WO2024095756A1

WO2024095756A1 - Welding state determination method and determination system

Info

Publication number: WO2024095756A1
Application number: PCT/JP2023/037507
Authority: WO
Inventors: 整稲村; 俊祐上垣
Original assignee: パナソニックエナジー株式会社
Priority date: 2022-10-31
Filing date: 2023-10-17
Publication date: 2024-05-10

Abstract

This welding state determination method comprises: a step for acquiring a first image which is an image of a welded spot; a step for measuring the intensity of light being emitted from a welded member that has been irradiated with a laser beam, along the moving direction of the laser beam; a step for creating a second image by converting, into an image, measurements of the intensity of light along the moving direction of the laser beam; and a step for determining the state of the welded spot on the basis of determination images that include the first image and the second image.

Description

Welding condition determination method and determination system

This disclosure relates to a method and system for determining the welding condition.

Traditionally, in order to determine the welding condition of a laser-welded object, a judgment is made as to the quality of the welding condition by estimating the welding strength, etc. based on image data of the welded part and the intensity of the reflected light or plasma light emitted when the object is irradiated with a laser (see, for example, Patent Documents 1 and 2). In recent years, the welding condition has also been judged based on machine learning by artificial intelligence (AI) (see, for example, Patent Document 3). Patent Document 3 describes a method of judging the welding condition by AI processing based on image data of the laser welded part of a plate material captured by a camera.

JP 2000-061672 A JP 2005-095942 A JP 2020-044546 A

When judging the welding condition of a laser-welded object, the accuracy of the judgment can be improved by judging based on the intensity of the reflected light of the irradiated laser, the plasma light generated at the weld, and other factors, along with image data of the welded area. However, conventional judgment methods have issues such as determining that a good product is defective depending on the threshold value, and further improvements in judgment accuracy are required.

The method of determining the welding condition according to the present disclosure is a method of determining the welding condition of a welded portion that has been irradiated with a laser moving relative to the welded members and melted and solidified, and includes the steps of acquiring a first image, which is an image of the welded portion; measuring the intensity of light emitted from the welded members irradiated with the laser along the direction of movement of the laser; imaging processing step of imaging the intensity of the light corresponding to the direction of movement of the laser and creating a second image; and judging the condition of the welded portion based on an image for judgment that includes the first image and the second image.

The welding condition determination system according to the present disclosure is a determination system for determining the welding condition of a welded portion that has been irradiated with a laser moving relative to the welded members and melted and solidified, and includes an image acquisition means for acquiring a first image that is an image of the welded portion, a measurement means for measuring the intensity of light emitted from the welded members irradiated with the laser along the direction of movement of the laser, an imaging processing means for imaging the intensity of the light in accordance with the direction of movement of the laser and creating a second image, and a determination means for determining the condition of the welded portion based on the first image and the second image.

According to the method disclosed herein, the intensity of light emitted from the welded parts irradiated with a laser is visualized and used to judge the welding condition together with an image of the weld, enabling highly accurate judgment. In addition, this visualization makes it easy to use machine learning in artificial intelligence (AI) to judge the weld from multiple perspectives.

1 is a block diagram showing a schematic configuration of a welding condition determination system according to an embodiment; FIG. 2 is a diagram showing an example of a welded portion. 1 is a flowchart illustrating a procedure of a determination method according to an embodiment. FIG. 11 is a schematic diagram for explaining a determination method according to an embodiment;

Below, an example of an embodiment of the welding condition determination method and determination system according to the present disclosure will be described in detail with reference to the drawings. Note that the present disclosure includes configurations that selectively combine multiple embodiments and modified examples described below.

In this embodiment, a battery is exemplified as the welded member. Specifically, the quality of the welded state is determined for the laser welded portion 102 between the sealing body 100 and the lead 101 of a cylindrical battery. However, the subject of the welded state determination method and determination system according to the present disclosure is not limited to the welded portion 102, but may be a laser welded portion of another member constituting a battery, or a laser welded portion other than a battery. The welded state determination method and determination system according to the present disclosure can be widely applied to laser welded portions formed by irradiating a laser (laser light) by moving it relative to the welded member.

