JP2005077180A - Cause estimation device and cause estimation program and cause estimation method of surface defect of metal plate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a cause estimation device and a cause estimation program and a cause estimation method capable of estimating a cause of surface defects of a metal plate rapidly and with high accuracy. <P>SOLUTION: The cause estimation device 10 comprises a storage part 13 receiving data concerning the surface defects of a cause estimation object and storing the data, a classification part 14 classifying the surface defects of the cause estimation object into cyclic defects and acyclic defects based on the data concerning the surface defects of the cause estimation object stored in the storage part 13, a retrieval part 15 retrieving an estimation cause corresponding to the data concerning the surface defects of the cause estimation object from a database 11 in which the data concerning the surface defects generated in the past is stored concerning the surface defects of the cause estimation object classified by the classification part 14, and an output part 16 outputting the estimation cause discovered by the retrieval part 15. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a cause estimation device, a cause estimation program, and a cause estimation method used when estimating the cause when a surface defect occurs in a metal plate such as an aluminum plate or a copper plate.

Conventionally, an appearance inspection of a metal plate material has been performed by visual observation during or after manufacture. However, since there are individual differences in the appearance inspection by visual observation, in recent years, the number of cases in which an inspection apparatus is introduced is increasing.
As the inspection device, for example, a device using a CCD camera, a device using a laser beam, or the like is used. FIG. 6 shows an example of an inspection apparatus using a CCD camera. When the inspection apparatus 60 inspects the metal plate material 61, an image of the surface of the metal plate material 61 illuminated by the illumination device 62 is captured by the CCD camera 63, and the image data is input to the defect determination device 64. Then, the quality is determined by comparing with the inspection standard set in advance, and the surface defect is detected.

  Since the inspection apparatus 60 detects surface defects based on image density data, for example, when wrinkles occur in the metal plate 61, two-dimensional data such as wrinkle size and contrast can be obtained. At present, data on depth cannot be obtained.

  On the other hand, when an inspection apparatus 70 using a laser beam as shown in FIG. 7 is used, a laser beam 73 from a laser light source 72 is scanned on a metal plate 71 via a rotating mirror 74, and further from the metal plate 71. The reflected light 75 is diffracted by the diffraction lens 76. Then, the diffracted light 77 is input to the defect determiner 79 via the photodetector 78. Then, the quality is determined by comparing with the inspection standard set in advance, and the surface defect is detected.

  In the inspection apparatus 70, the surface defect is detected by the change of the diffracted light 77. Since the diffracted light 77 changes due to the unevenness of the surface of the metal plate 71, data on the depth of the surface defect can be obtained. However, it is difficult to detect discoloration that is a two-dimensional surface defect. Although the above-described inspection device 60 and inspection device 70 are not shown in the drawing, the front surface side and the back surface side of the metal plate materials 61 and 71 so that both surfaces (front and back surfaces) of the metal plate materials 61 and 71 can be inspected. Respectively. Moreover, the conveyance speed of the metal plate materials 61 and 71 is about 50 to 1000 m / min.

  By the way, in the conventional visual inspection or visual inspection using the inspection device, when a surface defect of the metal plate material is detected, the defect is sampled and analyzed by using various analyzers (SEM, ESCA, etc.). I investigated and took measures. Therefore, when it takes a long time to investigate the cause, it is not possible to take measures promptly, and depending on the surface defect, the production line of the metal plate material may have to be stopped for several days.

On the other hand, in the process of manufacturing a semiconductor integrated circuit, conventionally, a defect (for example, a pattern defect, a foreign object, a flaw, etc.) that has occurred in the past has been databased by associating the data of the defective part with its cause. As a result, when a defect occurs in the semiconductor integrated circuit, an estimated cause corresponding to the data of the defective portion is retrieved from the database using a computer, and the cause of the defect is estimated. For example, in Patent Document 1, by inputting data of a defective portion of a semiconductor integrated circuit in which a defect has occurred, a defect keyword (for example, pattern defect, foreign matter, flaw, etc.) is inferred from the data, and the defect keyword and In addition, a failure analysis apparatus for estimating the cause of occurrence of a failure from a past case database is described.
JP-A-10-209230 (paragraphs 0026 to 0028, FIGS. 1 to 3)

  However, even if the above-described defect analysis apparatus is used to estimate the cause of the surface defect of the metal plate material, it is a highly accurate cause for the surface defect that occurs periodically, such as a transfer flaw caused by irregularities on the surface of the rolling roll. Cannot be estimated. In other words, even if it is estimated that the surface defects are caused by the rolling rolls, it is not possible to determine where the transfer defects are caused by the rolling rolls installed, so it is not possible to take quick measures. . As described above, it is not easy to divert a failure analysis apparatus for a single product to a continuous product such as a metal plate.

