Enameled wire surface state management device and enameled wire manufacturing device

Abstract

To provide a surface state management device of an enameled wire capable of easily estimating a cause of abnormality of the enameled wire, and a manufacturing device of the enameled wire.SOLUTION: The enameled wire surface state management device 1 includes the cause estimation processing unit 23 that estimates the cause from at least one of the conductor surface image 31 and the appearance image 32 to be estimated, using the learned model 34 created by machine learning in advance at least one of the conductor surface image 31 and the appearance image 32 when an abnormality occurs on the surface of the conductor 11 or the insulator 12 and the cause of the abnormality, using at least one of the conductor surface image 31 obtained by imaging the surface of the conductor 11 before the insulator 12 is formed and the appearance image 32 obtained by imaging the surface of the insulator 12 after the insulator 12 is formed.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an apparatus for controlling the surface condition of an enameled wire and an apparatus for manufacturing an enameled wire. [Background technology]

[0002] When manufacturing enameled wire, abnormalities (surface defects) can occur on the surface of the enameled wire for a variety of reasons. If there are abnormalities on the surface of the enameled wire, it can deteriorate its appearance and make the insulation more susceptible to peeling. For this reason, in the past, measures were taken such as inspecting the surface of enameled wire with a visual inspection machine and discarding any enameled wire that was found to have surface abnormalities.

[0003] Prior art document information related to the invention of this application includes Patent Document 1. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6775446 Summary of the Invention [Problem to be solved by the invention]

[0005] To produce enameled wire with a high yield, it is desirable to minimize surface defects on the enameled wire. Therefore, when an abnormality is detected on the surface of the enameled wire, it is desirable to identify the cause of the abnormality and take appropriate measures to improve the surface abnormality.

[0006] SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide an apparatus for managing the surface condition of an enameled wire, and an apparatus for manufacturing an enameled wire, which are capable of easily estimating the cause of an abnormality in an enameled wire. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, the present invention provides a surface condition management device for an enameled wire having a conductor and an insulator formed by applying and baking an enamel paint around the conductor, the surface condition management device including a cause estimation processing unit that uses at least one of a conductor surface image taken of the surface of the conductor before the insulator is formed and an appearance image taken of the surface of the insulator after the insulator is formed, and that estimates the cause of the abnormality from at least one of the conductor surface image and the appearance image that are the subject of estimation using a trained model created in advance by machine learning of at least one of the conductor surface image and the appearance image when an abnormality occurs on the surface of the conductor or the insulator, and the cause of the abnormality.

[0008] Furthermore, in order to solve the above-mentioned problems, the present invention provides an apparatus for manufacturing an enameled wire having a conductor and an insulator formed by applying and baking an enamel paint around the conductor, the apparatus comprising: a conductor imaging unit that images the surface of the conductor before the insulator is formed to obtain a conductor surface image; an appearance imaging unit that images the surface of the insulator after the insulator is formed to obtain an appearance image; and the enameled wire surface condition management device according to any one of claims 1 to 6. [Effects of the Invention]

[0009] According to the present invention, it is possible to provide an apparatus for managing the surface condition of an enameled wire and an apparatus for manufacturing an enameled wire, which are capable of easily estimating the cause of an abnormality in an enameled wire. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic diagram illustrating the configuration of an enameled wire manufacturing apparatus according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram showing an example of a conductor surface image. [Figure 3] 1A and 1B are diagrams showing examples of appearance images, where (a) is an example of a bright-field image and (b) is an example of a dark-field image. [Figure 4] 10A and 10B are diagrams showing an image of a conductor surface and an image of its appearance when an abnormality occurs, and an example of the cause of the abnormality. [Figure 5] 10A and 10B are diagrams showing an image of a conductor surface and an image of its appearance when an abnormality occurs, and an example of the cause of the abnormality. [Figure 6] FIG. 10 is a diagram illustrating an example of a countermeasure table. [Figure 7] FIG. 2 is a diagram showing a control flow of an enameled wire surface condition management device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0011] [Embodiment Mode] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

[0012] (Schematic configuration of enameled wire manufacturing apparatus 100) Fig. 1 is a schematic diagram of an enameled wire manufacturing apparatus according to one embodiment of the present invention. As shown in Fig. 1, the enameled wire manufacturing apparatus 100 processes a conductor 11 into a rectangular shape, then applies enamel paint around the conductor 11 and bakes it to form an insulator 12, thereby manufacturing an enameled wire 10.

