JP2023022814A - Generation method of wound shape determination model of steel strip coil, wound shape determination method, treatment step setting method and manufacturing method of steel strip coil - Google Patents

Abstract

To automatically and correctly determine whether or not there is need to correct the wound shape of a steel strip coil in each manufacturing process of a steel strip such that the steel strip coil satisfies a predetermined quality standard after becoming a final product.SOLUTION: A generation method of a wound shape determination model of a steel strip coil comprises the steps of: forming a determination performance database by accumulating a dataset including image data of the wound shape, quality standard data about the contents of the quality standard and determination result data about a result of the determination every time determination on whether or not there is need to correct the wound shape of a steel strip coil in manufacture is performed such that the steel strip coil satisfies the predetermined quality standard after becoming a final product; performing machine learning by using the determination performance database; and generating a wound shape determination model for outputting the determination result data when the image data and the quality standard data are input.SELECTED DRAWING: Figure 5

Description

The present invention relates to a method for generating a winding shape determination model for a steel strip coil, a method for determining a winding shape, a method for setting treatment steps, and a method for manufacturing a steel strip coil.

Steel sheets, which are steel products, are manufactured as steel strips with a thickness of about 0.1 to 25 mm by a hot-rolling line or a cold-rolling line, and conveyed between manufacturing processes in a coiled state by a winder. , is generally shipped as a steel strip product. A steel strip wound in such a coil shape is called a steel strip coil.

In inspections for quality assurance of thin steel sheets, in addition to the mechanical properties and surface quality of the thin steel sheets, the wound shape of the steel strip coil is also subject to inspection. A quality inspection of a thin steel sheet wound on a steel strip coil is performed by observing the side surface of the steel strip coil and determining whether or not quality defects such as cracks, abrasions, and bends have occurred. Such defects occurring on the sides of the steel strip coil are called "ear pain" or the like.

Here, when the steel strip manufactured in the hot rolling line is meandering in the width direction and wound by the winder, a part of the side surface of the steel strip coil may protrude in the width direction of the coil. Such a shape is called a “telescope”. This telescope is also subject to inspection because the quality of the steel strip coil is defective. The telescope has an inner periphery telescope shape in which the inner periphery of the steel strip coil protrudes in the coil width direction, an outer periphery telescope shape in which the outer periphery of the steel strip coil protrudes in the coil width direction, and an inner periphery telescope shape in which the outer periphery of the steel strip coil protrudes in the coil width direction. There is a form such as a bowl-shaped telescopic shape in which the peripheral part and the outer peripheral part protrude in the coil width direction. For example, Patent Literature 1 discloses a steel strip winding method for preventing the occurrence of telescope by winding a steel strip manufactured in a hot rolling line around a coiler while centering it so that it does not meander in the width direction. It is

If the steel strip coil has a telescope, when the steel strip coil is conveyed with the axis center of the coil facing up and down, defects such as cracks, rubbing, and bending may occur on the side of the coil that comes into contact with the conveying device. is likely to occur, and may eventually result in defects such as ear pain. Hereinafter, the form in which the steel strip is wound in a coil shape is referred to as the "wound shape", including aspects such as ear pain and telescope, which are determined to be defective in the quality inspection of the steel strip coil.

JP-A-2006-61934 Japanese Patent Application Laid-Open No. 2010-210388

Here, inspection of the winding shape of the steel strip coil is generally carried out by visual observation by a judge. Specifically, the judge approaches the steel strip coil and judges whether the winding shape is good or bad while irradiating the side surface of the steel strip coil with light from a flashlight or the like. However, since the steel strip coil is hot immediately after being wound on the hot rolling line, the inspector must judge the winding shape from a position some distance away from the steel strip coil, which causes large variations in inspection results. I had a problem.

In response to such problems, Patent Literature 2 discloses a coiled steel strip side imaging method and a side observation device for facilitating determination of whether the winding shape of the steel strip coil is good or bad. . In the technique disclosed in Patent Document 2, the irradiation condition for irradiating light from the side surface of the steel strip coil and the imaging condition for imaging the side surface of the steel strip coil are set to have a predetermined relationship. A clear image of the winding shape is obtained. By referring to the clear image and judging whether the winding shape of the steel strip coil is good or bad, the judge does not need to directly visually observe the high-temperature steel strip coil.

However, even if it is possible to obtain a clear image by imaging the winding shape of the steel strip coil using the technique disclosed in Patent Document 2, the quality of the winding shape is determined based on the obtained image. There was no change in the fact that it was done by visual observation by the person. For this reason, the judgment result of whether the winding shape is good or bad varies depending on the judge, and in order to improve the quality of the steel strip coil and prevent manufacturing troubles, there is still room for improvement in the judgment of the winding shape of the steel strip coil. was there.

The present invention has been made to solve the above-mentioned problems. A method for generating a steel strip coil winding shape determination model, a winding shape determination method, a treatment process setting method, and a steel strip coil manufacturing method that can automatically and accurately determine whether or not it is necessary to correct the winding shape of a steel strip coil The purpose is to provide a method.

In order to solve the above problems, the present invention has the following features.

[1] Every time it is determined whether or not it is necessary to correct the winding shape of the steel strip coil during manufacture in order to satisfy a predetermined quality standard after the steel strip coil is a final product, A data set including image data of the winding shape, quality standard data regarding the content of the quality standard, and determination result data regarding the result of the determination is accumulated to form a determination result database, and the determination result database is used. A method for generating a winding shape determination model for a steel strip coil, wherein machine learning is performed to generate a winding shape determination model that outputs the determination result data when the image data and the quality standard data are input.

[2] The quality standard data includes information on the product standard of the steel strip coil, information on the shape or weight of the steel strip coil, information on the shipping destination of the steel strip coil, and information on the steel strip coil after the winding shape has been determined. The method for generating a winding shape determination model for a steel strip coil according to [1], which includes at least one piece of information relating to the manufacturing process that the strip coil passes through until it becomes a final product.

[3] The method for generating a winding shape determination model for a steel strip coil according to [1], wherein at least one of neural network, decision tree learning, random forest, and support vector regression is used in the machine learning.

[4] The method for generating a winding shape determination model for a steel strip coil according to [2], wherein at least one of neural network, decision tree learning, random forest, and support vector regression is used in the machine learning.

[5] Acquiring image data of the winding shape of the steel strip coil being manufactured and quality standard data relating to quality standards to be met after the steel strip coil becomes the final product;
The acquired image data and the quality standard data are input into the winding shape determination model generated by the method for generating a winding shape determination model for a steel strip coil according to any one of [1] to [4], and the Output determination result data, and based on the output determination result data, the winding shape of the steel strip coil during manufacture in order to satisfy the quality standard after the steel strip coil is a final product A method for determining the winding shape of a steel strip coil, which determines whether or not correction is necessary.

[6] If it is determined that the winding shape of the steel strip coil in the process of manufacturing needs to be modified in order to ensure that the final product of the steel strip coil satisfies a predetermined quality standard, the winding shape is modified. When the treatment process is set, a data set including the image data of the winding shape, the quality standard data regarding the quality standard, and the treatment data regarding the content of the treatment process is accumulated to form a treatment result database. , the degree of similarity between the image data and the quality standard data of the steel strip coil in the process of production for which it is determined that the winding shape needs to be corrected by the steel strip coil winding shape determination method according to [5]. Treatment of the steel strip coil, wherein the data set containing the high image data and the quality standard data is retrieved from the treatment performance database, and the treatment process is set based on the treatment data included in the retrieved data set. Process setting method.

[7] Each time it is determined whether or not it is necessary to correct the winding shape of the steel strip coil during manufacture in order to satisfy a predetermined quality standard after the steel strip coil is a final product, accumulating a data set including the image data of the winding shape, the quality standard data regarding the content of the quality standard, and the determination result data regarding the result of the determination to form a determination performance database; Each time a treatment process for correcting the winding shape is set, a data set including image data of the winding shape, quality standard data regarding the quality standard, and treatment data regarding the content of the treatment process is accumulated. to form a treatment result database, and perform machine learning using the judgment result database and the treatment result database, so that when the image data and the quality standard data are input, the determination result data and the winding shape correction A method for generating a winding shape determination model for a steel strip coil, further generating a winding shape determination model for outputting the treatment data when it is determined that the treatment data is necessary.

[8] The quality standard data includes information on the product standard of the steel strip coil, information on the shape or weight of the steel strip coil, information on the shipping destination of the steel strip coil, and information on the steel strip after the winding shape has been determined. The method for generating a winding shape determination model for a steel strip coil according to [7], which includes at least one piece of information relating to the manufacturing process that the strip coil passes through until it becomes a final product.

[9] The method for generating a winding shape determination model for a steel strip coil according to [7], wherein at least one of neural network, decision tree learning, random forest, and support vector regression is used in the machine learning.