  The present invention provides a cause estimation device, a cause estimation program, and a cause estimation method capable of quickly and accurately estimating the cause when a surface defect occurs in a metal plate during production of the metal plate. Is an issue.

  The invention described in claim 1 of the present invention that has solved the above problems is a metal plate material including a database in which data relating to surface defects of a metal plate material that has occurred in the past and causes corresponding thereto are registered in advance. The apparatus for estimating the cause of surface defects of the present invention, which receives data related to the surface defect of the cause estimation target (hereinafter simply referred to as “estimation target defect”), stores the data, and stored by the storage unit Based on the data related to the estimation target defect, a classification unit that classifies the estimation target defect into a periodic defect and a non-periodic defect that occur periodically in the longitudinal direction of the metal plate material, and the classification unit that classifies the defect For the estimation target defect, a search unit that searches the database for an estimated cause corresponding to the data related to the estimation target defect, and outputs the estimated cause found by the search unit Characterized by comprising a that output section.

  In the cause estimation device according to claim 1, based on the data related to the estimation target defect stored in the storage unit, the classification unit determines that each of the estimation target defects is caused by the same cause based on a predetermined criterion. The surface defects are classified into periodic defects in which the surface defects are periodically generated in the longitudinal direction of the metal plate and other non-periodic defects. In this classification, for example, among the estimation target defects, the estimation target defects in which the predetermined data related to the respective estimation target defects match each other according to a predetermined criterion are extracted. When it is determined by the reference that the metal plate material is generated at a constant interval in the longitudinal direction, it is performed by attaching a periodic defect code to each of the estimation target defects. In this way, after classifying the estimation target defect into a periodic defect and a non-periodic defect, the search unit provides data related to the surface defect that has occurred in the past, that is, surface defect shape data, density data, etc. Since the estimated cause corresponding to the data related to the estimation target defect is searched from the database registered in association with the cause, the cause of the surface defect can be estimated with high accuracy and promptly.

  The invention described in claim 2 is the cause estimation device according to claim 1, wherein the search unit finds an estimated cause as a result of searching for an estimated cause corresponding to data related to the estimation target defect. In the case where it is not possible, data related to the estimation target defect is registered in the database.

  In the cause estimation device according to claim 2, since the data related to the unregistered surface defect is sequentially registered in the database so as to correspond to a predetermined determination item of the surface defect, the newly generated surface It is possible to save the trouble of registering data related to defects.

  The invention described in claim 3 is the cause estimation apparatus according to claim 1 or 2, further comprising a detection unit that detects a surface defect of the metal plate material to be inspected, and is detected by the detection unit. Data regarding the estimation target defect is configured to be transmitted to the storage unit.

  In the cause estimation apparatus according to claim 3, when the detection unit inspects the metal plate material to be inspected and determines that it is a surface defect based on a predetermined standard, the surface defect, that is, data on the estimation object defect. Is transmitted to the storage unit, and the storage unit that has received the data stores the data. Thereby, since all the data regarding the surface defect detected by the detection part are memorize | stored in a memory | storage part, a series of operation | work from the detection of a surface defect to the output of the presumed cause can be performed.

  Further, the invention described in claim 4 includes a database in which data on surface defects of a metal plate material generated in the past and a cause corresponding thereto are registered in advance, and receives data on defects to be estimated, A cause estimation program in a cause estimation device for a surface defect of a metal plate material having a storage unit for storing the data, based on the data on the estimation target defect stored by the storage unit, the estimation target defect, A classification step for classifying periodic defects and non-periodic defects that occur periodically in the longitudinal direction of the metal plate material, and data relating to the estimation target defects from the database for the estimation target defects classified by the classification step A search step for searching for an estimated cause corresponding to the output step, and an output step for outputting the estimated cause found by the search step. Characterized in that to execute the up to the computer.