[0013] The enameled wire manufacturing apparatus 100 includes a bobbin 101, a flat-angle processing section 102 that processes the conductor 11 fed from the bobbin 101 into a flat angle, an insulator forming section 103 that applies enamel paint around the flat-angle processed conductor 11 and bakes it to form an insulator 12 to form an enameled wire 10, and a winding machine 104 that winds up the enameled wire 10.

[0014] The rectangular processing unit 102 includes a rectangular rolling mill 102a, a rectangular wire drawing machine 102b, and a cleaning device 102c. The rectangular rolling mill 102a rolls the conductor 11 delivered from the bobbin 101 into a rectangular shape. The cross-sectional shape perpendicular to the longitudinal direction of the conductor 11 delivered from the bobbin 101 is, for example, circular. The rectangular rolling mill 102a rolls the conductor 11 from top to bottom and left to right so that the thickness and width of the conductor 11 reach predetermined dimensions, thereby processing the cross section of the conductor 11 into a rectangular shape. The rectangular rolling mill 102a includes, for example, a cassette roller die (CRD). The rectangular rolling mill 102a may perform rolling processing only once, or may be a multi-stage rolling mill that performs rolling processing multiple times in stages.

[0015] The rectangular wire drawing machine 102b draws the rectangular conductor 11. The rectangular wire drawing machine 102b processes the conductor 11 to the desired dimensions. The rectangular wire drawing machine 102b includes, for example, a wire drawing die. To smoothly pass the conductor 11 through the wire drawing die, for example, a lubricant containing an emulsion can be used. The rectangular wire drawing machine 102b also uses, for example, felt. As the felt, for example, a wiping cloth made of ultrafine fibers made of polyester, nylon, or the like can be used. By sandwiching the rectangular conductor 11 between the felts from above and below or from the left and right, copper powder and foreign matter adhering to the surface of the rectangular conductor 11 are removed.

[0016] The cleaning device 102c cleans the conductor 11 that has been drawn into a rectangular shape, removing lubricating oil, moisture, and the like adhering to the conductor 11. The cleaning can also remove copper powder (for example, copper powder that could not be completely removed by felt) adhering to the surface of the conductor 11. The cleaning device 102c is equipped with a cleaning liquid for cleaning the conductor 11 after drawing. For example, water such as pure water can be used as the cleaning liquid. For example, pure water can be RO water.

[0017] The insulator forming unit 103 includes a paint applicator 103a and a baking furnace 103b. The paint applicator 103a applies enamel paint to the cleaned conductor 11. The paint applicator 103a includes, for example, a paint tank for storing enamel paint, a paint applicator roll for drawing up the enamel paint from the paint tank and applying it to the outer surface of the conductor 11, and a paint die for adjusting the thickness of the enamel paint so that it is uniform. The enamel paint is, for example, a paint containing a resin and a solvent. Examples of solvents in enamel paint include N-methyl-2-pyrrolidone (NMP), cresol, N,N-dimethylacetamide (DMAc), and cyclohexanone. Examples of resins in enamel paint include polyamideimide, polyimide, and polyesterimide.

[0018] The baking furnace 103b evaporates the solvent in the enamel paint applied to the conductor 11 and hardens the resin in the enamel paint. After the solvent evaporates and the resin hardens, the insulator 12 is formed, resulting in the enameled wire 10. The baking furnace 103b may also include a drying furnace. The drying furnace only evaporates the solvent in the enamel paint and does not harden the resin in the enamel paint. The application of the enamel paint by the paint applicator 103a and the baking of the enamel paint by the baking furnace 103b may be repeated until the insulator 12 reaches a desired thickness.

[0019] The winder 104 winds the enameled wire 10 on which the insulator 12 is formed. The rectangular processing by the rectangular processing unit 102 and the formation of the insulator 12 by the insulator forming unit 103 may be performed on separate production lines.