FIG. 1 is a block diagram showing a schematic configuration of a welding condition judgment system 10, which is an example of an embodiment. As shown in FIG. 1, the judgment system 10 includes a first processing unit 11, a second processing unit 12, and a third processing unit 13 as a computer that executes processing to judge the quality of the welded portion 102. The judgment system 10 further includes a camera 20 and a sensor 21. The camera 20 is an image acquisition means that acquires an image of the welded portion 102. In this specification, the image captured by the camera 20 is referred to as the first image. The sensor 21 is a light intensity measurement means that measures the intensity of light emitted from the welded parts irradiated with a laser along the direction of movement of the laser.

Although described in more detail below, the judgment system 10 includes an imaging processing means that images the measured value of the intensity of light emitted from the welded parts irradiated with the laser in accordance with the direction of movement of the laser to create a second image, and a judgment means that judges the state of the welded part 102 based on the first and second images. In the judgment system 10, the second processing unit 12 is the imaging processing means, and the third processing unit 13 is the judgment means. The first processing unit 11 imports an image captured by the camera 20, and performs a trimming process to cut out only the portion necessary for judging the quality of the welded part 102.

The determination system 10 includes, for example, three computers. That is, the first processing unit 11, the second processing unit 12, and the third processing unit 13 are configured as separate computers. In the example shown in FIG. 1, the first processing unit 11 and the third processing unit 13 are connected so as to be able to communicate with each other, and the second processing unit 12 and the third processing unit 13 are connected so as to be able to communicate with each other. The computers that make up the determination system 10 may be connected via a local area network (LAN) or via a wide area network (WAN) such as the Internet. It is also possible to realize the functions of each processing unit using one computer. Alternatively, the functions of each processing unit may be realized using four or more computers.

The computer constituting each processing unit includes a storage unit that stores programs for executing the functions of the processing unit, parameters required for calculations and processing, acquired image data, etc., and a calculation unit that reads out the programs and executes the quality judgment of the welded parts and the processing required for said judgment. The calculation unit is composed of a processor such as a central processing unit (CPU). Note that the configuration of the computer constituting each processing unit is not particularly limited as long as it is capable of executing the quality judgment of the welded parts.

The judgment system 10 is attached to, for example, a laser welding device that welds the sealing body 100 and the lead 101. In this case, the welding state can be inspected immediately following the welding process. At least a part of the configuration of the judgment system 10 may be incorporated into the laser welding device, and in this embodiment, the light intensity measuring means is incorporated into the laser head 52 of the laser welding device. In addition, the second processing unit 12 is communicatively connected to the laser oscillator 50, and obtains from the laser oscillator 50 a measured value of the laser output irradiated to the workpieces to be welded.

The laser welding device includes a laser oscillator 50, an optical fiber 51, and a laser head 52. The laser output from the laser oscillator 50 is propagated to the laser head 52 via the optical fiber 51, and is irradiated to a welding point from the laser head 52 arranged in close proximity to the sealing body 100 and the lead 101, which are members to be welded. A fiber laser oscillator is generally used as the laser oscillator 50, but a YAG laser oscillator, a _CO2 laser oscillator, or the like may also be used.

The laser head 52 includes a mirror 53 that reflects the laser output from the laser oscillator 50 in the direction of the workpieces to be welded and transmits the light emitted from the workpieces to be welded. The laser head 52 is generally also provided with a focusing lens, a filter, etc. The laser welding device includes, for example, a drive device that scans at least one of the laser head 52 or a table on which the workpieces to be welded are placed, and a control device that controls the operation of the welding device including the laser oscillator 50, the drive device, etc.

