finishing mill 6. Therefore, even if there is a part lower than the lower limit value of the roughing target width in a part of the length direction in the area of 0.2 to 0.5 m at the front end or the tail end of the rough-rolled piece, it is not necessary to set the edger opening to be large. In this way, by controlling so as not to be lower than the lower limit value of the roughing target width over the entire length of the rough-rolled piece, it is possible to suppress the occurrence of width insufficiency in the hot-rolled steel sheet that is finish-rolled by the finishing mill and wound by the coiler 8, and manufacture a hot-rolled steel sheet with excellent width accuracy.

[0090] 〔Example 1〕

[0091] In this embodiment, the present invention is applied to a hot rolling line having a width reduction stamping device 4 on the downstream side of a heating furnace 2, and including a roughing mill 5 composed of 4 reversible rolling mills and 1 irreversible rolling mill, and a finishing mill 6 with 7 stands. In this embodiment, through the above hot rolling line, a slab with a thickness of 250 to 270 mm and a width of 600 to 1600 mm is heated by the heating furnace 2 to manufacture a steel plate with a slab thickness (roughing billet thickness) of 30 to 35 mm and a plate thickness on the outlet side of the finishing mill of 2 to 3 mm. The hot rolling line is equipped with a roughing exit side width gauge 11, a finishing exit side width gauge 12, and a width gauge 13 in front of the coiler. The controller 90 for controlling the hot rolling line, the control computer 91, and the host computer 92 collect the actual values of the operation parameters of the steel plates manufactured on the hot rolling line, and the actual data is obtained by the data acquisition unit 103.

[0092] In this embodiment, the data acquisition unit 103 obtains the width reduction amount of the stable part of the slab and the feed pitch of the slab for each pass of width reduction stamping as the operation actual data of the width reduction stamping device 4. As the operation actual data of the roughing mill 5, the set values of the width of the inlet side plate and the outlet side plate of the edger and the roll diameter of the edger roll of the stable part of the rolled piece are used, and they are obtained for all passes of the roughing passes. In addition, for the operation actual data of the roughing mill 5, the set values of the width of the inlet side plate and the outlet side plate of the horizontal rolling and the work roll diameter of the horizontal rolling mill of the stable part of the rolled piece are used, and they are obtained for all passes of the roughing passes. In this embodiment, in addition to the above, the slab thickness of the slab is also obtained from the host computer 92 as the actual data of the attribute information of the slab.

[0093] The data acquisition unit 103 obtains the actual data of the width distribution in the length direction of the roughing piece measured by using the roughing exit side width gauge 11 from the control computer 91. The information on the width distribution of the roughing piece normalizes the position data in the length direction using the length of the roughing piece, and is the width at each position divided into 1000 parts in the length direction of the roughing piece. Then, when 30,000 data sets are accumulated in the database unit 101, the width prediction model M is generated by the machine learning unit 102. In this case, the machine learning unit 102 divides the data sets accumulated in the database unit 101 into 20,000 learning data and 10,000 test data, and uses the width prediction model M generated using the learning data to evaluate the prediction accuracy of the test data. Machine learning is performed using a neural network. The middle layer of the neural network is set to 1 layer, the number of nodes is set to 500, and the sigmoid function is used as the activation function. The number of nodes in the output layer is set to 200, and the width at the positions divided into 200 parts in the length direction of the roughing piece is predicted.