[0020] (conductor imaging unit 61, external imaging unit 62, microscope 63) The enameled wire manufacturing apparatus 100 further includes a conductor imaging unit 61 that images the surface of the conductor 11 after flat angle processing by the flat angle processing unit 102 and before the formation of the insulator 12 to obtain a conductor surface image 31, and an appearance imaging unit 62 that images the surface of the insulator 12 after the insulator 12 is formed by the insulator forming unit 103 to obtain an appearance image 32.

[0021] 2 is a diagram showing an example of a conductor surface image 31 obtained by the conductor imaging unit 61. As shown in FIG. 2, the conductor imaging unit 61 continuously acquires conductor surface images 31, in which the height (depth) of the irregularities on the surface of the conductor 11 is expressed by shading, along the longitudinal direction of the conductor 11. The conductor imaging unit 61 may be configured to, for example, irradiate the surface of the conductor 11 with light and measure the height (depth) of the irregularities on the surface of the conductor 11 based on the difference in intensity (brightness, darkness) of the reflected light. The conductor imaging unit 61 may be, for example, a line scanner (such as an ultra-high-definition inline profile measuring instrument manufactured by Keyence Corporation).

[0022] 3(a) and 3(b) are diagrams showing an example of an appearance image 32 obtained by the appearance imaging unit 62. The appearance imaging unit 62 captures a bright-field image 321 shown in FIG. 3(a) and a dark-field image 322 shown in FIG. 3(b). The bright-field image 321 is an image of the flat and corner surfaces of the enameled wire 10 captured by irradiating the flat and corner surfaces with bright-field imaging light. The dark-field image 322 is an image of the flat and corner surfaces of the enameled wire 10 captured by irradiating the enameled wire 10 with dark-field imaging light along the longitudinal direction. The appearance imaging unit 62 that captures the bright-field image 321 and the dark-field image 322 can be, for example, an appearance inspection device described in JP 2016-102727 A.

[0023] Furthermore, in this embodiment, the enameled wire manufacturing apparatus 100 further includes a microscope 63 for obtaining an insulator-removed image 33, which is a micrograph of the enameled wire 10 after the insulator 12 has been removed. The microscope 63 may be an optical microscope or a scanning electron microscope (SEM). Using the insulator-removed image 33 enables detailed analysis of the irregularities present on the surface of the conductor 11, enabling more detailed estimation of the cause of the irregularities. However, since the insulator-removed image 33 cannot be obtained while manufacturing the enameled wire 10, it is recommended to use the insulator-removed image 33 to estimate the cause when it is determined that the estimation of the cause using the conductor surface image 31 or the appearance image 32 is insufficient (for example, when countermeasures taken in response to the estimated cause of the abnormality do not produce sufficient results).

[0024] (Enameled wire surface condition control device 1) The enameled wire manufacturing apparatus 100 is equipped with an enameled wire surface condition management device 1 according to this embodiment. The enameled wire surface condition management device 1 monitors abnormalities (surface defects) on the surfaces of the conductor 11 and the insulator 12, and when an abnormality occurs, it estimates the cause and proposes countermeasures. An abnormality (surface defect) on the surface of the insulator 12 refers to irregularities or bubbles that degrade the appearance. Furthermore, an abnormality (surface defect) on the surface of the conductor 11 refers to a defect that could cause bubbles or the like when the surface (e.g., a flat surface) of the finished enameled wire 10 is observed.

[0025] The enameled wire surface condition management device 1 has a control unit 2 and a storage unit 3. The enameled wire surface condition management device 1 is, for example, a computer such as a personal computer or a server device, and is equipped with a processing element such as a CPU, memories such as RAM and ROM, a storage device such as a hard disk, and a communication interface which is a communication device such as a LAN card.

[0026] The control unit 2 has a data acquisition processing unit 21, an abnormality detection processing unit 22, a cause estimation processing unit 23, a countermeasure presentation processing unit 24, and a second cause estimation processing unit 25. Details of each unit will be described later. The storage unit 3 is realized by a predetermined storage area of ​​a memory or a storage device.