As described above, the judgment system 10 measures the intensity of light emitted from the welded parts irradiated with a laser and visualizes the measured light intensity. It then uses the second image obtained by the visualization together with the first image of the welded part 102 captured by the camera 20 to judge the welding condition. The welding condition is judged for the presence or absence of welding defects, and can be judged by an inspector's visual inspection, but is preferably judged by computerized image analysis, and more preferably by using an artificial intelligence (AI) model. The judgment system 10 visualizes the intensity of light emitted from the welded part 102, thereby providing the third processing unit 13 with data suitable for AI processing.

The judgment system 10 includes a sensor 21 and a spectroscopic unit 22 as light intensity measuring means for measuring the intensity of the light. The light emitted from the workpieces irradiated with a laser includes, for example, plasma light and thermal radiation light. The light emitted from the workpieces irradiated with a laser includes reflected laser light. The judgment system 10 performs imaging processing on at least one, and more preferably all, selected from the plasma light, thermal radiation light, and reflected laser light, and uses the imaging processing to judge the welding condition. The judgment system 10 is provided with a spectroscopic unit 22 that separates these lights. In the example shown in FIG. 1, the sensor 21 and spectroscopic unit 22 are mounted on the laser head 52.

Plasma light generated during laser welding causes the laser to be absorbed and refracted. For this reason, it is thought that the plasma light reduces the energy of the laser irradiated to the welded parts, which affects the penetration depth of the weld, for example. Therefore, in welds where the state of the plasma light fluctuates greatly, there is a possibility that defects such as insufficient penetration depth have occurred. Similarly, the thermal radiation light (near-infrared light) generated from the weld and the reflected light of the laser reflected by the weld are useful for determining the welding condition, and large fluctuations in the thermal radiation light and reflected light allow for the estimation of defects in the weld.

In FIG. 1, one sensor 21 is shown, but the sensor 21 includes a first sensor that receives plasma light and measures its intensity, a second sensor that receives thermal radiation light and measures its intensity, and a third sensor that receives reflected laser light and measures its intensity. The sensor 21 is, for example, a photodiode with a sensitivity range in the wavelength range of the light to be detected, and outputs an electrical signal according to the intensity of the light as detection information. Note that the light intensity measuring means may be a device that integrates a spectroscopic unit and a sensor capable of detecting reflected light of each wavelength.

In the determination system 10, the detection information of the sensor 21 is transmitted to the second processing unit 12, which performs imaging processing of the measured values of the intensity of the plasma light, thermal radiation light, and reflected light. The second processing unit 12 further obtains the measured value of the laser output from the laser oscillator 50, and performs imaging processing of the measured value.

FIG. 2 is a diagram showing the welded portion 102 between the sealing body 100 and the lead 101. As shown in FIGS. 1 and 2, the welded portion 102 is formed by irradiating the surface of the lead 101 with a laser while the lead 101 is placed on the sealing body 100. The lead 101 is, for example, a strip-shaped conductive member connected to the positive electrode, and is made of a metal containing aluminum as a main component. The thickness and width of the lead 101 can be changed as appropriate depending on the size of the battery, etc., but as an example, the thickness is 50 μm to 500 μm and the width is 2 mm to 10 mm. The sealing body 100 is thicker than the lead 101 and includes a metal plate to which the lead 101 is welded. The metal plate is, for example, made of a metal containing aluminum as a main component.

In the example shown in FIG. 2, the welded portion 102 extends parallel to the width direction of the lead 101 and is formed in the shape of a thin line having a substantially constant width. The width of the welded portion 102 is, for example, 1 mm to 4 mm, or 1.5 mm to 3.5 mm. The welded portion 102 is formed when the laser is irradiated onto the welded members while moving relative to the members, and the irradiated portion of the laser melts and solidifies. The relative movement of the laser with respect to the members to be welded is achieved by scanning at least one of the laser and a table on which the members to be welded are placed.

