Warpage amount measurement method for rolled material, warpage amount reduction method, warpage amount measurement device, warpage amount reduction device, learning method for warpage measurement model, and warpage measurement model

Abstract

Provided is a technique for measuring the amount of warpage in a rolled material, where the amount of warpage in the rolled material can be accurately measured. Provided is a warpage amount measurement method for a rolled material, where the amount of warpage in a rolled material is measured before and after a rolling mill on a hot rolling line, wherein: a camera capable of measuring luminance in a wavelength band from visible light to near infrared is used to capture an image of the rolled material after rolling from diagonally above the rolled material; the captured image of the rolled material is input into a warpage measurement model created in advance by machine learning; regions where the plate widths of the tip and tail end parts of the rolled material are reduced are estimated; a measured image region to be used in the measurement of the warpage amount is identified; a plate width edge part of the rolled material is detected on the basis of the luminance value of the image of the rolled material present in the measured image region; the image of the rolled material is divided along the rolling direction of the rolled material; the shapes of the plate width edge parts in the divided images are approximated by a quadratic expression; and the warpage amount of the rolled material is quantified by the curvature or the warpage height thereof on the basis of the shapes of the plate width edge parts approximated by the quadratic expression.

Description

Method for measuring warpage of rolled material, method for reducing warpage, device for measuring warpage, device for reducing warpage, and learning method for warpage measurement model and warpage measurement model

[0001] The present invention relates to a method for measuring the amount of warp of rolled material before and after a rolling mill in a hot rolling line, a method for reducing the amount of warp, a warp measuring device and a warp reducing device, and further to a learning method for a warp measurement model and a warp measurement model used therein.

[0002] In a hot rolling line, slabs cast to a thickness of about 200 to 300 mm are rolled by a single or multiple roughing mills, and if necessary, reverse rolled to form steel plates with a thickness of about 20 to 80 mm, so-called sheet bars. The sheet bars are then finish-rolled to a thickness of about 0.6 to 30 mm by multiple finishing mills arranged in series. Note that the steel plates may be cooled with cooling water during hot rolling to suppress surface defects and control the structure.

[0003] During rough rolling or finish rolling, if there is asymmetry in the rolling conditions between the top and bottom surfaces of a steel sheet, warpage, in which the steel sheet curves in the vertical direction, may occur. Specifically, it is said that warpage occurs in steel sheets when there is asymmetry in factors such as the friction coefficient, roll speed, roll conditions, steel sheet temperature, and incidence angle into the rolling mill. In particular, the leading and trailing ends of a steel sheet correspond to the unsteady rolling region, making it difficult to control the aforementioned factors, and the amount of warpage tends to be large at the leading and trailing ends. If large warpage occurs at the leading end of a steel sheet, problems such as the steel sheet not being able to be caught in the next rolling mill and being unable to be rolled, or the steel sheet damaging surrounding equipment, may occur. In particular, thick steel sheets, i.e., sheet bars, are more likely to damage surrounding equipment due to their high rigidity. To prevent such problems, it is essential to measure the warpage of steel sheets occurring in the stand in question and the preceding stands and take measures to prevent it.

[0004] Patent Document 1 discloses a method for measuring the amount of warpage of a rolled material before and after a rolling mill on a hot rolling line. This measurement method uses an apparatus configuration as shown in Fig. 7, in which an image of the rolled material after rolling is taken from diagonally above the rolled material using a camera, and the widthwise edge portions of the rolled material are detected based on the brightness values ​​of the captured image of the rolled material. The shape of the widthwise edge portions is then approximated by a quadratic equation, and the amount of warpage of the rolled material is quantified in terms of curvature based on the shape of the widthwise edge portions approximated by the quadratic equation.