[10] The method for generating a winding shape determination model for a steel strip coil according to [8], wherein at least one of neural network, decision tree learning, random forest, and support vector regression is used in the machine learning.

[11] Acquiring image data of the winding shape of the steel strip coil being manufactured and quality standard data relating to quality standards to be met after the steel strip coil becomes a final product;
The acquired image data and the quality reference data are input to the winding shape determination model generated by the method for generating a winding shape determination model for a steel strip coil according to any one of [7] to [10]. outputting the determination result data;
Based on the output determination result data, whether it is necessary to correct the winding shape of the steel strip coil during production in order to satisfy the quality standards after the steel strip coil becomes the final product. A method for determining the winding shape of a steel strip coil.

[12] Applying the treatment step based on the treatment data to the steel strip coil whose winding shape is determined not to satisfy the quality standard by the steel strip coil winding shape determination method according to [11]. A method for setting a treatment process for a steel strip coil.

[13] A steel strip coil manufacturing method for manufacturing the steel strip coil whose winding shape is determined to satisfy the quality standard by the steel strip coil winding shape determining method according to [5].

[14] A steel strip coil manufacturing method for manufacturing the steel strip coil whose winding shape is determined to satisfy the quality standard by the steel strip coil winding shape determining method according to [11].

[15] A steel that is manufactured by correcting the winding shape of the steel strip coil in the process of being manufactured by the steel strip coil treatment process setting method described in [6] so as to satisfy the quality standards after becoming a final product. A steel strip coil manufacturing method for manufacturing a strip coil.

[16] A steel that is made to meet the quality standards after the final product is produced by correcting the winding shape of the steel strip coil in the process of manufacturing by the steel strip coil treatment process setting method according to [12]. A steel strip coil manufacturing method for manufacturing a strip coil.

According to the method for generating a winding shape determination model for a steel strip coil, the method for determining a winding shape, the method for setting a treatment process, and the method for manufacturing a steel strip coil according to the present invention, after the steel strip coil is a final product, a predetermined quality standard is obtained. It is possible to automatically and accurately determine whether or not it is necessary to correct the winding shape of the steel strip coil in each manufacturing process of the steel strip in order to satisfy the above.

FIG. 1 is a diagram showing an example of a winding process including winding of a steel strip coil by a winding machine. FIG. 2(a) and FIG. 2(b) are photographs showing earache, which is an example of a winding shape defect of a steel strip coil. FIGS. 3(a) to 3(d) are diagrams showing a telescope, which is another example of the winding shape defect of the steel strip coil. It is a figure which shows the flow of the winding shape determination method of the steel strip coil of 1st embodiment of this invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram showing the overall configuration of a manufacturing control system that executes a steel strip coil winding shape determining method and a steel strip coil manufacturing method according to a first embodiment of the present invention; FIG. 6(a) is a diagram showing an example of the measurement position of the side projection amount of the steel strip coil, and FIG. 6(b) is a chart representing the measured side projection amount of the steel strip coil. It is a figure which shows an example of image information. FIG. 4 is a diagram showing the flow of a method for generating a winding shape determination model for a steel strip coil according to the first embodiment of the present invention; FIG. 5 is a diagram showing the flow of a steel strip coil treatment process setting method according to a second embodiment of the present invention. FIG. 10 is a diagram showing the flow of a winding shape determination method for a steel strip coil according to a third embodiment of the present invention; FIG. 10 is a diagram showing the flow of a method for generating a winding shape determination model for a steel strip coil according to a third embodiment of the present invention; FIG. 4 is a diagram showing an example of a neural network that is machine learning in the method for determining the winding shape of a steel strip coil according to the present invention; It is a figure which shows the flow of the winding shape determination method of the conventional steel strip coil.

Embodiments of a method for generating a winding shape determination model for a steel strip coil, a method for determining a winding shape and a treatment process setting method, and a method for manufacturing a steel strip coil according to the present invention will be specifically described below with reference to the drawings. .
<Steel strip winding process>
First, a steel strip coil to which the steel strip coil winding shape determination model generation method, the winding shape determination method, the treatment process setting method, and the steel strip coil manufacturing method of the present embodiment are applied is manufactured. The coiling process of the steel strip in the manufacturing process will be described.

In the steel sheet manufacturing process, a slab is first hot-rolled in a hot rolling line to form a thin steel sheet. is wound into a coil to make a steel strip coil (hot rolled coil). In a hot-rolling line, when coiling thin steel strip after hot rolling, the coil shape of the steel strip coil tends to deteriorate. be.

In some cases, steel strip coils (hot-rolled coils) manufactured in hot-rolling lines are shipped as products as they are. , a surface treatment line, a temper rolling line, and the like. In each of these production lines, a winder installed at the end of the line rewinds the thin steel sheet after processing on the line into a coil shape to make a steel strip coil, which is then sent to the next production line. Convey or convey to a shipping yard for product shipment. For example, a steel strip coil wound in a cold-rolling line for thin steel sheets is conveyed to a continuous annealing line for performing a continuous annealing process, which is the next process. After the continuous annealing process is performed, a skin pass rolling process is performed in a skin pass rolling line that is continuously provided to the continuous annealing line. . At the shipping yard for product shipment, after inspecting whether or not the quality of the thin steel sheets, including the winding shape of the steel strip coils, satisfies the prescribed standards, the steel strip coils are packed, etc., and processed into steel strip products. Ship.

Here, when a steel strip coil wound on a certain production line has a defective winding shape, if it is transported to the next process as it is, it will cause manufacturing troubles and quality problems of the steel strip products in the next process. sell. Therefore, it is important from the viewpoint of quality control to appropriately determine whether the winding shape of the steel strip coil is acceptable before it is transported to the next production line or shipping yard.

For example, if a steel strip coil wound on a hot rolling line is telescoped and transported to the pickling line as it is for the next pickling process, the steel strip may meander during the pickling process. Manufacturing problems are more likely to occur. Therefore, in such a case, instead of conveying the steel strip coil to the pickling line according to the initial manufacturing plan, it is sent to the recoil process or temper rolling process before that, and the steel strip coil is rewound and wound. After improving the shape, it is preferable to shift to the pickling process. In this way, when a steel strip product is manufactured through a plurality of manufacturing processes, the winding shape of the steel strip coil in the intermediate process can also be appropriately determined for each manufacturing process. It is possible to suppress the occurrence of manufacturing troubles on the line.

The steel strip coil winding shape determination model generation method, winding shape determination method, treatment process setting method, and steel strip coil manufacturing method of the present embodiment manufacture a steel strip product through such a plurality of manufacturing processes. In this case, it is applied to the steel strip coil wound by the winding machine in each manufacturing process. Of these, the steel strip coil (hot-rolled coil) that is wound on the hot rolling line in particular tends to deteriorate the winding shape as described above. The determination method, the treatment process setting method, and the steel strip coil manufacturing method are preferably applied to steel strip coils (hot-rolled coils) manufactured in a hot-rolling line.

FIG. 1 shows an example of a winding process including winding of a steel strip coil by a winding machine. FIG. 1 shows an example of the winding process performed on the delivery side of the hot rolling line. As shown in FIG. 1, the steel strip S that leaves the finish rolling mill of the hot rolling line and is cooled on the run-out table is wound by a winder 71 to form a steel strip coil C. As shown in FIG. The wound steel strip coil C is transferred to a carrier (not shown) and moved, and the outer peripheral portion of the steel strip coil C is bound by a binder 72 . Further, on the outer surface of the steel strip coil C, information such as the manufacturing control number and product number of the steel strip coil is printed by the coil marker 73 . Then, the weighing machine 74 measures the weight of the steel strip coil C, and then the transfer machine 76 moves the steel strip coil C to a coil yard (not shown).

In the manufacturing process of the thin steel sheet, the thin steel sheet wound as the steel strip coil has a thickness of about 0.1 to 25 mm. Of these, the method for generating a winding shape determination model for a steel strip coil, the method for determining a winding shape, the method for setting a treatment process, and the method for manufacturing a steel strip coil according to the present embodiment are the main targets of the steel strip forming the steel strip coil. has a thickness of 0.1 to 8 mm. In addition, the sheet width of the steel strip coil, which is mainly targeted by the method for generating the winding shape determination model of the steel strip coil, the method for determining the winding shape, the method for setting the treatment process, and the method for manufacturing the steel strip coil of the present embodiment, is 600 to 600 mm. The length is about 2400 mm, and the weight of the steel strip coil is about 2-40 tons.
<Poor quality of winding shape>
FIG. 2(a) and FIG. 2(b) show photographs of ear pain, which is an example of a winding shape defect of a steel strip coil. Ear pain is caused by cracks, rubbing, bending, etc., occurring at the ends of the steel strip wound in the steel strip coil C in the width direction, and when the steel strip coil is observed from the side, the wound is caused. It is a winding shape defect in which disorder is observed in the layer shape of the steel strip.