[0027] The enameled wire surface condition management device 1 also has a display 4 and an input device 5. The display 4 is, for example, a liquid crystal display, and the input device 5 is, for example, a keyboard or a mouse. The display 4 may be configured as a touch panel so that it also serves as the input device 5. The display 4 and the input device 5 may also be configured separately from the enameled wire surface condition management device 1 and configured to be able to communicate with the enameled wire surface condition management device 1 via wireless communication or the like. In this case, the display 4 or the input device 5 may be configured as a mobile terminal such as a tablet or smartphone.

[0028] (Data acquisition processing unit 21) The data acquisition processing unit 21 acquires a conductor surface image 31 and an appearance image 32 (bright-field image 321 and dark-field image 322) from the conductor imaging unit 61 and the appearance imaging unit 62 in real time during the manufacture of the enameled wire 10. The acquired conductor surface image 31 and appearance image 32 (bright-field image 321 and dark-field image 322) are stored in the memory unit 3. Furthermore, the data acquisition processing unit 21 acquires an insulator-removed image 33 from the microscope 63 as necessary. The acquired insulator-removed image 33 is stored in the memory unit 3.

[0029] (Abnormality detection processing unit 22) The anomaly detection processing unit 22 performs an anomaly detection process to detect anomalies on the surfaces of the conductor 11 and the insulator 12 from the conductor surface image 31 and the appearance image 32 (bright-field image 321, dark-field image 322) acquired by the data acquisition processing unit 21. The anomaly detection processing unit 22 detects anomalies on the surface of the conductor 11 from the conductor surface image 31, and also detects anomalies on the surface of the insulator 12 from the appearance image 32 (bright-field image 321, dark-field image 322).

[0030] The specific method for detecting an abnormality is not particularly limited. For example, irregularities that are candidates for abnormality detection may be detected from the conductor surface image 31, and an abnormality on the surface of the conductor 11 may be detected when the depth of the irregularities is greater than a threshold value. Furthermore, as a method for detecting an abnormality on the surface of the insulator 12 from the appearance image 32 (bright-field image 321, dark-field image 322), a known appearance inspection method described in, for example, JP 2016-102727 A may be applied.

[0031] The anomaly detection processing unit 22 does not need to be integrally mounted on the enameled wire surface condition management device 1, and may be realized by a device external to the enameled wire surface condition management device 1 (for example, another personal computer). For example, when the appearance imaging unit 62 is configured as an existing appearance inspection device, part of the anomaly detection processing unit 22 may be mounted on the appearance inspection device. When the anomaly detection processing unit 22 is configured as an external device, the data acquisition processing unit 21 may be configured to acquire the conductor surface image 31 and appearance image 32 (bright-field image 321, dark-field image 322) in which the anomaly is detected and store them in the memory unit 3 only when the anomaly is detected.

[0032] (Cause estimation processing unit 23) The cause estimation processing unit 23 performs a cause estimation process using at least one of the conductor surface image 31 and the appearance image 32, to estimate the cause of the abnormality from at least one of the conductor surface image 31 and the appearance image 32 to be estimated, using a trained model 34 created in advance by machine learning at least one of the conductor surface image 31 and the appearance image 32 when an abnormality occurs on the surface of the conductor 11 or the insulator 12 and the cause of the abnormality. In this embodiment, when an abnormality is detected on the surface of the conductor 11 or the insulator 12 by the anomaly detection processing unit 22, the cause estimation processing unit 23 is configured to estimate the cause using as input three images: the conductor surface image 31 in which the abnormality was detected, a bright-field image 321, and a dark-field image 322.

[0033] The trained model 34 is created by repeatedly performing machine learning on a large number of samples in advance on a combination of three images, namely, the conductor surface image 31 at the time of an abnormality as training data, the bright-field image 321 as the appearance image 32, and the dark-field image 322, as well as the cause of the abnormality. As the algorithm of the trained model 34, for example, a so-called supervised learning classifier (support vector machine, deep forest, etc.) can be applied.

[0034] 4 and 5 show a conductor surface image 31 and an appearance image 32 when an abnormality occurs, along with an example of the cause of the abnormality. When unevenness occurs over a relatively large area, as in the upper part of FIG. 4, it is believed that a lump of copper powder (copper pieces) accumulated in the flat wire drawing machine 102b is pressed against and crimped onto the conductor 11. When continuous defects (unevenness) are present in the longitudinal direction, as in the middle part of FIG. 4, it is believed that the conductor 11 interferes with a part of the device before or after wire drawing, or that excessive force is being applied by the die during wire drawing. When concave and point-like defects are present, as in the lower part of FIG. 4, it is believed that a lump of copper powder (copper pieces) accumulated in the flat wire drawing machine 102b is pressed against the conductor 11, or that the defect occurred during wire drawing.