The method for determining the welding condition will be described in detail below with reference to Figures 3 and 4. Figure 3 is a flowchart showing the steps of the determination method of this embodiment.

As shown in FIG. 3, the determination method of this embodiment includes the following steps.
(1) An image acquisition step (S10) of acquiring a first image of the welded portion 102.
(2) A light intensity measurement step (S11) of measuring the intensity of light emitted from the workpieces (sealing body 100 and leads 101) irradiated with the laser along the direction of laser movement and acquiring measurement data.
(3) An imaging processing step (S14) of imaging the measured light intensity in accordance with the laser movement direction to create a second image.
(4) A determination step (S16) of determining the state of the welded portion 102 based on a determination image including the first image and the second image.

The judgment method of this embodiment further includes a step (S15) of combining the first image and the second image to create a combined image. The combined image is used as the judgment image in the judgment step. In the judgment step, it is preferable to judge the state of the welded portion 102 based on learning data obtained by machine learning of the judgment image. By imaging the measured values of the intensity of light emitted from the welded portion 102 and providing the combined image, which is a single piece of data obtained by combining multiple images, to the third processing unit 13, the accuracy of the machine learning is improved, and thus the judgment accuracy of the welded state is improved.

The determination method of this embodiment further includes a step (S13) of normalizing the measured light intensity to create parameters for imaging. In this case, in the imaging processing step of S14, the parameters are used to create a second image. Since the measured values of plasma light, thermal radiation light, and reflected light may differ greatly in scale, it is expected that the accuracy of machine learning will decrease or learning will take a long time if the values are used as is to create an image. For this reason, it is preferable to perform normalization to align the scale of the measurement data, for example.

The judgment method of this embodiment further includes a step (S12) of measuring the output of the laser irradiated to the workpiece along the direction of laser movement. The laser oscillator 50, for example, measures the output of the emitted laser. Alternatively, a laser power meter may be installed in the laser head 52, etc. The measured value of the laser output is normalized in the same manner as the measured value of plasma light, etc., and imaged in accordance with the direction of laser movement. In step S14, a third image based on the measured value of the laser output is created along with the second image. In this case, the judgment image includes the third image as well as the first and second images.

FIG. 4 is a schematic diagram showing the determination method of this embodiment. As shown in FIG. 4, in the determination method of this embodiment, an image (original image) of the welded parts including the welded part 102 is captured, and the intensity of light emitted from the welded part 102 is measured to obtain measurement data. The original image is captured by the camera 20, for example, after the welding process is completed, and the image data is transmitted to the first processing unit 11. Note that the image of the welded part 102 may be acquired in real time during the welding process. The first processing unit 11 performs a trimming process to remove unnecessary parts from the original image and leave only the image of the welded part 102, and creates a first image 31 of the welded part 102.

The measurement data is acquired, for example, in real time during the welding process. The acquired measurement data is then imaged through normalization processing. As described above, the measurement data imaged in the imaging processing step includes measured values of the intensity of plasma light, thermal radiation light, and reflected laser light obtained by dispersing the light emitted from the workpieces irradiated with the laser. The intensity of each light is measured in real time during welding by the sensor 21 and the spectroscopic unit 22 along the direction of laser movement, i.e., along the length of the welded portion 102. The measurement data further includes measured values of the laser output measured in real time along the direction of laser movement.

The normalization and imaging processes for each measurement data are outlined below.
(1) The maximum and minimum values are extracted from the measurement values that make up the measurement data.
(2) A conversion equation is calculated for the measurement values such that the maximum value is 255 (density value: white) and the minimum value is 0 (density value: black).
(3) According to the above conversion formula, each measurement value is converted into a numerical value within the range of 0 to 255. This numerical value within the range of 0 to 255 is a parameter for imaging in step S13 of FIG.
(4) The measurement data is visualized by repeatedly arranging each replaced numerical value (a value in the range of 0 to 255) as a single pixel by assigning a shade of color according to the magnitude of the numerical value. In other words, the measurement data consisting of the measured values of light intensity and laser output is converted into image data.

