Welding system and welding method

Abstract

To determine welding abnormalities. [Solution] The welding system 100 comprises a welding device 110 that performs welding, an imaging device 120 that images the welding area by the welding device 110, and a processing device 130 that determines welding abnormalities based on the image captured by the imaging device 120. The image is captured between the time when light emission begins at the welding area and before the material melts.

Description

Technical Field

[0001] The present disclosure relates to a welding system and a welding method.

Background Art

[0002] Welding such as arc welding, electron beam welding, and laser beam welding is widely used. Therefore, there is a desire to efficiently perform quality control of welding. For example, Patent Document 1 discloses a technique for reducing the oxygen concentration contained in a purge gas in order to suppress the generation of spatter in laser beam welding in which the purge gas is blown toward the irradiation point of laser light. In the technique of Patent Document 1, an area including the irradiation point is imaged, and in the obtained image, a fume area where fume is likely to be generated is specified, and based on the color of each pixel included in the fume area, it is determined whether the oxygen concentration contained in the purge gas is higher than a reference value.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the technique disclosed in the above Patent Document 1, an abnormality in welding is determined by paying attention to the fume generated along with the melting of the material. Here, a new proposal for determining an abnormality in welding is desired.

[0005] In view of such problems, an object of the present disclosure is to provide a welding system and a welding method capable of determining an abnormality in welding.

Means for Solving the Problems

[0006] To solve the above problems, a welding system according to one aspect of the present disclosure comprises a welding apparatus for performing welding, an imaging apparatus for imaging the welding area by the welding apparatus, and a processing apparatus for determining welding abnormalities based on the image captured by the imaging apparatus, wherein the image is captured between the time when light emission begins at the welding area and before the material melts.

[0007] The processing device may determine an anomaly based on the color of the image.

[0008] The processing device may also include a predictive model for predicting anomalies from images, and may determine anomalies based on the images.

[0009] The system may also include a control device that stops the welding device if the processing device detects an abnormality.

[0010] The system may also be equipped with a notification device that notifies of an abnormality when the processing unit determines that an abnormality has occurred.

[0011] The welding is arc welding, the image shows the plasma at the welding site, and the image may be taken between the start of arc generation at the welding site and before the material melts.

[0012] To solve the above problems, a welding method according to one aspect of the present disclosure includes performing welding with a welding device, imaging the welding site with an imaging device, and determining welding abnormalities based on the image captured by the imaging device, wherein the image is captured between the time when light emission begins at the welding site and before the material melts. [Effects of the Invention]

[0013] According to this disclosure, it becomes possible to determine welding abnormalities. [Brief explanation of the drawing]

[0014] [Figure 1]Figure 1 is a schematic diagram showing the configuration of a welding system according to the first embodiment of this disclosure. [Figure 2] Figure 2 shows an example of an image captured by the imaging device. [Figure 3] Figure 3 is a flowchart showing the processing flow of the welding method according to the first embodiment of this disclosure. [Figure 4] Figure 4 is a flowchart showing an example of an anomaly detection process according to the first embodiment of this disclosure. [Figure 5] Figure 5 shows an example of the spectral sensitivity characteristics of an imaging device. [Modes for carrying out the invention]

[0015] Embodiments of this disclosure will be described in detail below with reference to the attached drawings. The dimensions, materials, and other specific numerical values ​​shown in the embodiments are merely examples for the purpose of facilitating understanding and do not limit this disclosure unless otherwise specified. In this specification and in the drawings, elements having substantially the same function or configuration are denoted by the same reference numerals to avoid redundant explanations. Elements not directly related to this disclosure are omitted from the illustrations.

[0016] [1. First Embodiment] [1.1. Overview of the Welding System] First, an overview of the welding system 100 according to the first embodiment of this disclosure will be described with reference to Figure 1. Figure 1 is a schematic diagram showing the configuration of the welding system 100 according to the first embodiment.

[0017] As shown in Figure 1, the welding system 100 according to this embodiment includes, for example, a welding apparatus 110, an imaging apparatus 120, a processing apparatus 130, a control device 140, and a notification device 150. However, the welding system 100 does not necessarily have to include the control device 140 and the notification device 150.

