Power battery welding seam detection method and system
By combining two-dimensional and three-dimensional image detection, the reference plane is constructed and the area ratio at the intersection of the weld and the reference plane and the distance between the highest point and the positioning plane of the weld is solved, and the existing welding detection is ineffective and incomplete, and the welding is realized Efficient and comprehensive evaluation and inspection of seam quality.
Patent Information
- Application Number
- CN202411381615.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-30
- Publication Date
- 2024-12-06
AI Technical Summary
The existing welding inspection methods are time-consuming and labor-intensive, and it is difficult to comprehensively detect the welding quality, especially whether there are abnormal problems in the weld.
Using a combination of two-dimensional and three-dimensional image detection, welds are used to construct a reference plane on a three-dimensional image, and the area ratio at the intersection of welds and the distance between the highest point of welds and the positioning surface, and welds in combination with two-dimensional images. Seam width, black holes, white holes, broken welding, false welding and offset detection to achieve a comprehensive quality assessment of welds.
It improves the efficiency and comprehensiveness of weld detection, and can more accurately determine the defects of welds, including residual height, width, offset and hole defects, reducing the error rate of manual detection.
Smart Images

Figure CN119086594A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of product detection, and specifically relates to a power battery weld detection method and system. Background Art
[0002] With the gradual rise of new energy vehicles, the manufacturing technology of new energy vehicles is developing rapidly. The endurance of new energy vehicles is gradually becoming one of the issues of automobile energy. Therefore, power battery assembly has become a vital part of new energy vehicle manufacturing.
[0003] The key part of power battery manufacturing is the assembly and welding of multiple battery cells. Multiple battery cells are connected in parallel to increase the voltage and current to drive the operation of electric vehicles. Therefore, it is particularly important to detect the quality of power battery packs. Failure to prevent welding quality problems between battery cells will lead to cumbersome maintenance, which leads to the inspection and control of welding quality. The existing welding inspection uses ordinary 2D and manual inspection to inspect the welding quality, which is time-consuming and labor-intensive, and it is difficult to detect whether there are abnormalities in the welding using ordinary 2D. Summary of the invention
[0004] The object of the present invention is to provide a power battery weld detection method and system to solve the problem of laborious and incomplete weld detection in the prior art mentioned in the above background technology.
[0005] To achieve the above object, the present invention provides the following technical solution: a power battery weld detection method, comprising:
[0006] Acquire image data of a weld, wherein the weld includes an annular weld formed by welding a pole and a connecting piece, and the image data includes a two-dimensional image and a three-dimensional image;
[0007] Locating the upper end surface of the connecting piece as a positioning surface on the three-dimensional image and,
[0008] The second defect set of the weld is detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes:
[0009] Constructing a reference plane based on the three-dimensional image, wherein the reference plane is arranged parallel to the positioning plane, and the distance between the reference plane and the plane is a first set value;
[0010] In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
[0011] Furthermore, the residual height detection further includes:
[0012] The distance between the highest point of the weld and the positioning surface is calculated based on the three-dimensional image. When the distance between the highest point of the weld and the positioning surface is greater than a second set value, it is determined that the weld has a residual height defect, and the first set value is less than the second set value.
[0013] Furthermore, the method also includes detecting a first defect set of the weld based on the two-dimensional image and the three-dimensional image, and the first defect set detection of the weld includes weld width detection, black hole and white hole detection, broken weld and cold weld detection, and weld offset detection.
[0014] Furthermore, the weld width detection includes:
[0015] Based on the two-dimensional image, the weld is partitioned in a circumferential direction to form a plurality of partitioned welds;
[0016] The weld width of each partition is calculated and compared with the set reference range. When the weld width of any partition is not within the set reference range, it is determined that the weld has a size defect.
