A welding inspection system based on multi-point welding process

Abstract

The application provides a kind of welding detection system based on multi-point welding process, including manipulator and detection structure on manipulator, detection structure includes detection frame and ray source, film plate, cleaning piece and driving assembly in detection frame, detection frame is equipped with detection port and switching port, manipulator is also equipped with controller, and controller includes welding database, acquisition module for acquiring image to be detected, analysis module for issuing stop command and normal command, control module for controlling cleaning piece and comparison module for comparing weld coloration;The application has the advantages of accurately analyzing weld qualified grade according to the condition of weld.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of welding, in particular to a welding detection system based on a multi-point welding process. BACKGROUND

[0002] Welding is a manufacturing process that joins metals or other thermoplastic materials by heating, high temperature or high pressure. Current welding methods include gas flame, electric arc, laser, electron beam, friction and ultrasonic waves. The above welding is mainly operated by manual operation. Some manufacturers use automatic equipment to complete welding. However, manual welding mainly relies on work experience and varies from person to person, resulting in low welding efficiency and high cost. Moreover, welding can cause certain harm to the human body. The existing automatic welding, such as linear welding or weaving, is realized by structure intervention. The welding gun always moves in a certain position, which cannot realize accurate positioning or automatic recognition of welding. The welding quality is low, the weld stacking is serious, the welding wire is wasted, the weld of the cut or metal connection is not firm, and other problems exist.

[0003] Currently, there is a method of planning a welding route by budgeting welding points for welding technology, such as the invention patent with the patent name of "Multi-point budgeted welding process" (application number: 2022115970607). The method calculates the demand level of the welding surface or welding cut, and then budgets the point positions of each welding surface to obtain the welding movement trajectory of the welding gun by connecting the points. However, this multi-point budgeted welding method also needs to detect the weld. The usual detection is still original manual detection by coloring, ultrasonic detection or X-ray detection, and the detection is performed after welding. The multi-point budgeted welding of the cut requires n welds to be combined. Problems such as bubbles are easily caused between each weld. If one of the welds has a bubble or is not welded to the edge, the other welds need to be removed, causing waste and failing to meet the purpose of weld detection for multi-point welding. SUMMARY

[0004] In view of the deficiencies in the prior art, the present application aims to provide a welding detection system based on a multi-point welding process. The welding detection system can accurately analyze the weld qualification level according to the condition of the weld.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0006] A kind of welding detection system based on multi-point welding process, including manipulator and the detection structure on manipulator, the detection structure includes detection frame and the ray source in detection frame, film plate, cleaning piece and drive assembly, detection port and switching port are equipped on the detection frame, the manipulator drives detection frame rotation, to make the detection port align the welding piece, the manipulator drives detection frame to move, to make the welding piece enter detection frame, the drive assembly is used to drive the film plate and cleaning piece rotation, to make the film plate and cleaning piece to switching port rotate out of detection frame or to switching port rotate into detection frame, camera is also equipped on the detection frame;

[0007] Controller is also equipped on the manipulator, welding database is included in the controller, coloring reference information is included in the welding database, the coloring reference information reflects the color image after detection liquid is sprayed to qualified weld;

[0008] Welding database also includes acquisition module, analysis module, comparison module and control module;

[0009] The acquisition module obtains the image of weld on the film plate as the image to be detected;

[0010] The analysis module is analyzed according to the image to be detected, if there is abnormality in the image to be detected, more than threshold value, then stop command is sent to manipulator, if there is no abnormality in the image to be detected or there is abnormality, less than threshold value, then normal command is sent to manipulator;

[0011] The control module controls the cleaning piece to spray detection also to weld surface when obtaining normal command after welding the layer of weld surface, and the acquisition module is controlled to obtain the image of weld surface as comparison image after waiting for reflection time;

[0012] The comparison module is compared according to the comparison image and coloring reference information, if it is consistent, then complete command is sent to manipulator, if it is not consistent, then correction command is sent to manipulator.