[0094] Figure 14 、 Figure 15Indicates the width distribution in the length direction of the predicted rough-rolled piece in this embodiment. In Figure 14 , Figure 15 , the actual data showing the width distribution in the length direction of the rough-rolled piece included in the test data is compared with the information on the width distribution in the length direction of the rough-rolled piece predicted by the above width prediction model M. The vertical axis of the curve graph represents the deviation from the target value of the width set in the roughing width control. In the prediction results shown in Figure 14 and the prediction results shown in Figure 15 , the operating conditions of the width reduction stamping device 4 and the operating conditions of the roughing mill 5 are different. According to any of the results, it is confirmed that the width distribution in the length direction of the rough-rolled piece can be predicted with high accuracy, including the influence of the width reduction formed at the front end or the tail end of the rough-rolled piece and the intermittent stamping marks formed at the fixed part. 100 data sets with the same steel grade of the slab and the same size of the hot-rolled steel sheet after finish rolling are selected from 10,000 test data, and the actual data on the width distribution in the length direction of the rough-rolled piece included therein is evaluated. The average deviation from the target width with respect to the total length of the rough-rolled piece is 5.5 mm, and the standard deviation is 5.0 mm. For the rough-rolled piece with such width variation, the width prediction model M generated in this embodiment can predict with an accuracy of an average deviation of 0.0 mm and a standard deviation of 1.5 mm from the target width with respect to the total length of the rough-rolled piece.

[0095] 〔Example 2〕

[0096] In this embodiment, slabs with a thickness of 230 - 235 mm and a width of 600 - 2000 mm are heated by the heating furnace 2 through the same hot rolling line as in Example 1 to manufacture steel sheets with a slab thickness (roughing billet thickness) of 30 - 35 mm and a sheet thickness of 2 - 3 mm on the exit side of the finish rolling mill. In this embodiment, as the input data for the width prediction model M, in addition to the operating parameters used in Example 1, the operating parameters of the heating furnace 2 are also used. The operating parameters of the heating furnace 2 are selected as the slab temperature at the time of charging into the heating furnace, the residence time in the heating furnace 2, the charging position of the slab in the heating furnace 2, the charging interval from the preceding slab in the heating furnace 2, and the charging interval from the subsequent slab in the heating furnace 2. Then, in the same manner as in Example 1, the data acquisition unit 103 acquires the operating actual data of the heating furnace 2 together with the operating actual data of the width reduction stamping device 4, the operating actual data of the roughing mill 5, and the actual data of the slab attribute information, and accumulates the data sets in the database unit 101. In this case, 30,000 data sets are accumulated in the database unit 101, and the accumulated data sets are divided into 20,000 learning data and 10,000 test data, and the width prediction model M is generated using the learning data. The machine learning method used in the machine learning unit 102 is a neural network, and its structure is the same as that in Example 1.Figure 16 Indicates the width distribution in the length direction of the predicted rough-rolled piece in this embodiment. As Figure 16 shown, the characteristics of the width distribution of the rough-rolled piece caused by the skid mark on the furnace floor formed by the heating furnace 2 are also predicted, and the measured data can be predicted with high accuracy.

[0097] 〔Embodiment 3〕

[0098] In this embodiment, using the data set stored in the database unit 101 in Embodiment 1, the prediction accuracy of the width distribution when the input data of the width prediction model M is changed was evaluated. In addition to the operation performance data of the width reduction stamping device 4 and the rough rolling mill 5 used for the input data of the width prediction model M in Embodiment 1, the database unit 101 also stores the operation performance data of other width reduction stamping devices 4 and rough rolling mills 5. In this embodiment, 30,000 data sets stored in the database unit 101 were divided into 20,000 learning data and 10,000 test data, and using the width prediction model M generated using the learning data, the prediction accuracy for the test data was evaluated. Machine learning was performed using a neural network. The middle layer of the neural network was set to 1 layer, the number of nodes was set to 500, and the sigmoid function was used as the activation function. The number of nodes in the output layer was set to 200, and the width at the positions where the length direction of the rough-rolled piece was divided into 200 parts was predicted. Table 1 shown below represents the operation performance data used as the input of the width prediction model M in this embodiment.