[0017] Furthermore, the weld width detection further includes:
[0018] Determine the starting and ending areas of the weld based on the two-dimensional image, and calculate the width of the weld in the starting and ending areas and the ratio of the length of the weld in the starting and ending areas to the total length of the weld;
[0019] The width of the weld in the starting and ending area partitions and the ratio of the weld length in the starting and ending area partitions to the total weld length are compared with the set reference value. When the width of the weld in the starting and ending area partitions is not less than the set reference value or the ratio of the weld length in the starting and ending area partitions to the total weld length is greater than the set reference value, it is determined that there is a weld size defect.
[0020] Further, the weld offset detection includes:
[0021] Determine the pole center position and the weld center position on the two-dimensional image, and calculate the distance between the pole center position and the weld center position accordingly;
[0022] The calculated distance between the pole center position and the weld center position is compared with the set reference value. When the distance between the pole center position and the weld center position is greater than the set reference value, it is determined that a weld offset defect exists.
[0023] Furthermore, black hole and white hole detection includes:
[0024] Based on the two-dimensional image, the grayscale value is used to determine whether black holes and / or white holes exist. When black holes or white holes exist, the ratio of the black hole and / or white hole area to the overall weld area is calculated accordingly. When the ratio of the black hole and / or white hole area to the overall weld area is greater than the set reference value, it is determined that white hole and / or black hole defects exist in the weld.
[0025] On the other hand, the present application discloses a power battery weld detection system, comprising:
[0026] A camera module configured to acquire image data of a weld, wherein the weld includes an annular weld formed by welding a pole and a connecting piece, and the image data includes a two-dimensional image and a three-dimensional image;
[0027] A task module, which is configured to locate the upper end surface of the connecting piece as a positioning surface on the three-dimensional image and
[0028] The first defect set and the second defect set of the weld are detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes:
[0029] Constructing a reference plane based on the three-dimensional image, wherein the reference plane is arranged parallel to the positioning plane, and the distance between the reference plane and the plane is a first set value;
[0030] In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
[0031] Furthermore, the system also includes:
[0032] The light source module is configured to provide the camera module with required light source, and the light source module includes a coaxial light source and an annular light source.
[0033] Furthermore, the system also includes a data saving module, and the data saving module is configured to save system operation data and log data.
[0034] Compared with the prior art, the present invention has the following beneficial effects:
[0035] The present application adds the acquisition of three-dimensional images and a detection process based on three-dimensional images to the traditional 2D detection. Specifically, a reference plane parallel to the upper end face of the connecting piece is constructed based on the three-dimensional image, and the ratio between the weld area outside the range and the overall weld area is calculated based on the range defined by the reference plane and the upper end face of the connecting piece. This enables comprehensive detection of the weld residual height, thereby improving the detection efficiency and ensuring the comprehensiveness of the detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Figure 1 This is a flowchart of the application method;
[0037] Figure 2 It is a schematic diagram of the two-dimensional image of the weld;
[0038] Figure 3 It is a schematic diagram of the three-dimensional image of the weld;
[0039] Figure 4 The cross-sectional view of the welding seam between the connecting piece and the pole;
[0040] Figure 5a It is a schematic diagram of weld partitioning;
[0041] Figure 5b This is a schematic diagram of zoned weld processing;
[0042] Figure 5c It is a line chart of the zoned weld width;
[0043] Figure 6 This is a schematic diagram of weld offset detection;
[0044] Figure 7a Schematic diagram of residual height detection results;
[0045] Figure 7b Select a schematic diagram for the positioning surface;
[0046] Figure 7c It is the three-dimensional image positioning map of the weld;
[0047] Figure 7d This is a schematic diagram of the three-dimensional image of the weld after processing. DETAILED DESCRIPTION
[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0049] A method for detecting a power battery weld, in particular a method for detecting a power battery Busbar weld, referring to Figure 1 , the method comprising:
[0050] Acquire image data of a weld, wherein the weld includes an annular weld formed by welding a pole and a connecting piece, and the image data includes a two-dimensional image (refer to Figure 2 ) and 3D images (refer to Figure 3 );
[0051] Locating the upper end surface of the connecting piece on the three-dimensional image as a positioning surface and
[0052] The second defect set of the weld is detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes:
[0053] Construct a reference surface based on the three-dimensional image (refer to Figure 7b and 7c , respectively, are a schematic diagram of a positioning surface selected based on a three-dimensional image and a schematic diagram of a detection area), the reference plane is arranged parallel to the positioning surface, and a distance between the reference plane and the plane is a first set value;
[0054] In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
[0055] Reference Figure 4 , is a cross-sectional schematic diagram of the annular weld 12 formed by welding the pole 11 and the connecting piece 10, and the positions of the positioning surface 13 and the reference surface 14 are exemplarily given in the figure, that is, the positional relationship between the positioning surface 13 and the reference surface 14 in the thickness direction of the connecting piece is shown. Specifically, the above-mentioned positioning surface 13 is constituted as the upper end surface of the connecting piece 10. Correspondingly, the above-mentioned first setting value can be exemplarily 0.5mm. Further, after constructing the reference surface 14, the height pixels outside the positioning surface 13 and the reference surface 14 can be extracted by the 3D calibration tool, and the weld prediction defect can be judged by calculating the ratio of the height data and the overall weld data (refer to Figure 7a and 7d , which are the detection results and the three-dimensional image of the weld respectively.