Further, a telescope, which is another example of a winding shape defect of a steel strip coil, will be described with reference to FIGS. 3(a) to 3(d). 3(a) is a side view of the steel strip coil C, and FIGS. 3(b) to 3(d) are cross-sectional views of the steel strip coil C shown in FIG. 3(a) taken along line III-III.

When the winding shape of the steel strip coil is good, the cross-sectional shape of the steel strip coil C is substantially rectangular as shown in FIG. 3(b). On the other hand, when the steel strip coil C has a telescope, the cross-sectional shape of the steel strip coil C is uneven in the width direction of the steel strip, as shown in FIG. have As shown in FIG. 3(c), the inner peripheral telescope has a configuration in which unevenness is generated in the cross section of the steel strip coil C on the inner peripheral side. As shown in FIG. 3(d), the outer telescope has a configuration in which unevenness is generated in the cross section of the steel strip coil C on the outer peripheral side. In addition, there is also a form such as a bowl-shaped telescope in which unevenness is generated in the cross section of the inner peripheral portion and the outer peripheral portion of the steel strip coil C. The size of these projections and recesses of the telescope can be easily visually determined by observing the side surface of the steel strip coil C from an oblique direction.
<Conventional winding shape determination method for steel strip coil>
Here, with reference to FIG. 12, the procedure of the conventional method for determining the winding shape of a steel strip coil will be described. FIG. 12 shows an example of determining the winding shape of a steel strip coil wound by a winder provided at the terminal end of the production line where process A is performed.

First, the steel strip coil wound in the process A undergoes primary determination of the winding shape in this production line. The primary determination of the winding shape of the steel strip coil determines the quality of the winding shape, such as "pass" indicating that the winding shape is good, or "fail" indicating that the winding shape is not good. It is. However, in some cases, it may not always be possible to clearly determine whether the winding shape is "passed" or "failed" by visual evaluation by the judge. Therefore, in place of the two-stage pass/fail judgment, a three-stage evaluation such as "passed", "suspended", "failed" or "good", "acceptable", "failed", or an evaluation in four stages or more sometimes do.

For example, in the case of a steel strip coil wound on a hot rolling line, the primary determination of the winding shape is performed by measuring the weight of the steel strip coil C by the weighing machine 74 shown in FIG. This is often done by visually observing the side of C. At this time, as disclosed in Patent Document 2, the side surface of the steel strip coil C is imaged while being irradiated with light, and the light irradiation condition and the imaging condition are set to a predetermined relationship, whereby the steel strip coil In some cases, the winding shape of C is acquired as a clear image. In this case, an image is projected on a monitor screen provided in an examination room or the like, and the image is visually observed by the judge to make the primary judgment.

In addition, in a state in which the steel strip coil C after winding is loaded on the weighing machine 74, the distance from the reference wall surface to one end face of the steel strip coil is measured by a winding shape meter such as a non-contact rangefinder. 75 may be used to acquire the uneven shape of the side surface of the steel strip coil C as numerical information. In this case, when the numerical value of the uneven shape obtained by the winding shape meter 75 is larger than a preset threshold value, the primary determination result of the winding shape may be "failed". However, it is difficult to detect earache from the numerical information of the uneven shape of the side surface of the steel strip coil, which is obtained by automatic measurement by such a winding shape meter 75 . Therefore, even when the numerical value of the uneven shape of the side surface of the steel strip coil C is obtained by the winding shape meter 75, the primary determination by visual observation by the determiner is not omitted.

A steel strip coil determined to be "acceptable" in the primary determination of the winding shape of the steel strip coil is transported to the next process or shipping process according to the originally scheduled manufacturing plan. On the other hand, a secondary judgment is performed for the steel strip coil judged to be "failed" in the primary judgment. A secondary judgment is often made by a different judge than the primary judgment.

In the secondary judgment, the results of the primary judgment of the winding shape of the steel strip coil, the manufacturing plan for the next and subsequent processes, and the product standards that should be met when the steel strip coil becomes the final product are taken into consideration, and the initially planned production Comprehensively judge whether or not it is acceptable to continue the production of steel strip coils according to the plan. If it is judged that the steel strip coil can meet the prescribed quality standards when the steel strip coil becomes the final product even if the production of the steel strip coil is continued according to the originally planned manufacturing process, If not, it is determined as "failed". For the steel strip coils judged to be "failed" in the secondary judgment, the judgment person who made the secondary judgment asked for a treatment process (addition of a process, In some cases, it may be decided to partially change the process during the manufacturing plan that was originally planned). By performing such a treatment process, the winding shape of the steel strip coil is corrected, and then the steel strip coil is conveyed to the next process or the shipping process. Additional treatment processes include, for example, a recoil process for rewinding the steel strip coil, a trimming process for cutting off the width end portion of the steel strip, and a cutting process for cutting and removing a part of the steel strip coil in the longitudinal direction. etc.
(First embodiment)
A method for determining the winding shape of a steel strip coil, a method for generating a winding shape determination model, and a method for manufacturing a steel strip coil according to the first embodiment will be described below.

FIG. 4 shows the processing flow of the steel strip coil winding shape determination method of the present embodiment. FIG. 5 shows manufacturing control including a winding shape determination unit 1 that executes the steel strip coil winding shape determining method and the steel strip coil manufacturing method of the present embodiment, and a host computer 6 that controls the steel strip coil manufacturing plan. The overall configuration of the system is shown schematically.
<How to determine the winding shape of the steel strip coil>
The winding shape determination model generated by the method of generating a winding shape determination model for a steel strip coil according to the present embodiment is used to satisfy predetermined quality standards after the steel strip coil is a final product. It is used to automatically determine whether or not it is necessary to modify the winding shape of the steel strip coil.

As shown in FIGS. 4 and 5, the method for determining the winding shape of a steel strip coil according to the present embodiment is executed by a winding shape determining section 1 comprising a general-purpose computer or the like. As shown in FIG. 5, the winding shape determination unit 1 is configured by, for example, a general-purpose computer or a tablet terminal that can communicate with the host computer 6 . Alternatively, instead of this, the function of the winding shape determining section 1 may be provided inside the host computer 6 .

The processing of the steel strip coil winding shape determination method executed by the winding shape determination unit 1 will be described below with reference to FIGS. 4 and 5 .

First, as shown in FIG. 5, the steel strip is wound by a winder provided at the end of the process A where the winding shape of the steel strip coil is determined to form a steel strip coil. Then, the side surface of the steel strip coil is photographed by a camera (not shown), image data of the winding shape of the steel strip coil (details will be described later) is acquired, and transmitted to the host computer 6 .

The high-level computer 6 stores in advance quality standard data (specifically described later) regarding quality standards to be met after various steel strip coils become final products. When the host computer 6 receives the image data of the winding shape of the steel strip coil, the host computer 6 stores the quality standard data related to the quality standards to be met after the steel strip coil becomes the final product, such as the coil number displayed on the steel strip coil. Identify using identification information. The identification information of the steel strip coil can be acquired, for example, from the image data of the steel strip coil. Then, the host computer 6 transmits the image data of the steel strip coil and the identified quality standard data to the winding shape determining section 1 . Thus, when the winding shape determination unit 1 receives the image data of the winding shape of the steel strip coil and the quality standard data of the steel strip coil, the processing of the winding shape determination method of the steel strip coil is started.

The winding shape determination unit 1 stores in advance a winding shape determination model that, when inputting image data and quality reference data, outputs determination result data (specifically described later) corresponding to the input values. When the winding shape determination unit 1 receives the image data of the winding shape of the steel strip coil and the quality standard data of the steel strip coil from the host computer 6, the winding shape determination unit 1 inputs the image data and the quality standard data into the winding shape determination model. In response to this input value, the winding shape determination model outputs determination result data relating to the result of determination as to whether or not the winding shape of the steel strip coil satisfies a predetermined quality standard after the final product.

The determination result data output from the winding shape determination model is output to an output device such as a monitor provided in the winding shape determination unit 1 together with the image data of the winding shape of the steel strip coil and the quality standard data of the steel strip coil. . Based on the determination result data output from the winding shape determination unit 1, the determiner determines whether or not it is necessary to correct the winding shape of the steel strip coil. The determiner may determine whether or not it is necessary to correct the winding shape of the steel strip coil, directly following the content of the determination result contained in the output determination result data. Alternatively, the judge needs to comprehensively consider the image data of the steel strip coil winding shape and the quality standard data of the steel strip coil output together with the judgment result data, and correct the winding shape of the steel strip coil. A secondary determination may be made as to whether there is.