By performing the above processes (1) to (4) on all the measurement data, a second image 33 based on the measured value of the plasma light, a second image 34 based on the measured value of the thermal radiation light, a second image 35 based on the measured value of the reflected light of the laser, and a third image 36 based on the measured value of the laser output are created. Then, the first image 31, second images 33, 34, 35, and third image 36 of the welded part 102 captured by the camera 20 are combined to create a combined image 37, which is a single image for judgment. When creating an image as wide-range data, it is preferable to perform a pooling process and compress the number of pixels.

Normalization and imaging of the measurement data are performed in the second processing unit 12. The second processing unit 12 obtains the measurement values of each light intensity for measurement points set along the laser movement direction, for example. That is, the measurement values of each light intensity are obtained at the same measurement points as each other. The number of measurement points is not particularly limited, and several thousand to tens of thousands of points may be set in a row along the laser movement direction. Similarly, the measurement values of the laser output are measured at the same points as the measurement points of each light intensity.

In the judgment method of this embodiment, the combined image 37 is sent to the third processing unit 13, and a judgment is made as to whether the welding condition is good or bad. This judgment step is preferably performed based on learning data obtained by machine learning of the combined image 37. The third processing unit 13 judges the quality of the welded part 102 from the acquired combined image 37, for example, using an AI model for images. That is, the combined image 37 is provided as an input value for the AI model. Note that commercially available software for images can be used for the AI model, and conventional AI models can be used as they are. In other words, according to the judgment method of this embodiment, various data with different scales and dimensions are visualized, making it easier to detect abnormalities from a broad perspective, which is AI's specialty.

As shown in FIG. 4, the combined image 37 is configured as a single piece of data in which the first image 31, second images 33, 34, 35, and third image 36 are aligned so that they do not overlap one another. As described above, each image is configured with a number of pixels 38 lined up in a row along the direction of laser movement, and has a length direction and a width direction. FIG. 4 shows the combined image 37 in a schematic manner, and the second images 33, 34, 35, and third image 36 each have, for example, several thousand to tens of thousands of pixels 38 in the length direction of each image that correspond to the measurement points described above. Meanwhile, the width direction is configured with one pixel 38.

The combined image 37 is configured by arranging each image so that the width direction of each image faces the direction X in which the images are lined up. In addition, pixels 38 (imaging parameters) based on numerical data corresponding to the same measurement point of each image are arranged to be lined up in the direction X. In the example shown in FIG. 4, the multiple images that make up the combined image 37 are arranged in the order of a first image 31 captured by the camera 20, a third image 36 of the measured value of the laser output, a second image 35 of the reflected laser light, a second image 33 of the plasma light, and a second image 34 of the thermal radiation light, but the order of the images is not limited to this as long as the AI model can recognize them as a single data. It is also possible to construct the combined image 37 by overlapping each image data.

As described above, it is preferable that the quality of the welded portion 102 is judged using the combined image 37 based on learning data obtained by machine learning of the AI. Any conventionally known method can be applied to the machine learning, and there is no particular limitation to the method. Machine learning also includes deep learning. For example, patterns of good and bad welded portions 102 are input to the third processing unit. Then, the AI's machine learning enables it to judge the quality of the welded state. In the judgment step, the combined image 37, which is one piece of data, is input to the AI model, and the presence or absence of an abnormality is judged based on the learning data.

The AI model sets parameters used in machine learning, such as the image size and number for the convolution process, the selection of the convolution process filter, the selection of the activation process filter, the selection of the thinning process filter, the size and number of thinning processes, the size and number of intermediate layers for the perceptron combination process, the dropout rate, and the number of epochs.Then, the welding condition is judged in real time based on the learning data and the combined image 37 that are the results of the machine learning.

