Flaw detection system

Abstract

To perform flaw detection regardless of the cross-sectional shape of a wire.SOLUTION: The flaw detection system 3 for the wire rod 4 to be subjected to wire drawing includes a sensor 21 for measuring the vibration acceleration generated by the wire drawing of the wire rod 4, and a processing device 2 for performing the flaw detection of the wire rod 4 by using the measurement result of the sensor 21, and the processing device 2 includes an acquisition unit 31 for acquiring the measurement result of the vibration acceleration from the sensor 21, an arithmetic unit 32 for calculating a frequency spectrum of the measurement result to obtain a determination target spectrum, and a determination unit 34 for determining whether or not a flaw is generated in the wire rod 4 by using the determination target spectrum.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a flaw detection system. [Background technology]

[0002] Wire drawing can be performed by drawing the wire through a die.

[0003] For example, Patent Document 1 discloses that eddy currents are used to detect flaws in an object to be inspected. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2014-178200 A Summary of the Invention [Problem to be solved by the invention]

[0005] When conducting flaw detection tests on wire rods, rotary eddy current testing or penetration testing using eddy currents is used, as in Patent Document 1. Rotary eddy current testing and penetration testing require the wire rod to pass through a coil, but this requires preparing a coil that matches the cross-sectional shape of the wire rod, making it difficult to perform flaw detection depending on the cross-sectional shape of the wire rod.

[0006] In view of the above problems, an object of the present invention is to provide a flaw detection system that can perform flaw detection regardless of the cross-sectional shape of a wire. [Means for solving the problem]

[0007] In order to solve the above problems, the flaw detection system of the present invention is a flaw detection system for wire that is subjected to wire drawing processing, and includes a sensor that measures vibration acceleration generated by the wire drawing processing of the wire, and a processing device that performs flaw detection on the wire using the measurement results of the sensor, and the processing device includes an acquisition unit that acquires the measurement results of the vibration acceleration from the sensor, a calculation unit that calculates the frequency spectrum of the measurement results to obtain a spectrum to be judged, and a judgment unit that uses the spectrum to be judged to judge whether or not a flaw has occurred in the wire.

[0008] In addition, in the flaw detection system, the judgment unit determines whether the spectrum to be judged is in an abnormal state relative to the normal spectrum, assuming that the frequency spectrum of the vibration acceleration corresponding to the case where no flaws have occurred in the wire, and if it determines that the spectrum to be judged is in an abnormal state, it determines that a flaw has occurred in the wire.

[0009] In addition, in the flaw detection system, the determination unit determines that a flaw has occurred in the wire when the proportion of the determination target spectra determined to be in an abnormal state within a predetermined period is equal to or greater than a threshold value.

[0010] In addition, in the flaw detection system, the judgment unit determines that a plurality of the target spectra that are consecutively judged to be in an abnormal state are a group of abnormal spectra, and if the number of the group of abnormal spectra within a specified period is equal to or greater than a threshold value, it judges that a flaw has occurred in the wire.

[0011] In the flaw detection system, the sensor is installed in a housing having a die for drawing the wire. [Effects of the Invention]

[0012] According to the flaw detection system of the present invention, flaw detection can be performed regardless of the cross-sectional shape of the wire. [Brief explanation of the drawings]

[0013] [Figure 1]1 is a schematic diagram showing an example of the overall configuration of a wire processing system having a flaw detection system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram showing an example of a functional configuration of the processing device. [Figure 3] FIG. 10 is a diagram showing an example of a frequency spectrum corresponding to a case where no flaws are present on the wire. [Figure 4] FIG. 10 is a diagram showing an example of a frequency spectrum corresponding to a case where a flaw occurs in a wire rod. [Figure 5] 10 is a flowchart showing an example of the flow of a learning process by the processing device. [Figure 6] 10 is a flowchart showing an example of the flow of a flaw detection process by the processing device. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. To facilitate understanding of the description, the same components in the drawings will be denoted by the same reference numerals as much as possible, and duplicate descriptions will be omitted where appropriate.

[0015] === Implementation form === <Overall structure> FIG. 1 is a schematic diagram showing an example of the overall configuration of a wire processing system 1 having a flaw detection system 3 according to this embodiment.

[0016] As shown in FIG. 1, the wire processing system 1 includes a wire drawing device 2 and a flaw detection system 3 as main components.

