Method and system for detecting weld quality

Abstract

This application is applicable to the field of weld detection technology, and provides a method for detecting the quality of a weld. The method includes: obtaining an image of a detection area of the welding component by using a visual recognition device; performing a scan along the weld by using an ultrasonic detection device to obtain an ultrasonic detection area of the detection area; wherein, the ultrasonic detection area is the welding area between the second welding piece and the bottom wall of the depression; comparing the ultrasonic detection area with a dynamic area threshold; if the ultrasonic detection area is greater than or equal to the dynamic area threshold, it is determined that the weld quality is qualified; if the ultrasonic detection area is less than the dynamic area threshold, it is determined that the weld quality is unqualified; wherein, the dynamic area threshold is dynamically determined according to the statistical parameters of historical qualified samples and the current weld length. According to the ultrasonic detection area, the predicted value of the penetration depth and the defect recognition confidence level, the present invention calculates a weld quality index, which can further improve the accuracy, adaptability and intelligent level of weld quality detection.

Description

Technical Field

[0001] This application belongs to the field of weld inspection technology, and in particular relates to a method and system for inspecting weld quality. Background Technology

[0002] For products assembled by welding, weld seams are usually left on the product. Taking battery modules as an example, after the end plate and side plate are welded, a weld seam is formed between the end plate and the side plate. The reliability of the connection at the weld seam is one of the important factors affecting product performance.

[0003] Currently, for welded products with edge-sealing structures, weld inspection methods cannot directly measure weld penetration, resulting in low accuracy in weld quality testing. Existing technologies employ a combination of visual recognition and ultrasonic testing, determining weld quality by acquiring the weld area between the second weldment and the recessed bottom wall and comparing it to a fixed area threshold. However, this method uses an area threshold that is the product of a fixed width threshold and the weld length, which cannot adapt to variations in materials, plate thickness, and welding processes, leading to a high false positive rate. Furthermore, existing methods cannot directly quantify weld penetration and do not fully utilize the complementary information from visual and ultrasonic data, leaving room for improvement in both accuracy and efficiency. Summary of the Invention

[0004] This application provides a method for detecting weld quality, which can further improve the accuracy, adaptability, and intelligence of weld quality detection.

[0005] In a first aspect, embodiments of this application provide a method for detecting weld quality, used to detect the welding quality of a welding assembly, the welding assembly including a first welded component and a second welded component welded together, the first welded component having a recessed step, a portion of the second welded component being located within the recessed step, the recessed step including a recessed sidewall and a recessed bottom wall connected to each other, the second welded component being welded to the recessed sidewall and the recessed bottom wall respectively; the detection method includes:

[0006] An image of the detection area of ​​the welding assembly is acquired using a visual recognition device;

[0007] An ultrasonic testing device is used to scan along the weld seam to obtain the ultrasonic testing area of ​​the testing area; wherein, the ultrasonic testing area is the welding area between the second welded part and the bottom wall of the depression;

[0008] The ultrasonic detection area is compared with the dynamic area threshold; if the ultrasonic detection area is greater than or equal to the dynamic area threshold, the weld quality is determined to be qualified.

[0009] If the ultrasonic detection area is less than the dynamic area threshold, the weld quality is deemed unqualified.

[0010] The dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld length.

[0011] Optionally, the dynamic area threshold satisfy:

[0012]

[0013] in, The statistical average width of historical qualified samples. The width standard deviation of historical qualified samples. Here is the confidence coefficient. This represents the weld length.

[0014] Optionally, it also includes: extracting at least one ultrasonic feature during the ultrasonic scanning process, and calculating the predicted melt depth based on the ultrasonic feature using a pre-trained regression model; the ultrasonic feature is selected from one or more of the following: bottom wall echo amplitude, signal attenuation rate, waveform correlation coefficient, and bottom wall echo broadening.

[0015] Optionally, the regression model is:

[0016]

[0017] in, This is the predicted melting depth. This represents the bottom wall echo amplitude. For signal attenuation rate, The waveform correlation coefficient, To broaden the bottom wall echo, These are the model coefficients fitted by metallographic calibration.

[0018] Optionally, the step of determining the weld quality as qualified further includes: when the ultrasonic detection area is less than the dynamic area threshold but the predicted penetration depth is greater than or equal to the target penetration depth, it is still determined to be qualified.