The AI model reads image data for learning, and initially performs machine learning based on preset initial parameters, and then performs machine learning based on parameters calculated by learning. For example, it varies the weighting coefficient, threshold value, offset value, etc. for the numerical value of each pixel 38 of the image, and ultimately determines the weighting coefficient that maximizes the probability of predicting the welding condition. In the case of deep learning, it may include a layer that performs weighting locally, or a layer that performs weighting globally.

The AI model, for example, compares the learning data with the combined image 37 to determine the degree of match with various welding conditions. Then, by weighting the results of the match determination or evaluating them using a threshold, the degree of match with various welding conditions is determined comprehensively, and the welding condition with the highest degree of match is determined.

The effects of the determination method of this embodiment include the following effects.
(1) By visualizing each piece of numerical data, there is no need to use a complex AI model corresponding to each piece of data, and an AI model corresponding only to the image data can be used.
(2) Multiple numerical data can be viewed on the image with uniform density values (black, gray, white). By comparing with surrounding values on the image, it is possible to focus only on the amount of change in each data.
(3) By performing a pooling process on image data, it is possible to narrow down the range of data analysis from a broad range such as monthly, daily, or hourly.
(4) By visualizing numerical data, the data has a two-dimensional positional relationship, making it possible to perform analysis focusing on the information contained in the positional relationship.
(5) As a method for processing image data, various image processing methods have already been developed in addition to AI models, and analysis can be performed using these methods. Depending on the analysis method, it is possible to focus only on gradual changes, only on steep changes, or to remove noise data.
(6) Image data is easily visible to humans, so a wide range of data can be confirmed at a glance.

As described above, according to the judgment method of this embodiment, numerical data such as the intensity of light emitted from the welded parts irradiated with a laser is visualized and used to judge the welding condition together with an image of the weld, making it possible to make a highly accurate pass/fail judgment. In addition, this visualization makes it easy to apply various analytical methods, including AI models. When an AI model is used to make a pass/fail judgment, it is easy to judge the welding condition from multiple perspectives.

In addition, rapid temperature changes and changes in material state occur in laser welds, making it difficult to analyze the weld condition. For this reason, optimization of parameters related to laser welding is based on an accumulation of trial and error. Furthermore, during mass production, welding defects such as spattering and holes can occur due to a variety of complex factors. For this reason, in the judgment method of this embodiment, a wide range of data is acquired and used as input values for the judgment step. On the other hand, one problem with using various data as input values is that the data formats and scales are different. However, by converting this numerical data into images and inputting it into an AI model as image data, it becomes possible to make a multifaceted and highly accurate judgment of the quality of the weld condition.

The present disclosure is further illustrated by the following embodiments.
Configuration 1: A method for determining the welding condition of a welded portion that has been irradiated with a laser that moves relative to the welded members and melted and solidified, the method comprising the steps of: acquiring a first image that is an image of the welded portion; measuring the intensity of light emitted from the welded members irradiated with the laser along the movement direction of the laser; imaging processing step of imaging the measured value of the light intensity corresponding to the movement direction of the laser and creating a second image; and judging the condition of the welded portion based on an image for judgment that includes the first image and the second image.
Configuration 2: The method for determining a welding condition described in Configuration 1, further comprising a step of normalizing the measured values of the light intensity to create parameters for imaging, and in the imaging processing step, the parameters are used to create the second image.
Configuration 3: The determination method according to configuration 1 or 2, further comprising a step of combining the first image and the second image to create the determination image.
Configuration 4: A determination method described in any one of configurations 1 to 3, wherein the measured values of the light intensity imaged in the imaging processing step include measured values of the intensity of plasma light obtained by dispersing the light emitted from the welded members.
Configuration 5: A determination method described in any one of configurations 1 to 4, wherein the measured values of the light intensity imaged in the imaging processing step include measured values of the intensity of thermal radiation light obtained by dispersing the light emitted from the welded members.
Configuration 6: A determination method described in any one of configurations 1 to 5, wherein the measured values of the light intensity imaged in the imaging processing step include measured values of the intensity of the reflected light of the laser obtained by dispersing the light emitted from the welded members.
Configuration 7: The determination method according to any one of configurations 1 to 6, further comprising a step of measuring the output of the laser irradiated to the welded parts along a moving direction of the laser, and in the imaging processing step, the measured value of the laser output is imaged corresponding to the moving direction of the laser to create a third image, and the determination image includes the third image as well as the first image and the second image.
Configuration 8: A judgment method according to any one of configurations 1 to 7, in which the judgment step judges the state of the welded portion based on learning data obtained by machine learning of the judgment image.
Configuration 9: A judgment system for judging the welding condition of a welded portion that has been irradiated with a laser that moves relative to the welded members and melted and solidified, the judgment system for the welding condition comprising: an image acquisition means for acquiring a first image that is an image of the welded portion; a light intensity measuring means for measuring the intensity of light emitted from the welded members irradiated with the laser along the movement direction of the laser; an imaging processing means for imaging the measured value of the light intensity in accordance with the movement direction of the laser and creating a second image; and a judgment means for judging the condition of the welded portion based on the first image and the second image.