[0017] The wire drawing device 2 is a device that performs wire drawing (plastic processing to reduce the diameter of the wire 4) on the wire 4. The wire 4 is a material to be inspected and processed by the wire drawing process. For example, the wire 4 is pre-hardened. The wire 4 has a cross section that is, for example, a round shape (a perfect circle), but may have an irregular shape other than a round shape. An irregular shape is, for example, an elliptical shape.

[0018] The wire drawing device 2 mainly includes a first shaft 10, a second shaft 11, a die 12, a drive unit 13, and a control panel 14.

[0019] The wire 4 before being drawn is wound around the first shaft 10. The wire 4 is coiled on the first shaft 10. One end of the wire 4 wound around the first shaft 10 passes through a die 12 and extends to the second shaft 11.

[0020] The drawn wire 4 is wound around the second shaft 11. One end of the wire 4 extending from the first shaft 10 is fixed to the second shaft 11. The wire 4 is coiled on the second shaft 11. The first shaft 10 and the second shaft 11 cause the wire 4 to pass through the die 12 continuously.

[0021] The die 12 is a component that performs wire drawing by passing the wire 4 through it. The die 12 has a hole formed therein, through which the wire 4 passes. The hole has a large diameter at the entrance and gradually narrows in diameter. The die 12 is provided within the housing 23 and is fixed in position. The die 12 is fixed relative to the housing 23.

[0022] A lubricant is also supplied to the die 12. The lubricant forms a coating between the die 12 and the wire 4 passing through the holes in the die 12, and smoothly plastically deforms the wire 4. The lubricant is supplied to the die 12, for example, through a supply path extending from the outside to the inside of the housing 23.

[0023] The drive unit 13 drives the second shaft 11. Specifically, the drive unit 13 rotates the second shaft 11 in a direction in which the wire rod 4 is wound. As a result, the wire rod 4 is wound from the first shaft 10 onto the second shaft 11, passing through the die 12 during this process. The first shaft 10 rotates as the wire rod 4 is wound by the second shaft 11. As the wire rod 4 is drawn out of the die 12, the diameter of the wire rod 4 becomes smaller to fit the hole in the die 12, and the shape of the diameter can also be changed. In this way, the wire rod 4 is drawn. After the drawing process, the wire rod 4 is subjected to, for example, annealing.

[0024] The control panel 14 is a device that controls the drive unit 13 to control the wire drawing process. The control panel 14 is connected to, for example, an operation panel, and the user can use the operation panel to make various settings for the wire drawing process. The control panel 14 executes the wire drawing process based on the information set by the user.

[0025] The flaw detection system 3 performs flaw detection (flaw detection inspection) on the wire 4 to be drawn as the material to be inspected. Specifically, the flaw detection system 3 detects flaws (damage) that have occurred in the wire 4 due to the wire drawing.

[0026] The flaw detection system 3 mainly includes a sensor 21 and a processing device 22 .

[0027] The sensor 21 is a vibration acceleration sensor that measures vibration acceleration caused by the drawing of the wire rod 4. The sensor 21 is installed corresponding to the die 12 that draws the wire rod 4. Specifically, the sensor 21 is installed in a housing 23 that has the die 12. The sensor 21 is preferably provided in a position close to the die 12. The sensor 21 may also be installed in the die 12. In this way, the sensor 21 detects the acceleration of vibration caused when the wire rod 4 is drawn in the die 12. The measurement result of the vibration acceleration by the sensor 21 is output to the processing device 22.

[0028] The processing device 22 is an information processing device that performs flaw detection processing on the wire 4. Specifically, the processing device 22 performs flaw detection on the wire 4 using the measurement results of the sensor 21. The processing device 22 includes, for example, a control device, a communication device, a storage device, an operation device, and a display device. The control device is mainly configured with a CPU (Central Processing Unit) and a memory. In the control device, the CPU executes predetermined programs stored in the memory or storage device, etc., thereby functioning as various functional components described below. The communication device includes a communication interface for communicating with external devices. The storage device includes a hard disk, etc., and stores various programs and information required for executing processing in the control device, measurement results, and processing result information. The operation device is a device for operating the processing device 22, such as a keyboard, mouse, or touch panel. The display device is a device for displaying various information and processing results, such as a display or touch panel. The processing device 22 may be configured with a single information processing device or multiple information processing devices. The processing device 22 may also include other components. The processing device 22 can be configured using a CPU unit or an arithmetic unit such as a PC or PLC.