[0019] Optionally, after acquiring an image of the detection area using a visual recognition device, visual features are also extracted. These visual features include one or more of the following: weld width uniformity, weld edge straightness, and indentation step integrity. When determining the weld quality, the visual features are fused with the ultrasonic detection area and the dynamic area threshold for determination.

[0020] Optionally, the step of scanning along the weld seam using an ultrasonic testing device includes:

[0021] First, a coarse scan is performed along the weld centerline extracted by the visual recognition device to calculate the local welding area at each local location of the weld; then, areas with local welding areas lower than a local threshold are identified, and these areas are subjected to a slower and more intensive fine scan; the local threshold is determined based on a dynamic area threshold and the weld length.

[0022] Optionally, the local threshold satisfy:

[0023]

[0024] in, These are the position coordinates along the weld centerline. The midpoint of the weld. The attenuation coefficient is... .

[0025] Optionally, it also includes: calculating the weld quality index (WQI) based on the ultrasonic detection area, the predicted penetration depth, and the defect identification confidence level.

[0026]

[0027] in, The area measured by the ultrasonic wave. This is the predicted melting depth. For the target melting depth, To determine the confidence level for defect identification, To achieve the highest confidence level, These are the weighting coefficients, and their sum is 1;

[0028] Weld quality is classified into multiple grades based on WQI values.

[0029] Secondly, embodiments of this application provide a weld quality inspection system. Applying the inspection method of any of the above embodiments, the weld quality inspection system includes: a visual recognition device for acquiring an image of the inspection area of ​​the welded assembly; an ultrasonic inspection device for scanning along the weld to acquire the ultrasonic inspection area of ​​the inspection area; and a data processing module for calculating a dynamic area threshold, comparing the ultrasonic inspection area with the dynamic area threshold, and outputting a judgment result.

[0030] The above-described solution of this application has the following beneficial effects: A visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second welded component and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention can further improve the accuracy, adaptability, and intelligence level of weld seam quality inspection.

[0031] Other beneficial effects of this application will be described in detail in the following detailed description section. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 A flowchart illustrating a weld quality inspection method provided in one embodiment of this application;

[0034] Figure 2 This is a schematic diagram of a weld quality inspection system provided in one embodiment of this application. Detailed Implementation

[0035] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0036] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0037] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0038] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0039] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0040] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0041] This application provides a method for inspecting weld quality in a welded assembly. The welded assembly includes a first welded component and a second welded component welded together. The first welded component has a recessed step, and a portion of the second welded component is located within the recessed step. The recessed step includes interconnected recessed sidewalls and a recessed bottom wall. The second welded component is welded to both the recessed sidewalls and the recessed bottom wall. The second welded component is bent and includes interconnected first and second connecting portions, with the first connecting portion forming a ferrule structure within the recessed step.

[0042] The first welded component can be the end plate of the battery module, and the second welded component can be the side plate of the battery module.

[0043] Example 1:

[0044] See Figure 1 The detection method in this embodiment includes:

[0045] Step S10: Use a visual recognition device to acquire an image of the detection area of ​​the welding assembly.

[0046] The visual recognition device may include a CCD camera and an image processing unit. The welding assembly is fixed to the inspection platform, and the CCD camera captures multiple images of the recessed step area of ​​the welding assembly from various angles to obtain a high-resolution image containing the weld contour. Image processing algorithms are used to identify the geometric features of the weld, including the junction of the recessed sidewall and the recessed bottom wall, and the edge contour of the embedded portion of the second weldment.

[0047] Step S20: An ultrasonic testing device is used to scan along the weld seam to obtain the ultrasonic testing area of ​​the testing region. The ultrasonic testing area is the welding area between the second weldment and the recessed bottom wall.

[0048] First, the ultrasonic testing device is calibrated by selecting a probe that matches the sound velocity of the welding material and performing distance-amplitude curve calibration using a standard test block. Based on the detection area boundary provided by the visual recognition device, the probe's scanning path and speed are set. Coupling agent is evenly applied to the surfaces of the first and second welded parts, and the probe is driven to move along the scanning path, emitting ultrasonic waves in real time and receiving reflected signals. The effective welding area is identified by analyzing the amplitude and position of the reflected signals.

[0049] Step S30: Compare the ultrasonic testing area with the dynamic area threshold. If the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld quality is deemed acceptable; if the ultrasonic testing area is less than the dynamic area threshold, the weld quality is deemed unacceptable.