[0018] The welding device 110 welds the object W. Welding includes, for example, arc welding, electron beam welding, and laser beam welding. Note that welding may also include gouging that melts the object W to form a groove, metal additive manufacturing that stacks melted metal, and the like. In the present embodiment, gas shielded arc welding among arc weldings is taken as an example.

[0019] The welding device 110 includes, for example, a robot 112 and a tool 114.

[0020] The robot 112 holds the tool 114. The robot 112 moves the tool 114 based on a signal from a control device 140 described later. The robot 112 is, for example, a multi-axis robot. For example, the multi-axis robot may be a multi-degree-of-freedom articulated robot. For example, the robot 112 may move the tool 114 with six degrees of freedom including the X direction, Y direction, Z direction, pitch direction, yaw direction, and roll direction.

[0021] The tool 114 is used for welding the object W. The tool 114 includes, for example, a welding torch 116 and a gas supply device 118.

[0022] The welding torch 116 generates an arc discharge and melts the welding point P of the object W. The welding torch 116 includes, for example, an electrode. The tool 114 may be configured to be able to send a filler metal from the welding torch 116 to the welding point P. Note that the tool 114 may be configured not to send a filler metal from the welding torch 116 to the welding point P.

[0023] The gas supply device 118 supplies shielding gas to the welding location P of the object W. The shielding gas is, for example, a gas that does not contain oxygen and nitrogen. The shielding gas consists of one or more of the following: for example, an inert gas such as argon or helium, carbon dioxide, and other gases. In this embodiment, the gas supply device 118 injects the shielding gas in an annular manner from around the welding torch 116. By supplying shielding gas to the welding location P of the object W with the gas supply device 118, the molten metal at the welding location P can be shielded from the atmosphere (air) by the shielding gas.

[0024] The imaging device 120 images the welding location P by the welding device 110. The imaging device 120 is, for example, a camera. When welding is performed by the welding device 110, light is emitted, and the imaging range of the imaging device 120 includes the location where the light is emitted. The imaging range of the imaging device 120 includes, for example, the space between the location where the arc discharge occurs and the welding location P of the object W. The imaging range may also include the location where the arc discharge occurs and the welding location P of the object W. In this embodiment, the imaging range includes, for example, the space between the tip of the welding torch 116 and the welding location P of the object W.

[0025] Figure 2 shows an example of an image captured by the imaging device 120. In arc welding, light emission (arc) occurs when an arc discharge starts around the welding point P. Light emission also occurs around the welding point P in welding methods other than arc welding. For example, in laser welding, although no arc discharge occurs, the gas around the welding point P is excited by the energy from the laser and emits light when it transitions to a lower energy level. Also, in laser welding, scattered light is generated when the gas between the laser source and the welding point P scatters the laser. In electron beam welding, although the welding point P is kept under vacuum, if residual gas is present around the welding point P, accelerated electrons collide with the residual gas molecules, causing the electrons to scatter, and characteristic X-rays (light) are emitted at this time.

[0026] As shown in Figure 2, for example, when an arc is initiated at welding point P, before the material begins to melt, a region IA extending from the welding torch 116 toward the object W emits light between the welding torch 116 and the object W. This region IA is also called the inner arc IA. The inner arc IA is, for example, a roughly conical region whose cross-sectional area widens toward the object W from the welding torch 116. The inner arc IA is formed when the shielding gas is turned into plasma by the arc discharge. If the shielding gas contains impurities, the inner arc IA will be formed when the shielding gas and impurities are turned into plasma by the arc discharge. Therefore, when an arc is initiated at welding point P, the inner arc IA is captured in the image 122A taken by the imaging device 120.

[0027] Subsequently, after the material begins to melt at the welding point P, in addition to the emission of light from the inner arc IA, the region OA surrounding the inner arc IA also emits light. This region OA is also called the outer arc OA. The outer arc OA is, for example, a roughly conical region whose cross-sectional area expands from the welding torch 116 toward the object W. The inner arc IA is located within the outer arc OA. The outer arc OA is formed by the plasma of fumes in addition to the shielding gas. Therefore, after the material begins to melt at the welding point P, the outer arc OA is captured in the image 122B taken by the imaging device 120.