[0005] Japanese Patent Application Laid-Open No. 2019-181562

[0006] However, the above-mentioned conventional technologies have the following problems that must be solved. For example, an example of a warpage measurement system described in Patent Document 1 is shown in FIG. 7 as a schematic diagram. FIG. 8 shows an example of an image in which the warpage amount was measured correctly using this device configuration. FIG. 9 shows an example of an image in which the warpage amount measured using the conventional method was abnormal. As shown in FIG. 8, when the shape of the leading edge of a steel plate is close to a rectangle and has an upward camber, the warpage amount can be quantified relatively accurately from the image of the steel plate. The width edge of the steel plate on the operator side (OP) can be accurately extracted so that the warpage amount can be measured. Note that the width edge on the motor side (DR) cannot be used for measuring the warpage amount because it is obstructed by surrounding equipment. On the other hand, when the leading edge of the steel plate has a complex shape, such as a downward camber or a fishtail, there is room for improvement in the accuracy of the warpage measurement. In particular, when equipment such as side guides or descaling equipment is present around the exit side of the rolling mill, which obstructs the capture of an image of the steel plate, the steel plate must be captured from an obliquely upward angle. In such cases, it becomes difficult to extract the shape of the width direction end of the steel plate from the image depending on the capture angle. For example, as shown in FIG. 9, the amount of warpage may show an abnormal value, making it difficult to measure with accuracy.

[0007] The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a technique for measuring the amount of warpage of a rolled material, which can accurately measure the amount of warpage of the rolled material.

[0008] The method for measuring the amount of warpage of a rolled material according to the present invention, which advantageously solves the above-mentioned problems, is a method for measuring the amount of warpage of a rolled material before and after a rolling mill on a hot rolling line, and is characterized in that an image of the rolled material after rolling is taken from diagonally above the rolled material using a camera capable of measuring luminance in wavelength bands from visible light to near-infrared, the image of the rolled material taken is input into a warpage measurement model created in advance by machine learning, regions where the width of the head and tail ends of the rolled material decreases are estimated, a measurement image region to be used for measuring the amount of warpage is specified, width edge portions of the rolled material are detected based on the luminance values ​​of the image of the rolled material present in the measurement image region, the image of the rolled material is divided along the rolling direction of the rolled material, the shape of the width edge portions in each divided image is approximated by a quadratic equation, and the amount of warpage of the rolled material is quantified in terms of curvature or warpage height based on the shape of the width edge portion approximated by the quadratic equation.

[0009] In addition, the method for measuring the amount of warpage of rolled material according to the present invention is more preferably: (a) the warpage measurement model is machine-learned using images of the rolled material taken in the past by the camera and the manufacturing conditions of the rolled material taken in the past as explanatory variables, and using regions in the images of the rolled material taken in the past where the plate width at the leading and trailing ends of the rolled material decreases as objective variables; and (b) the warpage measurement model is machine-learned using any of logistic regression analysis, decision tree, neural network, and deep learning.

[0010] The method for reducing the amount of warpage of a rolled material according to the present invention, which advantageously solves the above-mentioned problem, is characterized in that it uses any of the above-mentioned methods for measuring the amount of warpage of a rolled material to perform control to suppress the amount of warpage of the rolled material based on the amount of warpage quantified.

[0011] The warpage measuring device for rolled material according to the present invention, which advantageously solves the above-mentioned problems, is a rolled material warpage measuring device that measures the warpage of rolled material before and after a rolling mill on a hot rolling line, and is characterized by comprising: a camera that takes an image of the rolled material after rolling from diagonally above the rolled material and is capable of measuring luminance in wavelength bands from visible light to near-infrared; and an information processing device that uses a warpage measurement model created in advance by machine learning to estimate from the taken image of the rolled material an area where the plate width at the head and tail ends of the rolled material decreases, and identifies a measurement image area to be used for measuring the amount of warpage, detects a plate width edge portion of the rolled material based on the luminance value of the image of the rolled material present in the measurement image area, divides the image of the rolled material along the rolling direction of the rolled material, quadratically approximates the shape of the plate width edge portion in each divided image, and quantifies the amount of warpage of the rolled material in terms of curvature or warpage height based on the quadratically approximated shape of the plate width edge portion.

[0012] The warpage reduction device for rolled material according to the present invention, which advantageously solves the above-mentioned problems, is characterized by being configured with a means for suppressing warpage, which includes changing either the rolling conditions or the crop cutting amount based on the warpage quantified using the warpage measurement device.