[0035] When there are numerous linear defects, as in the upper part of Figure 5, it is believed that excessive force was applied to the die during wire drawing, or that foreign matter was stuck in the wire drawing die, damaging the conductor 11. When there is no abnormality on the conductor surface and only an abnormality in appearance, as in the middle part of Figure 5, it is believed that there is a problem in the painting process (a problem with the paint applicator 103a). Conversely, when there is no abnormality in appearance and only an abnormality in the conductor surface, as in the lower part of Figure 5, it is believed that there is a defect in the conductor 11 with relatively small irregularities that are not visible to the eye. Note that in this embodiment, even in such cases, countermeasures will be presented as preventive measures to improve quality.

[0036] In the present embodiment, the cause estimation process is configured to estimate the cause using one trained model 34 and inputting three images, namely, the conductor surface image 31, the bright-field image 321, and the dark-field image 322. However, the present invention is not limited to this. For example, a trained model that estimates the cause of an abnormality using the conductor surface image 31 as input and a trained model that estimates the cause of an abnormality using the bright-field image 321 and the dark-field image 322 as input may be separately configured, and cause estimation may be performed for each. Furthermore, a trained model that estimates the cause of an abnormality using the bright-field image 321 as input and a trained model that estimates the cause of an abnormality using the dark-field image 322 as input may also be separately configured. In this case, the causes of three abnormalities are estimated using three trained models, and countermeasures for all three causes of the abnormality are presented in the countermeasure presentation process described below.

[0037] (Countermeasure presentation processing unit 24) The countermeasure presentation processing unit 24 performs a countermeasure presentation process to present a countermeasure for the abnormality that has occurred based on the cause estimated by the cause estimation processing unit 23. In the countermeasure presentation process, the countermeasure presentation processing unit 24 refers to a countermeasure table 35 stored in advance in the storage unit 3, extracts a countermeasure corresponding to the cause estimated by the cause estimation processing unit 23, and displays the extracted countermeasure on the display unit 4 or the like, thereby presenting a countermeasure according to the cause.

[0038] Fig. 6 is a diagram showing an example of the countermeasure table 35. As shown in Fig. 6, the countermeasure table 35 is a database that stores causes of abnormalities and corresponding countermeasures in association with each other. The countermeasure presentation processing unit 24 refers to the countermeasure table 35, extracts a countermeasure corresponding to the cause estimated by the cause estimation processing unit 23, and displays it on the display 4 or the like.

[0039] Note that Figure 6 is merely an example, and more specific countermeasures may be presented. For example, for No. 1 in the figure, "copper powder is compressed," more specific countermeasures may be presented, such as replacing the felt, replacing the lubricating oil, or replacing both.

[0040] (Second cause estimation processing unit 25) The second cause estimation processing unit 25 performs second cause estimation processing to estimate the cause from the insulator-removed image 33, which is the subject of estimation, using a second trained model 36 that has previously undergone machine learning to learn about the cause and the insulator-removed image 33, which is a micrograph of the enameled wire 10 after the insulator 12 has been removed. Note that the second cause estimation processing is not processing that is performed in real time while the enameled wire 10 is being manufactured, but processing that is performed as needed after the enameled wire 10 has been manufactured in order to verify the cause in more detail.

[0041] (Control Flow) 7 is a diagram showing a control flow of the enameled wire surface condition management device 1. The control flow of FIG. 7 is executed, for example, at the start of the production of the enameled wire 10.

[0042] First, in step S1, a data acquisition process is performed. In the data acquisition process, the data acquisition processing unit 21 acquires a conductor surface image 31 from the conductor imaging unit 61, and also acquires an appearance image 32 (a bright-field image 321 and a dark-field image 322) obtained by the appearance imaging unit 62. The acquired conductor surface image 31 and appearance image 32 are stored in the storage unit 3.