[0099] In Invention Example 1, the width reduction amount and feed pitch of the stable part of the slab were obtained as representative values as the operation performance data of the width reduction stamping device 4. As the operation performance data of the rough rolling mill 5, the set value of the edger opening and the edger roll diameter at the stable part of the rolled piece in all rolling passes of rough rolling were used as the operation performance data of the edger 52. In addition, as the operation performance data of the rough rolling mill 5, the set values of the inlet side plate width and the outlet side plate width of the horizontal rolling mill and the working roll diameter of the horizontal rolling mill 51 at the stable part of the rolled piece in all rolling passes of rough rolling were used as the operation performance data of the horizontal rolling mill 51.

[0100] In Invention Example 2, the width reduction amount and feed pitch of the front end part of the slab were obtained as representative values as the operation performance data of the width reduction stamping device 4. The same data as in Invention Example 1 was used as the operation performance data of the rough rolling mill 5.

[0101] In Invention Example 3, the same data as in Invention Example 1 was used as the operation performance data of the width reduction stamping device 4. As representative values for the operation performance data of the roughing mill 5, the edger opening and edger roll diameter set for the front end of the rolled piece in all rolling passes of roughing were used as the operation performance data of the edger 52. In addition, as representative values for the operation performance data of the roughing mill 5, the set values of the inlet side plate width and outlet side plate width of the horizontal rolling mill at the stable part of the rolled piece and the work roll diameter of the horizontal rolling mill 51 in all rolling passes of roughing were used as the operation performance data of the horizontal rolling mill 51.

[0102] In this embodiment, using the width prediction model M generated as described above, 100 data sets in the same partition with the same steel type of the slab and the dimensions of the hot-rolled steel sheet after finish rolling were selected from 10,000 test data, and the prediction accuracy of their width distributions was evaluated. Specifically, for the width at each position where the rough-rolled piece was divided into 200 parts in the length direction, the deviation between the predicted width and the actual width based on the width prediction model M was calculated, and the average value and standard deviation of the calculated deviations were obtained. Table 1 shows the results of the prediction accuracy of the width distribution.

[0103] As shown in Table 1, in Invention Example 1, the average value of the width deviation with respect to the total length of the rough-rolled piece was 0.0 mm, and the standard deviation was 1.7 mm, confirming that a high prediction accuracy could be obtained with respect to the total length of the rough-rolled piece. In Invention Example 2, the average value of the width deviation with respect to the total length of the rough-rolled piece was 0.0 mm, and the standard deviation was 2.3 mm. Invention Example 2 used the width reduction amount and feed pitch at the front end of the slab, which were the operation performance data of the width reduction stamping device 4, as the input to the width prediction model M. In this case, it was confirmed that even if different settings were made for the width reduction amount and feed pitch of the width reduction stamping device at the front end and the stable part of the slab, a width distribution with relatively high accuracy could be predicted. In Invention Example 3, the average value of the width deviation with respect to the total length of the rough-rolled piece was 0.0 mm, and the standard deviation was 2.6 mm. Invention Example 3 used the edger opening set for the front end of the rolled piece in all rolling passes of roughing as the input to the width prediction model M. In this case, it was confirmed that even if different settings were made for the edger opening at the front end and the stable part of the slab, a width distribution with relatively high accuracy could be predicted.

[0104] [Table 1]

[0105]

[0106] As described above, embodiments of the invention accomplished by the inventors have been explained, but the present invention is not limited by the descriptions and drawings that are part of the disclosure of the present invention based on these embodiments. That is, all other embodiments, examples, and application techniques etc. accomplished by those skilled in the art based on these embodiments are included in the scope of the present invention.

[0107] Industrial Applicability

[0108] According to the present invention, it is possible to provide a width prediction method for a rough rolled product that can accurately predict the width distribution in the length direction of the rough rolled product. In addition, according to the present invention, it is possible to provide a width control method for a rough rolled product that can accurately control the width in the length direction of the rough rolled product. In addition, according to the present invention, it is possible to provide a method for manufacturing a hot rolled steel sheet that can manufacture a hot rolled steel sheet with excellent width accuracy in the length direction. In addition, according to the present invention, it is possible to provide a method for generating a width prediction model for a rough rolled product that can generate a width prediction model that accurately predicts the width distribution in the length direction of the rough rolled product.