[0056] The residual height detection also includes: calculating the distance between the highest point of the weld and the positioning surface. When the distance between the highest point of the weld and the positioning surface is greater than a second set value (for example, 1.5 mm), it is determined that the weld has a residual height defect, and the first set value is less than the second set value.
[0057] In some embodiments, the method further comprises detecting a first defect set of the weld based on the two-dimensional image and the three-dimensional image, wherein the detection of the first defect set comprises weld width detection, black hole and white hole detection, broken weld and false weld detection, and weld offset detection;
[0058] The black hole and white hole detection include:
[0059] Based on the two-dimensional image, the grayscale value is used to determine whether black holes and / or white holes exist. When black holes or white holes exist, the ratio of the black hole and / or white hole area to the overall weld area is calculated accordingly. When the ratio of the black hole and / or white hole area to the overall weld area is greater than the set reference value, it is determined that white hole and / or black hole defects exist in the weld.
[0060] Reference Figure 5a-5c , where the weld width detection process is as follows:
[0061] Based on the 2D image, the weld is partitioned in the circumferential direction to form multiple partitioned welds (refer to Figure 5a and 5b , respectively, schematic diagrams of weld partitioning and schematic diagrams of processing of partitioned welds after partitioning). For example, the weld is divided into 150 partitioned welds, and the width of each partitioned weld is compared with a set reference value (refer to Figure 5c is a line graph of the width of the partitioned weld). When the width of any partitioned weld is not within the set reference value range, the weld size is determined to be unqualified, that is, there is a weld size defect. By way of example, the reference values set for the above weld width are 2.5 mm and 4.5 mm. At this time, when the width of any partitioned weld is not within the range of 2.5 to 4.5 mm, the weld size is determined to be unqualified.
[0062] Furthermore, the above-mentioned weld width detection also includes:
[0063] Determine the starting and ending areas of the weld, compare the width of the weld in the starting and ending area partition and the ratio of the weld length in the starting and ending area partition to the total weld length with the set reference value, when the weld width in the starting and ending area partition is not less than the set value (for example, 3mm) or the ratio of the weld length in the starting and ending area partition to the total weld length is greater than the set reference value (for example, 0.1), it is determined that there is a weld size defect, that is, the weld size is unqualified.
[0064] In some embodiments, reference Figure 6 , the above weld offset detection process is as follows:
[0065] The pole center position and the weld center position are determined on the two-dimensional image, and the distance between the pole center position and the weld center position is compared with the set reference value. When the distance between the pole center position and the weld center position is greater than the set reference value, it is determined that a weld offset defect exists.