[0013] The warpage measurement model learning method of the present invention, which advantageously solves the above-mentioned problems, is a method for learning a warpage measurement model used to measure the amount of warpage of a rolled material before and after a rolling mill on a hot rolling line, and is characterized in that it performs machine learning using an image of the rolled material after rolling, taken from diagonally above the rolled material using a camera capable of measuring brightness in the wavelength band from visible light to near-infrared, and the manufacturing conditions of the rolled material as explanatory variables, and using the area in the image of the rolled material where the plate width at the leading and trailing ends of the rolled material decreases as a target variable.

[0014] The warpage measurement model of the present invention, which advantageously solves the above-mentioned problems, is a model trained by the above-mentioned warpage measurement model training method, characterized in that at least photographed images of the rolled material are used as input variables, and the regions where the plate width decreases at the head and tail ends of the rolled material are used as output variables.

[0015] According to the present invention, even if there is a region where the plate width is reduced at the end of the rolled material, the amount of warpage can be accurately measured, which contributes to reducing the amount of cutting crop and to reducing equipment damage accidents, making it industrially useful.

[0016] FIG. 1 is a schematic block diagram showing the system configuration of a rolled material warpage measuring device according to one embodiment of the present invention. FIG. 2 is a schematic diagram showing an example of constructing a warpage measurement model. FIG. 3 is a schematic diagram explaining a method for specifying an image to be used for measuring the amount of warpage from photographed images using a warpage measurement model. FIG. 4 is a flow diagram explaining a method for measuring the amount of warpage using the rolled material warpage measuring device according to the above embodiment. FIG. 5 is a schematic diagram showing an overview of a hot rolling line to which the above embodiment is applied. FIG. 6 is a schematic diagram explaining an example of equipment trouble caused by upward warpage of a steel plate. FIG. 7 is a schematic conceptual diagram showing an example of a conventional warpage measuring system. FIG. 8 is an image of the tip of a rolled material where the amount of warpage was normally observed by a conventional method. FIG. 9 is an image of the tip of a rolled material where the amount of warpage was abnormal by a conventional method. FIG. 10 is a graph comparing the amount of warpage measured by the method of the present invention with the amount of warpage measured by a conventional method.

[0017] The following describes in detail embodiments of the present invention. Note that the drawings are schematic and may differ from the actual embodiments. Furthermore, the following embodiments exemplify devices and methods for embodying the technical concept of the present invention, and are not intended to limit the configuration to those described below. In other words, the technical concept of the present invention can be modified in various ways within the technical scope described in the claims.

[0018] [Hot Rolling Line] First, with reference to FIG. 5, the configuration of a hot rolling line 100 to which a method for measuring the amount of warpage of a rolled material according to one embodiment of the present invention is applied will be described.

[0019] In the hot rolling line 100 to which the method for measuring the amount of warpage of a rolled material according to this embodiment is applied, first, a slab having a thickness of about 200 to 300 mm, a width of about 600 to 2200 mm, and a length of about 5 to 15 m, which has been produced in the preceding casting process, is heated to a temperature of about 1000 to 1250°C and then extracted from the heating furnace 110.

[0020] Next, in rough rolling, the steel sheet S is rough rolled to a thickness of approximately 20 to 80 mm using multiple rough rolling mills R1 to R5. If necessary, a cooling device for the steel sheet may be installed during rough rolling. Examples of cooling devices include a device that cools the steel sheet S using rod-shaped water, known as laminar cooling, and a device that powerfully cools the steel sheet S by forcefully spraying a large amount of water. Note that, in order to control the amount of warping of the steel sheet S, it is preferable to configure the cooling device so that it can adjust the temperature of the upper and lower surfaces of the steel sheet S, i.e., the deformation resistance of the upper and lower surfaces of the steel sheet S. Specifically, it is preferable to configure the cooling device so that it can adjust the flow rate of cooling water on the upper and lower surfaces of the steel sheet S or so that it can stop cooling on either the upper or lower surface of the steel sheet S.