[0050] The dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld length. Specifically, the dynamic area threshold satisfies:

[0051]

[0052] In the formula, The statistical average width of historical qualified samples. The width standard deviation of historical qualified samples. This is the confidence coefficient (recommended value is 1.0~1.5). This represents the weld length.

[0053] Implementation steps: Collect ultrasonic testing area data from no fewer than 30 qualified samples under the same production conditions, calculate the equivalent width of each sample, and obtain the statistical average and standard deviation. The threshold is automatically recalculated before each batch of production, or a sliding window is set in the testing system for real-time updates. This dynamic threshold can adapt to fluctuations in different materials, plate thicknesses, and welding processes, significantly reducing the false judgment rate.

[0054] In this embodiment, a visual recognition device is used to obtain an image of the detection area of the welding assembly; an ultrasonic detection device is used to scan along the weld to obtain the ultrasonic detection area of the detection area; wherein, the ultrasonic detection area is the welding area between the second welding piece and the bottom wall of the depression; the ultrasonic detection area is compared with the dynamic area threshold; if the ultrasonic detection area is greater than or equal to the dynamic area threshold, it is determined that the weld quality is qualified; if the ultrasonic detection area is less than the dynamic area threshold, it is determined that the weld quality is unqualified; wherein, the dynamic area threshold is dynamically determined according to the statistical parameters of historical qualified samples and the current weld length. The present invention can further improve the accuracy, adaptability and intelligent level of weld quality detection.

[0055] Embodiment 2: [[ID=...]]

[0056] This embodiment further includes a penetration prediction step on the basis of Embodiment 1.

[0057] During the ultrasonic scan, multiple ultrasonic feature quantities are extracted, including the amplitude of the bottom wall echo, the signal attenuation rate, the waveform correlation coefficient, and the broadening of the bottom wall echo. Based on these feature quantities, the penetration prediction value is calculated through a pre-trained regression model.

[0058] The regression model is:

[0059]

[0060] Calibration process: Select 5 - 10 typical samples, conduct ultrasonic detection and record the above feature values; perform metallographic sectioning at the same position to measure the actual penetration; use the least squares method to fit the model coefficients. After the model is embedded in the detection system, the penetration prediction value can be output in real time during detection.

[0061] When determining the weld quality, if the ultrasonic detection area is less than the dynamic area threshold but the penetration prediction value is greater than or equal to the target penetration, it is still determined to be qualified. This solution can avoid misjudgment caused by a slightly smaller welding area but sufficient penetration, and improve the rationality of detection.

[0062] In this embodiment, the ultrasonic detection area is compared with the dynamic area threshold; if the ultrasonic detection area is greater than or equal to the dynamic area threshold, it is determined that the weld quality is qualified; if the ultrasonic detection area is less than the dynamic area threshold, it is determined that the weld quality is unqualified; wherein, the dynamic area threshold is dynamically determined according to the statistical parameters of historical qualified samples and the current weld length. The present invention can further improve the accuracy, adaptability and intelligent level of weld quality detection.

[0063] Embodiment 3:

[0064] This embodiment introduces the fusion determination of visual features and ultrasonic features on the basis of Embodiment 1.

[0065] After acquiring images of the inspection area using a visual recognition device, visual features are extracted, including weld width uniformity, weld edge straightness, and the integrity of recessed steps. When determining weld quality, these visual features are fused with the ultrasonic inspection area and dynamic area threshold.

[0066] The fusion determination rules are shown in the table below:

[0067] The ultrasonic testing area is greater than or equal to the dynamic area threshold and all visual features are qualified. qualified pass The ultrasonic testing area is less than the dynamic area threshold, but all visual features are qualified. qualified Record characteristic deviations, allowing for... Any visual feature is severely out of bounds (e.g., a deviation greater than 0.3 mm). Unqualified Marked as "Abnormal weld appearance"

[0068] By using multimodal fusion, the limitations of a single indicator are reduced, the reliability of judgment is improved, and specific non-compliance categories are output.

[0069] In this embodiment, a visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second welded component and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention introduces the fusion judgment of visual features and ultrasonic features, which can further improve the accuracy of weld seam quality detection.

[0070] Example 4:

[0071] This embodiment optimizes the ultrasonic scanning path by introducing real-time feedback scanning path planning.

[0072] First, a coarse scan is performed along the weld centerline extracted by the visual recognition device at a relatively high speed (e.g., 20 mm / s), calculating the local weld area at each location in real time. After the coarse scan, areas with local weld areas below a local threshold are identified. The local threshold satisfies the following conditions:

[0073]

[0074] in, These are the position coordinates along the weld centerline. The midpoint of the weld. The attenuation coefficient is... We recommend a value of 0.2.

[0075] For areas below a local threshold, the scanning speed is automatically reduced to 5 mm / s and encrypted sampling is performed to re-extract the ultrasonic features of that area. This adaptive scanning method can shorten the detection cycle by approximately 20%-30% while ensuring detection quality.

[0076] In this embodiment, a visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second welded component and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention introduces real-time feedback scanning path planning, which can further improve the accuracy, adaptability, and intelligence level of weld seam quality inspection.

[0077] Example 5:

[0078] This embodiment provides a method for continuous quantitative evaluation of weld quality.

[0079] The weld quality index is calculated based on the ultrasonic testing area, predicted penetration depth, and defect identification confidence level.

[0080]

[0081] in, The area measured by the ultrasonic wave. This is the predicted melting depth. For the target melting depth, To determine the confidence level for defect identification, This represents the maximum confidence level (usually set to 1). The weights are coefficients and their sum is 1. Recommended values ​​are 0.4, 0.4, and 0.2.

[0082] Weld quality is classified into multiple grades based on weld quality index:

[0083] Greater than or equal to 0.90 high quality Normal release Between 0.75 and 0.90 qualified Record it, then you can proceed. Between 0.60 and 0.75 suspicious Manual re-inspection or downgraded use Less than 0.60 Unqualified Repair or scrap

[0084] This quality index enables numerical and refined evaluation of weld quality, facilitating statistics and traceability.

[0085] In this embodiment, a visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second weldment and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention calculates a weld seam quality index based on the ultrasonic testing area, the predicted penetration depth, and the defect identification confidence level, which can further improve the accuracy, adaptability, and intelligence level of weld seam quality inspection.

[0086] Example 6:

[0087] This application also provides a weld quality inspection system that applies the inspection method of any of the above embodiments. See also Figure 2 The weld quality inspection system includes:

[0088] The visual recognition device 201 is used to acquire images of the detection area of ​​the welding assembly.

[0089] The ultrasonic testing device 202 is used to scan along the weld seam to obtain the ultrasonic testing area of ​​the testing region.

[0090] The data processing module 203 is used to calculate the dynamic area threshold, compare the ultrasonic detection area with the dynamic area threshold, and output the judgment result.

[0091] The data storage module is used to store test data and equipment calibration records;

[0092] The alarm device and automatic marking device are used to trigger an alarm when the weld quality is determined to be unqualified, and to control the automatic marking device to mark the defect location at the unqualified weld.

[0093] The data processing module is also configured to: extract ultrasonic feature quantities, calculate the predicted melt depth value through a regression model, and correct the output based on the comparison between the predicted melt depth value and the target melt depth. The data processing module can be built on an embedded processor or industrial computer, and includes the algorithm code for the aforementioned model.

[0094] In this embodiment, a visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second welded component and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention can further improve the accuracy, adaptability, and intelligence level of weld seam quality inspection.

[0095] Example 7:

[0096] Based on the above embodiments, a complete detection process is as follows:

[0097] After welding is completed, the welded components are fixed to the testing platform;

[0098] The visual recognition device images the detection area and extracts the weld centerline and visual features.

[0099] The ultrasonic testing device performs a coarse scan along the centerline of the weld and calculates the local weld area.

[0100] Identify low-area areas and perform adaptive encrypted fine scanning;

[0101] During the scanning process, ultrasonic characteristic quantities (echo amplitude, attenuation rate, correlation coefficient, and broadening) are extracted.

[0102] The data processing module calculates the dynamic area threshold and predicts the melt depth using the melt depth inversion model.

[0103] A fusion judgment is made by combining visual features, ultrasonic detection area, and predicted melt depth.

[0104] If the overall assessment is satisfactory, the weld quality index is recorded; if it is unsatisfactory, an alarm is triggered and the defect location is automatically marked.

[0105] Store all test data for quality traceability and process optimization.