[0028] In this context, the material melted during welding refers to a metallic material, including, for example, the object W (base material) and the filler material. The material melted during welding may also include electrodes.

[0029] Returning to Figure 1, the processing unit 130 includes one or more processors 132 and one or more memories 134 connected to the processors 132. The processors 132 include, for example, a CPU (Central Processing Unit). The memories 134 include, for example, ROM (Read Only Memory) and RAM (Random Access Memory). ROM is a memory element that stores programs and arithmetic parameters used by the CPU. RAM is a memory element that temporarily stores data such as variables and parameters used in processing executed by the CPU.

[0030] The processing device 130 determines welding abnormalities based on images captured by the imaging device 120. In this embodiment, the processing device 130 determines welding abnormalities based on images 122A captured at the welding location P from the time the arc starts until the material melts. The determination of welding abnormalities includes, for example, whether or not an abnormality exists and the nature of the abnormality.

[0031] The processing unit 130 may, for example, determine an anomaly based on the color of the image 122A. In this case, the processing unit 130 may also calculate the emission intensity from the brightness of the pixels in the image 122A, taking into account the camera sensitivity of the imaging device 120. The brightness of an image pixel is correlated with the emission intensity of the object corresponding to that pixel. Therefore, the emission intensity of the object corresponding to that pixel can be estimated based on the brightness of the pixel. However, the relationship between the brightness of a pixel and the emission intensity of the object corresponding to that pixel changes depending on the camera sensitivity. Therefore, in order to accurately estimate the emission intensity of an object, it is necessary to consider the camera sensitivity. For example, the brightness of a pixel is obtained by multiplying the emission intensity of the object corresponding to that pixel by the camera sensitivity. Here, the camera sensitivity differs depending on the wavelength (frequency) of light. In other words, even if the emission intensity is the same, the brightness of the pixel will differ if the frequency is different. Therefore, it is preferable to accurately estimate the emission intensity of an object by taking into account the camera sensitivity according to the frequency. Details of the welding anomaly determination process by the processing unit 130 will be described later.

[0032] The functions of the processing unit 130 may be divided among multiple devices, or multiple functions may be realized by a single device.

[0033] The control device 140 includes one or more processors 142 and one or more memories 144 connected to the processors 142. The processors 142 include, for example, a CPU. The memories 144 include, for example, ROM and RAM. ROM is a memory element that stores programs and arithmetic parameters used by the CPU. RAM is a memory element that temporarily stores data such as variables and parameters used in processing performed by the CPU.

[0034] In this embodiment, the control device 140 controls the entire welding apparatus 110. For example, the control device 140 may stop the welding apparatus 110 if the processing device 130 determines that there is a welding abnormality. Alternatively, the control device 140 may control the welding apparatus 110 and change the welding conditions if the processing device 130 determines that there is a welding abnormality. The welding conditions include, for example, the flow rate of the shielding gas from the gas supply device 118, the injection pressure of the shielding gas from the gas supply device 118, the distance between the welding torch 116 and the object W, the angle of the welding torch 116 relative to the object W, the current value, the voltage value, etc.

[0035] The functions of the control device 140 may be divided among multiple devices, or multiple functions may be realized by a single device.

[0036] Furthermore, the functions of the processing unit 130 and the control unit 140 may be realized by a single device.

[0037] The notification device 150 may also notify of welding abnormalities when the processing device 130 determines that there is an abnormality. The notification device 150 may be, for example, a display device, an audio output device, or a lamp. The display device, which functions as the notification device 150, displays information indicating a welding abnormality when the processing device 130 determines that there is an abnormality. The audio output device, which functions as the notification device 150, outputs information indicating a welding abnormality as an audio when the processing device 130 determines that there is an abnormality. The lamp, which functions as the notification device 150, lights up or flashes when the processing device 130 determines that there is an abnormality.

[0038] [1.2. Overview of Welding Methods] Next, an overview of the welding method according to this embodiment will be described with reference to Figure 3. The welding method according to this embodiment uses, for example, the welding system 100 described above. Figure 3 is a flowchart showing the processing flow of the welding method according to this embodiment. The welding method according to this embodiment is started, for example, when a welding start instruction corresponding to the operator's input is input to the control device 140.