[0021] Generally, the shape of the leading and tailing ends of the steel sheet S after rough rolling is called a fishtail, in which only the widthwise end portion extends in the longitudinal direction of the steel sheet S. If the fishtail shape differs in the width direction of the steel sheet S, the steel sheet S may meander during finish rolling, which may lead to problems. For this reason, from the viewpoint of operational stability, it is common to cut the leading and tailing ends of the steel sheet S after rough rolling using a crop shear FSC, which cuts off part or all of them to form a rectangle.

[0022] After rough rolling, the surface temperature of the steel plate is high, at around 700 to 1000°C, causing iron oxides (scale) to form. To prevent scratches remaining on the final product, the iron oxides are removed using a finisher scale breaker (FSB) under water pressure before finish rolling.

[0023] In the finish rolling, multiple finishing mills F1 to F7 are arranged in series at intervals of 2000 to 4000 mm, and each finishing mill is equipped with a large frame called a housing and multiple rolling rolls arranged in the vertical direction. Finishing mills come in several types, such as a four-high type in which two pairs of rolling rolls are arranged above and below the steel plate, and a six-high type in which three pairs of rolling rolls are arranged above and below the steel plate. Figure 5 shows configuration examples of four-high and six-high finishing mills F1 to F7. In the finishing mills F1 to F7 shown in Figure 5, the steel plate S is transported by a transport device (not shown) and rolled from right to left in Figure 5. At this time, the steel plate S is rolled by a pair of upper and lower work rolls arranged inside the housing, and backup rolls are arranged above and below the work rolls.

[0024] Generally, the rolling load of the steel sheet S is reduced by applying tension to the steel sheet S. Therefore, in order to apply tension between the finishing rolling mills, the tension of the steel sheet S is controlled by increasing or decreasing the angle of a looper provided between the finishing rolling mills. In the finish rolling, the steel sheet S is finish-rolled to a thickness of about 0.6 to 30 mm.

[0025] Although it differs depending on the design specifications of each factory, thermometers FET and FDT for measuring the temperature of the steel sheet S are installed at least on the entry side and exit side of the finish rolling. The thermometer FET on the entry side of the finish rolling may be placed before the finisher scale breaker FSB in order to eliminate the influence of the FSB, or may be placed before cutting by the crop shear FCS. In particular, the thermometer FET on the entry side of the finish rolling is an important thermometer for controlling the amount of warpage of the steel sheet S, and therefore it is preferable that it be a thermometer that can measure both the top and bottom surfaces of the steel sheet S.

[0026] In hot rolling using the above-mentioned equipment, when measuring and quantifying warpage, it is necessary to optimize optical conditions such as radiation from the heated steel plate. At the same time, it is required to quantitatively indicate warpage in a region of about 1.0 m or more from the leading and trailing ends of the steel plate while dealing with the reflection of the scale breaker on the finishing entry side, the cooling water sprayed between stands, the looper, and other equipment.

[0027] In order to grasp the warpage behavior of the steel sheet S due to rolling, it is necessary to grasp the amount of warpage of the steel sheet S before and after rolling and the conditions for entering the rolling mill, i.e., the position of the rolling rolls and the height and width directions of the steel sheet S when the steel sheet S enters the rolling mill. Before finish rolling, a leveler may be provided or the leading edge of the steel sheet may be cut with a crop shear. Therefore, it is possible to consider the steel sheet S as being straight and without warpage before finish rolling. On the other hand, the steel sheet S after rough rolling has a region called a crop where the sheet width decreases at the leading and trailing edges. It is preferable to actually measure the amount of warpage of the steel sheet S and use it for practical purposes. Furthermore, since the shape of the leading edge and trailing edge of the steel sheet S may trigger an increase in warpage during subsequent rolling, it is preferable to grasp the warpage shape in a length region of about 200 to 400 mm in the rolling direction during measurement. The thickness and rigidity of the steel sheet S increases after rough rolling or in the stage before the finishing rolling mill, and there is a higher risk of serious trouble such as equipment damage. Therefore, it is particularly important to measure the amount of warpage of the steel sheet S after rough rolling or in the stage before the finishing rolling mill.