[0029] <Functional configuration> 2 shows an example of various functions of the processing device 22. The processing device 22 includes an acquisition unit 31, a calculation unit 32, a recording unit 33, a determination unit , a control unit 35, and a display unit .

[0030] The acquiring unit 31 acquires the measurement results of the vibration acceleration from the sensor 21. Specifically, the acquiring unit 31 acquires time-series data of the vibration acceleration. For example, the acquiring unit 31 performs A / D conversion (analog-to-digital conversion) on the measurement results (analog data) of the sensor 21 to acquire the digital data. When performing flaw detection processing, it is preferable that the acquiring unit 31 acquires the vibration acceleration in real time while the wire drawing processing is being performed.

[0031] The calculation unit 32 calculates the frequency spectrum of the measurement results obtained from the sensor 21. Specifically, the calculation unit 32 obtains the frequency spectrum of the measurement results of the vibration acceleration for each unit time t. For example, the calculation unit 32 converts the vibration acceleration into a frequency spectrum by a short-time Fourier transform (STFT).

[0032] Fig. 3 is a diagram showing an example of a frequency spectrum corresponding to vibration acceleration when no scratches have been generated in the wire 4 due to the wire drawing process. In Fig. 3, the horizontal axis represents time and the vertical axis represents frequency, and the intensity of each frequency is indicated by shading. For example, when no scratches have been generated in the wire 4, the frequency spectrum has a distribution of high intensity at frequency H1 or lower, and a distribution of high intensity at multiple frequencies (frequencies H2, H3, and H4) in the frequency range higher than frequency H1.

[0033] FIG. 4 is a diagram showing an example of a frequency spectrum corresponding to vibration acceleration when a scratch is generated in the wire rod 4 due to the wiredrawing process. In FIG. 4, the horizontal axis represents time and the vertical axis represents frequency, and the intensity of each frequency is indicated by shading. When a scratch is generated in the wire rod 4, as in FIG. 3, there is a distribution of high intensity below frequency H1, and there is also a distribution of high intensity at multiple frequencies (frequencies H2, H3, and H4). Furthermore, when a scratch is generated in the wire rod 4, the intensity is high over a wide range of frequencies at times T1 and T2. This is due to an increase in vibration acceleration at times T1 and T2. As such, there is a difference in the frequency spectrum between when no scratch is generated in the wire rod 4 due to the wiredrawing process and when a scratch is generated in the wire rod 4 due to the wiredrawing process.

[0034] 2, the recording unit 33 records the frequency spectra calculated by the calculation unit 32. For example, the recording unit 33 records each frequency spectrum in a storage device of the processing device 22. Note that the recording unit 33 may also record the vibration acceleration acquired by the acquisition unit 31.

[0035] The determination unit 34 determines whether or not a flaw has occurred in the wire 4. The determination unit 34 has a preparation mode and a flaw detection mode.

[0036] The preparation mode is a mode in which advance preparation is made for determining the presence or absence of flaws in the flaw detection mode. Specifically, the determination unit 34 performs learning (machine learning) for flaw detection processing in the preparation mode. In the preparation mode, the determination unit 34 acquires a frequency spectrum corresponding to a case in which no flaws have occurred in the wire rod 4 due to the wire drawing process. In the present embodiment, a frequency spectrum corresponding to a vibration acceleration in a case in which no flaws have occurred in the wire rod 4 due to the wire drawing process is referred to as a "normal spectrum." The normal spectrum is a spectrum obtained by acquiring vibration acceleration in the acquisition unit 31 in response to a case in which the determination unit 34 performs the preparation mode and converting the vibration acceleration into a frequency spectrum in the calculation unit 32. Note that the normal spectrum is created using vibration acceleration confirmed by an operator or the like to indicate that no flaws have occurred in the wire rod 4 due to the wire drawing process. The normal spectrum is, for example, a spectrum as shown in FIG. 3. Then, the determination unit 34 performs learning using the normal spectrum. Note that the determination unit 34 preferably performs learning using a plurality of normal spectra. In this way, the determination unit 34 learns a normal pattern of a normal frequency spectrum. When the determining unit 34 has completed learning on the normal spectrum, the preparation mode ends.