  According to the cause estimation program according to claim 4, the computer performs an operation of classifying the estimation target defect into a periodic defect and an aperiodic defect based on data on the estimation target defect stored in the storage unit. Let In this classification, for example, among the estimation target defects, the estimation target defects in which the predetermined data related to the respective estimation target defects match each other according to a predetermined criterion are extracted. When it is determined by the reference that the metal plate material is generated at a constant interval in the longitudinal direction, it is performed by attaching a periodic defect code to each of the estimation target defects. In this way, after classifying the estimation target defect into a periodic defect and a non-periodic defect, data relating to a surface defect that has occurred in the past is related to the cause from the database registered in association with the cause. Since the computer is caused to search for an estimated cause corresponding to the data, the cause of the surface defect can be estimated with high accuracy and promptly.

  Further, in the invention described in claim 5, in the cause estimation program according to claim 4, as a result of searching the estimated cause corresponding to the data relating to the estimation target defect in the search step, the estimated cause is found. In a case where the data cannot be registered, the computer is caused to execute a registration step of registering data relating to the estimation target defect in the database.

  According to the cause estimation program according to claim 5, by causing the computer to execute the registration step, the data relating to the unregistered surface defect is stored in the database so as to correspond to a predetermined determination item of the surface defect. Since registration is performed sequentially, it is possible to save the trouble of registering data relating to newly generated surface defects.

  Further, the invention described in claim 6 includes a database in which data on surface defects of a metal plate material generated in the past and a cause corresponding thereto are registered in advance, and receives data on defects to be estimated, A cause estimation method in a surface defect cause estimation device for a metal plate material, including a storage unit that stores the data, based on the data related to the estimation target defect stored by the storage unit, the estimation target defect, A classification step for classifying periodic defects and non-periodic defects that occur periodically in the longitudinal direction of the metal plate material, and data relating to the estimation target defects from the database for the estimation target defects classified by the classification step A search step for searching for an estimated cause corresponding to the output, and an output step for outputting the estimated cause found by the search step; Characterized by comprising.

  According to the cause estimation method of the sixth aspect, in the classification step, the estimation target defect is classified into the periodic defect and the non-periodic defect based on the data related to the estimation target defect stored in the storage unit. In this classification, for example, among the estimation target defects, the estimation target defects in which the predetermined data related to the respective estimation target defects match each other according to a predetermined criterion are extracted. When it is determined by the reference that the metal plate material is generated at a constant interval in the longitudinal direction, it is performed by attaching a periodic defect code to each of the estimation target defects. In this way, after classifying the estimation target defect into a periodic defect and a non-periodic defect, the data relating to the surface defect that occurred in the past by the search step is registered from the database registered in association with the cause. Since the estimated cause corresponding to the data related to the estimation target defect is searched, the cause of the surface defect can be estimated with high accuracy and promptly.

  The invention described in claim 7 is the cause estimation method according to claim 6, wherein the search step finds the estimated cause as a result of searching the estimated cause corresponding to the data related to the estimation target defect in the search step. In the case where the data could not be obtained, a registration step of registering data related to the estimation target defect in the database is provided.

  According to the cause estimation method of claim 7, the registration step sequentially registers the data related to the unregistered surface defect in the database so as to correspond to the predetermined determination item of the surface defect. Thus, it is possible to save the trouble of registering data relating to newly generated surface defects.

  The invention described in claim 8 is the cause estimation method according to claim 6 or claim 7, wherein a detection step of detecting a surface defect of the metal plate material to be inspected and an estimation detected by the detection step A transmission step of transmitting data relating to the target defect to the storage unit.

  According to the cause estimation method according to claim 8, when the metal plate material to be inspected is inspected by the detection step and is determined to be a surface defect according to a predetermined standard, the surface defect, that is, the estimation object defect Is transmitted to the storage unit, and the data is stored in the storage unit. Thereby, since all the data regarding the surface defect detected by the detection step are memorize | stored in a memory | storage part, a series of operation | work from the detection of a surface defect to the output of the presumed cause can be performed.

  According to such a cause estimation apparatus, cause estimation program, and cause estimation method for a surface defect of a metal plate material, the cause of the surface defect can be estimated with high accuracy and promptly. This makes it possible to take measures promptly.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings as appropriate, taking as an example the manufacturing process of an aluminum plate material.