[0028] FIG. 6 shows a schematic example of an equipment trouble caused by warpage at the leading edge of the steel sheet S. For example, if the top surface of the steel sheet S before entering the upstream rolling mill is the low-temperature side (LT) and the bottom surface is the high-temperature side (HT) ( FIG. 6( a)), the deformation resistance of the top surface of the steel sheet S is greater and the deformation resistance of the bottom surface of the steel sheet S is smaller. As a result, the leading edge of the steel sheet S will warp upward at the exit side of the upstream rolling mill shown in FIG. 6( b). If this warpage becomes excessive, the steel sheet S may be unable to enter the downstream rolling mill, or may be caught in the work rolls 12 or backup rolls 13 at the downstream rolling mill, causing equipment trouble, as shown in FIG. 6( c). FD in FIG. 6 indicates the conveying direction of the steel sheet S.

[0029] [Method for Measuring Warpage of Rolled Material] When photographing the state of warpage of the steel sheet S before and after rolling from the side of the steel sheet S, it is difficult to capture the state of warpage of the steel sheet S immediately after rolling or immediately before rolling because the roll bite located between the work rolls 12 is inside the housing. Therefore, in this embodiment, the imaging field of view 21 of the imaging device 3 is selected to be immediately after the rolling mill or crop shear, or between the housings of the rolling mill. FIG. 7 is a schematic diagram for explaining the imaging angle θ of the imaging device 3. As shown in FIG. 7, the imaging angle θ of the imaging device 3 is preferably set to an angle at which an image of the steel sheet S is captured from diagonally above, specifically, within a range of 10 to 60 degrees with respect to the horizontal direction. More preferably, θ is within a range of 30 to 60 degrees with respect to the horizontal direction. If the imaging angle θ is less than the lower limit, when there is a difference in the height direction position of the steel sheet S in the width direction of the steel sheet S, the steel sheet S is photographed at an angle, making it difficult to detect the edge portion of the steel sheet S. On the other hand, if the imaging angle θ is greater than the upper limit, it becomes difficult to obtain information about the height direction of the steel plate S.

[0030] The distance L between the steel sheet S and the imaging device 3, i.e., the distance from the tip of the imaging device 3 to the center of the steel sheet S in the width direction, is not set within a particularly preferable range. However, if the distance L is extremely short, the imaging device 3 may be heated by radiant heat from the steel sheet S, which may cause a malfunction. A heat shield may be installed between the imaging device 3 and the steel sheet S. The accuracy of measuring the amount of warp of the steel sheet S is also affected by the clarity of the image. On the other hand, the measurement accuracy also varies depending on the distance L and the number of pixels. If the distance L is extremely long, the accuracy of measuring the amount of warp of the steel sheet S may decrease. It is possible to enlarge the captured image using a telephoto lens. On the other hand, in consideration of the accuracy of measuring the amount of warp of the steel sheet S, it is preferable to set the imaging length per pixel to at least 5 mm or less. In Figure 7, the height H of the imaging device 3 relative to the pass line of the steel sheet S is preferably set to a height that ensures the field of view width W in the width direction of the steel sheet S. Depending on the size of the steel sheet S, the field of view width W is preferably approximately 2000 to 3200 mm. Although it depends on the steel sheet temperature and the equipment configuration, the distance L from the camera 3 to the widthwise center of the steel sheet S is preferably 3 to 10 m, and the height of the camera 3 from the pass line is preferably 800 to 4000 mm.

[0031] The imaging device 3 may be a color camera or a monochrome camera. In the case of a color camera, the edge portion of the steel sheet S may be detected using any one of R (red), G (green), and B (blue), or a grayscale value.