[0066] On the other hand, the present application provides a power battery weld detection system, the system comprising:
[0067] A camera module configured to acquire weld image data, wherein the image data includes a two-dimensional image and a three-dimensional image,
[0068] A task module, which is configured to locate the upper end surface of the connecting piece as a positioning surface on the three-dimensional image and
[0069] The first defect set and the second defect set of the weld are detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes:
[0070] Constructing a reference plane based on the three-dimensional image, wherein the reference plane is arranged parallel to the positioning plane, and the distance between the reference plane and the plane is a first set value;
[0071] In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
[0072] In some embodiments, the above-mentioned image data is acquired through an industrial camera, that is, the camera module is an industrial camera. The industrial camera can convert the light signal generated by the lens into an ordered electrical signal through a CCD or CMOS imaging sensor, and receive and process the electrical signal through, for example, an industrial computer to obtain weld image data. Exemplarily, the above-mentioned industrial camera includes a 2D plane camera and a 3D line scan camera, which are used to acquire two-dimensional images and three-dimensional images respectively.
[0073] In some embodiments, the above-mentioned task module uses VP operators and tools to combine logic and functions, and can flexibly link external parameter adjustment interfaces and communication sending interfaces through standard input and output interfaces. It can also receive images transmitted from the camera module, process, calculate, and output the final results, and return them to external devices (PLC, robots, etc.). At the same time, the above-mentioned task module program editing methods are diversified. It can be called using scripts using nested ToolBlock methods, or it can use C# to call tasks for a single ToolBlock. The output data can also be freely edited as needed. Furthermore, the above-mentioned task module sends results and completion signals to the outside by calling the communication interface, and can also interact with external devices (PLC, robots) before and after the program is run according to the specific operating conditions.
[0074] In some embodiments, the system further includes a light source module, which is configured to provide the required light source to the camera module to improve the effect of the acquired image data, that is, based on the adjustment of the light source shape and angle, the weld can present different effects to obtain image data in different states. In some examples, the light source module is an LED light source. Based on the advantages of long life, stable lighting and fast response of the LED light source, the display effect of the weld image can be improved. In some embodiments, the light source module includes a coaxial light source and a low-angle annular light source. The coaxial light source eliminates the surface of the object to be measured through a spectroscope design, which can improve the imaging clarity, that is, improve the clarity of the two-dimensional image; the annular light source highlights the three-dimensional information of the weld through 360 degrees and low angles, and highlights characteristic information such as fried welds, that is, it can improve the clarity of the three-dimensional image.
[0075] In some embodiments, the light source module includes a coaxial light source and a low-angle annular light source. The coaxial light source eliminates the surface of the object being measured through a spectroscope design to improve imaging clarity; the annular light source highlights the three-dimensional information of the weld through 360 degrees and low angles, highlighting characteristic information such as fried welds.
[0076] In some embodiments, the system further includes a main interface module, which is an integrated display, parameter adjustment, program modification and other functional interface based on Cognex's VisionPro (hereinafter referred to as VP) algorithm tool and its custom controls, and constitutes an interactive interface. At the same time, the main interface module is configured to have at least the following functions:
[0077] 1. Display the process running results and the process running log.
[0078] 2. Function of modifying and saving operator parameters or logic in task flow.
[0079] 3. Adjust camera parameters, adjust imaging effects and other functions.
[0080] 4. Open external process parameter or tool parameter adjustment interface to users based on actual process.
[0081] 5. Modify the storage path of folders such as logs and pictures, set picture storage rules, and switch plans, etc.
[0082] 6. Manage user account names, passwords, login permissions and other functions.
[0083] In some embodiments, the above system also includes a communication module, which is an interactive module written based on specific needs and is used to interact with external devices such as PLC / robot and control the action of the camera module to take pictures. Specifically, the communication module can initialize communication variables and parameters, cyclically monitor related variables such as PLC triggers, and trigger the camera module to take pictures.
[0084] In some embodiments, the system further comprises a parameter interface module, which is designed according to the task flow input and standard process parameters.
[0085] In some embodiments, the above system also includes a data saving module, which includes two parts: an operation data saving sub-module (result data and image data) and a log data saving sub-module. The operation data is saved and named according to fixed standards, and the log is displayed and saved according to the actual software operation. It is fixedly saved as a local folder for easy viewing and tracing.