Furthermore, when the determiner determines that it is necessary to correct the winding shape of the steel strip coil based on the determination result data output from the winding shape determination unit 1, the steel strip output together with the determination result data By referring to the image data of the winding shape of the coil and the quality standard data of the steel strip coil, it is possible to set a treatment process for correcting the winding shape of the steel strip coil. Specifically, the judge inputs treatment data to the host computer 6, thereby specifying the details of the treatment process to be set for the steel strip coil determined to require modification of the winding shape. Can be specified.
・Image data of the winding shape of the steel strip coil The image data of the winding shape of the steel strip coil, which is one of the data input to the winding shape determination model of the winding shape determination unit 1, is the winding shape of the steel strip coil. Image information about the appearance of a steel strip coil that can be used for judgment. Specifically, the image data of the winding shape of the steel strip coil is, for example, a photograph obtained by capturing a still image of the appearance of the steel strip coil with a camera, or a still image cut from a moving image obtained by capturing a video. , image data obtained by directly imaging the appearance of the steel strip coil can be used.

When acquiring image data by photographing the appearance of the steel strip coil, the field of view and the resolution of the image data necessary for determining the winding shape of the steel strip coil are ensured. The image data of the winding shape of the steel strip coil does not necessarily have to be acquired in one shot. may be a combination of

Further, the image data of the winding shape of the steel strip coil does not necessarily have to be captured such that the entire side surface of the steel strip coil is included in one image data. For example, focusing on the disorder of the layer shape of the steel strip in the steel strip coil, image data obtained by photographing part of the side surface of the steel strip coil at a high magnification may be used. Using such image data makes it easier to identify disturbances in the layer shape of the steel strip. Pain determination can be performed easily and with high accuracy.

In addition, as the image data of the winding shape of the steel strip coil, it is preferable to use the image data of the side surface of the steel strip coil taken not from the front but from an oblique direction. By using such image data, it is possible to easily and highly accurately judge telescopes as shown in FIGS. can. Furthermore, by applying an imaging method as disclosed in Patent Document 2 and photographing the appearance of the steel strip coil while irradiating the side surface of the steel strip coil with light, it is possible to obtain clear image data. More preferred.

Further, it is more preferable to use a combination of image data obtained by photographing the side surface of the steel strip coil from the front and image data obtained by photographing the side surface of the steel strip coil from an oblique direction. In this way, the image data obtained by photographing the side surface of the steel strip coil from the front and the image data obtained by photographing the side surface of the steel strip coil from an oblique direction can be used to detect defects in the winding shape of the steel strip coil. Of these, both earache and telescopic determinations can be performed easily and with high accuracy.

In addition, the outer shape of the winding shape of the steel strip coil is measured by a measuring device or the like, and the measurement results expressed in the form of image data such as diagrams and graphs also include feature amounts related to the winding shape of the steel strip coil. It can be used as image data of the winding shape of the steel strip coil.

For example, as indicated by straight lines in FIG. 6(a), the side surface of the steel strip coil is scanned in the radial direction. From this measurement result, the amount of protrusion of the side surface of the steel strip coil (displacement in the plate width direction of the steel strip), that is, the size of the unevenness of the telescope of the steel strip coil, is shown in FIG. , in the form of image data such as graphs. The image data created in this way can also be used as the image data of the winding shape of the steel strip coil.

In addition, as image data of the winding shape of the steel strip coil, image data obtained by photographing the appearance of the steel strip coil, and measurement results of the outer shape of the winding shape of the steel strip coil are included in the image data such as diagrams and graphs. It is more preferable to use it in combination with what is expressed in the form. In this way, the amount of information about the winding shape of the steel strip coil that is input to the winding shape determination model is increased, and the determination of the winding shape of the steel strip coil can be performed with even higher accuracy.
Steel strip coil quality standard data The steel strip coil quality standard data, which is data input to the winding shape determination model of the winding shape determination unit 1 together with the image data of the winding shape of the steel strip coil, is the data of the steel strip coil. Information on product standards, information on the shape or weight of the steel strip coil, information on the shipping destination of the steel strip coil, and the manufacturing process through which the steel strip coil after the winding shape is determined becomes the final product. shall contain at least one of the information relating to

Information on product standards for steel strip coils specifically refers to the product standards (Japanese Industrial Standards JIS, American Society for Testing and Materials Standards ASTM, German Industrial Standard DIN, etc.). Since the rigor of quality required for the final product differs depending on the product standard of the steel strip coil, the winding shape of the steel strip coil in the process of production can be determined by using the information on the product standard of the steel strip coil as the quality standard data of the steel strip coil. It is possible to appropriately determine whether or not correction of is necessary.

The information about the shape or weight of the steel strip coil specifically means the thickness, width, and length of the steel strip forming the steel strip coil, the weight of the steel strip coil, and the like. As the thickness of the steel strip coil becomes smaller, cracks and bends are more likely to occur at the ends of the steel strip in the width direction, and defects in the winding shape of the steel strip coil are more likely to occur. In addition, the narrower the width of the steel strip coil, the more easily meandering occurs in the production line through which the steel strip coil passes until it becomes the final product, even if the amount of unevenness on the side surface of the steel strip coil is the same as measured by the telescope. As a result, defects in the winding shape tend to occur after the final product is produced. Furthermore, the longer the length of the steel strip coil, the more likely it is that telescoping will occur, and the greater the weight of the steel strip coil, the more likely ear pain will occur due to the effect of its own weight when the steel strip coil is transported. Therefore, by using the information about the shape or weight of the steel strip coil as the quality standard data of the steel strip coil, it can be appropriately determined whether or not the winding shape of the steel strip coil in the process of production needs to be corrected.

The information on the delivery destination of the steel strip coil specifically refers to the delivery destination of the steel strip coil, the delivery consignee of the steel strip coil, and the like. The quality requirements for the steel strip coil as the final product differ depending on the delivery destination of the steel strip coil, and the processing equipment for the steel strip coil differs depending on the coil center to which the steel strip coil is shipped. different levels. Therefore, by using the information about the delivery destination of the steel strip coil as the quality standard data of the steel strip coil, it can be appropriately determined whether or not the winding shape of the steel strip coil in the process of production needs to be corrected.

The information on the manufacturing process that the steel strip coil after the determination of the winding shape passes through until it becomes the final product is, specifically, the steel strip coil after the determination of the winding shape is the final product It refers to the number and type of production lines that pass through until the product reaches a certain level, the order in which the production lines pass, or the identification symbols related to these. During the period from the manufacturing process in which the winding shape is determined until the steel strip coil becomes the final product, the winding shape defect of the steel strip coil may be corrected to eliminate the winding shape defect. In addition, there is a case where the defective winding shape that occurs in the manufacturing process in which the determination is made becomes the cause of the defective winding shape in the subsequent passing process. Therefore, by using the information on the manufacturing process through which the steel strip coil passes until the final product after the determination of the winding shape is used as the quality standard data of the steel strip coil, the winding of the steel strip coil in the process of production It is possible to appropriately determine whether or not shape correction is necessary.
・Determination result data In the present embodiment, the determination result data output from the winding shape determination model is a steel strip in the process of manufacturing in order to satisfy a predetermined quality standard after the steel strip coil becomes the final product. Refers to information on the content of determination as to whether or not it is necessary to correct the winding shape of the coil. The judgment result data is, for example, "accepted" indicating that the winding shape of the steel strip coil under production does not need to be corrected, or "failed" indicating that the winding shape needs to be corrected. , are output in two sections. In this case, the determiner determines whether or not it is necessary to correct the winding shape of the steel strip coil in the process of production based on the determination result data output in two categories from the winding shape determination unit 1. become. Therefore, the judge does not necessarily need to make the secondary judgment, and the two-stage judgment consisting of the primary judgment and the secondary judgment like the above-described conventional steel strip coil winding shape judgment method can be simplified to one step. can be

Alternatively, the determination result data output from the winding shape determination model may be subdivided into three stages such as "accepted", "suspended", and "failed", or output in four or more stages. In this case, each category of 3 or 4 or more stages is classified in advance into those that are not subject to the treatment process of correcting the winding shape and those that are subject to it.

If judgment result data is output for a category that is not subject to application of the treatment process for correcting the winding shape among the categories of 3 or 4 or more stages, the assessor shall, based on this judgment result data , the steel strip coil is instructed to be conveyed to the next process or the shipping process according to the originally scheduled manufacturing plan. Specifically, the judge performs an operation of inputting the next process instruction information to the host computer 6, thereby instructing the steel strip coil to be conveyed to the next process or the shipping process in accordance with the originally scheduled manufacturing plan. .

In addition, when the judgment result data of the division to which the treatment process for correcting the winding shape is applied among the three stages or four stages or more is output, the judge shall use this judgment result data. Then, for the steel strip coil, a treatment process that is not included in the originally scheduled manufacturing plan (addition of a process, partial change of the originally scheduled process during the manufacturing plan, etc.) is set. Specifically, the judge inputs treatment data to the host computer 6, thereby specifying the details of the treatment process to be set for the steel strip coil determined to require modification of the winding shape. Can be specified.