[0039] First, in step S110, the control device 140 determines whether or not a welding stop instruction has been input in response to the operator's input. If it is determined that a stop instruction has been input (YES in step S110), the control device 140 terminates the welding method. On the other hand, if it is determined that no stop instruction has been input (NO in step S110), the control device 140 proceeds to step S120.

[0040] In step S120, the control device 140 first operates the robot 112 of the welding apparatus 110 to move the tool 114 to the welding location P of the object W. Then, the control device 140 operates the tool 114 to start welding at the welding location P of the object W.

[0041] In step S130, the welding location P by the welding device 110 is imaged by the imaging device 120.

[0042] In step S140, the processing device 130 determines welding abnormalities based on the image captured by the imaging device 120. The processing device 130 determines welding abnormalities based on the image 122A captured between the start of light emission at the welding location P and before the material melts. Furthermore, as described above, in this embodiment, the processing device 130 determines abnormalities based on the color of the image 122A.

[0043] Figure 4 is a flowchart showing an example of the abnormality detection process (step S140) according to this embodiment.

[0044] In step S140-1, the processing unit 130 first identifies the region of the inner arc IA in image 122A.

[0045] Then, in step S140-3, the processing device 130 detects the color of one or more pixels contained in the region of the inner arc IA. For example, in step S140-3, the processing device 130 detects the color space value of the pixel. Various existing color spaces can be used, such as the RGB color system, L*a*b* color system, HSV color system, HSL color system, and XYZ color system. Note that if the shielding gas contains impurities above a predetermined value, the color space value of the pixel changes compared to when the impurities are below a predetermined value. Furthermore, the color space value that changes differs depending on the type of shielding gas and the type of impurities. Examples of impurities include oxygen, nitrogen, air, and water (including liquid water and water vapor).

[0046] Furthermore, when using the RGB color system, the processing device 130 may detect at least one of the R component value, G component value, and B component value of one or more pixels included in the inner arc IA region. This makes it possible to efficiently determine welding abnormalities, such as those caused by the inclusion of impurities in the shielding gas.

[0047] For example, when using the RGB color system and argon as the shielding gas, the processing device 130 preferably detects the R component value and the G component value of one or more pixels included in the inner arc IA region, and more preferably detects the R component value.

[0048] Figure 5 shows an example of the spectral sensitivity characteristics of the imaging device 120. In Figure 5, the vertical axis represents relative sensitivity [au], and the horizontal axis represents wavelength [nm]. Also in Figure 5, the solid line represents the spectral sensitivity characteristics of the R component, the dashed line represents the spectral sensitivity characteristics of the G component, and the dashed line represents the spectral sensitivity characteristics of the B component. Assume that the imaging device 120 has, for example, the spectral sensitivity characteristics shown in Figure 5, and that the emission spectra of oxygen and nitrogen coincide with the wavelengths that respond to the R and G components in the imaging device 120. In the image 122A captured by such an imaging device 120, when oxygen is mixed with argon as the shielding gas, the values ​​of the R component and the G component, especially the R component, change in the inner arc IA region. For example, the higher the concentration of oxygen in argon, the larger the values ​​of the R component and the G component. Also, when nitrogen is mixed with argon as the shielding gas, the value of the R component changes in the inner arc IA region. For example, the higher the concentration of nitrogen in argon, the larger the value of the R component. When air is mixed with argon as a shielding gas, the value of the R component changes in the inner arc IA region. For example, the higher the concentration of air in the argon, the larger the value of the R component. Therefore, when argon is used as a shielding gas, the processing device 130 can more efficiently determine the presence of oxygen, nitrogen, and air in the shielding gas by detecting the R component and G component values ​​of one or more pixels contained in the inner arc IA region.

[0049] Furthermore, when using the L*a*b* color system, the processing device 130 may detect at least one of the L* component value, a* component value, and b* component value of one or more pixels included in the inner arc IA region. This allows for efficient determination of welding abnormalities, such as those caused by the inclusion of impurities in the shielding gas. Moreover, when using the L*a*b* color system, it is preferable for the processing device 130 to detect at least one of the a* component value and b* component value of one or more pixels included in the inner arc IA region. This allows for even more efficient determination of welding abnormalities, such as those caused by the inclusion of impurities in the shielding gas.