[0037] The flaw detection mode is a mode in which the trained determination unit 34 determines whether or not a flaw has occurred in the wire 4 due to the wire drawing process. In the flaw detection mode, the determination unit 34 acquires the frequency spectrum calculated by the calculation unit 32. Then, the determination unit 34 sets the acquired frequency spectrum as the "spectrum to be determined." The spectrum to be determined is a spectrum obtained by acquiring vibration acceleration in the acquisition unit 31 during the wire drawing process and converting it into a frequency spectrum in the calculation unit 32. The spectrum to be determined is a frequency spectrum corresponding to a state in which it is unknown whether or not a flaw has occurred in the wire 4. The spectrum to be determined is, for example, a spectrum as shown in FIG. 3 or a spectrum as shown in FIG. 4.

[0038] In the flaw detection mode, the determination unit 34 uses the target spectrum to determine whether or not a flaw has occurred in the wire 4. Specifically, the determination unit 34 determines (infers) whether or not the target spectrum is in an abnormal state relative to a normal spectrum (normal pattern). If the target spectrum is in an abnormal state, it means that the target spectrum is an outlier, indicating that the target spectrum is in an abnormal state because it is far from the normal spectrum. Note that the determination criteria for determining whether or not the target spectrum is in an abnormal state relative to the normal spectrum can be set appropriately. For example, if the target spectrum is not similar to the normal spectrum learned in the learning mode (if the similarity is equal to or less than a threshold), or if the target spectrum is outside the range set by the multiple learned normal spectra, the determination unit 34 determines that the target spectrum is in an abnormal state.

[0039] In the flaw detection mode, the determination unit 34 determines whether or not a flaw has occurred in the wire 4 based on the determination result of whether or not the target spectrum is in an abnormal state. Specifically, the determination unit 34 determines whether or not a flaw has occurred in the wire 4 using at least one of the determination method M1, the determination method M2, and the determination method M3. Note that the determination unit 34 may determine whether or not a flaw has occurred by combining the determination method M1, the determination method M2, and the determination method M3.

[0040] (Judgment method M1) When the determination unit 34 determines that the spectrum to be determined is in an abnormal state, it determines that a flaw has occurred in the wire 4. That is, the determination unit 34 determines that a flaw has occurred in the wire 4 when the vibration acceleration (frequency spectrum) caused by the wire drawing process is in a state different from the normal vibration acceleration (frequency spectrum).

[0041] (Judgment method M2) The determination unit 34 determines that a flaw has occurred in the wire 4 when the proportion P of determination target spectra determined to be in an abnormal state within the predetermined period T is equal to or greater than a threshold value α. The predetermined period T is set in advance as a period longer than a unit time t. That is, the predetermined period T includes a plurality of determination target spectra. The determination unit 34 determines whether or not each of the plurality of determination target spectra included in the predetermined period T is in an abnormal state. Then, the determination unit 34 calculates the proportion P of determination target spectra determined to be in an abnormal state within the predetermined period T. The proportion P is the ratio (abnormality occurrence rate) of the number of determination target spectra determined to be in an abnormal state to the number (total number) of determination target spectra within the predetermined period T. Then, the determination unit 34 determines that a flaw has occurred in the wire 4 when the calculated proportion P is equal to or greater than a threshold value α.

[0042] (Judgment method M3) The determination unit 34 determines that a flaw has occurred in the wire 4 when the number G of abnormal spectrum groups within the predetermined period T is equal to or greater than a threshold value β. An abnormal spectrum group is a plurality of target spectra that have been consecutively determined to be in an abnormal state. The number of consecutive spectra (the number of consecutive spectra) is set in advance. For example, the number of consecutive spectra is 3. The determination unit 34 determines whether or not each of the target spectra included in the predetermined period T is in an abnormal state. Then, when three consecutive target spectra that are consecutive in time series are determined to be in an abnormal state, the determination unit 34 sets the three target spectra as an abnormal spectrum group. Then, the determination unit 34 determines that a flaw has occurred in the wire 4 when the number G of abnormal spectrum groups (the number of consecutive abnormalities) within the predetermined period T is equal to or greater than a threshold value β.