In the winding shape determination method of the present embodiment, image data of the winding shape of the steel strip coil and quality standard data of the steel strip coil are used in combination as input data to the winding shape determination model. By doing so, even if the winding shape of the steel strip coil is the same, the steel strip coil after the product standard, shape or weight, shipping destination, and winding shape of the steel strip coil is determined is the final product. It is possible to automatically and accurately determine whether or not it is necessary to correct the winding shape of the steel strip coil in the process of being manufactured, depending on the manufacturing processes, etc., that are passed through until the end of the manufacturing process.

For example, even if the ear pain observed in the steel strip coil during manufacturing is minor, the shape or weight of the steel strip coil may affect the manufacturing processes that the steel strip coil passes through before it becomes the final product, and the final product. Cracks may develop at the ends of the steel strip in the width direction during shipping after it has become. In this case, manufacturing troubles occur in each manufacturing process that goes through until the final product is made, and defects occur before the steel strip coil is actually used after it is made into the final product. You may not be able to meet the standards. Or, conversely, if the ear pain observed in the steel strip coil during manufacture is minor, depending on the type and sequence of further manufacturing steps that the steel strip coil undergoes before it becomes the final product, In some cases, the winding shape of the steel strip coil is corrected and improved. That is, even if the earache observed in the steel strip coil is of the same degree, the winding shape of the steel strip coil during production is necessary to ensure that the steel strip coil satisfies the predetermined quality standards after the steel strip coil becomes the final product. The determination of whether or not it is necessary to modify the steel strip coil is passed until the steel strip coil becomes the final product after the product standard, shape or weight, shipping destination, and winding shape have been determined. It varies depending on the manufacturing process.

In the winding shape determination method of the steel strip coil of the present embodiment, even if the ear pain observed in the steel strip coil during manufacturing is the same, the steel strip coil satisfies the predetermined quality standard after becoming the final product. Therefore, it is possible to automatically and accurately determine whether or not it is necessary to correct the winding shape of the steel strip coil according to the various conditions described above.

A case will be described below in which the determination result data output from the winding shape determination model is output in two categories of "accepted" and "failed".
<Manufacturing method of steel strip coil>
When the winding shape determination model of the winding shape determination unit 1 outputs the determination result data of "accepted" by the winding shape determination method of the steel strip coil described above, the determination result data is input to the host computer 6 by the determiner as described above. After receiving the next process instruction information, the steel strip coil is transported to the next process or the shipping process according to the originally scheduled manufacturing plan. In this way, a steel strip coil that satisfies a predetermined quality standard is manufactured after the steel strip coil has become a final product.

Further, in the winding shape determination method of the steel strip coil described above, when the determination result data of "fail" is output from the winding shape determination model of the winding shape determination unit 1, as described above, the determination person , a treatment process not included in the originally scheduled manufacturing plan is set for the steel strip coil. For the content of the treatment process, the judge refers to the image data of the winding shape of the steel strip coil and the quality standard data of the steel strip coil output together with the judgment result data, and judges the content of the treatment process to be set. Specified by In this case also, the judge corrects the winding shape for the steel strip coil for which the judgment result data of "fail" is output based on the judgment result data output in two categories from the winding shape judgment unit 1. It suffices to set the treatment data for performing Therefore, the judge does not necessarily need to make the secondary judgment, and the two-stage judgment consisting of the primary judgment and the secondary judgment like the above-described conventional steel strip coil winding shape judgment method can be simplified to one step. It is possible to set an appropriate treatment process for the steel strip coil by one-step judgment by the judge. In this way, the winding shape of the steel strip coil being manufactured is corrected, and after the modified steel strip coil is the final product, a steel strip coil that satisfies a predetermined quality standard is manufactured.

As described above, in the steel strip coil manufacturing method of the present embodiment, the image data of the winding shape of the steel strip coil and the quality standard data of the steel strip coil are combined and input to the winding shape determination model to obtain the steel strip coil. By determining the winding shape of the steel strip coil, the manufacturing plan of the steel strip coil can be changed appropriately to modify the winding shape of the steel strip coil.
<Method for Generating Winding Shape Determination Model of Steel Strip Coil>
The method of generating a winding shape determination model for a steel strip coil according to the present embodiment corrects the winding shape of a steel strip coil in the process of being manufactured in order to satisfy a predetermined quality standard after the steel strip coil has become a final product. It is to generate a winding shape determination model for automatically determining whether or not it is necessary to perform

FIG. 7 shows the flow of the steel strip coil winding shape determination model generation method of the present embodiment, and the winding shape determination model generation unit 3 that executes the steel strip coil winding shape determination model generation method of the present embodiment. The configuration is shown schematically. The winding shape determination model generation unit 3 is composed of a general-purpose computer or the like, and includes a database unit 31 and a machine learning unit 32 as shown in FIG.
Judgment result database In the method of generating a winding shape judgment model for a steel strip coil according to the present embodiment, first, in order to satisfy a predetermined quality standard after the steel strip coil is a final product, a steel strip in the process of production When it is determined whether or not it is necessary to correct the winding shape of the coil, the image data of the winding shape of the steel strip coil, the quality standard data of the steel strip coil, and the judgment result data regarding the result of the above judgment are used. Get the containing dataset. Then, a data set obtained each time this determination is made is accumulated to form a determination result database in the database unit 31 .

Specifically, the determination record database is formed as follows. First, those that were judged to be “passed” by the primary judgment by the judge’s visual inspection, etc., and those that were judged to be “failed” by the primary judgment by the judge’s visual observation, etc., were judged to be “failed” by the primary judgment by the judge’s visual observation, etc. The judgment result data of "final pass" is defined as "accepted" by the judging person through the secondary judgment. In addition, if it is judged as "failed" by the primary judgment by the judge's visual observation, etc., and the second judgment is made by the judge of the second judgment based on this judgment result, and it is judged to be "failed" "Final rejection" judgment result data.

Then, as shown in FIG. 7, the winding shape determination model generating unit 3 generates the image data of the winding shape of the steel strip coil, the quality standard data of the steel strip coil, and the "final pass" each time the above-described determination is performed. Alternatively, obtain a data set that includes the judgment result data of "final failure". Then, a data set acquired each time the above-described determination is performed is accumulated to form a determination result database in the database unit 31 .

The image data of the steel strip coil included in each data set of the determination result database is acquired as described above in the manufacturing process in which the winding shape of the steel strip coil is determined, and is stored in the host computer 6 (see FIG. 5). is used. As described above, the steel strip coil quality standard data contained in each data set of the determination record database is stored in advance in the host computer 6 and specified by the host computer 6 is used.

The determination result database may be stored in the host computer 6 instead of the database section 31 of the winding shape determination model generation section 3 . In this case, the winding shape determination model generation unit 3 accesses the determination result database stored in the host computer 6 by transmitting and receiving data to and from the host computer 6 when necessary.

In the determination result database in the database unit 31, the product standard, shape or weight of the steel strip coil, shipping destination, and the manufacturing process that the steel strip coil that has been determined for the winding shape passes through until it becomes the final product is common. It is preferable to accumulate 50 or more data sets for each type of steel strip coil used. In this way, machine learning, which will be described later, can be performed appropriately. The number of data sets accumulated in the determination result database in the database unit 31 is more preferably 100 sets or more, and still more preferably 500 sets or more. In addition, it is preferable to accumulate 2000 or more sets of data sets for each manufacturing process in which the winding shape of the steel strip coil is determined.

In addition, among the data sets accumulated in the determination result database in the database unit 31, the image data of the winding shape of the steel strip coil is screened as necessary to remove unclear image data. good too. In this way, data with low reliability is excluded from the data set accumulated in the determination result database, and machine learning, which will be described later, can be performed appropriately.

Further, a certain upper limit may be set for the number of data sets accumulated in the judgment result database in the database unit 31, and the data sets accumulated in the judgment result database may be appropriately updated within this range.

Furthermore, as shown in FIG. 7, the machine learning unit 32 of the winding shape determination model generation unit 3 performs machine learning using the determination result database accumulated in the database unit 31 to obtain the image data and the quality standard data. When input, a winding shape determination model is generated that outputs determination result data corresponding to this input value.

The machine learning performed by the machine learning unit 32 is any machine learning that can output determination result data having practically sufficient determination accuracy from the winding shape determination model generated by this machine learning. A model may be used. For example, neural networks generally used in machine learning (including deep learning, convolutional neural networks, etc.), decision tree learning, random forest, support vector regression, etc. can be used. Also, an ensemble model combining a plurality of learning models may be used. Classification models such as the k-neighborhood method and logistic regression may also be used. However, since the image data of the winding shape of the steel strip coil is used as input, it is preferable to use the convolutional neural network method, which has an excellent image recognition function and includes a convolutional layer and a pooling layer in the network structure. This is because the ability to extract feature amounts from image data is excellent, and it becomes easy to identify ear pain, telescope, and other winding shapes.
(Second embodiment)
Hereinafter, a method for setting treatment steps for a steel strip coil according to a second embodiment of the present invention will be described.
<How to set the treatment process for the steel strip coil>
The steel strip coil treatment process setting method of the present embodiment corrects the winding shape of the steel strip coil for which it is determined that the winding shape needs to be corrected by the steel strip coil winding shape determination method of the first embodiment. It automatically sets the action steps to be applied to bring the product to meet quality standards.

FIG. 8 shows the flow of the steel strip coil treatment process setting method according to the present embodiment.

As shown in FIG. 8, in the steel strip coil treatment process setting method of the present embodiment, in the winding shape determination method of the first embodiment, the determination result data of "pass" from the winding shape determination model of the winding shape determination unit 1 is output, the steel strip coil is conveyed to the next process or shipping process according to the originally scheduled manufacturing plan.

On the other hand, when the winding shape determination model of the winding shape determining unit 1 outputs “failed” determination result data, this determination result data is sent to the treatment data setting unit 5 . Upon receiving the determination result data, the treatment data setting unit 5 acquires from the host computer 6 the image data and the quality standard data of the steel strip coil corresponding to the “failed” determination result data. Then, the treatment data setting unit 5 obtains treatment data relating to the content of the treatment process for correcting the winding shape of the steel strip coil from a treatment performance database (specifically described later) formed in the treatment data setting unit 5. search for.

Specifically, the treatment data setting unit 5 compares the image data and quality standard data of the steel strip coil to be determined with the image data and quality standard data included in the data set accumulated in the treatment record database. In contrast, search the data set containing the image data and quality reference data with the highest degree of similarity.

The degree of similarity is a numerical representation of how similar image data and quality reference data to be compared and contrasted are, specifically, distance information, a correlation coefficient, and the like. Various techniques can be used to evaluate similarity. For example, a method using the sum of squared differences of the luminance value of each corresponding pixel of the image data to be compared or the data value of the corresponding item of the quality standard data to be compared, a method using the sum of absolute differences, normalization A method using cross-correlation or the like can be used. Further, as a similarity evaluation method, a method of performing feature point matching by extracting feature points and feature amounts of image data or quality reference data to be compared and contrasted may be used. Alternatively, the degree of similarity may be evaluated using a commonly available similar image search tool that has been trained by deep learning.

The similarity in the present embodiment is also the distance information and the correlation coefficient between the image data or quality standard data of the steel strip coil corresponding to the "fail" determination result data and the image data or quality standard data in the treatment record database. Etc. can be used for evaluation. If the quality standard data is character information instead of numerical information, the similarity may be calculated by converting the character information into numerical information.

Then, based on the treatment data included in the data set retrieved from the treatment record database as described above, the winding shape is corrected for the steel strip coil corresponding to the "failed" judgment result data. set.
- Treatment results database The treatment results database is formed as follows. First, in the winding shape determination method of the steel strip coil described above, the winding shape determination model outputs the determination result data of "failed", and the judge makes a secondary determination based on this determination result data to make a "failed" determination. ” is regarded as “final rejection”. Then, for the steel strip coil judged to be "finally rejected", the judge provides the content of the treatment process for correcting the winding shape together with the judgment result data, the output image data of the winding shape of the steel strip coil, and Determine and set with reference to the quality standard data of the steel strip coil.

In this way, every time the judging person sets the treatment process for a steel strip coil whose winding shape is "finally rejected", the image data of the winding shape of the steel strip coil and the image data of the steel strip coil A data set including quality standard data and treatment data relating to the content of the treatment process set by the judge is acquired, and a treatment performance database is formed by accumulating this data set.

The steel strip coil quality standard data contained in the treatment record database includes information such as steel strip product standards and shipping destinations. When a shipping destination is set, it becomes necessary to update the contents of the data set accumulated in the treatment record database. In particular, some quality standard data for steel strip coils change in a relatively short period of time.

As described above, when it is necessary to update the contents of the data set, in a configuration in which treatment data is output together with the determination result data by the winding shape determination model, as in the third embodiment described later, each time the data set is updated, Therefore, it becomes necessary to redo the machine learning of the winding shape determination model. Therefore, in the present embodiment, when the winding shape determination model is generated, it is generated without including such a treatment result database, and the treatment process is set using the treatment result database generated separately from this. I'm trying

Further, the treatment information database is classified (also called clustering) into groups having similar attribute information such as the product standard, shape or weight of the steel strip coil, the shipping destination, and the manufacturing process to be passed through, and the treatment information database is created for each group. may be generated. These attribute information, treatment data, and the like may be character information, or may be stored in the treatment record database in the form of numerical values or codes assigned in advance.
・Treatment data The treatment data related to the contents of the treatment process set for the steel strip coil whose winding shape was "finally rejected" refers to the steel strip coil that is not included in the originally planned manufacturing plan. This is data for setting the content of the treatment process for correcting the winding shape. Specifically, the treatment data includes, for example, data specifying the content of the recoil process for rewinding the steel strip coil, data specifying the content of the trimming process for cutting off the width end of the steel strip, It is data etc. which designates the content of the cutting process which cut|disconnects and removes a part of longitudinal direction. In this way, the treatment data can be codes, numerical values, etc. for setting the details of the additional steps that are not included in the initially planned manufacturing steps.

In addition, the content of the treatment data is not necessarily limited to newly setting additional processes that are not included in the originally scheduled manufacturing plan. It may be changed or the like. For example, when judging the winding shape of a hot-rolled coil wound by a winder installed at the end of a hot-rolling line, the initially planned manufacturing plan states that the width of the steel strip in the next pickling line is Suppose that the production instruction is given not to perform the trimming process for cutting off the ends. In such a case, the treatment process may also include the addition of a production instruction to perform the strip trimming process using the trimming device in the pickling line. That is, in the treatment data, whether or not to execute a specific treatment in the initial production plan was set in advance. shall include
(Third embodiment)
A method for determining the winding shape of the steel strip coil and a method for generating the winding shape determination model of the steel strip coil according to the third embodiment of the present invention will be described below.
<How to determine the winding shape of the steel strip coil>
FIG. 9 shows the flow of the method for determining the winding shape of the steel strip coil according to the present embodiment.

As shown in FIG. 9, in the steel strip coil winding shape determination method of the present embodiment, similarly to the steel strip coil winding shape determination method of the first embodiment, the winding shape determination unit 2 determines the steel to be determined. The image data of the winding shape of the strip coil and the quality standard data of the steel strip coil are received from the host computer 6 and input to the winding shape determination model stored in advance in the winding shape determination unit 2 . In response to this input value, the winding shape determination model outputs determination result data relating to the result of determination as to whether or not the winding shape of the steel strip coil satisfies a predetermined quality standard after the final product.

In the present embodiment, the winding shape determination model further generates treatment data relating to the content of the treatment process for correcting the winding shape for the steel strip coil for which the determination result data indicating that the winding shape does not meet the quality standards described above is output. and output. As described above, in this embodiment, in addition to determination result data, as output values from the winding shape determination model, measures regarding the content of the treatment process for the steel strip coil whose winding shape has been determined not to satisfy the above-described quality standards The difference from the first embodiment is that data is included. Such a winding shape determination model is generated as follows.
<Method for Generating Winding Shape Determination Model of Steel Strip Coil>
The steel strip coil winding shape determination model generation method of the present embodiment combines the steel strip coil winding shape determination model generation method of the first embodiment with the steel strip coil treatment process setting method of the second embodiment. It is a thing. That is, when the image data and the quality standard data are input to the winding shape determination model, the winding shape determination model outputs the determination result data and further outputs the treatment data when it is determined that the winding shape needs to be corrected. Generate a model.

FIG. 10 shows the flow of the method for generating the winding shape determination model of the steel strip coil according to the present embodiment. As shown in FIG. 10, in the present embodiment, similarly to the first embodiment, the winding shape determination model generation unit 4 generates a Each time the secondary judgment is performed, a data set is obtained that includes image data of the winding shape of the steel strip coil, quality standard data of the steel strip coil, and judgment result data of "final pass" or "final rejection". Then, a data set acquired each time the above-described determination is performed is accumulated to form a determination result database in the database unit 41 . In addition, the database unit 41 obtains the determination result data of “final pass” or “final rejection” output from the winding shape determination model generated by the winding shape determination model generation method of the first embodiment. A determination performance database may be formed within the .