  First, the manufacturing process of an aluminum plate will be described with reference to FIG. For example, an aluminum plate is obtained by homogenizing and heat-treating a slab 1 in which a desired composition is dissolved and cast, and hot hot rolling with hot rough rolling rolls 2 and 2 as shown in FIG. As shown in FIG. 1 (b), the resulting rough bar 3 is hot finish-rolled by a plurality of hot finish rolling rolls 4, and subsequently, a plurality of cold rolling rolls as shown in FIG. 1 (c). After cold rolling with 5, as shown in FIG. 1 (d), it is obtained by intermediate annealing through an annealing furnace 6 and further cold rolling with a plurality of cold rolling rolls 5 [see FIG. 1 (c)]. It is done. Although not shown, a straightening process, a chromate treatment process, and the like may be provided after the final cold rolling process.

  The cause estimation apparatus 10 which is one embodiment of the present invention is installed in front of the winding unit 7 in the final cold rolling step, as shown in FIG. In addition, although the installation position of the cause estimation apparatus 10 is not specifically limited, When using an inspection apparatus as a detection part of the cause estimation apparatus 10, the installation position of the inspection apparatus should be installed before the coil final winding process. preferable.

  Next, details of the cause estimating apparatus 10 will be described with reference to FIG. As shown in FIG. 2, the cause estimating apparatus 10 includes a database 11 in which data related to surface defects of a metal plate material generated in the past and causes corresponding thereto are registered in advance, and the surface of the aluminum plate material to be inspected. A detection unit 12 that detects a defect, a surface defect detected by the detection unit 12, that is, a storage unit 13 that receives data related to an estimation target defect, stores the data, and an estimation target defect stored in the storage unit 13 Based on the data relating to, the classification unit 14 that classifies the estimation target defect into a periodic defect and a non-periodic defect, and the estimation target defect classified by the classification unit 14 from the database 11 A search unit 15 for searching for an estimated cause corresponding to the data and an output unit 16 for outputting the estimated cause found by the search unit 15 are provided. The cause estimation apparatus 10 also includes a database registration unit 17 that registers data on surface defects that have occurred in the past in the database 11 and a control unit 18 that controls the operation of each unit.

  The detection unit 12 detects a surface defect of the aluminum plate material to be inspected and transmits the data to the storage unit 13. When the data transmission is completed, a transmission completion notification is transmitted to the control unit 18. As the detection unit 12, the inspection device 60 (see FIG. 6) described in the background art, the inspection device 70 (see FIG. 7), or the like can be used. For example, when the inspection device 60 is used as the detection unit 12, the inspection standard set in advance is registered in the defect determination unit 64, and the quality is determined by comparing with the surface of the aluminum plate material to be inspected. Perform defect detection. The items for determining surface defects differ depending on the inspection apparatus used, but when the inspection apparatus 60 is used, the position, length, area, shading, etc. of the surface defects are determined as determination items. In addition, about the position and length of a surface defect, it judges about the width direction and longitudinal direction of an aluminum plate, respectively.

  The storage unit 13 receives and stores data on the estimation target defect transmitted from the detection unit 12. The memory 13 can be a computer memory or the like. In addition, when detecting a surface defect by visual observation without using the detection unit 12, the data may be manually input into the storage unit 13 using an input means (not shown) such as a keyboard or a mouse. .

  The classification unit 14 classifies the estimation target defect into a periodic defect and an aperiodic defect based on the data related to the estimation target defect stored in the storage unit 13. When the classification is completed, a classification completion notification is transmitted to the control unit 18.

  The search unit 15 searches the database 11 for an estimated cause corresponding to the data related to the estimation target defect with respect to the estimation target defect classified by the classification unit 14. If the estimated cause can be found, the found estimated cause is transmitted to the output unit 16 together with the data related to the estimation target defect, and the search completion notification is transmitted to the control unit 18. On the other hand, when the estimation cause cannot be found, the data related to the estimation target defect is registered in the database 11 so as to correspond to the predetermined determination item of the surface defect, and the search is completed in the control unit 18. Send a notification. In addition, when the estimation cause cannot be found, data regarding the estimation target defect may be separately registered by the database registration unit 17.

  The output unit 16 outputs data on the estimation target defect transmitted from the search unit 15 and its estimation cause. Then, when the output is completed, an output completion notification is transmitted to the control unit 18. The output unit 16 can be a display, a printer, or the like.