[0043] In this way, in the flaw detection mode, the determination unit 34 uses the frequency spectrum obtained from the vibration acceleration to perform flaw detection on the wire 4 undergoing wire drawing. The determination result by the determination unit 34 is recorded by the recording unit 33.

[0044] The control unit 35 controls the drawing process of the wire rod 4 based on the determination result by the determination unit 34. Specifically, the control unit 35 stops the wire drawing process when the determination unit 34 determines that a scratch has occurred in the wire rod 4 during the wire drawing process. Specifically, the control unit 35 outputs a stop signal to the control panel 14, and stops the driving of the second shaft 11 by the drive unit 13.

[0045] The display unit 36 ​​displays various pieces of information processed by the processing device 22 on a display device or the like. In particular, the display unit 36 ​​displays the determination results (presence or absence of flaws) of the flaw detection process on the wire 4. The user can recognize the state of flaws in the wire 4 by checking the display of the determination results.

[0046] <Processing flow> FIG. 5 is a flowchart showing an example of the flow of the learning process according to this embodiment. The processes of the following steps are started when a user or the like issues an instruction to start the learning process. The following steps are executed in parallel with the wire drawing process of the wire 4. In the following steps, the unit time t, the predetermined period T (>t), and the number of times N are set in advance. The order and content of the following steps can be changed as appropriate.

[0047] (Step SP10) The acquisition unit 31 starts measuring time, and the process then proceeds to step SP11.

[0048] (Step SP11) The acquisition unit 31 acquires the measurement result of the vibration acceleration from the sensor 21. Then, the process proceeds to step SP12.

[0049] (Step SP12) The calculation unit 32 calculates a frequency spectrum for each unit time t based on the measurement result of the vibration acceleration, and then the process proceeds to step SP13.

[0050] (Step SP13) The acquisition unit 31 determines whether the elapsed time has reached a predetermined time. The predetermined time is the time obtained by multiplying a predetermined period T by the number of times N. If the elapsed time has not reached the predetermined time, the process returns to step SP11. If the elapsed time has reached the predetermined time, the process proceeds to step SP14. As a result, time series data of vibration acceleration and frequency spectrum for the predetermined time (T×N) is acquired.

[0051] (Step SP14) In the preparation mode, the determination unit 34 acquires the calculated frequency spectrum as a normal spectrum and performs learning. As a result, the normal state of the frequency spectrum is learned using a plurality of normal spectra for a predetermined time (T×N).

[0052] The learning process is performed as described above. Note that since the learning is performed for a predetermined time (T×N) of wire drawing, an operator or the like checks that no damage has occurred to the wire 4 during the wire drawing.

[0053] FIG. 6 is a flowchart showing an example of the flow of the flaw detection process according to this embodiment. The processes of the following steps are started when a command to start the flaw detection process is given by a user or the like. The following steps are executed in parallel with the wire drawing process of the wire 4. The following steps may be started automatically when the wire drawing process of the wire 4 is started. Note that the following steps will be described as an example of a case where the determination process is performed by combining the determination method M2 and the determination method M3. The order and content of the following steps can be changed as appropriate.

[0054] (Step SP20) The acquisition unit 31 acquires the measurement results of the vibration acceleration from the sensor 21. The acquisition unit 31 acquires time-series data of the vibration acceleration for a predetermined period of time T. Then, the process proceeds to step SP21.

[0055] (Step SP21) The calculation unit 32 calculates a frequency spectrum for each unit time t for the measurement result of the vibration acceleration. As a result, a frequency spectrum for a predetermined period T is calculated for each unit time t. Then, the process proceeds to step SP22.

[0056] (Step SP22) In the flaw detection mode, the determination unit 34 sets the calculated frequency spectrum as a target spectrum and determines whether or not each target spectrum is in an abnormal state. As a result, the determination unit 34 determines whether or not each of the target spectra for the predetermined period T is in an abnormal state. Then, the process proceeds to step SP23.

[0057] (Step SP23) The determination unit 34 determines whether the proportion P of the determination target spectra determined to be in an abnormal state within the predetermined period T is equal to or greater than a threshold α. If the proportion P is equal to or greater than the threshold α, the process proceeds to step SP27. If the proportion P is not equal to or greater than the threshold α, the process proceeds to step SP24.