[0109] Explanation of Reference Numerals

[0110] 1... hot rolling line; 2... heating furnace; 3... descaling device; 4... width reduction stamping device; 5... rough rolling mill; 6... finish rolling mill; 7... cooling device; 8... coiler (reeler); 11... rough rolling exit side width gauge; 12... finish rolling exit side width gauge; 13... width gauge before coiler (coiler inlet side width gauge); 14... rolling line; 15a, 15b, 15c, 15d... cameras; 21... thermometer; 22... walking beam (fixed slideway); 23... moving slideway; 41... die for width reduction; 41a... parallel part; 41b... inclined part; 42... driving device; 43... pinch roll; 90... control controller; 91... control computer; 92... host computer; 100... width prediction model generation unit; 101... database unit; 102... machine learning unit; 103... data acquisition unit; 110... width prediction unit; 111... input data acquisition unit; M... width prediction model; SA... slab; SB... steel sheet.

Claims

1. A method for predicting the width of a rough rolling piece, which predicts the width of the rough rolling piece in a hot rolling line, and the hot rolling line includes: A heating furnace that heats a slab; A width reduction stamping device that intermittently performs width reduction on the heated slab; a rough rolling mill that rough rolls the width-reduced slab to produce a rough rolled product; and a finishing mill that finish rolls the rough rolled product to produce a hot rolled steel sheet, The width prediction method for the rough rolled product is characterized by including the following steps: Using a width prediction model learned through machine learning, predict the width distribution in the length direction of the rough rolled product. The width prediction model includes, as input data, one or more operation parameters selected from the operation parameters of the width reduction stamping device and one or more operation parameters selected from the operation parameters of the rough rolling mill, and uses the information on the width distribution in the length direction of the rough rolled product as output data.

2. The method for predicting the width of a rough rolling piece according to claim 1, characterized in that The operation parameters of the width reduction stamping device and the operation parameters of the rough rolling mill are operation parameters determined as representative values in the length direction of the rough rolled product.

3. The method for predicting the width of a rough rolling piece according to claim 1 or 2, characterized in that The width prediction model includes, as the input data, one or more operation parameters selected from the operation parameters of the heating furnace.

4. The method for predicting the width of a rough rolling piece according to any one of claims 1 to 3, characterized in that The width prediction model includes, as the input data, one or more attribute parameters selected from the attribute information of the slab.

5. A method for controlling the width of a rough rolling piece, characterized in that Using the width prediction method for the rough rolled product according to any one of claims 1 to 4, predict the width distribution in the length direction of the rough rolled product, and set the operation parameters of the rough rolling mill based on the predicted width distribution.

6. A method for manufacturing a hot rolled steel plate, characterized in that With respect to the rough rolled product whose width has been controlled using the width control method for the rough rolled product according to claim 5, perform finish rolling using the finishing mill to produce a hot rolled steel sheet.

7. A method for generating a width prediction model of a rough rolling piece, which generates a width prediction model for predicting the width of the rough rolling piece in a hot rolling line, and the hot rolling line includes: A heating furnace that heats a slab; A width reduction stamping device that intermittently performs width reduction on the heated slab; a rough rolling mill that rough rolls the width-reduced slab to produce a rough rolled product; and a finishing mill that finish rolls the rough rolled product to produce a hot rolled steel sheet, The method for generating the width prediction model for the rough rolled product is characterized by including the following steps: Obtain a plurality of learning data, and generate the width prediction model through machine learning using the obtained plurality of learning data. Among them, the plurality of learning data includes, as input actual data, one or more operation actual data selected from the operation parameters of the width reduction stamping device and one or more operation actual data selected from the operation parameters of the rough rolling mill, and uses the information on the width distribution in the length direction of the rough rolled product as output actual data.

Images