[0086] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A method for detecting a power battery weld. It is characterized in that include: Acquire image data of the weld, wherein the image data includes a 2D image and a 3D image; The upper end surface of the positioning connecting piece on the three-dimensional image is used as the positioning surface and, The second defect set of the weld is detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes: Constructing a reference plane based on the three-dimensional image, wherein the reference plane is arranged parallel to the positioning plane, and the distance between the reference plane and the positioning plane is a first set value; In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
2. A power battery weld detection method according to claim 1, Features: The residual height detection further includes: The distance between the highest point of the weld and the positioning surface is calculated based on the three-dimensional image. When the distance between the highest point of the weld and the positioning surface is greater than a second set value, it is determined that the weld has a residual height defect, and the first set value is less than the second set value.
3. A power battery weld detection method according to claim 1, Features: The method also includes detecting a first defect set of the weld based on the two-dimensional image and the three-dimensional image, wherein the first defect set detection of the weld includes weld width detection, black hole and white hole detection, broken weld and false weld detection, and weld offset detection.
4. A power battery weld detection method according to claim 3, Features: The weld width detection includes: Based on the two-dimensional image, the weld is partitioned in a circumferential direction to form a plurality of partitioned welds; The weld width of each partition is calculated and compared with the set reference range. When the weld width of any partition is not within the set reference range, it is determined that the weld has a size defect.
5. A power battery weld detection method according to claim 4, Features: The weld width measurement also includes: Determine the starting and ending areas of the weld based on the two-dimensional image, and calculate the width of the weld in the starting and ending areas and the ratio of the length of the weld in the starting and ending areas to the total length of the weld; The width of the weld in the starting and ending area partitions and the ratio of the weld length in the starting and ending area partitions to the total weld length are compared with the set reference value. When the width of the weld in the starting and ending area partitions is not less than the set reference value or the ratio of the weld length in the starting and ending area partitions to the total weld length is greater than the set reference value, it is determined that there is a weld size defect.
6. A power battery weld detection method according to claim 3, Features: The weld offset detection includes: Determine the pole center position and the weld center position on the two-dimensional image, and calculate the distance between the pole center position and the weld center position accordingly; The calculated distance between the pole center position and the weld center position is compared with the set reference value. When the distance between the pole center position and the weld center position is greater than the set reference value, it is determined that a weld offset defect exists.
7. A power battery weld detection method according to claim 3, Features: Black hole and white hole detection includes: Based on the two-dimensional image, the grayscale value is used to determine whether black holes and / or white holes exist. When black holes or white holes exist, the ratio of the black hole and / or white hole area to the overall weld area is calculated accordingly. When the ratio of the black hole and / or white hole area to the overall weld area is greater than the set reference value, it is determined that white hole and / or black hole defects exist in the weld.
8. A power battery weld detection system, It is characterized in that include: A camera module configured to acquire image data of a weld, wherein the weld includes an annular weld formed by welding a pole and a connecting piece, and the image data includes a two-dimensional image and a three-dimensional image; A task module, which is configured to locate the upper end surface of the connecting piece as a positioning surface on the three-dimensional image and The first defect set and the second defect set of the weld are detected based on the two-dimensional image and the three-dimensional image, wherein the second defect set detection includes residual height detection, and the residual height detection includes: Constructing a reference plane based on the three-dimensional image, wherein the reference plane is arranged parallel to the positioning plane, and the distance between the reference plane and the plane is a first set value; In the three-dimensional image, it is determined whether the reference plane and the weld intersect. When the reference plane and the weld intersect, the weld is segmented based on the reference plane, and the ratio of the weld area outside the reference plane and the positioning plane to the overall weld area is calculated in combination with the two-dimensional image. When the ratio is greater than the set reference value, it is determined that the weld has a residual height defect.
9. A power battery weld detection system according to claim 8, Features: The system further comprises: The light source module is configured to provide the camera module with required light source, and the light source module includes a coaxial light source and an annular light source.
10. A power battery weld detection system according to claim 8, It is characterized in that The system also includes a data saving module, which is configured to save system operation data and log data.