[0032] In hot rolling mills, fine iron oxide scatters during rolling and descaling. For this reason, it is preferable to protect the imaging device 3 in a dustproof case. When photographing a steel sheet S in an environment where a lot of water vapor is generated, it is effective to collect only the luminance of wavelengths longer than visible light, such as near-infrared wavelengths, which reduce the diffuse reflection of water vapor. An imaging device 3 capable of measuring the luminance of wavelengths ranging from visible light (e.g., 360 to 830 nm in JIS Z 8120:2001) to near-infrared light (e.g., 700 to 2500 nm in JIS 0134:2002) may be equipped with a filter that cuts visible light, thereby using information only in the near-infrared wavelength band. Similar effects can be achieved using a mid-infrared or far-infrared camera. However, the imaging device 3 may be expensive. However, because the luminance of infrared wavelengths decreases as the temperature of the steel sheet S decreases, it is preferable to select the wavelength band to be measured depending on the environment.

[0033] When calculating the amount of warpage of the steel sheet S from the images captured by the imaging device 3, the relationship between pixels and actual dimensions and the selection of the captured images can be determined using the method described in Patent Document 1. In this embodiment, as shown in FIG. 1, the leading edge of the steel sheet S is photographed by the imaging device 3 installed at the exit of the roughing rolling mills R1 to R5, the crop shear FSC, and the upstream finishing rolling mills F1 and F2. Then, as shown in FIG. 2, the captured images are labeled with the shape of the range (crop) where the sheet width narrows at the leading and trailing ends of the steel sheet, and a warpage measurement model for detecting the crop is constructed by machine learning. The warpage measurement model is constructed, for example, on the server 6. Here, the labeled crops from a variety of images that show different aspects such as the width, thickness, presence or absence of warpage, and crop shape are treated as training data for the warpage measurement model (FIG. 2(a)). Here, labeling is performed by defining the range from the point where the boundary (edge) between the steel sheet S and the background in the image changes from the width direction to the length direction as a crop, until the point where the boundary changes back to the width direction (FIG. 2(b)). Then, the constructed warpage measurement model is applied to the training data for inference, and the crop is estimated (FIG. 2(c)). At this time, it is preferable to learn the crop shape again for image data for which inference could not be made, thereby improving the accuracy of the warpage measurement model.

[0034] FIG. 3 is a schematic diagram illustrating a method for identifying an image to be used for warpage measurement from captured images using a warpage measurement model. FIG. 4 is a flow diagram illustrating a method for measuring the warpage amount using a rolled material warpage measurement device. In step 1 (S01), a camera 3 installed on the exit side of a rolling mill or the like captures an image of the leading edge of the steel sheet S. The newly captured image of the steel sheet S is then sent to a server 6, which operates the constructed warpage measurement model, via a machine-side relay panel 4 and a warpage meter 5. In step 2 (S02), a crop is estimated from the image constructed in the server 6. The image after applying the warpage measurement model and data on the detected crop range are then sent to the warpage meter 5. The method described in Patent Document 1, for example, can be used to measure the warpage amount. In step 3 (S03), the warpage meter 5 masks the cropped portion from the image based on the masked data ( FIG. 3(c) ). In step 4 (S04), the shape of the widthwise edge of the steel sheet S is extracted from the remaining images, and the warpage amount is measured. In step 5 (S05), the measurement result of the amount of warpage is sent to the host computer 7, for example, a rolling process computer, and the amount of warpage is calculated. The host computer 7 transmits the amount of warpage to the warpage suppressing device 8.

[0035] It is preferable to approximate the shape of the measured edge portion by quadratic approximation, and quantify the amount of warpage of the steel sheet S by curvature or warpage height. Specifically, a general quadratic formula is first defined as shown in Equation (1) below. The height in the thickness direction of the edge portion at the position farthest from the pass line in the conveying direction FD of the steel sheet S is defined as the warpage height, with upward warpage being positive and downward warpage being negative. Furthermore, the curvature of the edge portion is calculated by Equation (2) below using coefficients a, b, and c calculated by the least squares method. Note that although Equation (2) can calculate the curvature of the edge portion, the value changes depending on the rolling direction. Therefore, a representative value can be obtained by taking the average value in the calculated rolling direction or by taking the curvature at the center of the rolling direction. The value calculated by Equation (2) can be positive or negative. If the warpage of the steel sheet S is upward, the value is positive, and if it is downward, the value is negative. The shape of the edge portion may also be approximated by other polynomials, but the higher the degree, the greater the influence of noise, making it difficult to accurately calculate the curvature. Also, approximating the shape of the edge portion by the least squares method using a general circular formula is not preferable because it may result in a solution with a large curvature if the warp is close to a straight line. ２ +bx+c (1) 1 / r=2a / {1+(2ax+b) ２} １．５ (2)