[0043] Thereafter, in step S2, an abnormality detection process is performed. In the abnormality detection process, the abnormality detection processing unit 22 detects abnormalities on the surfaces of the conductor 11 and the insulator 12 from the conductor surface image 31 and the appearance image 32 obtained in step S1.

[0044] Thereafter, in step S3, it is determined whether or not an abnormality has been detected as a result of the anomaly detection process in step S3. If the determination in step S3 is No (N), the process proceeds to step S6. If the determination in step S3 is Yes (Y), then in step S4, cause estimation processing is performed. In the cause estimation processing, the cause estimation processing unit 23 inputs the conductor surface image 31 and appearance image 32 at the time of detecting the abnormality into a trained model 34 that has undergone machine learning in advance, and estimates the cause of the abnormality. Note that if no countermeasure is stored (if the cause cannot be estimated), it is more preferable to allow a person to input the estimation result.

[0045] Thereafter, in step S5, a countermeasure presentation process is performed. In the countermeasure presentation process, the countermeasure presentation processing unit 24 refers to the countermeasure table 35, extracts a countermeasure corresponding to the cause estimated in step S4, and displays the extracted countermeasure on the display 4 or the like. The worker refers to the countermeasure displayed on the display 4 or the like and performs, for example, the presented countermeasure (for example, work to remove foreign matter in the rectangular processing unit 102 or to strengthen foreign matter suppression (such as replacing felt or lubricating oil)). Then, the process proceeds to step S6.

[0046] In step S6, it is determined whether the production of the enameled wire 10 is completed. If the determination in step S6 is No (N), the process returns to step S1 and continues. If the determination in step S6 is Yes (Y), the process ends.

[0047] (Variation) In this embodiment, the trained model 34 is created by machine learning both the conductor surface image 31 and the appearance image 32 and the cause, but the trained model 34 may be created using only either the conductor surface image 31 or the appearance image 32. In this case, the cause of the abnormality is estimated and a countermeasure is presented, targeting only either the abnormality on the surface of the conductor 11 or the abnormality in the appearance (surface of the insulator 12).

[0048] Furthermore, in this embodiment, the anomaly detection process and the cause estimation process are performed as separate processes, but they can also be configured to be performed simultaneously. In this case, the trained model 34 may include the conductor surface image 31 and the appearance image 32 when no anomaly is present in the training data, and perform machine learning using the corresponding label (when an anomaly is present, the cause of the anomaly is used as the label) as "no anomaly." The conductor surface image 31 and the appearance image 32 acquired in real time during the production of the enameled wire 10 are then input to the trained model 34, which then performs processing to output whether or not there is an anomaly and, if there is an anomaly, the cause.

[0049] Furthermore, in the present embodiment, the cause of the anomaly is estimated based on the conductor surface image 31 and the appearance image 32 in the cause estimation process. However, the cause may also be estimated by taking into consideration manufacturing parameters that indicate the manufacturing state, such as temperature and wire speed. In this case, when creating the pre-trained model 34, machine learning may be performed on the conductor surface image 31, the appearance image 32, the cause, and manufacturing parameters such as temperature and wire speed. That is, the pre-trained model 34 may be created by pre-training at least one of the conductor surface image 31 and the appearance image 32, the manufacturing parameters that indicate the manufacturing state, and the cause. Then, the cause estimation processing unit 23 may use the pre-trained model 34 to estimate the cause from at least one of the conductor surface image 31 and the appearance image 32 that are the estimation targets and the manufacturing parameters.

[0050] 7, a step may be included between steps S5 and S6 to issue a warning when the number of surface defects (surface abnormalities of the insulator 12) detected in step S2 exceeds a predetermined number (e.g., 25 defects / 3200 m). This makes it possible to warn the operator that some kind of countermeasure is necessary, for example, when the occurrence of foreign matter (e.g., copper powder) increases during rectangular wire processing, and thus makes it possible to prevent a decrease in yield.