[0058] (Step SP24) The judgment unit 34 judges whether the number G of abnormal spectrum groups within the predetermined period T is equal to or greater than the threshold value β. If the number G of abnormal spectrum groups is equal to or greater than the threshold value β, the process proceeds to step SP27. If the number G of abnormal spectrum groups is not equal to or greater than the threshold value β, the process proceeds to step SP25.

[0059] (Step SP25) The determining unit 34 determines that no flaws have occurred in the wire 4. The display unit 36 ​​displays the determination result. Then, the process proceeds to step SP26.

[0060] (Step SP26) The determination unit 34 determines whether or not to continue the flaw detection process. For example, if the wire drawing process has not ended and is continuing, the flaw detection process is continued. If the flaw detection process is to be continued, the process returns to step SP20 and is repeated. If the flaw detection process is not to be continued, the process ends.

[0061] (Step SP27) The determining unit 34 determines that a flaw has occurred in the wire 4. The display unit 36 ​​displays the determination result. Then, the process proceeds to step SP28.

[0062] (Step SP28) The control unit 35 stops the wire drawing process, and the process ends.

[0063] The flaw detection process is performed as described above. In the above example, the determination is made by the determination method M2 in step SP23, and the determination is made by the determination method M3 in step SP24. However, either step SP23 or step SP24 may be omitted.

[0064] By performing the process as described above, the flaw detection process is performed in conjunction with the wire drawing process of the wire 4. Therefore, if a flaw occurs in the wire 4, the device can be stopped early.

[0065] <Action and effect> As described above, in this embodiment, the flaw detection system 3 for a wire 4 undergoing wire drawing processing comprises a sensor 21 that measures the vibration acceleration generated by the wire drawing processing of the wire 4, and a processing device 22 that performs flaw detection on the wire 4 using the measurement results of the sensor 21. The processing device 22 comprises an acquisition unit 31 that acquires the measurement results of the vibration acceleration from the sensor 21, a calculation unit 32 that calculates the frequency spectrum of the measurement results to obtain a spectrum to be judged, and a judgment unit 34 that judges whether or not a flaw has occurred in the wire 4 using the spectrum to be judged.

[0066] According to this configuration, the vibration acceleration caused by the wire drawing process of the wire 4 is measured, and the frequency spectrum is calculated to perform flaw detection. Therefore, it is possible to detect flaws not only in the cross-sectional shape of the wire 4 that is round but also in the cross-sectional shape of an irregular shape. In other words, flaw detection can be performed regardless of the cross-sectional shape of the wire 4. Since flaw detection is possible for various wires 4, flaws caused by the wire drawing process can be efficiently found. This makes it possible to reduce the amount of wire 4 that is discarded due to flaws (scratches rate).

[0067] In addition, in the flaw detection system 3 of this embodiment, the judgment unit 34 determines whether the spectrum to be judged is in an abnormal state relative to the normal spectrum, taking the frequency spectrum of the vibration acceleration corresponding to the case where no flaw has occurred in the wire 4 as the normal spectrum, and if it judges that the spectrum to be judged is in an abnormal state, it judges that a flaw has occurred in the wire 4.

[0068] According to this configuration, it is possible to efficiently perform flaw detection by determining whether or not the spectrum to be determined is in an abnormal state relative to the normal spectrum.

[0069] Furthermore, in the flaw detection system 3 according to this embodiment, the determination unit 34 determines that the wire 4 has a flaw when the proportion P of the determination target spectra determined to be in an abnormal state within a predetermined period is equal to or greater than a threshold value.

[0070] According to this configuration, even if external disturbances or the like affect the spectrum to be judged, the influence of the disturbances or the like can be suppressed and erroneous judgment of flaw detection can be suppressed by basing the judgment on the proportion P of the spectrum to be judged to be in an abnormal state within a specified period.

[0071] Furthermore, in the flaw detection system 3 according to this embodiment, the judgment unit 34 regards a plurality of target spectra that have been consecutively judged to be in an abnormal state as a group of abnormal spectra, and judges that a flaw has occurred in the wire 4 if the number G of groups of abnormal spectra within a specified period is equal to or greater than a threshold value.

[0072] According to this configuration, even when, for example, a disturbance or the like affects the determination target spectrum, the influence of the disturbance or the like can be suppressed and false determination of flaw detection can be suppressed by basing on the number G of abnormal spectrum groups within a predetermined period.