[0106] In this embodiment, a visual recognition device is used to acquire an image of the detection area of ​​the welding assembly; an ultrasonic testing device is used to scan along the weld seam to acquire the ultrasonic testing area of ​​the detection area; wherein, the ultrasonic testing area is the welding area between the second welded component and the recessed bottom wall; the ultrasonic testing area is compared with a dynamic area threshold; if the ultrasonic testing area is greater than or equal to the dynamic area threshold, the weld seam quality is determined to be qualified; if the ultrasonic testing area is less than the dynamic area threshold, the weld seam quality is determined to be unqualified; wherein, the dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld seam length. This invention can further improve the accuracy, adaptability, and intelligence level of weld seam quality inspection.

[0107] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the weld quality detection method provided in this invention and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0108] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for inspecting weld quality, characterized in that, For detecting the welding quality of a welding assembly, the welding assembly includes a first welded component and a second welded component welded together. The first welded component has a recessed step, and a portion of the second welded component is located within the recessed step. The recessed step includes a recessed sidewall and a recessed bottom wall connected to each other. The second welded component is welded to both the recessed sidewall and the recessed bottom wall. The detection method includes: An image of the detection area of ​​the welding assembly is acquired using a visual recognition device; An ultrasonic testing device is used to scan along the weld seam to obtain the ultrasonic testing area of ​​the testing area; wherein, the ultrasonic testing area is the welding area between the second welded part and the bottom wall of the depression; The ultrasonic detection area is compared with the dynamic area threshold; if the ultrasonic detection area is greater than or equal to the dynamic area threshold, the weld quality is determined to be qualified. If the ultrasonic detection area is less than the dynamic area threshold, the weld quality is deemed unqualified. The dynamic area threshold is dynamically determined based on the statistical parameters of historical qualified samples and the current weld length.

2. The detection method according to claim 1, characterized in that, The dynamic area threshold satisfy: in, The statistical average width of historical qualified samples. The width standard deviation of historical qualified samples. Here is the confidence coefficient. This represents the weld length.

3. The detection method according to claim 1, characterized in that, Also includes: At least one ultrasonic feature is extracted during ultrasonic scanning, and the predicted melt depth is calculated based on the ultrasonic feature using a pre-trained regression model; the ultrasonic feature is selected from one or more of the following: bottom wall echo amplitude, signal attenuation rate, waveform correlation coefficient, and bottom wall echo broadening.

4. The detection method according to claim 3, characterized in that, The regression model is as follows: in, This is the predicted melting depth. This represents the bottom wall echo amplitude. For signal attenuation rate, The waveform correlation coefficient, To broaden the bottom wall echo, These are the model coefficients fitted by metallographic calibration.

5. The detection method according to claim 3, characterized in that, The step of determining the quality of the weld seam further includes: when the ultrasonic detection area is less than the dynamic area threshold but the predicted penetration depth is greater than or equal to the target penetration depth, it is still determined to be qualified.

6. The detection method according to claim 1, characterized in that, After acquiring an image of the detection area using a visual recognition device, visual features are extracted. These visual features include one or more of the following: weld width uniformity, weld edge straightness, and integrity of recessed steps. When determining the weld quality, the visual features are fused with the ultrasonic detection area and the dynamic area threshold for judgment.

7. The detection method according to claim 1, characterized in that, The step of using an ultrasonic testing device to scan along the weld seam includes: First, a coarse scan is performed along the weld centerline extracted by the visual recognition device to calculate the local welding area at each local location of the weld; then, areas with local welding areas lower than a local threshold are identified, and these areas are subjected to a slower and more intensive fine scan; the local threshold is determined based on a dynamic area threshold and the weld length.

8. The detection method according to claim 7, characterized in that, The local threshold satisfy: in, These are the position coordinates along the weld centerline. The midpoint of the weld. The attenuation coefficient is... .

9. The detection method according to claim 1, characterized in that, Also includes: Based on the ultrasonic testing area, predicted penetration depth, and defect identification confidence level, the weld quality index (WQI) is calculated: in, The area measured by the ultrasonic wave. This is the predicted melting depth. For the target melting depth, To determine the confidence level for defect identification, To achieve the highest confidence level, These are the weighting coefficients, and their sum is 1; Weld quality is classified into multiple grades based on WQI values.

10. A weld quality inspection system, characterized in that, The weld quality inspection system, using the inspection method of any one of claims 1 to 9, comprises: A visual recognition device for acquiring images of the inspection area of ​​the welding assembly; An ultrasonic testing device is used to scan along the weld seam to obtain the ultrasonic testing area of ​​the testing region; The data processing module is used to calculate the dynamic area threshold, compare the ultrasonic detection area with the dynamic area threshold, and output the judgment result.

Images