[0051] (Variation) In this embodiment, the cause estimation process is performed using both the conductor surface image 31 and the appearance image 32. However, the cause estimation process may also be performed using only one of them. For example, the cause estimation processing unit 23 may be configured to estimate the cause of the abnormality from the conductor surface image 31, which is the estimation target, using only the conductor surface image 31, by using a trained model 34 created in advance through machine learning of the conductor surface image 31 when an abnormality occurs on the surface of the conductor 11 and the cause of the abnormality. Furthermore, the cause estimation processing unit 23 may be configured to estimate the cause of the abnormality from the appearance image 32, which is the estimation target, using only the appearance image 32, by using a trained model 34 created in advance through machine learning of the appearance image 32 when an abnormality occurs on the surface of the insulator 12 and the cause of the abnormality.

[0052] (Actions and Effects of the Embodiments) As described above, the enameled wire surface condition management device 1 according to this embodiment is equipped with a cause estimation processing unit 23 that uses at least one of the conductor surface image 31 and the appearance image 32, and estimates the cause from at least one of the conductor surface image 31 and the appearance image 32 that are the subject of estimation, using a trained model 34 that has been created in advance by machine learning at least one of the conductor surface image 31 and the appearance image 32 at the time of an abnormality and the cause of the abnormality.

[0053] This makes it possible to easily and quickly identify the cause of an abnormality when it occurs, and to take appropriate measures according to the identified cause. As a result, it is possible to improve the quality of the enameled wire 10 and increase the yield. That is, according to this embodiment, it is possible to efficiently manufacture enameled wire 10 with good surface quality (i.e., with few surface defects). More specifically, it is possible to maintain the number of surface defects measured on the enameled wire 10 during product inspection at 25 defects / 3200 m or less.

[0054] (Summary of the embodiment) Next, the technical ideas grasped from the above-described embodiments will be described by using the reference numerals and the like in the embodiments. However, the reference numerals and the like in the following description do not limit the components in the claims to the members and the like specifically shown in the embodiments.

[0055] [1] A surface condition management device for an enameled wire (10) having a conductor (11) and an insulator (12) formed by applying and baking an enamel coating around the conductor (11), the surface condition management device (1) comprising: a cause estimation processing unit (23) that uses at least one of a conductor surface image (31) obtained by capturing an image of the surface of the conductor (11) before the insulator (12) is formed and an appearance image (32) obtained by capturing an image of the surface of the insulator (12) after the insulator (12) is formed, and that estimates a cause from at least one of the conductor surface image (31) and the appearance image (32) that are the subject of estimation, using a trained model (34) that has been created by machine learning in advance on at least one of the conductor surface image (31) and the appearance image (32) when an abnormality occurs on the surface of the conductor (11) or the insulator (12) and the cause of the abnormality.

[0056] [2] The surface condition management device (1) of enameled wire described in [1], wherein the trained model (34) is created by machine learning both the conductor surface image (31) and the appearance image (32) and the cause, and the cause estimation processing unit (23) estimates the cause from both the conductor surface image (31) and the appearance image (32) using the trained model (34).

[0057] [3] The surface condition management device (1) of enameled wire according to [1], further comprising an anomaly detection processing unit (22) that detects an anomaly on the surface of the conductor (11) from the conductor surface image (31) and detects an anomaly on the surface of the insulator (12) from the appearance image (32), and the cause estimation processing unit (23), when the anomaly detection processing unit (22) detects an anomaly on the surface of the conductor (11) or the insulator (12), estimates the cause using at least one of the conductor surface image (31) and the appearance image (32) at the time the anomaly was detected as the estimation target.

[0058] [4] The surface condition management device (1) of [1], further comprising a countermeasure suggestion processing unit (24) that suggests a countermeasure for the abnormality that has occurred based on the cause estimated by the cause estimation processing unit (23).

[0059] [5] The surface condition management device (1) of [1], further comprising a second cause estimation processing unit (25) that estimates the cause from the insulator-removed image (33) that is the subject of estimation, using a second trained model (36) that has previously undergone machine learning to learn about the cause and the insulator-removed image (33), which is a micrograph of the enameled wire (10) after the insulator (12) has been removed.

[0060] [6] The surface condition management device (1) for enameled wire according to [1], wherein the cause estimation processing unit (23) estimates the cause from at least one of the conductor surface image (31) and the appearance image (32) that are the estimation targets and the manufacturing parameters, using the trained model (34) created in advance by machine learning of at least one of the conductor surface image (31) and the appearance image (32), manufacturing parameters that are parameters indicating the manufacturing state of the enameled wire (10), and the cause.