REFERENCE SIGNS LIST 10 Determination system, 11 First processing unit, 12 Second processing unit, 13 Third processing unit, 20 Camera, 21 Sensor, 22 Spectroscopic unit, 31 First image, 33, 34, 35 Second image, 36 Third image, 37 Combined image, 38 Pixel, 50 Laser oscillator, 51 Optical fiber, 52 Laser head, 53 Mirror, 100 Sealing body, 101 Lead, 102 Welded portion

Claims

A method for determining a welding state of a welded portion that is melted and solidified by irradiating a laser onto the welded members while the laser moves relative to the welded members, comprising:
acquiring a first image of the weld;
measuring the intensity of light emitted from the workpiece irradiated with the laser along a moving direction of the laser;
an imaging process step of imaging the light intensity measurements in a direction corresponding to the laser movement to produce a second image;
a determination step of determining a state of the welded portion based on a determination image including the first image and the second image;
A method for determining a welding condition comprising:
The method further comprises normalizing the light intensity measurements to generate imaging parameters;
The method for determining a welding condition according to claim 1 , wherein in said imaging processing step, said second image is created using said parameters.
The method of claim 1 further comprises a step of combining the first image and the second image to create the image for determination.
The method of claim 1, wherein the measured values of the light intensity imaged in the imaging process step include measured values of the plasma light intensity obtained by dispersing the light emitted from the welded members.
The method of claim 1, wherein the measured values of the light intensity imaged in the imaging process step include measured values of the intensity of thermal radiation light obtained by dispersing the light emitted from the welded members.
The method of claim 1, wherein the measured values of the light intensity imaged in the imaging process step include measured values of the intensity of the reflected light of the laser obtained by dispersing the light emitted from the workpiece.
The method further includes a step of measuring an output of the laser irradiated to the workpiece along a moving direction of the laser,
In the imaging process, the measured values of the laser output are imaged in a direction corresponding to the laser movement to create a third image;
The method according to claim 1 , wherein the determination images include the third image in addition to the first image and the second image.
The method according to any one of claims 1 to 7, wherein in the judgment step, the condition of the weld is judged based on learning data obtained by machine learning of the judgment image.
A judgment system for judging a welding state of a welded portion that is melted and solidified by irradiating a laser to the welded members while moving the laser relative to the welded members,
an image acquisition means for acquiring a first image which is an image of the welded portion;
a light intensity measuring means for measuring the intensity of light emitted from the workpiece irradiated with the laser along a moving direction of the laser;
an image processing means for imaging the measured light intensity in a direction corresponding to the laser movement direction to generate a second image;
a determination means for determining a state of the welded portion based on the first image and the second image;
A welding condition determination system comprising: