Welding speed detection method, storage medium, control device and manual welding machine

Abstract

The invention relates to a welding speed detection method, a storage medium, control equipment and a manual welding machine. The method comprises the steps that a real-time image of a welding target in the welding process is obtained; the real-time image is analyzed, and the contour edge of the molten pool and the length-width ratio of the contour edge of the molten pool are determined; determining the position deviation between the center of the molten pool and the end part of the welding point according to the real-time image and the contour edge of the molten pool; and the welding speed is determined according to the length-width ratio and the position deviation. According to the technical scheme, the instantaneous welding speed can be accurately detected.

Description

Technical Field

[0001] The present invention relates to the technical field of welding machines, and in particular to a welding speed detection method, a storage medium, a control device and a manual welding machine. Background Art

[0002] During the welding process, heat input is a key factor affecting weld quality. It's typically characterized by heat input Q: Q = (U × I) / v, where U is the welding voltage; I is the welding current; and v is the welding speed. When calculating heat input, welding current and voltage are electrical signals, allowing for accurate measurement using sensors. However, speed, a relative physical quantity, is more difficult to measure, and currently, no sensor can directly measure welding speed.

[0003] Existing methods for measuring welding speed mainly include: 1. Using acceleration sensors to sense and detect the speed, and then converting it into a speed value. However, due to the relatively low welding speed, the accuracy of the acceleration conversion method is poor. 2. Manually measuring the total length of the weld, recording the welding time, and dividing the total length by the time to obtain the welding speed. This method has large errors and the obtained welding speed is an average speed. If a certain section of the weld is welded too fast or too slow during the welding process, it will also affect the welding quality. However, the welding speed of the entire weld may be acceptable. 3. Using arc light or temperature sensing and converting the light / temperature signal into an electrical signal to calculate the welding speed. However, these methods still measure the average welding speed. However, the welding speed value of a smaller distance can be obtained by adjusting the spacing of the sensors. These methods also have high requirements for sensor placement and require additional space in the welding area. This is not convenient for small welding locations or welding workpieces with special shapes. The additional tools and devices may even affect the welder's operation. Currently, there is an urgent need for a solution that can detect instantaneous welding speed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is that, in view of the lack of a technical solution capable of detecting instantaneous welding speed in the prior art, a welding speed detection method, a storage medium, a control device and a manual welding machine are provided.

[0005] The technical solution adopted by the present invention to solve the technical problem is to construct a welding speed detection method, including:

[0006] Acquire real-time images of welding targets during the welding process;

[0007] Analyzing the real-time image to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour;

[0008] Determine the positional deviation between the center of the molten pool and the end of the welding point according to the real-time image and the edge of the molten pool contour;

[0009] A welding speed is determined according to the aspect ratio and the position deviation.

[0010] Preferably, the step of analyzing the real-time image to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour includes:

[0011] The real-time image is recognized by a geometric shape recognition algorithm, and a contour edge that conforms to a preset geometric shape is extracted to obtain a molten pool contour edge;

[0012] Fitting a circumscribed rectangle of the edge of the molten pool contour;

[0013] Determine the length and width of the circumscribed rectangle as the length and width of the edge of the molten pool contour respectively;

[0014] The ratio of the length of the edge of the molten pool contour to the width of the edge of the molten pool contour is calculated to obtain the aspect ratio of the edge of the molten pool contour.

[0015] Preferably, the step of determining the positional deviation between the center of the molten pool and the end of the welding point according to the real-time image and the edge of the molten pool contour comprises:

[0016] Determine the center coordinates of the circumscribed rectangle of the molten pool outline edge as the molten pool center coordinates;

[0017] Acquire the coordinates of the end of the welding point on the real-time image to obtain the end coordinates; wherein the camera device that captures the real-time image is configured to move along with the welding torch of the welding machine;

[0018] The position deviation between the center of the molten pool and the end of the welding point is calculated according to the coordinates of the center of the molten pool and the coordinates of the end.

[0019] Preferably, the step of determining the welding speed according to the aspect ratio and the position deviation comprises:

[0020] Calculating the welding speed according to a set expression, the aspect ratio and the position deviation;

[0021] Wherein, the setting expression is expressed as:

[0022] V=η×f(a / b)+Ψ×f(δ), where V represents the welding speed, η represents the first weighting coefficient, a / b represents the aspect ratio, f(a / b) represents a first preset function for calculating the welding speed based on the aspect ratio, Ψ represents a second weighting coefficient, δ represents the position deviation, and f(δ) represents a second preset function for calculating the welding speed based on the position deviation.

[0023] Preferably, the step of determining the first preset function includes:

[0024] Obtaining the aspect ratios of the molten pools formed when the welding machine operates at multiple welding speeds, and obtaining an aspect ratio speed data set;

[0025] The first preset function is fitted according to the aspect ratio velocity data set.

[0026] Preferably, the step of determining the second preset function includes:

[0027] Obtaining positional deviations between the center of a molten pool and the end of a weld point formed when the welder operates at multiple welding speeds, and obtaining a deviation speed data set;

[0028] The second preset function is fitted according to the deviation speed data set.

[0029] In addition, the present invention also constructs a computer storage medium storing a computer program, which implements the steps of the welding speed detection method described above when the computer program is executed.

[0030] In addition, the present invention also constructs a control device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the welding speed detection method described above when executing the computer program.

[0031] In addition, the present invention also constructs a manual welding machine, comprising:

[0032] A camera device, used to capture real-time images of the welding target during the welding process;

[0033] The control device described above.

[0034] Preferably, the manual welding machine further includes a welding torch, and the camera device is fixed on the welding torch.

[0035] The implementation of the present invention has the following beneficial effects: a welding speed detection method is provided, which first obtains a real-time image of the welding target during the welding process, and then analyzes the real-time image to determine the edge of the molten pool contour and the aspect ratio of the molten pool contour edge. Then, the position deviation between the center of the molten pool and the end of the welding point is determined based on the real-time image and the molten pool contour edge. Finally, the welding speed is determined based on the aspect ratio and the position deviation, thereby realizing the detection of the instantaneous welding speed of the welding machine. It also has the advantages of high detection accuracy and a wide range of applications, and plays a positive role in improving the quality of the welding machine. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0037] Figure 1 is a flowchart of a welding speed detection method in some embodiments of the present invention;

[0038] Figure 2 is a flowchart of step S20 in some embodiments of the present invention;

[0039] Figure 3 is a schematic diagram of the edge of the molten pool contour in some embodiments of the present invention;

[0040] Figure 4 yes Figure 3 Schematic diagram of the two-dimensional coordinate system after rotation by θ;

[0041] Figure 5 is a flowchart of step S30 in some embodiments of the present invention;

[0042] Figure 6 Schematic diagram of the structure of a manual welding machine in some embodiments of the present invention. DETAILED DESCRIPTION

[0043] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described in detail with reference to the accompanying drawings.

[0044] It should be noted that the flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all content and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0045] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0046] Figure 1 This is a flowchart of a welding speed detection method according to some embodiments of the present invention. The welding speed detection method is applied to a welding machine (such as a manual welding machine) and can accurately detect the instantaneous welding speed. Figure 1 The welding speed detection method may include step S10, step S20, step S30 and step S40.

[0047] Step S10 includes: obtaining a real-time image of the welding target during the welding process. In this step, a real-time image of the welding target during the welding process can be captured by a camera device.

[0048] Step S20 includes: analyzing the real-time image to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour.

[0049] In some embodiments, by executing Figure 2 Steps S201 to S204 are shown to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour.

[0050] Step S201 includes: using a geometric shape recognition algorithm to identify the real-time image, extracting the contour edge that conforms to the preset geometric shape, and obtaining the contour edge of the molten pool. It should be noted that during normal welding, the shape of the molten pool is usually elliptical or approximately elliptical, so the preset geometric shape can be elliptical.

[0051] In some embodiments, the geometric shape recognition algorithm may be an existing recognition algorithm that can recognize a preset geometric shape such as an ellipse in an image.

[0052] In another embodiment, the contour edge can be extracted from the real-time image in the following manner: obtaining elliptical images of various shapes; training an elliptical recognition model capable of recognizing elliptical contours based on the elliptical images of various shapes as a training set and an existing neural network algorithm; inputting the real-time image into the trained elliptical recognition model to obtain the contour edge of the molten pool.

[0053] Step S202 includes: fitting a circumscribed rectangle of the edge of the molten pool contour.

[0054] In some embodiments, the circumscribed rectangle of the molten pool contour edge can be fitted by performing the following steps: take the line between the two contour points with the largest distance on the molten pool contour edge as the translation line; determine two points on the molten pool contour edge with the largest vertical distance to the translation line and located on both sides of the translation line to obtain the upper translation point and the lower translation point; translate the translation line to the straight line perpendicular to the translation line to the upper translation point and the lower translation point respectively to obtain two long sides; connect the two end points of one of the long sides with the end point with a closer distance on the other long side to form a circumscribed rectangle.

[0055] Specifically, see Figure 3 , a two-dimensional coordinate system can be established based on the real-time image. Therefore, after determining the edge of the molten pool contour, the coordinates of each contour point on the edge of the molten pool contour can be determined. According to the coordinates of each contour point, the distance between each contour point and other contour points is calculated, and the line between the two contour points corresponding to the largest distance is taken as the translation line L. The inclination angle θ of the translation line L relative to the horizontal axis of the current two-dimensional coordinate system is determined, and the current two-dimensional coordinate system is rotated according to the inclination angle θ so that the translation line L is parallel to the horizontal axis of the rotated two-dimensional coordinate system (see Figure 4Calculate the rotated coordinates of each contour point based on the inclination angle θ; determine the contour point with the smallest ordinate among the rotated coordinates as the lower translation point C, and the contour point with the largest ordinate among the rotated coordinates as the upper translation point D. Then, translate the translation line L along the negative longitudinal axis to the lower translation point C and along the positive longitudinal axis to the upper translation point D, respectively, to obtain the two long sides. Next, connect the two endpoints with equal abscissas in the two long marks to obtain the circumscribed rectangle.

[0056] It should be noted that, given the coordinates of the two contour points connecting the translation line L, tanθ can be calculated using the Pythagorean theorem, and the inclination angle can then be calculated using the inverse tangent function. The method for calculating the rotated coordinates from the inclination angle and the original coordinates can also be referenced by existing algorithms and will not be elaborated here.

[0057] Step S203 includes: determining the length and width of the circumscribed rectangle as the length and width of the edge of the molten pool contour respectively.

[0058] In some embodiments, the translation line is actually equal to the length of the circumscribed rectangle, so the length of the translation line can be determined as the length of the edge of the molten pool outline. For the width, the difference between the ordinate of the upper translation point D and the ordinate of the lower translation point C is determined as the width of the molten pool outline.

[0059] Step S204 includes calculating the ratio of the length of the edge of the molten pool contour to the width of the edge of the molten pool contour to obtain the aspect ratio of the edge of the molten pool contour.

[0060] Step S30 includes: determining the position deviation between the center of the molten pool and the end of the welding point according to the real-time image and the edge of the molten pool contour.

[0061] In some embodiments, by executing Figure 5 Steps S301 to S303 are shown to determine the positional deviation between the center of the molten pool and the end of the welding point.

[0062] Step S301 includes: determining the center coordinates of the circumscribed rectangle of the edge of the molten pool contour as the center coordinates of the molten pool.

[0063] It should be noted that, since the molten pool may be an asymmetric ellipse, that is, the distance of the upper and lower translations of the translation line may be different and the line connecting the lower translation point and the upper translation point is not necessarily perpendicular to the longitudinal axis of the rotated coordinate system, in some embodiments, the coordinates of the center of the molten pool can be calculated by the following method: connect the two contour points (respectively) forming the translation line L and calculate the center coordinates of the molten pool. Figure 4 The average value of the horizontal coordinates of the rotated coordinates of points E and F in the figure is determined as the horizontal coordinate of the center coordinate of the molten pool; the average value of the vertical coordinates of the lower translation point and the vertical coordinates of the upper translation point is determined as the vertical coordinate of the center coordinate of the molten pool.

[0064] Step S302 includes obtaining the coordinates of the weld end on the real-time image to obtain the end coordinates. The camera device capturing the real-time image is configured to move with the welding torch. Specifically, the camera device can be fixed to the welding torch to ensure that the camera device moves with the torch during welding.

[0065] In some embodiments, the welding torch carries the electrode, so the position of the electrode relative to the welding torch is fixed, and the position of the camera device relative to the welding torch is also fixed. If the shooting angle of the camera device remains unchanged, then the position coordinates of the end of the welding point (equivalent to the end of the electrode close to the welding target) in the real-time image are fixed, so the end coordinates can be recorded in a pre-stored manner.

[0066] Furthermore, in other embodiments, the welding torch clamps the welding rod, and the angle at which the welding rod is clamped is uncertain, and the welding rod is consumed during the welding process. Therefore, in this embodiment, the end of the weld point (i.e., the end of the welding rod closest to the welding target) is uncertain. Accordingly, the exact coordinates of the weld point end can be determined by: obtaining a welding rod end image using a pre-trained end recognition model capable of identifying the image of the welding rod end in real-time images; and determining the coordinates of the welding rod end image in a (rotated) two-dimensional coordinate system established by the real-time image based on the pixel coordinates corresponding to the welding rod end image to obtain the end coordinates. Specifically, the neural network model can be trained using a large number of welding rod end images, so that the trained model can accurately extract the welding rod end image in the real-time image.

[0067] Of course, in scenarios where the welding speed detection accuracy is low, the influence of electrode consumption can be ignored. Therefore, when clamping the electrode, the end of the electrode can be made to coincide with the shooting focus of the camera device as much as possible, and the pre-stored focus coordinates can be determined as the center coordinates of the molten pool.

[0068] Step S303 includes calculating a positional deviation between the center of the molten pool and the end of the weld based on the coordinates of the center of the molten pool and the coordinates of the end. Specifically, the abscissa and ordinate differences between the coordinates of the center of the molten pool and the coordinates of the end are calculated, and the positional deviation is calculated based on the Pythagorean theorem, the abscissa and ordinate differences.

[0069] Step S40 includes determining a welding speed according to the aspect ratio and the position deviation.

[0070] In some embodiments, the welding speed may be determined by performing the following steps: calculating the welding speed according to a set expression, the aspect ratio, and the position deviation.

[0071] The setting expression is expressed as:

[0072] V=η×f(a / b)+Ψ×f(δ), where V represents the welding speed, η represents the first weighting coefficient, a / b represents the aspect ratio (a represents the length of the edge of the molten pool contour, and b represents the width of the edge of the molten pool contour), f(a / b) represents the first preset function for calculating the welding speed based on the aspect ratio, Ψ represents the second weighting coefficient, δ represents the position deviation, and f(δ) represents the second preset function for calculating the welding speed based on the position deviation.

[0073] In some embodiments, the first preset function can be determined by: obtaining the aspect ratios of the molten pool formed by the welder at multiple welding speeds to obtain an aspect ratio-speed dataset; and fitting the first preset function based on the aspect ratio-speed dataset. Specifically, each welding speed and the molten pool aspect ratio formed at the corresponding welding speed are combined to form a single first coordinate, i.e., multiple first coordinates can be obtained based on the aspect ratio-speed dataset; and curve fitting is performed on the multiple first coordinates using the least squares method to obtain a first preset function that can characterize the variation pattern of the welding speed and aspect ratio.

[0074] In some embodiments, the second preset function can be determined by obtaining a deviation speed dataset of the positional deviations between the center of the molten pool and the end of the weld point formed at various welding speeds, and fitting the second preset function based on the deviation speed dataset. Specifically, the positional deviations between the center of the molten pool and the end of the weld point formed at each welding speed are used to form a single second coordinate, i.e., multiple second coordinates can be obtained based on the deviation speed dataset. The multiple second coordinates are then curve-fitted using the least squares method to obtain a second preset function that can characterize the variation patterns of welding speed and positional deviations.

[0075] In addition, the first weighting coefficient and the second weighting coefficient can be customized by operating the human-computer interaction unit according to actual conditions. Usually, the first weighting coefficient and the second weighting coefficient can be determined through debugging.

[0076] In another embodiment, the welding speed can also be determined by performing the following steps: inputting the aspect ratio and position deviation into a pre-trained speed calculation model; wherein the speed calculation model is configured to calculate the welding speed based on the input aspect ratio and position deviation. The speed calculation model training step may include: establishing an initial model using a multi-layer perceptron; obtaining a training set, wherein the training set includes multiple welding speeds and the corresponding molten pool center aspect ratio and position deviation between the molten pool center and the weld end; and training the initial model using the training set until a preset training target is achieved, thereby obtaining the speed calculation model. As will be readily understood, this embodiment utilizes a multi-layer perceptron to construct a model capable of accurately calculating the welding speed based on the aspect ratio and position deviation. Compared to the above embodiment, its detection accuracy is higher, but the hardware requirements and cost are relatively higher.

[0077] The present invention also provides a computer storage medium storing a computer program, which implements the steps of the welding speed detection method provided in an embodiment of the present invention when the computer program is run.

[0078] The present invention also provides a control device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the welding speed detection method provided in an embodiment of the present invention are implemented.

[0079] See also Figure 6 The present invention further provides a manual welding machine, comprising a camera 1 and a control device provided by an embodiment of the present invention. The camera 1 is used to capture a real-time image of a welding target during the welding process.

[0080] In some embodiments, the manual welding machine also includes a welding torch 3, and the camera device 1 is fixed on the welding torch. The welding torch 3 is used to carry the electrode or clamp the welding rod, so that the shooting direction of the camera device 1 is coaxial with the welding torch 3, and the direction is basically consistent with the arc direction. Compared with other installation positions, this shooting position basically does not need to consider the changes in the shape of the molten pool caused by the tilt of the electrode or welding rod, which is more convenient for subsequent image processing.

[0081] In some embodiments, the camera device 1 preferably uses an existing miniature camera device with arc resistance and electromagnetic interference resistance to avoid strong arc light and strong electromagnetic signals during welding interfering with the shooting work, which helps to improve the accuracy of image recognition. It should be noted that the miniature camera device has the advantages of small size, good imaging effect, and simple installation process. Therefore, it does not require complex modification of the original welding torch. Compared with other welding speed detection methods, it has the least interference with the original welding equipment and system, and will not cause adverse effects on the welder's operation. Furthermore, due to the small degree of modification, the present invention can be applied to the welding of components with complex shapes and is suitable for welding speed measurement in narrow spaces. Basically, as long as the welder can operate the welding, the welding speed can be detected, and it has a wide range of applications.

[0082] In some embodiments, the manual welding machine may further include a human-computer interaction device for displaying the real-time welding speed and setting the first weighting coefficient and / or the second weighting coefficient.

[0083] It can be understood that the technical solution of the present invention adopts an imaging method to detect the welding speed, and determines the welding speed based on the characteristic that the shape of the molten pool changes with the welding speed. Moreover, the detected welding speed is the instantaneous speed, not the average speed, which can help workers observe the welding speed in real time and is of great significance to improving the welding quality.

[0084] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0085] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0086] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0087] It is understandable that the above embodiments only express the preferred implementation modes of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the patent scope of the present invention. It should be pointed out that for ordinary technicians in this field, without departing from the concept of the present invention, the above technical features can be freely combined, and several deformations and improvements can be made, all of which fall within the scope of protection of the present invention. Therefore, all equivalent changes and modifications made to the scope of the claims of the present invention should fall within the scope of coverage of the claims of the present invention.

Claims

1. A welding speed detection method, characterized in that: include: Acquire real-time images of welding targets during the welding process; Analyzing the real-time image to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour; Determine the positional deviation between the center of the molten pool and the end of the welding point according to the real-time image and the edge of the molten pool contour; A welding speed is determined according to the aspect ratio and the position deviation.

2. The welding speed detection method according to claim 1, characterized in that: The step of analyzing the real-time image to determine the edge of the molten pool contour and the aspect ratio of the edge of the molten pool contour includes: The real-time image is recognized by a geometric shape recognition algorithm, and a contour edge that conforms to a preset geometric shape is extracted to obtain a molten pool contour edge; Fitting a circumscribed rectangle of the edge of the molten pool contour; Determine the length and width of the circumscribed rectangle as the length and width of the edge of the molten pool contour respectively; The ratio of the length of the edge of the molten pool contour to the width of the edge of the molten pool contour is calculated to obtain the aspect ratio of the edge of the molten pool contour.

3. The welding speed detection method according to claim 2, characterized in that: The step of determining the positional deviation between the center of the molten pool and the end of the welding point according to the real-time image and the edge of the molten pool contour comprises: Determine the center coordinates of the circumscribed rectangle of the molten pool outline edge as the molten pool center coordinates; Acquire the coordinates of the end of the welding point on the real-time image to obtain the end coordinates; wherein the camera device that captures the real-time image is configured to move along with the welding torch of the welding machine; The position deviation between the center of the molten pool and the end of the welding point is calculated according to the coordinates of the center of the molten pool and the coordinates of the end.

4. The welding speed detection method according to any one of claims 1 to 3, characterized in that: The step of determining the welding speed according to the aspect ratio and the position deviation comprises: Calculating the welding speed according to a set expression, the aspect ratio and the position deviation; Wherein, the setting expression is expressed as: V=η×f(a / b)+Ψ×f(δ), where V represents the welding speed, η represents the first weighting coefficient, a / b represents the aspect ratio, f(a / b) represents a first preset function for calculating the welding speed based on the aspect ratio, Ψ represents a second weighting coefficient, δ represents the position deviation, and f(δ) represents a second preset function for calculating the welding speed based on the position deviation.

5. The welding speed detection method according to claim 4, characterized in that: The step of determining the first preset function includes: Obtaining the aspect ratios of the molten pools formed when the welding machine operates at multiple welding speeds, and obtaining an aspect ratio speed data set; The first preset function is fitted according to the aspect ratio velocity data set.

6. The welding speed detection method according to claim 4, characterized in that: The step of determining the second preset function includes: Obtaining positional deviations between the center of a molten pool and the end of a weld point formed when the welder operates at multiple welding speeds, and obtaining a deviation speed data set; The second preset function is fitted according to the deviation speed data set.

7. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed, the steps of the welding speed detection method according to any one of claims 1 to 6 are implemented.

8. A control device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the welding speed detection method according to any one of claims 1 to 6 are implemented.

9. A manual welding machine, characterized in that: include: A camera device, used to capture real-time images of the welding target during the welding process; A control device as claimed in claim 8.

10. The manual welding machine according to claim 9, characterized in that: The manual welding machine further comprises a welding torch, and the camera device is fixed on the welding torch.

Images