[0061] [7] An apparatus for manufacturing an enameled wire (10) having a conductor (11) and an insulator (12) formed by applying and baking an enamel coating around the conductor (11), the apparatus comprising: a conductor imaging unit (61) that images the surface of the conductor (11) before the insulator (12) is formed to obtain a conductor surface image (31); an appearance imaging unit (62) that images the surface of the insulator (12) after the insulator (12) is formed to obtain an appearance image (32); and an enameled wire surface condition management device (1) according to any one of [1] to [6].

[0062] (Addendum) Although the embodiments of the present invention have been described above, the invention according to the claims is not limited to the above-described embodiments. It should be noted that not all of the combinations of features described in the embodiments are necessarily essential to the means for solving the problems of the invention. Furthermore, the present invention can be appropriately modified and implemented within the scope of its spirit. [Explanation of symbols]

[0063] 1...Enameled wire surface condition control device 10...Enameled wire 11...conductor 12...Insulator 2...Control unit 21...Data acquisition processing unit 22...Abnormality detection processing unit 23...Cause estimation processing section 24...Countermeasure presentation processing section 25...Cause estimation processing section 3...Storage section 31...Conductor surface image 32...Exterior image 321...Bright-field image 322…Dark field image 33...Insulator removal image 34…Trained model 35...Countermeasures table 36…Second trained model 61...conductor imaging unit 62...External imaging unit 100...Enameled wire manufacturing equipment

Claims

1. An apparatus for managing the surface condition of an enameled wire having a conductor and an insulator formed by applying and baking an enamel coating around the conductor, and a cause estimation processing unit that uses at least one of a conductor surface image obtained by capturing an image of the surface of the conductor before the formation of the insulator and an appearance image obtained by capturing an image of the surface of the insulator after the formation of the insulator, and estimates the cause from at least one of the conductor surface image and the appearance image to be estimated using a trained model that is created by machine learning in advance on at least one of the conductor surface image and the appearance image when an abnormality occurs on the surface of the conductor or the insulator, and the cause of the abnormality. Enameled wire surface condition control device.

2. The trained model is created by machine learning both the conductor surface image and the appearance image and the cause, The cause estimation processing unit estimates the cause from both the conductor surface image and the appearance image using the trained model.

2. The surface condition control device for an enameled wire according to claim 1.

3. an abnormality detection processing unit that detects an abnormality on the surface of the conductor from the conductor surface image and detects an abnormality on the surface of the insulator from the appearance image; When the abnormality detection processing unit detects an abnormality on the surface of the conductor or the insulator, the cause estimation processing unit estimates the cause using at least one of the conductor surface image and the appearance image at the time the abnormality is detected as the estimation target.

2. The surface condition control device for an enameled wire according to claim 1.

4. a countermeasure suggestion processing unit that suggests a countermeasure for the abnormality that has occurred based on the cause that has been guessed by the cause guess processing unit, 2. The surface condition control device for an enameled wire according to claim 1.

5. a second cause estimation processing unit that estimates the cause from the insulator-removed image that is an estimation target, using a second trained model that has been machine-learned in advance to learn about the cause and an insulator-removed image, which is a micrograph of the enameled wire after the insulator has been removed; 2. The surface condition control device for an enameled wire according to claim 1.

6. the cause estimation processing unit estimates the cause from at least one of the conductor surface image and the appearance image, which are estimation targets, and the manufacturing parameters, using the trained model created in advance by machine learning on at least one of the conductor surface image and the appearance image, manufacturing parameters that are parameters indicating the manufacturing state of the enameled wire, and the cause; 2. The surface condition control device for an enameled wire according to claim 1.

7. An apparatus for manufacturing an enameled wire having a conductor and an insulator formed by applying and baking an enamel coating around the conductor, a conductor imaging unit that images a surface of the conductor before the insulator is formed to obtain a conductor surface image; an appearance imaging unit that captures an image of the surface of the insulator after the insulator is formed to obtain an appearance image; and the surface condition management device for an enameled wire according to any one of claims 1 to 6. Enameled wire manufacturing equipment.

Images