Further, in the present embodiment, every time the treatment process setting person sets the treatment process for the steel strip coil that has been determined to be "finally rejected", the image data of the winding shape of the steel strip coil, the steel strip coil A data set is obtained that includes quality criteria data for the process and treatment data for the content of the treatment process. Then, a data set acquired each time a treatment process is set is accumulated to form a treatment record database in the database unit 41 .

Specifically, the treatment data included in the data set accumulated in the treatment record database is obtained as follows. For steel strip coils that have been determined to be "finally rejected" by the judge of the secondary judgment based on the result of the primary judgment by the judge's visual observation, etc., the treatment process setter sets the treatment process, and the treatment data is obtained. In addition, the treatment process setting person performs the treatment process for the steel strip coil for which the determination result data of "final rejection" is output from the winding shape determination model of the winding shape determination unit 1 described in the first embodiment. Treatment data is acquired by setting. Specifically, the person who sets the treatment process is set in the host computer 6 by transmitting instruction data for setting the treatment process to the host computer 6 shown in FIG. is stored as treatment data in the database unit 41 of the. This treatment data is used for the treatment record database stored in the database unit 41 . Here, the treatment process setter may be the same person as the judge who performs the secondary judgment, or may be a person different from the judge who makes the secondary judgment. However, since the setting of the treatment process affects the production plan of steel strip coils, A person who is capable of making judgments and decisions is preferred.

Then, as shown in FIG. 10 , the machine learning unit 42 of the winding shape determination model generation unit 4 performs machine learning using the determination result database and the treatment result database accumulated in the database unit 41 to obtain image data and When the quality standard data is input, the "acceptable" judgment result data, or the "fail" judgment result data and this "fail" judgment result data are output to the steel strip coil corresponding to this input value. A winding shape determination model for outputting treatment data for correcting the winding shape is generated.

As in the first embodiment, the determination result data included in the data set accumulated in the determination result database includes the steel strip coils that were "passed" in the primary determination and "failed" in the secondary determination. Data that do not satisfy the criteria are regarded as "accepted", and data judged as "unacceptable" by the judge are regarded as "final unacceptable".

The determination result database and treatment result database may be stored in the host computer 6 instead of the database section 41 of the winding shape determination model generation section 4 . In this case, the winding shape determination model generation unit 4 transmits and receives data to and from the host computer 6 when necessary so as to access the determination result database and treatment result database stored in the host computer 6. to

At this time, the output of treatment data from the winding shape determination model may be performed together with the "failed" determination result data output from the winding shape determination model. This is because if the determination result data of "accepted" is output, no treatment process to be set for this steel strip coil occurs.

In this embodiment, the treatment process for the steel strip coil is set based on the treatment data output from the winding shape determination model as described above. Specifically, the treatment data output from the winding shape determination model of the winding shape determination unit 2 is transmitted from the winding shape determination unit 2 to the host computer 6, and the host computer 6 performs the treatment process for the steel strip coil. is set.

In the steel strip coil treatment process setting method of the present embodiment, the treatment record database includes at least the quality standard data of the steel strip coil, but the quality standard data included in the judgment record database does not necessarily match the quality standard data. good. Further, the treatment record database can be a tabular database or the like.

The treatment data output from the winding shape determination model of the winding shape determination unit 2 is preferably used after a certain amount of treatment data corresponding to new product standards, shipping destinations, etc. is accumulated in the treatment record database.

As described above, in this embodiment, when the determination result data of "fail" is output from the winding shape determination model, it is possible to output the treatment data related to the details of the treatment process for correcting the winding shape. Therefore, it is possible to appropriately perform the treatment process for correcting the winding shape while minimizing the progress delay of the manufacturing process due to the occurrence of the winding shape defect of the steel strip coil.

By applying the winding shape determination model generation method, the winding shape determination method, the treatment process setting method, and the steel strip coil manufacturing method of the present invention, a test for actually manufacturing a steel strip coil was performed. The effects of the invention have been verified. This result will be explained below.

In the test of this example, a winding shape determination model was generated by the method of generating a winding shape determination model for a steel strip coil according to the first embodiment shown in FIG. A data set including image data of the steel strip coil, quality standard data of the steel strip coil, and determination result data of the winding shape was accumulated in the database section of the winding shape determination model generation section. As the image data of the steel strip coil, image data obtained by photographing the entire side surface of the steel strip coil was used. A color image obtained directly by camera photography is converted into a 256-gradation grayscale image and stored in the database unit 31, thereby compressing the amount of information in the image data. In addition, when photographing a steel strip coil, the steel strip coil is loaded on a weighing machine, and a lighting device capable of providing sufficient illumination is placed on the side of the steel strip coil at a predetermined position. It was taken with a camera fixed to the The weight of the steel strip coil was used as quality standard data for the steel strip coil. Also, in the conventional winding shape determination method shown in FIG. Although it was judged as "failed" by the judgment, the judge of the second judgment made a second judgment based on this judgment result, and the one that was judged to be "passed" was used as the judgment result data of "final pass". . In addition, if it is judged as "failed" by the primary judgment by the judge's visual observation, etc., and the second judgment is made by the judge of the second judgment based on this judgment result, and it is judged to be "failed" It was used as the judgment result data of "Final failure". Then, the determination result data used for generating the winding shape determination model is determined by two categories of "pass" and "failure", respectively, for the "final pass" and "final rejection" obtained in this way. Judgment result data was used.

In this test, at the stage where 48 sets of the above data sets are accumulated in the database unit 31 of the winding shape determination model generation unit 3, machine learning is performed by the machine learning unit 32 of the winding shape determination model generation unit 3, and the winding shape A judgment model was generated. A neural network was used as the machine learning algorithm, and the neural network had three intermediate layers and five nodes each. A sigmoid function was used as the activation function.

Furthermore, determination result data was output from the winding shape determination model of the winding shape determination unit 1 by the winding shape determination method of the steel strip coil of the first embodiment shown in FIG. Specifically, the winding shape determination model generated as described above is stored in a tablet terminal that can communicate with the host computer 6, and by using this tablet terminal as the winding shape determination unit 1, the winding shape determination model The judgment result data was output. As quality standard data to be input to the winding shape determination model, the weight of the steel strip coil was obtained from the host computer 6 based on the coil number of the steel strip coil to be determined.

Regarding the image data of the steel strip coil, photographs were taken at the same shooting position and under the same shooting conditions as the image data acquired when forming the judgment result database, and a tablet in which the winding shape judgment model was stored as the photograph data. It was sent to the winding shape determination unit 1 consisting of a terminal. Then, the obtained image data was converted into a 256-gradation grayscale image by a tablet terminal constituting the winding shape determination unit 1, and input to the winding shape determination model together with the quality standard data of the steel strip coil. Based on this input data, the winding shape determination model was used to output the determination result data of "pass" or "fail" for the earache of the steel strip coil, which was displayed on the screen of the tablet terminal. A series of operations from acquiring a color photograph of the steel strip coil on the tablet terminal to displaying the winding shape determination result data was automatically performed by software set in the tablet terminal.

On the other hand, for the steel strip coil to be determined by the steel strip coil winding shape determination method of the first embodiment shown in FIG. It was compared with the judgment result data displayed on the tablet terminal.

The winding shape was determined for 20 steel strip coils by the method described above. As a result, the 16 coils that were judged to be "acceptable" by the steel strip coil winding shape judging method of the first embodiment shown in FIG. On the other hand, the four coils that were judged to be "failed" by the steel strip coil winding shape judgment method of the first embodiment shown in FIG. , the pass/fail judgment by the judge was consistent. As a result, it was confirmed that the method for determining the winding shape of the steel strip coil of the present invention can automatically and accurately determine the winding shape with the same accuracy as the determination by the determiner.

In this example, similarly to Example 1 described above, a winding shape determination model was generated by the method of generating a winding shape determination model for a steel strip coil according to the first embodiment shown in FIG. A data set including image data of the steel strip coil, quality standard data of the steel strip coil, and determination result data of the winding shape was accumulated in the database section of the winding shape determination model generation section. As the image data of the steel strip coil, two types of image data obtained by photographing the entire side surface of the steel strip coil and image data obtained by photographing the side surface of the steel strip coil from the front were used. The former mainly contributes to the determination of the telescope of the steel strip coil, and the latter mainly contributes to the determination of ear pain of the steel strip coil. The method of photographing the steel strip coil and the method of compressing the amount of information in the image data were the same as in Example 1. As quality standard data for the steel strip coil, the thickness, width and weight of the steel strip coil were used. Also, in the conventional winding shape determination method shown in FIG. Although the primary judgment by the person's visual inspection, etc., was judged to be "failed", the second judge made a secondary judgment based on this judgment result and made a "pass". was used as the judgment result data. In addition, if it is judged as "failed" by the primary judgment by the judge's visual observation, etc., and based on this judgment result, the judge of the second judgment makes a second judgment and judged that it is "failed" It was used as the judgment result data of "Final failure". Then, the determination result data used for generating the winding shape determination model is determined by two categories of "pass" and "failure", respectively, for the "final pass" and "final rejection" obtained in this way. Judgment result data was used.