  The control unit 18 includes a central processing unit (CPU) and the like, and controls operations of the detection unit 12, the classification unit 14, the search unit 15, and the output unit 16.

  The database 11 can be a magnetic disk, an optical disk, a magnetic tape, or the like. The database registration unit 17 can be a keyboard, a mouse, or the like. In the database 11, the first table 11 a used for searching for the rolling roll (installation position) estimated to be the cause and the type of surface defect are estimated for the estimation target defect classified as the periodic defect. A second table 11b used when searching for causes and countermeasures is stored. Registration of data in the first table 11a and the second table 11b, change of registered data, and the like are performed by the database registration unit 17.

  Next, the first table 11a and the second table 11b will be described with reference to FIG. FIG. 3A shows an example of the first table 11a, and FIG. 3B shows an example of the second table 11b. 3B shows a case where an inspection apparatus 60 using a CCD camera (see FIG. 6) is used as the detection unit 12 (hereinafter, the case where the inspection apparatus 60 is used as the detection unit 12 will be described. To do).

  As shown in FIG. 3A, the first table 11a has a lower limit value (P1, P3, P5...) And an upper limit value (P2, P4, P6...) Of the periodic defect occurrence interval. A numerical data area 111a to be registered and a text data area 112a to register the name (installation position) of the rolling roll presumed to be the cause corresponding thereto are provided. The generation intervals (P1, P2, P3...) Are not shown in FIG. 3A, but actual numerical values are registered. This generation | occurrence | production space | interval (P1, P2, P3 ...) is decided by the diameter of each rolling roll, and the rolling reduction by each rolling roll. When registering the generation intervals (P1, P2, P3...), Actual numerical values may be registered, or by registering the diameter of each rolling roll and the rolling reduction by each rolling roll. The generation intervals (P1, P2, P3...) May be calculated. In addition, when registering the lower limit value (P1, P3, P5...) And the upper limit value (P2, P4, P6...) Of the generation interval, a certain amount of width is given in consideration of slip of the rolling roll. Preferably.

  As shown in FIG. 3B, the second table 11b includes a numerical data area 111b in which the lower limit value and the upper limit value are registered for the determination items of the surface defects, and the types of surface defects corresponding thereto. And a text data area 112b in which estimated causes and countermeasures are registered. In the numerical data area 111b, among the determination items of the surface defect, the length (width direction and longitudinal direction), area, light intensity, and dark intensity of the surface defect are registered. Note that the light intensity and the dark intensity are both determination items indicating the density of the surface defect, and in the image data of the surface defect, indicate the maximum value and the minimum value of the gradation (0 to 255) indicating the brightness level, respectively. . Further, as will be described later, codes C to G are attached to the respective determination items.

  Next, a cause estimation method using the cause estimation apparatus 10 according to an embodiment of the present invention will be described with reference to FIGS. FIG. 4 is a flowchart showing the operation of the cause estimating apparatus 10. FIG. 5 is a table showing an example of data related to the estimation target defect stored in the storage unit 13. FIG. 5 shows only the position data and the length data among the data relating to the estimation target defect, but other data is actually stored. Although not shown in FIG. 5, numerical values actually measured by the detection unit 12 are stored for each determination item.

[Detection step]
As shown in FIG. 4, the cause estimating apparatus 10 first detects a surface defect of the aluminum plate material to be inspected by the detection unit 12 (see FIG. 2) (detection step S <b> 30). At this time, it is preferable to control in the control unit 18 so that a length (for example, about 100 m) that can detect a periodic defect can be inspected.

[Sending step]
After inspecting the predetermined length of the aluminum plate by the detection unit 12, the position (width direction and longitudinal direction), length (width direction and longitudinal direction), length (width direction and longitudinal direction) of all detected surface defects (defects to be estimated), The area, light intensity, and dark intensity data are transmitted to the storage unit 13 (see FIG. 2) (transmission step S31). At this time, if each judgment item is transmitted with a code (alphabet, etc.), the data can be reliably distinguished in the classification step and the search step described later, so that the cause estimation time can be shortened. I can plan. In the present embodiment, the position of the surface defect in the width direction, the position in the longitudinal direction, the length in the width direction, the length in the longitudinal direction, the area, the light intensity, and the dark intensity are in this order A to G. It shall be sent with a code.

  An example of data relating to the estimation target defect stored in the storage unit 13 in the transmission step S31 will be described with reference to FIG. As shown in FIG. 5A, a table 50 showing an example of data relating to the estimation target defect includes a surface defect number area 51 in which surface defect numbers are accumulated and a periodic defect code area in which periodic defect codes are accumulated. 52, a position data area 53 for storing position data, and a length data area 54 for storing length data. In the surface defect number area 51, numbers assigned to the respective estimation target defects are accumulated. This number is given, for example, in the order detected by the detection unit 12. In the periodic defect code area 52, as will be described later, a periodic defect code that can recognize the generation interval of each of the estimation target defects classified as the periodic defect is accumulated. The position data area 53 includes a width direction position data area 55 in which position data in the width direction (code A) is stored, and a length direction position data area 56 in which position data in the longitudinal direction (code B) is stored. Is provided. Further, in the length data area 54, a width direction length data area 57 in which length direction length data (code C) is stored and a length direction length in which length direction length data (code D) is stored. A data area 58 is provided.

[Classification step]
Subsequently, for all estimation target defects stored in the storage unit 13, the classification unit 14 (see FIG. 2) classifies each estimation target defect into a periodic defect and an aperiodic defect (classification step S32). . The classification is based on a predetermined standard for the data in any of the judgment items of the position in the width direction, the length (width direction and longitudinal direction), the area, the light intensity, and the dark intensity among all the estimation target defects stored. Thus, it is determined that the estimation target defects that match each other are caused by the same cause, and the estimation target defects are extracted. At this time, after collating the codes (A, C to G) attached to the respective determination items, the respective data are compared. Next, regarding the extracted estimation target defects, if it is determined from the respective longitudinal positions (code B) that the estimation target defects are periodically generated according to a predetermined criterion, Classify and attach a periodic defect code to each of the estimated defects.

  For example, in the case of FIG. 5A, for the surface defect numbers 1, 3, 5 and 7, the position in the width direction is X1, the length in the width direction is W1, and the length in the longitudinal direction is L1. Furthermore, if the area, light intensity, and dark intensity also match each other according to a predetermined standard, surface defect numbers 1, 3, 5, and 7 are extracted as shown in FIG. The case where surface defect numbers 1, 3, 5, and 7 are periodic defects and surface defect numbers 2, 4, and 6 are non-periodic defects will be described). Since the longitudinal positions of the surface defect numbers 1, 3, 5, and 7 are periodically generated at the interval P5, the periodic defect code area 52 has a period as shown in FIG. For example, “P5a” is assigned as the sex defect code. On the other hand, the periodic defect code is not attached to the data related to the estimation target defect classified as the non-periodic defect like the surface defect numbers 2, 4, and 6. In the above-described extraction and classification, a certain range is given to the reference (predetermined reference) for comparing the data related to the respective estimation target defects in consideration of the measurement error of the detection unit 12 and the like. Better.

  Subsequently, the search unit 15 (see FIG. 2) selects any one of the estimation target defects stored in the storage unit 13 and searches for an estimated cause. In the search for the estimated cause, first, the search unit 15 determines whether or not the selected estimation target defect is a periodic defect (step S33). If it is a periodic defect (Yes), the process proceeds to the first search step S34, and if it is a non-periodic defect (No), the process proceeds to the second search step S37. Whether or not the defect is a periodic defect is determined based on the presence or absence of the periodic defect code.

[First search step]
In step S33, for the estimation target defect determined to be a periodic defect, the search unit 15 searches for the rolling roll (installation position) estimated to be the cause (first search step S34). The search is performed by comparing the occurrence interval recognized from the periodic defect code with the numerical data registered in the numerical data area 111a of the first table 11a [see FIG. 3 (a)]. And when the rolling roll estimated to be a cause can be discovered (Yes of step S35), it progresses to 1st output step S36. On the other hand, when the rolling roll estimated to be the cause cannot be found (No in step S35), the process proceeds to the second search step S37.

[First output step]
In the first search step S34, the rolling roll estimated to be the cause could be found (Yes in step S35), and the output target 16 (see FIG. 2), together with the data related to the estimation target defect. The name (installation position) of the rolling roll estimated to be the cause is output (first output step S36). For example, in the case of the aforementioned surface defect numbers 1, 3, 5, and 7 (see FIG. 5), since the generation interval is P5, as the rolling roll estimated to be the cause, “hot finish rolling roll (second stage)” Is output. After the output, the process proceeds to the second search step S37.

[Second search step]
Subsequently, each of the length (width direction and longitudinal direction), area, light intensity, and dark intensity data of the selected estimation target defect by the search unit 15 is stored in the second table 11b [see FIG. 3B]. The estimated cause is searched by collating with the numerical data registered in the numerical data area 111b (second search step S37). When collating, after collating the code (CG) attached | subjected to each of the judgment item, each data is collated. If the estimated cause can be found (Yes in step S38), the process proceeds to the second output step S39. On the other hand, if the estimated cause cannot be found (No in step S38), the process proceeds to registration step S40. The “first search step S34” and the “second search step S37” correspond to the “search step” in the claims, and the “rolling roll estimated to be the cause” described above. And “estimated cause” correspond to “estimated cause” in the claims.

[Second output step]
In the second search step S37, the estimated cause was found (Yes in step S38), and the output unit 16 estimated the type of surface defect along with the data related to the estimated target defect. Cause and countermeasure are output (second output step S39). For example, in the data relating to the defect to be estimated, when the length in the width direction and the longitudinal direction are both 1 mm, the area is 1 mm ² , the light intensity is 250, and the dark intensity is 80, the type of surface defect is “dent dimple”. As a presumed cause, “defect transferred by protrusions on the surface of the rolling roll” and “roll replacement” are output as countermeasures. After output, the process proceeds to step S41. The “first output step S36” and the “second output step S39” described above correspond to the “output step” in the claims.

[Registration step]
In the second search step S37, for the estimation target defect for which the estimated cause could not be found (No in step S38), the search unit 15 stores data related to the estimation target defect in the second table 11b. Are registered so as to correspond to the respective determination items (registration step S40), and the process proceeds to step S41. At this time, for example, when a display is used as the output unit 16, the control unit 18 may perform control so that “cause cannot be estimated” can be output on the display screen. In addition, if the cause to be estimated was found by sampling various presumed defects for which the cause could not be found, and the cause was clarified by various analyzers, the cause is stored in the database registration unit 17 (see FIG. 2). It may be registered in the text data area 112b of the table 11b.

  Subsequently, the search unit 15 determines whether or not there is an estimation target defect that has not been searched for an estimated cause (step S41), and if it does not exist, the cause estimation is completed (Yes) and the operation ends. To do. On the other hand, if there is an estimation target defect that has not been searched for the estimation cause (No), the process returns to step S33 and the same operation is performed. When searching for an estimated cause of a periodic defect, control is performed so as to search for only one estimated cause among the estimation target defects classified as the periodic defect in order to shorten the search time. You may control in the part 18. FIG.

  By operating the cause estimation apparatus 10 based on the above steps, a series of operations from detection of surface defects to output of the estimation cause can be performed with high accuracy and speed. In particular, by providing the classification unit 14, it is possible to estimate the cause with high accuracy for the periodic surface defects. In addition, it is also possible to operate the cause estimation apparatus 10 by a program that causes the computer to execute each of the steps described above.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment. For example, in the above-described embodiment, the cause estimation apparatus that receives numerical data related to surface defects and searches for the estimated cause corresponding to the numerical data has been described. However, in addition to numerical data related to surface defects, image data related to surface defects is also received. In addition, a cause estimation device that searches for an estimated cause corresponding to the numerical data and the image data may be used. Moreover, in the said embodiment, although the manufacturing process of the aluminum plate material was demonstrated as an example, even if it uses for the manufacturing process of another metal plate material, for example, a copper plate material, the effect of this invention is exhibited similarly.

  Of the types of surface defects shown in the second table 11b [see FIG. 3 (b)], those that are influenced by seasonal factors, those that occur only in a specific device, etc. In the second table 11b in association with the device or the like, the control unit 18 searches for the type of surface defect (cause of surface defect) by narrowing down and searching for the cause and the device or the like as keywords when estimating the cause. You may control. Thereby, the search time can be shortened.

It is a schematic diagram which shows the manufacturing process of the aluminum plate material to which the cause estimation apparatus which concerns on one Embodiment of this invention was applied. It is a block diagram of the cause estimation apparatus which concerns on one Embodiment of this invention. 2 shows an example of a database used in the present invention, where (a) shows an example of a first table 11a, and (b) shows an example of a second table 11b. It is a flowchart which shows operation | movement of the cause estimation apparatus which concerns on one Embodiment of this invention. (A) is a table which shows an example of the data regarding the estimation object defect memorize | stored in the memory | storage part, (b) is the position of a width direction, length (width direction and longitudinal direction), area in the table of (a). FIG. 5B is a table after extracting estimation target defects that match each other for data of any judgment item according to a predetermined criterion for light intensity and dark intensity, and FIG. This is a table after classification into periodic defects and non-periodic defects. It is a schematic diagram which shows an example of the test | inspection apparatus using a CCD camera. It is a schematic diagram which shows an example of the test | inspection apparatus using a laser beam.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Cause estimation apparatus 11 Database 12 Detection part 13 Storage part 14 Classification part 15 Search part 16 Output part

Claims (8)

A device for estimating the cause of a surface defect of a metal plate, comprising a database in which data related to the surface defect of the metal plate material generated in the past and a cause corresponding thereto are registered in advance,
A storage unit that receives data related to the surface defect of the cause estimation target and stores the data;
Based on the data related to the surface defect of the cause estimation object stored by the storage unit, the surface defect of the cause estimation object is classified into a periodic defect and a non-periodic defect that are periodically generated in the longitudinal direction of the metal plate material. A classification unit to
For the surface defect of the cause estimation target classified by the classification unit, a search unit that searches the database for an estimated cause corresponding to data related to the surface defect of the cause estimation target;
A cause estimation apparatus, comprising: an output unit that outputs an estimated cause discovered by the search unit.   As a result of searching the estimated cause corresponding to the data related to the surface defect of the cause estimation target by the search unit, if the estimated cause cannot be found, the data related to the surface defect of the cause estimation target is the database. The cause estimation apparatus according to claim 1, wherein the cause estimation apparatus is configured to be registered in the system. Provided with a detection unit that detects surface defects of the metal plate material to be inspected,
The cause estimation apparatus according to claim 1 or 2, wherein data relating to a surface defect of a cause estimation target detected by the detection unit is transmitted to the storage unit. It has a database in which data related to surface defects of metal plate materials that occurred in the past and causes corresponding to the data are registered in advance, and a storage unit that receives data related to surface defects to be estimated and stores the data In addition, a cause estimation program in the cause estimation device of the surface defect of the metal plate material,
Based on the data related to the surface defect of the cause estimation object stored by the storage unit, the surface defect of the cause estimation object is classified into a periodic defect and a non-periodic defect that are periodically generated in the longitudinal direction of the metal plate material. A classification step to
For the surface defect of the cause estimation target classified by the classification step, a search step of searching an estimated cause corresponding to data related to the surface defect of the cause estimation target from the database;
A cause estimation program causing a computer to execute an output step of outputting an estimated cause discovered by the search step.   As a result of searching the estimated cause corresponding to the data related to the surface defect of the cause estimation target by the search step, if the estimated cause cannot be found, the data related to the surface defect of the cause estimation target is stored in the database. The cause estimation program according to claim 4, wherein a registration step for registration is executed by a computer. It has a database in which data related to surface defects of metal plate materials that occurred in the past and causes corresponding to the data are registered in advance, and a storage unit that receives data related to surface defects to be estimated and stores the data In addition, a cause estimation method in the cause estimation device of the surface defect of the metal plate material,
Based on the data related to the surface defect of the cause estimation object stored by the storage unit, the surface defect of the cause estimation object is classified into a periodic defect and a non-periodic defect that are periodically generated in the longitudinal direction of the metal plate material. A classification step to
For the surface defect of the cause estimation target classified by the classification step, a search step of searching an estimated cause corresponding to data related to the surface defect of the cause estimation target from the database;
An output step of outputting the estimated cause discovered by the search step.   As a result of searching the estimated cause corresponding to the data related to the surface defect of the cause estimation target by the search step, if the estimated cause cannot be found, the data related to the surface defect of the cause estimation target is stored in the database. The cause estimation method according to claim 6, further comprising a registration step of registration. A detection step for detecting surface defects of the metal plate material to be inspected;
The cause estimation method according to claim 6, further comprising: a transmission step of transmitting data related to the surface defect of the cause estimation target detected by the detection step to the storage unit.

Images