[0050] For example, when using the L*a*b* color system and using argon as the shielding gas, it is preferable for the processing device 130 to detect the a* component value of one or more pixels included in the inner arc IA region. When oxygen is mixed with the argon used as the shielding gas, the a* component value changes in the inner arc IA region. For example, the higher the concentration of oxygen in the argon, the larger the a* component value. When nitrogen is mixed with the argon used as the shielding gas, the a* component value changes in the inner arc IA region. For example, the higher the concentration of nitrogen in the argon, the larger the a* component value. When air is mixed with the argon used as the shielding gas, the a* component value changes in the inner arc IA region. For example, the higher the concentration of air in the argon, the larger the a* component value. Therefore, when using argon as the shielding gas, the processing device 130 can more efficiently determine the presence of oxygen, nitrogen, and air in the shielding gas by detecting the a* component value of one or more pixels included in the inner arc IA region.

[0051] Furthermore, when using the HSV color system, the processing device 130 may detect at least one of the H component value, S component value, and V component value of one or more pixels included in the inner arc IA region. This allows for efficient determination of welding abnormalities, such as those caused by the inclusion of impurities in the shielding gas. Furthermore, when using the HSL color system, the processing device 130 may detect at least one of the H component value, S component value, and L component value of one or more pixels included in the inner arc IA region. This allows for efficient determination of welding abnormalities, such as those caused by the inclusion of impurities in the shielding gas.

[0052] For example, when using the HSV or HSL color system and using argon as the shielding gas, it is preferable for the processing device 130 to detect the H component value of one or more pixels included in the inner arc IA region. When oxygen is mixed with the argon used as the shielding gas, the H component value changes in the inner arc IA region. For example, the higher the concentration of oxygen in the argon, the closer the H component value approaches 0 degrees. When nitrogen is mixed with the argon used as the shielding gas, the H component value changes in the inner arc IA region. For example, the higher the concentration of nitrogen in the argon, the closer the H component value approaches 0 degrees. When air is mixed with the argon used as the shielding gas, the H component value changes in the inner arc IA region. For example, the higher the concentration of air in the argon, the closer the H component value approaches 0 degrees. Therefore, when using argon as the shielding gas, the processing device 130 can efficiently determine the mixing of oxygen, nitrogen, and air into the shielding gas by detecting the H component value of one or more pixels included in the inner arc IA region.

[0053] Next, in step S140-5, the processing unit 130 compares the color of the pixel detected in step S140-3 with a threshold value related to the color. The threshold value is, for example, expressed as a value in the color space.

[0054] The threshold may be set in advance based on a first image of the inner arc IA before the outer arc OA occurs, obtained using, for example, experiments or simulations, when welding is performed normally. For example, the average value of the color space values ​​of each pixel in the first image may be used as the threshold. In this case, the processing device 130 determines that there is an abnormality in the welding if, for example, the color of the pixel detected in step S140-3 is outside the range of the threshold, and proceeds to step S150.

[0055] Furthermore, the threshold may be set in advance based on a second image of the inner arc IA before the outer arc OA occurs, obtained, for example, using experiments or simulations, in cases where there is an abnormality in the welding. For example, the average value of the color space of each pixel in the second image may be used as the threshold. In this case, the processing device 130 determines that the welding was performed normally if, for example, the color of the pixel detected in step S140-3 is within the threshold range, and proceeds to step S150.

[0056] In step S140-5, the processing unit 130 may compare the average value of the colors of all pixels detected in step S140-3 with a threshold value. Alternatively, the processing unit 130 may compare the average value of the colors of multiple pixels extracted from all pixels detected in step S140-3 with a threshold value. Alternatively, the processing unit 130 may compare the color of a single pixel extracted from all pixels detected in step S140-3 with a threshold value.

[0057] Returning to Figure 3, in step S150, the control device 140 determines whether or not the processing device 130 has determined that there is an abnormality. If the processing device 130 determines that there is an abnormality (YES in step S150), the control device 140 proceeds to step S160. On the other hand, if the processing device 130 does not determine that there is an abnormality, that is, if it determines that there is no abnormality (NO in step S150), the control device 140 returns to step S110.

[0058] In step S160, the control device 140 stops the welding apparatus 110. In step S170, the notification device 150 notifies of the abnormality. Then, the control device 140 terminates the welding method.

[0059] [1.3. Summary] The welding system 100 and welding method according to this embodiment have been described above.

[0060] The welding system 100 according to this embodiment includes a welding apparatus 110 that performs welding, an imaging apparatus 120 that images the welding location P by the welding apparatus 110, and a processing apparatus 130 that determines welding abnormalities based on the image 122A captured by the imaging apparatus 120, wherein the image 122A is captured between the time when light emission begins at the welding location P and before the material melts.

[0061] As described above, when the material melts, an outer arc OA is generated in addition to the inner arc IA. When the welding location P is imaged by the imaging device 120 when the outer arc OA is generated, the resulting image 122B captures not only the light emitted due to the plasma formation of the shielding gas and impurities contained therein, but also the light emitted due to the plasma formation of fumes. Therefore, in the welding system 100 according to this embodiment, the image 122A captured between the start of light emission at the welding location P and before the material melts is used for abnormality detection. In other words, in the welding system 100 according to this embodiment, the image 122A captured in a state where only the inner arc IA, which mainly contains light emitted from the shielding gas and contains almost no light emitted from fumes, is generated is used for abnormality detection. To put it another way, in the welding system 100 according to this embodiment, the image 122A captured of the inner arc IA before the outer arc OA is generated is used for abnormality detection. As a result, the welding system 100 according to this embodiment can detect welding abnormalities, such as welding abnormalities caused by abnormalities in the shielding gas.

[0062] Furthermore, the welding system 100 according to this embodiment can detect abnormalities before the material melts. Therefore, the welding system 100 according to this embodiment can prevent welding that may result in defects.

[0063] Furthermore, the welding system 100 according to this embodiment can also determine welding abnormalities by analyzing image 122A after welding. This allows the welding system 100 according to this embodiment to contribute to improved traceability. In addition, the welding system 100 according to this embodiment can identify the cause of welding abnormalities.

[0064] The processing unit 130 may determine an abnormality based on the color of the image 122A.

[0065] As a result, the welding system 100 according to this embodiment can easily determine welding abnormalities, particularly those caused by abnormalities in the shielding gas. Furthermore, as described above, the color of image 122A, specifically the color of the inner arc IA region in image 122A, changes depending on the type of shielding gas and the type of impurities contained in the shielding gas. Therefore, the welding system 100 according to this embodiment can estimate the type of impurities contained in the shielding gas and the concentration of those impurities by determining the abnormality based on the color of image 122A.

[0066] The welding system 100 may also include a control device 140 that stops the welding apparatus 110 when the processing device 130 determines that there is an abnormality.

[0067] As a result, the welding system 100 according to this embodiment can avoid continuing welding that may result in malfunctions.

[0068] The welding system 100 may also include a notification device 150 that notifies of an abnormality when the processing device 130 determines that an abnormality exists.

[0069] As a result, the welding system 100 according to this embodiment can notify the operator of any abnormalities.

[0070] The welding is arc welding, and image 122A shows the plasma at the welding site P. Image 122A may be taken between the start of arc generation at the welding site P and before the material melts.

[0071] As a result, the welding system 100 according to this embodiment can determine abnormalities in arc welding, particularly welding abnormalities caused by abnormalities in the shielding gas.

[0072] The welding method according to this embodiment includes performing welding with a welding device 110, imaging the welding location P by the welding device 110 with an imaging device 120, and determining welding abnormalities based on the image 122A captured by the imaging device 120, wherein the image 122A is captured between the time when light emission begins at the welding location P and before the material melts.

[0073] The welding method according to this embodiment can achieve the same effects as those achieved by the welding system 100 described above.

[0074] [2. Second Embodiment] In the first embodiment described above, the processing device 130 was shown as an example in which it determines an abnormality based on the color of image 122A. However, if the processing device 130 can determine an abnormality based on image 122A taken between the start of light emission at the welding location P and before the material melts, it may also determine an abnormality based on something other than the color of image 122A.

[0075] In the second embodiment, the processing device 130 determines an anomaly based on a prediction model for predicting an anomaly from an image and the image 122A.

[0076] In this case, the processing unit 130 creates a prediction model in advance and stores it in the memory 134.

[0077] For example, a large amount of data is acquired that links images obtained during a welding procedure with information indicating whether the welding is normal or abnormal. This large amount of data is then used as training data to construct a predictive model using machine learning. By inputting the image 122A obtained during the actual welding procedure into this constructed predictive model, information indicating whether the welding is normal or abnormal is output. In this way, the processing device 130 can determine whether the welding is normal or abnormal based on the predictive model and the image 122A.

[0078] Furthermore, for example, in the training data collected in advance, images may be linked with information indicating whether the welding is normal, abnormal, or potentially abnormal. In this case, the processing device 130 can determine whether the welding is normal, abnormal, or potentially abnormal based on the prediction model and the image 122A.

[0079] Furthermore, for example, in pre-collected training data, images may be linked to welding anomalies (e.g., a value indicating the likelihood that the welding is abnormal). In this case, the processing unit 130 can determine the welding anomaly based on the prediction model and the image 122A.

[0080] Furthermore, existing technologies such as support vector machines can be used as algorithms for creating predictive models.

[0081] Thus, the processing apparatus 130 according to the second embodiment determines an anomaly based on a prediction model for predicting an anomaly from an image and the image 122A.

[0082] As a result, the processing apparatus 130 according to the second embodiment can easily determine welding abnormalities, such as welding abnormalities caused by abnormalities in the shielding gas. Furthermore, if the types of impurities and their concentrations contained in the shielding gas are linked to the training data used to create the prediction model, the processing apparatus 130 according to the second embodiment can estimate the types of impurities and their concentrations contained in the shielding gas.

[0083] While embodiments have been described above with reference to the attached drawings, it goes without saying that this disclosure is not limited to the embodiments described above. It will be obvious to those skilled in the art that various modifications or alterations can be conceived within the scope of the claims, and these will naturally also fall within the technical scope of this disclosure.

[0084] For example, in the first embodiment described above, the welding is arc welding, and image 122A shows the shielding gas at the welding location P. Image 122A is taken between the start of arc generation at the welding location P and before the material melts. However, welding is not limited to this. For example, if the welding is a type of welding other than arc welding (e.g., electron beam welding, laser beam welding), image 122A shows the gas around the welding location P. Image 122A is taken between the start of gas emission around the welding location P and before the material melts.

[0085] Furthermore, in the first embodiment described above, an example was given in which the welding system 100 is equipped with a notification device 150. However, the notification device 150 is not an essential component.

[0086] Furthermore, in the first embodiment described above, an example was given in which the welding system 100 is equipped with a control device 140. However, the control device 140 is not an essential component. For example, if an abnormality is detected by the processing device 130, the welding apparatus 110 may be stopped by the operator. [Explanation of symbols]

[0087] 100 welding systems 110 Welding equipment 120 Imaging device 122A Image 130 Processing Unit 140 Control device 150 Notification device

Claims

1. A welding machine that performs welding, An imaging device for imaging the welding area by the welding device, A processing device that determines the welding abnormality based on the image captured by the imaging device, Equipped with, The aforementioned image is a welding system captured between the time when light emission begins at the welding location and before the material melts.

2. The welding system according to claim 1, wherein the processing apparatus determines the abnormality based on the color of the image.

3. The welding system according to claim 1, wherein the processing apparatus comprises a predictive model for predicting the abnormality from the image, and determines the abnormality based on the image.

4. The welding system according to any one of claims 1 to 3, further comprising a control device that stops the welding apparatus when the processing device determines that the abnormality exists.

5. The welding system according to any one of claims 1 to 3, further comprising a notification device that notifies of the abnormality when the processing device determines that the abnormality exists.

6. The aforementioned welding is arc welding, The aforementioned image shows the plasma at the welding site. The welding system according to any one of claims 1 to 3, wherein the image is captured at the welding location between the start of arc generation and before the material melts.

7. Performing welding using welding equipment, The welding area by the welding apparatus is imaged by an imaging device, Based on the image captured by the aforementioned imaging device, an abnormality in the welding is determined, Includes, The aforementioned image is captured in the welding area between the start of light emission and before the material melts, in a welding method.

Images