[0013] Further, receiving module and calculation module are also included in the controller;

[0014] The receiving module obtains the three-dimensional model of welding arm to establish welding piece;

[0015] The calculation module extracts the coordinates of each boundary point of welding piece as first touch point position information according to the three-dimensional model of welding piece, obtains the position of detection frame after rotation angle as undetermined position according to the first touch point position information through avoidance algorithm, and obtains the actual rotation angle of detection frame as undetermined angle value by comparing the undetermined position with actual position;

[0016] The control module controls the detection frame to rotate to the angle value and then to be sleeved on the welding part.

[0017] Further, the first touch point information includes a first welding plate A point coordinate and a second welding plate B point coordinate, and the actual position includes a top C point coordinate of the detection frame.

[0018] The avoidance algorithm is configured to:

[0019]

[0020]

[0021] |α2-α1|=α

[0022] Wherein, α1 is an angle of the detection frame rotating to the to-be-determined position, α2 is an angle between the actual position and the initial position of the detection frame, α is the angle value, hy A , hx A is a height in a y-axis direction and a length in an x-axis direction of the first welding plate A point coordinate, hy B , hx B is a height in a y-axis direction and a length in an x-axis direction of the second welding plate B point coordinate, hy c , hx c is a height in a y-axis direction and a length in an x-axis direction of the C point coordinate.

[0023] Further, the controller further includes a verification module, the verification module obtains coordinates of each boundary point after the detection frame rotates to the to-be-determined position as second touch point information, and compares the second touch point information with the first touch point information through a coincidence calculation formula, if any point coincides, a touch command is sent, and if no point coincides, a normal command is sent.

[0024] The coincidence calculation formula is configured to:

[0025] T1(x n ,y n ,z n )∩T2(x m ,y m ,z m )

[0026] Wherein, T1 is the first touch point, T2 is the second touch point, x n , y n , z n is a coordinate of the first touch point, x m , y m , z m is a coordinate of the second touch point.

[0027] Further, the driving assembly comprises a screw rod, a gear set and a motor, the screw rod is vertically arranged in the detection frame, the motor is arranged on the outer sidewall of the detection frame, the motor is in transmission connection with the screw rod through the gear set, the radiation source is sleeved on the upper part of the screw rod through a bearing, the film plate is fixedly connected to the lower part of the screw rod, the cleaning piece is fixedly connected to the screw rod between the radiation source and the film plate, and the film plate and the cleaning piece are arranged staggeredly.

[0028] Further, the cleaning piece comprises a cleaning strip, the lower surface of the cleaning strip is provided with a plurality of cleaning nozzles which are evenly distributed, one end of the cleaning strip close to the screw rod is provided with a pipeline which is in communication with the external detection liquid through the detection frame, the inner bottom of the detection frame is provided with a collecting groove for collecting the detection liquid, and a sewage pipe for discharging the detection liquid is arranged on the collecting groove.

[0029] Further, the analysis module comprises an analysis submodule, the analysis submodule obtains the actual values of the bubbles, the cracks and the waste residues according to the images to be detected, the welding database further comprises welding defect threshold values and welding defect information, the welding defect threshold values reflect the welding seam grades, the welding defect information comprises bubble reference information, crack reference information and waste residue reference information, the calculation module indexes the corresponding weight proportions in the welding database according to the actual values of the bubbles, the cracks and the waste residues respectively, and obtains the integral through a weight calculation formula according to the weight proportions, and the analysis module compares the integral with the welding defect threshold values, if the integral is greater than or equal to the welding defect threshold values, the analysis module sends a trimming instruction to the control center, and if the integral is less than the welding defect threshold values, the analysis module sends a normal instruction to the control center.

[0030] Further, the weight calculation formula is configured as:

[0031] ∑m i p+m j r+mk s =l

[0032] Wherein, m i , m j , m k --weight proportion, l--integral, p--actual value of bubbles, r--actual value of cracks, s--actual value of waste residues.

[0033] Further, the acquisition module obtains the images of each welding seam after the welding gun welds as the detection images, and the analysis module analyzes the gap between the welding seam and the welding plate or the welding seam and the welding seam according to the detection images, if there is a gap between the two, the analysis module sends a welding point re-budgeting instruction to the welding arm, and if there is no gap between the two, the analysis module sends a normal instruction to the welding arm.