[0036] In step 6 (S06), when the warpage amount is a predetermined value, the warpage suppression device 8 first issues an instruction to change the rolling conditions when rolling in a downstream rolling mill, particularly the rolling mill immediately following the target rolling mill. By rolling in accordance with the instruction, warpage is suppressed during subsequent rolling. The rolling conditions to be changed include the cooling conditions of the cooling device caused by the temperature difference between the top and bottom surfaces of the steel sheet, the cutting amount by the crop shear, the line speed, etc., and the system automatically controls the amount of change in these rolling conditions based on the magnitude of the warpage amount. Also, in step 6 (S06), when the warpage amount is a predetermined value, the warpage suppression device 8 issues an instruction to change the rolling conditions when rolling in the target rolling mill from the next time onwards. By rolling in accordance with the instruction, warpage is suppressed during subsequent rolling by the target rolling mill. The rolling conditions to be changed include the cooling conditions of the cooling device caused by the temperature difference between the top and bottom surfaces of the steel sheet, the line speed, etc., and the system automatically controls the amount of change in these rolling conditions based on the magnitude of the warpage amount.

[0037] In this embodiment, the rough warp meter, finish warp meter, server, host computer, warp suppression device, etc. may be configured as separate computers. Alternatively, multiple functions may be integrated into one computer.

[0038] The effects of the present invention were verified using the hot rolling equipment shown in Figure 5. A continuously cast slab was heated and subjected to a sizing press, rough rolling, crop shearing, and a finisher scale breaker, followed by hot finish rolling. The warpage amount was calculated offline from images of the leading edge of the steel sheet S taken with the steel sheet leading and tail end warpage measurement system shown in Figure 7 installed on the outlet side of the roughing rolling mill R5. A telephoto lens was attached to the imaging device and adjusted so that the pixel size of the captured image was 2 mm. The frame rate of the imaging device was set to 20 fps, and images that captured the steel sheet as long as possible were selected and analyzed. The rolling direction length of the steel sheet obtained as the imaging field of view was 1.7 m. The warpage amount was evaluated in mm by quadratically approximating the edge shape and defining the height in the thickness direction of the edge at the position farthest from the pass line in the conveying direction FD of the steel sheet S as the warpage height. The imaging angle θ was set to 50° with respect to the horizontal direction.

[0039] FIG. 10(a) shows the relationship between the measured warpage calculated by the conventional method described in Patent Document 1 and the actually measured warpage. FIG. 10(b) shows the relationship between the warpage and the actually measured warpage when the leading edge crop is estimated from the captured image using a warpage measurement model, masked, and the widthwise edge portion of the steel sheet S is extracted from the remaining image. The warpage measurement model was created to estimate the crop shape using 499 leading edge images with different leading edge shapes, widths, and thicknesses as training data. The leading edge images that could not be estimated were then trained as training data. The number of training data is preferably 100 or more, and more preferably 499 or more, to improve the accuracy of the warpage measurement model characterized by the present invention. There is no particular upper limit, but 100,000 or less is preferred.

[0040] 10(a) and 10(b), it can be seen that the application of the present invention improves the measurement accuracy as the amount of warpage increases. Furthermore, even in the vicinity of 0 mm where the amount of warpage is small, the variation in the difference from the actual measured value is small, and it can be seen that more accurate measurements are possible.

[0041] According to the present invention, even if there is a region where the plate width is reduced at the end of the rolled material, the amount of warpage can be accurately measured, which contributes to reducing the amount of cutting crop and to reducing equipment damage accidents, making it industrially useful.