In the present embodiment, when 100 data sets described above are accumulated in the database unit 31 of the winding shape determination model generation unit 3, the machine learning unit 32 of the winding shape determination model generation unit 3 performs machine learning. A shape judgment model was generated. A neural network was used as a machine learning algorithm, and a configuration including a convolutional neural network as shown in FIG. 11 was used. At this time, the convolution layer and the pooling layer are applied to the first input layer for inputting the image data obtained by photographing the entire side surface of the steel strip coil and the image data obtained by photographing the side surface of the steel strip coil from the front, respectively. constructed a convolutional neural network with one layer each. Maximum pooling was used for the pooling layer. As a result, the feature amount of each image data is extracted and aggregated into one-dimensional array data in the fully connected layer. On the other hand, in the second input layer, in addition to the array data aggregated from the image data obtained by photographing the entire side surface of the steel strip coil and the image data obtained by photographing the side surface of the steel strip coil from the front, the quality standard Numerical data such as the thickness, width and weight of the steel strip coil were input and connected to a general neural network structure. The neural network at this time had three intermediate layers and five nodes each, and a sigmoid function was used as an activation function.

Then, determination result data was output from the winding shape determination model of the winding shape determination unit 1 by the winding shape determination method for the steel strip coil of the first embodiment shown in FIG. That is, the generated winding shape determination model was stored in the winding shape determining unit 1, and it was determined whether or not the winding shape of the steel strip coil needed to be corrected. The image data of the winding shape of the steel strip coil to be input to the winding shape determination model is image data of the entire side surface of the steel strip coil and image data of the side surface of the steel strip coil taken from the front in the same manner as the learning data described above. We used the image data obtained by The quality standard data of the steel strip coil was obtained from the host computer based on the coil number of the steel strip coil to be determined, and input into the winding shape determination model together with the image data of the steel strip coil. Based on this input data, the winding shape determination model was used to output determination result data of "pass" or "fail" for the steel strip coil telescope and ear pain.

On the other hand, for 20 steel strip coils to be determined by the steel strip coil winding shape determination method of the first embodiment shown in FIG. Pain was also evaluated at the same time and compared with the above-described evaluation result data.

The winding shape was determined for 20 steel strip coils by the method described above. As a result, the 18 coils that were judged to be "acceptable" by the steel strip coil winding shape judging method of the first embodiment shown in FIG. On the other hand, for the two coils that were "failed" by the steel strip coil winding shape determination method of the first embodiment shown in FIG. Thus, the pass/fail judgment according to the present embodiment and the pass/fail judgment by the judge of the secondary judgment coincided. As a result, it was confirmed that the method for determining the winding shape of the steel strip coil of the present invention can automatically and accurately determine the winding shape with the same accuracy as the determination by the person making the secondary determination.

1, 2 winding shape determination unit 3, 4 winding shape determination model generation unit 5 treatment data setting unit 31, 41 database unit 32, 42 machine learning unit 6 host computer 71 winding machine 72 binder 73 coil marker 74 weighing machine 75 winding shape Total 76 Transfer machine C Steel strip coil S Steel strip

Claims (16)

Each time it is determined whether or not it is necessary to modify the winding shape of the steel strip coil during manufacture in order to satisfy a predetermined quality standard after the steel strip coil is a final product, the winding shape is accumulating a data set containing the image data of the above, quality standard data related to the content of the quality standard, and determination result data related to the result of the determination to form a determination result database;
Generating a winding shape determination model for a steel strip coil that generates a winding shape determination model that outputs the determination result data when the image data and the quality standard data are input by performing machine learning using the determination result database. Method. The quality standard data includes information on the product standard of the steel strip coil, information on the shape or weight of the steel strip coil, information on the shipping destination of the steel strip coil, and information on the steel strip coil after the winding shape has been determined. 2. The method of generating a winding shape determination model for a steel strip coil according to claim 1, which includes at least one piece of information relating to the manufacturing process through which the final product is produced. The method for generating a winding shape determination model for a steel strip coil according to claim 1, wherein the machine learning uses at least one of neural network, decision tree learning, random forest, and support vector regression. 3. The method for generating a winding shape determination model for a steel strip coil according to claim 2, wherein the machine learning uses at least one of neural network, decision tree learning, random forest, and support vector regression. Acquiring image data of the winding shape of the steel strip coil in the process of manufacture and quality standard data relating to quality standards to be met after the steel strip coil becomes the final product,
The acquired image data and the quality reference data are input into the winding shape determination model generated by the method for generating a winding shape determination model for steel strip coils according to any one of claims 1 to 4, and the Output the judgment result data,
Based on the output determination result data, whether or not the winding shape of the steel strip coil under production needs to be modified in order to satisfy the quality standards after the steel strip coil becomes the final product. A method for determining the winding shape of a steel strip coil. When it is determined that the winding shape of the steel strip coil in the process of manufacturing needs to be modified in order to ensure that the final product of the steel strip coil satisfies a predetermined quality standard, a treatment step of modifying the winding shape. accumulating a data set including the image data of the winding shape, the quality standard data regarding the quality standard, and the processing data regarding the contents of the processing process when the setting is performed to form a processing result database;
The degree of similarity between the image data and the quality standard data of the steel strip coil in the process of production, which is determined to require correction of the winding shape by the steel strip coil winding shape determination method according to claim 5, is the highest searching the data set containing the image data and the quality reference data from the treatment history database;
A treatment process setting method for a steel strip coil, wherein the treatment process is set based on the treatment data contained in the searched data set. Each time it is determined whether or not it is necessary to modify the winding shape of the steel strip coil during manufacture in order to satisfy a predetermined quality standard after the steel strip coil is a final product, the winding shape is accumulating a data set containing the image data of the above, quality standard data related to the content of the quality standard, and determination result data related to the result of the determination to form a determination result database;
Each time a treatment step for correcting the winding shape required by the determination is set, the image data of the winding shape, the quality standard data about the quality standard, and the treatment data about the contents of the treatment step are generated. to form a treatment performance database by accumulating a data set containing
By performing machine learning using the determination result database and the treatment result database, when the image data and the quality standard data are input, the determination result data and when it is determined that the winding shape needs to be corrected A method for generating a winding shape determination model for a steel strip coil, further generating a winding shape determination model for outputting the treatment data. The quality standard data includes information on the product standard of the steel strip coil, information on the shape or weight of the steel strip coil, information on the shipping destination of the steel strip coil, and information on the steel strip coil after the winding shape has been determined. 8. The method for generating a winding shape determination model for a steel strip coil according to claim 7, which includes at least one piece of information about the manufacturing processes that pass through until the final product is obtained. The method for generating a winding shape determination model for a steel strip coil according to claim 7, wherein the machine learning uses at least one of neural network, decision tree learning, random forest, and support vector regression. The method for generating a winding shape determination model for a steel strip coil according to claim 8, wherein the machine learning uses at least one of neural network, decision tree learning, random forest, and support vector regression. Acquiring image data of the winding shape of the steel strip coil in the process of manufacture and quality standard data relating to quality standards to be met after the steel strip coil becomes the final product,
The obtained image data and the quality reference data are input into the winding shape determination model generated by the method for generating a steel strip coil winding shape determination model according to any one of claims 7 to 10. outputting the determination result data;
Based on the output determination result data, whether it is necessary to correct the winding shape of the steel strip coil during production in order to satisfy the quality standards after the steel strip coil becomes the final product. A method for determining the winding shape of a steel strip coil. 12. The steel strip coil, wherein the treatment step is applied based on the treatment data to the steel strip coil whose winding shape has been determined not to satisfy the quality standard by the steel strip coil winding shape determination method according to claim 11. A method of setting a treatment process for a band coil. A steel strip coil manufacturing method for manufacturing the steel strip coil whose winding shape is determined to satisfy the quality standard by the steel strip coil winding shape determining method according to claim 5 . A steel strip coil manufacturing method for manufacturing the steel strip coil whose winding shape is determined to satisfy the quality standard by the steel strip coil winding shape determination method according to claim 11 . According to the steel strip coil treatment process setting method according to claim 6, the winding shape of the steel strip coil in the process of production is corrected, and the steel strip coil is manufactured so as to satisfy the quality standard after becoming a final product. A method for manufacturing a steel strip coil. A steel strip coil is produced by correcting the winding shape of the steel strip coil in the process of manufacturing by the steel strip coil treatment process setting method according to claim 12 so as to satisfy the quality standard after the final product. A method for manufacturing a steel strip coil.

Images