[0034] Further, the controller further comprises an image processing module, the welding database further comprises preset image information, preset brightness values are calculated according to the preset image information, the to-be-inspected image, the comparison image and the detection image in the acquisition module are acquired, corresponding actual brightness values are calculated according to the to-be-inspected image, the comparison image and the detection image respectively, if the actual brightness value is same as the preset brightness value, an image normal signal is sent, if the actual brightness value is different from the preset brightness value, an image abnormal signal is sent and a brightness compensation value is calculated, and the brightness of the actual image information is adjusted according to the brightness compensation value.

[0035] The present application has the advantages that: the detection mechanical arm and the matched detection frame are arranged, the mechanism for ray detection and coloring detection is arranged in the detection frame, the matched detection of multi-point budget welding can be realized, specifically, the detection frame is accurately sleeved on the to-be-welded part through detection calculation, the ray detection can accurately analyze whether the weld of each pass welded by the welding gun has defects, whether the weld is qualified is judged according to the calculation of the defect weight, if not, timely repair or destruction is performed to avoid waste caused by subsequent welding, when the welding of one layer of welding surface is completed, the coloring detection can be performed integrally, if not qualified, timely repair or destruction is performed, and the detection precision of the weld is improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 It is the overall structure diagram of the present application;

[0037] Figure 2 It is the control diagram of the present application;

[0038] Figure 3 It is the schematic view of the detection frame and the to-be-welded part in the present application;

[0039] Figure 4 It is the schematic view of the cleaning part in the present application;

[0040] Figure 5 It is the avoidance principle diagram when the included angle of the to-be-welded part is a right angle in the present application;

[0041] Figure 6 It is the avoidance principle diagram when the included angle of the to-be-welded part is an acute angle in the present application.

[0042] The reference signs: 1, mechanical arm; 2, detection frame; 3, ray source; 4, film plate; 5, cleaning part; 6, screw rod; 7, detection port; 8, switching port; 9, collection groove; 101, acquisition module; 102, analysis module; 103, comparison module; 104, control module; 105, receiving module; 106, calculation module; 107, verification module; 108, analysis sub-module; 109, image processing module. DETAILED DESCRIPTION

[0043] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Identical components are denoted by the same reference numerals. It should be noted that the terms "front," "rear," "left," "right," "upper," and "lower" used in the following description refer to directions in the accompanying drawings, and the terms "bottom surface," "top surface," "inner," and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0044] Existing multi-point budget welding estimates the welding trajectory of the welding torch by connecting the points on each welding surface. However, this method still requires weld inspection, typically done manually using dye penetrant testing, ultrasonic testing, or radiographic testing, all performed after welding is complete. Since multi-point budget welding requires n weld passes, issues such as air bubbles can easily arise between each pass. If one weld pass has air bubbles or fails to reach the edge, other weld passes need to be removed, resulting in waste and failing to meet the weld inspection requirements of multi-point welding. Therefore, this invention designs a welding inspection system based on multi-point welding technology, the specific structure of which is as follows... Figure 1 , Figure 3 and Figure 4 As shown, the system includes a robotic arm 1 and a detection structure mounted on the robotic arm 1. The detection structure includes a detection frame 2 and a radiation source 3, a film plate 4, a cleaning component 5, and a drive assembly located within the detection frame 2. The detection frame 2 has a detection port 7 and a switching port 8. The robotic arm 1 drives the detection frame 2 to rotate so that the detection port 7 is aligned with the workpiece to be welded. The robotic arm 1 also drives the detection frame 2 to move so that the workpiece to be welded enters the detection frame 2. The drive assembly is used to drive the film plate 4 and the cleaning component 5 to rotate so that the film plate 4 and the cleaning component 5 rotate out of the detection frame 2 or into the detection frame 2 via the switching port 8. The detection frame 2 is also equipped with a camera. Specifically, the drive assembly includes a screw 6, a gear set, and a motor. The screw 6 is vertically located inside the detection frame 2. The motor is located on the outer wall of the detection frame 2. The motor and the screw 6 are connected by a gear set. The X-ray source 3 is sleeved on the upper part of the screw 6 through a bearing. The film plate 4 is fixedly connected to the lower part of the screw 6. The cleaning component 5 is fixedly connected to the screw 6 between the X-ray source 3 and the film plate 4. The film plate 4 and the cleaning component 5 are staggered. The cleaning component 5 includes a cleaning strip. The lower surface of the cleaning strip is provided with several evenly distributed cleaning nozzles. The end of the cleaning strip near the screw 6 is provided with a pipe that passes through the detection frame 2 and communicates with the external detection liquid. The inner bottom of the detection frame 2 is provided with a collection tank 9 for collecting the detection liquid. The collection tank 9 is provided with a drain pipe for discharging the detection liquid.

[0045] Inspection Principles: 1. X-ray Inspection: When the robotic arm 1 drives the inspection frame 2 to be aligned and fitted over the workpiece to be welded, the welding torch uses a multi-point reference to calculate the weld seam points, and the lines connecting these points form the weld seam trajectory. X-ray inspection is mainly for the inspection of single weld seams. The X-ray source 3 inside the inspection frame 2 projects the inspection light perpendicularly onto the weld seam, which is then received by the film plate 4. An image is formed on the film plate 4, and this image is analyzed by the system. If a defect occurs in the weld seam, it will appear as an anomaly in the image. 2. Colorimetric Inspection: Since section welding often requires multiple layers of welding, the welding torch uses multiple... The point budget has been budgeted for the location of the cut and the number of welding layers of the cut has been planned. After the welding of one layer is completed, the motor drives the gear set to rotate, and the screw 6 rotates to rotate the film plate 4 to the outside of the switching port 8. At this time, the cleaning strip is transferred from the switching port 8 into the detection frame 2. First, the penetrant is sprayed on the weld surface. After a certain time, clean water is sprayed on the weld for cleaning. Then, the adsorbent is sprayed on the cleaned weld surface for a certain time. The camera takes pictures of the weld surface and records them, and compares them with the system library. The cleaning fluid and detection fluid on the weld will fall into the collection tank 9 and be discharged by the drain pipe.

[0046] Control section such as Figure 2 As shown:

[0047] The robotic arm 1 is also equipped with a controller, which includes a welding database. The welding database includes color reference information, which reflects the color image of the qualified weld after the detection liquid is sprayed.

[0048] The welding database also includes an acquisition module 101, an analysis module 102, a comparison module 103, and a control module 104;

[0049] The acquisition module 101 acquires an image of the weld seam on the film plate 4 as the image to be inspected;

[0050] The analysis module 102 analyzes the image to be inspected. If there is an abnormality in the image to be inspected that exceeds the threshold, a stop command is sent to the robot arm 1. If there is no abnormality in the image to be inspected or if there is an abnormality but it does not exceed the threshold, a normal command is sent to the robot arm 1.

[0051] When the control module 104 receives the normal command after the welding of the layer, it controls the cleaning component 5 to spray the detection onto the welding surface. After waiting for the reaction time, it controls the acquisition module 101 to acquire the image of the welding surface as a comparison image.

[0052] The comparison module 103 compares the comparison image with the coloring reference information. If they match, it sends a completion command to the robot arm 1. If they do not match, it sends a correction command to the robot arm 1.

[0053] This invention uses radiographic testing to accurately analyze whether there are defects in each weld after welding. Based on the calculation of defect weight, it determines whether the weld is qualified. If it is not qualified, it is repaired or destroyed in time to avoid waste caused by subsequent welding. After one layer of weld is completed, it can be inspected as a whole by colorimetric testing. Based on the colorimetric judgment, if it is not qualified, it is repaired or destroyed in time, thus improving the inspection accuracy of weld.

[0054] Since it is difficult to accurately fit the detection frame 2 onto the workpiece to be welded, and the angle needs to be estimated, the controller also includes a receiving module 105 and a calculation module 106.

[0055] The receiving module 105 acquires the welding arm to establish a three-dimensional model of the workpiece to be welded;

[0056] The calculation module 106 extracts the coordinates of each boundary point of the workpiece to be welded based on the three-dimensional model of the workpiece to be welded as the first contact point information. Based on the first contact point information, it calculates the position of the detection frame 2 after rotating by an angle through an avoidance algorithm as the undetermined position. Based on the comparison between the undetermined position and the actual position, it obtains the actual rotation angle of the detection frame 2 as the undetermined rotation angle value.

[0057] The control module 104 controls the detection frame 2 to rotate according to the angle to be rotated and then fits onto the workpiece to be welded.

[0058] The first touch point information includes the coordinates of point A on the first board to be welded and the coordinates of point B on the second board to be welded. The actual position includes the coordinates of point C at the top of the detection frame.

[0059] The avoidance algorithm is configured as follows:

[0060]

[0061]

[0062] |α2-α1|=α

[0063] Where α1 -- the angle at which the detection frame rotates to the desired position, α2 -- the angle between the actual position and the initial position of the detection frame, α -- the value of the angle to be rotated, and hy A hx A --The height of point A on the first board to be welded along the y-axis and the length along the x-axis, hy B hx B --The height of point B on the second weld plate along the y-axis and the length along the x-axis, hy c hx c --The height of point C along the y-axis and the length along the x-axis.

[0064] Avoidance principle: such as Figure 5 andFigure 6 As shown, since the workpiece to be welded includes a first welding plate and a second welding plate, the first welding plate is generally placed horizontally, and the second welding plate can be placed at a certain angle to the first welding plate for welding depending on the angle of the workpiece to be welded. The coordinate of point A is taken at the bottom edge of the first welding plate away from the second welding plate, and the coordinate of point B is taken at the end of the second welding plate away from the first welding plate. The position of the workpiece to be welded is calculated, and then the angle that the detection frame 2 needs to rotate from the first position to the desired position is calculated. The first position is when the detection frame 2 is upright, but the actual position of the detection frame 2 may have already rotated a certain angle relative to the first position. Therefore, it is necessary to calculate the angle value of the actual position relative to the first position through the coordinate of point C. Based on the difference between the two, the actual angle value to be rotated of the detection frame 2 can be calculated.

[0065] The controller also includes a verification module 107. The verification module obtains the coordinates of each boundary point after the detection frame 2 is rotated to the pending position as the second touch point information. The second touch point information is compared with the first touch point information by an overlap formula. If any point overlaps, a touch command is issued; if they do not overlap, a normal command is issued.

[0066] The overlapping formula is configured as follows:

[0067] T1(x n ,y n ,z n )∩T2(x m ,y m ,z m )

[0068] Where T1 is the first touch point, T2 is the second touch point, and x n y n , z n --Coordinates of the first touch point, x m y m , z m --Coordinates of the second touch point.

[0069] Calibration principle: After the detection frame 2 has been estimated to be in the desired position, the coordinates of all boundary points of the detection frame 2 need to be extracted. Then, the coordinates of each boundary point of the workpiece to be welded are extracted. If the coordinates of any point extracted from the workpiece to be welded coincide with the left side of any point extracted from the detection frame 2, it means that the desired position needs to be adjusted.

[0070] The analysis module 102 includes an analysis submodule 108, which analyzes the image to be inspected to obtain the actual values ​​of bubbles, cracks, and slag. The welding database also includes weld defect thresholds and weld defect information. The weld defect thresholds reflect the weld grade, and the weld defect information includes bubble reference information, crack reference information, and slag reference information. The calculation module 106 indexes the corresponding weight ratios in the welding database according to the actual values ​​of bubbles, cracks, and slag, and calculates the integrals according to the weight ratios using a weight formula. The analysis module 102 compares the integrals with the weld defect thresholds. If the integral is greater than or equal to the weld defect threshold, a correction command is issued to the control center; if the integral is less than the weld defect threshold, a normal command is issued to the control center.

[0071] The weighting formula is configured as follows:

[0072]

[0073] Where, m i m j m k --Weight ratio, l--Integral, p--Actual value of bubble, r--Actual value of crack, s--Actual value of waste residue; Since only a small number of minor defects are detected in the weld, the integral calculated by the weight is relatively small. If the integral exceeds the weld defect threshold, it indicates that the weld strength level is low and needs to be repaired, ground, or scrapped.

[0074] The acquisition module 101 acquires images of each weld seam after welding with a welding torch, captured by a camera, as detection images. The analysis module 102 analyzes the gap between the weld seam and the welding plate or between weld seams based on the detection images. If a gap exists between the two, a re-calculation command for the weld point is sent to the welding arm. If no gap exists between the two, a normal command is sent to the welding arm.

[0075] The controller also includes an image processing module 109, and the welding database includes preset image information. The preset brightness value is calculated based on the preset image information. The controller acquires the image to be inspected, the comparison image, and the detection image from the acquisition module 101. The corresponding actual brightness value is calculated based on the image to be inspected, the comparison image, and the detection image. If the actual brightness value is the same as the preset brightness value, an image normal signal is issued. If the actual brightness value is different from the preset brightness value, an image abnormal signal is issued and a brightness compensation value is calculated. The brightness of the actual image information is adjusted based on the brightness compensation value.

[0076] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A welding inspection system based on multi-point welding process, characterized in that: The device includes a robotic arm (1) and a detection structure located on the robotic arm (1). The detection structure includes a detection frame (2) and a radiation source (3), a film plate (4), a cleaning component (5), and a drive assembly located within the detection frame (2). The detection frame (2) is provided with a detection port (7) and a switching port (8). The robotic arm (1) drives the detection frame (2) to rotate so that the detection port (7) is aligned with the workpiece to be welded. The robotic arm (1) drives the detection frame (2) to move so that the workpiece to be welded enters the detection frame (2). The drive assembly is used to drive the film plate (4) and the cleaning component (5) to rotate so that the film plate (4) and the cleaning component (5) are rotated out of the detection frame (2) through the switching port (8) or rotated into the detection frame (2) through the switching port (8). The detection frame (2) is also provided with a camera. The robotic arm (1) is also equipped with a controller, which includes a welding database and coloring reference information. The coloring reference information reflects the color image of the qualified weld after the detection liquid is sprayed. The welding database also includes a data acquisition module (101), an analysis module (102), a comparison module (103), and a control module (104); The acquisition module (101) acquires the image of the weld on the film plate (4) as the image to be inspected. The weld image is formed by the detection light emitted by the X-ray source (3) penetrating the weld perpendicularly and being received and imaged by the film plate (4). The analysis module (102) analyzes the image to be inspected. If there is an abnormality in the image to be inspected that exceeds the threshold, it sends a stop command to the robot (1). If there is no abnormality in the image to be inspected or if there is an abnormality that does not exceed the threshold, it sends a normal command to the robot (1). When the control module (104) receives a normal command after a layer of welding is completed, it controls the cleaning component (5) to spray the detection onto the welding surface. After waiting for the reaction time, it controls the acquisition module (101) to acquire the image of the welding surface as a comparison image. The comparison module (103) compares the comparison image with the coloring reference information. If they match, it sends a completion command to the robot (1). If they do not match, it sends a correction command to the robot (1).

2. The welding inspection system based on multi-point welding process according to claim 1, characterized in that: The controller also includes a receiving module (105) and a computing module (106); The receiving module (105) acquires the three-dimensional model of the workpiece to be welded by the welding arm; The calculation module (106) extracts the coordinates of each boundary point of the workpiece to be welded as the first touch point information based on the three-dimensional model of the workpiece to be welded. Based on the first touch point information, it calculates the position of the detection frame (2) after rotating by an angle through an avoidance algorithm as the undetermined position. Based on the comparison between the undetermined position and the actual position, it obtains the actual rotation angle of the detection frame (2) as the undetermined rotation angle value. The control module (104) controls the detection frame (2) to rotate according to the rotation angle value and then fits it onto the workpiece to be welded.

3. The welding inspection system based on multi-point welding process according to claim 2, characterized in that: The first touch point information includes the coordinates of point A on the first board to be welded and the coordinates of point B on the second board to be welded, and the actual position includes the coordinates of point C at the top of the detection frame; The avoidance algorithm is configured as follows: |α2-α1|＝α Where α1 -- the angle at which the detection frame rotates to the desired position, α2 -- the angle between the actual position and the initial position of the detection frame, α -- the value of the angle to be rotated, and hy A hx A --The height of point A on the first board to be welded along the y-axis and the length along the x-axis, hy B hx B --The height of point B on the second weld plate along the y-axis and the length along the x-axis, hy c hx c --The height of point C along the y-axis and the length along the x-axis.

4. The welding inspection system based on multi-point welding process according to claim 3, characterized in that: The controller also includes a verification module (107). The verification module (107) obtains the coordinates of each boundary point after the detection frame (2) rotates to the pending position as the second touch point information. The second touch point information is compared with the first touch point information by an overlap formula. If any point overlaps, a touch command is issued. If they do not overlap, a normal command is issued. The overlapping formula is configured as follows: T1(x n ,y n ,z n )∩T2(x m ,y m ,z m ) Where T1 is the first touch point, T2 is the second touch point, and x n y n , z n --Coordinates of the first touch point, x m y m , z m --Coordinates of the second touch point.

5. The welding inspection system based on multi-point welding process according to claim 1, characterized in that: The drive assembly includes a screw (6), a gear set, and a motor. The screw (6) is vertically located inside the detection frame (2). The motor is located on the outer wall of the detection frame (2). The motor and the screw (6) are connected by a gear set. The X-ray source (3) is mounted on the upper part of the screw (6) through a bearing. The film plate (4) is fixedly connected to the lower part of the screw (6). The cleaning component (5) is fixedly connected to the screw (6) between the X-ray source (3) and the film plate (4). The film plate (4) and the cleaning component (5) are staggered.

6. The welding inspection system based on multi-point welding process according to claim 5, characterized in that: The cleaning component (5) includes a cleaning strip. The lower surface of the cleaning strip is provided with a number of equally distributed cleaning nozzles. One end of the cleaning strip near the screw (6) is provided with a pipe that passes through the detection frame (2) and communicates with the external detection liquid. The inner bottom of the detection frame (2) is provided with a collection tank (9) for collecting the detection liquid. The collection tank (9) is provided with a drain pipe for discharging the detection liquid.

7. The welding inspection system based on multi-point welding process according to claim 2, characterized in that: The analysis module (102) includes an analysis submodule (108). The analysis submodule (108) analyzes the image to be inspected to obtain the actual values ​​of bubbles, cracks, and slag. The welding database also includes weld defect thresholds and weld defect information. The weld defect thresholds reflect the weld grade. The weld defect information includes bubble reference information, crack reference information, and slag reference information. The calculation module (106) indexes the corresponding weight ratios in the welding database according to the actual values ​​of bubbles, cracks, and slag. It calculates the integral according to the weight ratios using a weight formula. The analysis module (102) compares the integral with the weld defect threshold. If the integral is greater than or equal to the weld defect threshold, it sends a correction command to the control center. If the integral is less than the weld defect threshold, it sends a normal command to the control center.

8. The welding inspection system based on multi-point welding process according to claim 7, characterized in that: The weighting formula is configured as follows: Σm i p+m j r+m k s=l Where, m i m j m k --Weight ratio, l--Integral, p--Actual value of bubbles, r--Actual value of cracks, s--Actual value of waste residue.

9. The welding inspection system based on multi-point welding process according to claim 1, characterized in that: The acquisition module (101) acquires the image of each weld seam after welding by the welding gun captured by the camera as the detection image. The analysis module (102) analyzes the gap between the weld seam and the welding plate or between weld seams based on the detection image. If there is a gap between the two, a re-calculation instruction for the weld point is issued to the welding arm. If there is no gap between the two, a normal instruction is issued to the welding arm.

10. The welding inspection system based on multi-point welding process according to claim 9, characterized in that: The controller also includes an image processing module (109), and the welding database also includes preset image information. A preset brightness value is calculated based on the preset image information. The image to be inspected, the comparison image, and the detection image in the acquisition module (101) are acquired. The corresponding actual brightness value is calculated based on the image to be inspected, the comparison image, and the detection image. If the actual brightness value is the same as the preset brightness value, an image normal signal is issued. If the actual brightness value is different from the preset brightness value, an image abnormal signal is issued and a brightness compensation value is calculated. The brightness of the actual image information is adjusted based on the brightness compensation value.

Images