[0042] 100 Hot rolling line 110 Heating furnace 1 Warpage measurement system 3 Imaging device (camera) 4 Machine-side relay panel 5 Warpage meter 6 Server 7 Host computer 8 Warpage suppression device 12 Work roll 13 Backup roll W Imaging field of view R1 to R5 Roughing mill F1 to F7 Finishing mill S Steel plate FSC Crop shear FSB Finisher scale breaker FET, FDT Thermometer HT High temperature side LT Low temperature side FD Transport direction OP Operator side DR Motor side

Claims

1. A method for measuring the warp amount of a rolled material, which measures the warp amount of the rolled material before and after a rolling mill in a hot rolling line, comprising: taking an image of the rolled material after rolling from obliquely above the rolled material using a camera capable of measuring the luminance in the wavelength band from visible light to near infrared; inputting the image of the rolled material taken into a warp measurement model created in advance by machine learning; estimating a region where the sheet width of the leading and trailing ends of the rolled material decreases; specifying a measurement image region to be used for measuring the warp amount; detecting the sheet width edge portion of the rolled material based on the luminance value of the image of the rolled material existing in the measurement image region; dividing the image of the rolled material along the rolling direction of the rolled material; approximately fitting the shape of the sheet width edge portion in each of the divided images with a quadratic equation; and quantifying the warp amount of the rolled material in terms of curvature or warp height based on the shape of the sheet width edge portion approximately fitted with the quadratic equation.

2. The method for measuring the warp amount of a rolled material according to claim 1, wherein the warp measurement model is obtained by machine learning using, as explanatory variables, the past captured images of the rolled material by the camera and the manufacturing conditions of the past rolled material, and using, as the target variable, a region where the sheet width of the leading and trailing ends of the rolled material in the past captured images of the rolled material decreases.

3. The method for measuring the warp amount of a rolled material according to claim 1, wherein the warp measurement model is obtained by machine learning using any one of logistic regression analysis, decision tree, neural network, and deep learning.

4. A method for reducing the warp amount of a rolled material, which performs control to suppress the warp amount of the rolled material based on the quantified warp amount, using the method for measuring the warp amount of a rolled material according to any one of claims 1 to 3.

5. A measuring device for measuring the amount of warpage of a rolled material before and after a rolling mill in a hot rolling line, comprising: a camera capable of measuring the luminance in a wavelength band from visible light to near infrared, which takes an image of the rolled material obliquely from above the rolled material; and an information processing device that estimates a region where the plate width of the leading and trailing ends of the rolled material decreases from the taken image of the rolled material using a warpage measurement model created in advance by machine learning, specifies a measurement image region to be used for measuring the amount of warpage, detects the plate width edge portion of the rolled material based on the luminance value of the image of the rolled material existing in the measurement image region, divides the image of the rolled material along the rolling direction of the rolled material, approximately represents the shape of the plate width edge portion in each divided image by a quadratic equation, and quantifies the amount of warpage of the rolled material in terms of curvature or warpage height based on the shape of the plate width edge portion approximately represented by the quadratic equation. Measuring device for the amount of warpage of a rolled material.

6. Based on the amount of warpage quantified using the measuring device for the amount of warpage of a rolled material according to claim 5, a device for reducing the amount of warpage of a rolled material, which is configured by means for suppressing warpage, including any one of the rolling conditions to be changed and the amount of crop cutting to be changed.

7. A learning method for a warpage measurement model used for measuring the amount of warpage of a rolled material before and after a rolling mill in a hot rolling line, comprising: using a camera capable of measuring the luminance in a wavelength band from visible light to near infrared to take a captured image of the rolled material after rolling obliquely from above the rolled material and the manufacturing conditions of the rolled material as explanatory variables, and performing machine learning with the region where the plate width of the leading and trailing ends of the rolled material decreases in the captured image of the rolled material as the target variable. Learning method for a warpage measurement model.

8. A model learned by the learning method for a warpage measurement model according to claim 7, which has at least the captured image of the rolled material as an input variable and the region where the plate width of the leading and trailing ends of the rolled material decreases as an output variable. Warpage measurement model.

Images