[0073] Further, in the flaw detection system 3 according to the present embodiment, the sensor 21 is installed in a housing 23 having a die 12 for performing wire drawing of the wire 4.

[0074] According to this configuration, the sensor 21 can efficiently measure the vibration acceleration generated by the wire drawing of the wire 4 by being provided in the housing 23 of the die 12.

[0075] <Modification> Note that the present invention is not limited to the above-described embodiment. That is, what those skilled in the art appropriately modify in design in the above-described specific examples is also included in the scope of the present invention as long as it has the features of the present invention. Further, each element included in the above embodiment and the following modification can be combined as far as technically possible, and the combination thereof is also included in the scope of the present invention as long as it includes the features of the present invention.

[0076] For example, in the above embodiment, the determination unit 34 was described by taking as an example the case where the normal spectrum is machine-learned in the preparation mode and it is determined whether the determination target spectrum is in an abnormal state with respect to the normal spectrum in the flaw detection mode. However, the determination unit 34 is not limited to the method using machine learning. For example, it may be determined whether the determination target spectrum is in an abnormal state with respect to the normal spectrum by a method other than machine learning.

[0077] Also, in the flow of FIG. 6, the case where determination is made every predetermined period T was taken as an example, but it may be determined by the determination method M1 every unit time t (<T). For example, when the measurement result of the vibration acceleration is acquired every unit time t (<T) and the frequency spectrum is calculated to obtain the determination target spectrum, and it is determined that the determination target spectrum is in an abnormal state, it is determined that the wire 4 is damaged.

[0078] In the above embodiment, the vibration acceleration caused by the wire drawing process of the wire 4 is measured by the sensor 21 and acquired by the acquisition unit 31. However, an external sensor for measuring the vibration acceleration of the external environment may be further provided. The vibration acceleration of the external environment is the vibration acceleration transmitted to the wire drawing process apparatus 2 from sources other than the wire drawing process apparatus 2. In other words, the vibration acceleration of the external environment is a disturbance to the vibration acceleration caused by the wire drawing process of the wire 4. The external sensor for measuring the vibration acceleration of the external environment is provided, for example, at a position farther from the die 12 than the sensor 21. The acquisition unit 31 acquires the vibration acceleration caused by the wire drawing process from the sensor 21 and also acquires the vibration acceleration of the external environment from the external sensor, and suppresses the vibration acceleration of the external environment from the vibration acceleration caused by the wire drawing process by noise cancellation. Note that the noise cancellation may be performed on the vibration acceleration or after conversion to a frequency spectrum. [Explanation of symbols]

[0079] 3: Flaw detection system 4:Wire rod 21: Sensor 22: Processing equipment 31: Acquisition part 32: Arithmetic section 34: Judgment section

Claims

1. A flaw detection system for a wire rod to be drawn, a sensor for measuring vibration acceleration caused by the wire drawing process of the wire; a processing device that performs flaw detection on the wire rod using the measurement results of the sensor; Equipped with The processing device includes: an acquisition unit that acquires the measurement result of the vibration acceleration from the sensor; a calculation unit that calculates the frequency spectrum of the measurement result and sets it as a spectrum to be determined; a determination unit that determines whether or not the wire is damaged by using the target spectrum; A flaw detection system comprising:

2. the determination unit determines whether the target spectrum is in an abnormal state relative to the normal spectrum, assuming that the frequency spectrum of the vibration acceleration corresponding to the case where no flaw has occurred in the wire, and determines that the wire is damaged when the target spectrum is determined to be in an abnormal state. The flaw detection system according to claim 1 .

3. the determination unit determines that a flaw has occurred in the wire when a ratio of the spectrums to be determined as being in an abnormal state within a predetermined period is equal to or greater than a threshold value. The flaw detection system according to claim 2 .

4. the determination unit determines that a flaw has occurred in the wire rod when the number of the abnormal spectrum groups within a predetermined period of time is equal to or greater than a threshold value, by determining that a plurality of the determination target spectra that have been consecutively determined to be in an abnormal state are an abnormal spectrum group. The flaw detection system according to claim 2 .

5. The sensor is installed in a housing having a die for drawing the wire. The flaw detection system according to claim 1 or 2.

Images