A visual weld detection system for automotive parts welding

Abstract

The present application relates to a kind of visual weld detection systems for automobile parts welding, belong to automobile manufacturing industry weld detection technical field, including scanning sensor, control module, identification module and database, the scanning sensor is for scanning each weld surface of part and uploads scanning data to control module;The identification module is for identifying the type of the part to be detected and reporting control module, the control module identifies each weld according to part type, control module is pre-inputed with the structural strength requirement of each part site, and according to the structural strength requirement of each part site, the reflectivity of part surface and ambient light parameter, adjust the judging standard of each weld, the control module receives the scanning data of several welds, and judges whether several welds are qualified according to adjusted judging standard;It gives consideration to guarantee operation efficiency and finished product quality.

Description

Technical Field

[0001] This invention belongs to the field of weld inspection technology in the automotive manufacturing industry, and specifically relates to a visual weld inspection system for automotive parts welding. Background Technology

[0002] A visual weld inspection system is a system that uses image processing and computer vision technology to automatically inspect and evaluate welded joints and welds. This system acquires images of the weld through cameras or other image sensors and uses analytical algorithms to determine the quality, integrity, and compliance of the weld.

[0003] Currently, the commonly used method is to inspect welds visually, classify the workpieces based on the inspection results, and then proceed to the next step. However, discontinuous operations, large workpiece size and weight, and human error can affect the work of upstream and downstream stations, thus reducing production efficiency. Furthermore, the accuracy of the operation largely depends on the individual operator's work performance. Therefore, there is an urgent need for a more automated and accurate weld surface visual recognition system. To this end, Chinese patent application CN118641542A discloses a weld visual inspection system and its control method. The system includes a workpiece conveying assembly, a frame, a first position sensor and a second position sensor for identifying the location of the weld to be inspected, and multiple cameras for capturing images of the weld. It also includes a main controller. During the process of the workpiece conveying assembly transporting the workpiece, the main controller receives a first identification signal from the first position sensor. When the first identification signal indicates that the weld to be inspected has been identified, the main controller controls each camera to capture an image including the weld. The main controller analyzes the image to determine whether the shape of the weld is abnormal. The main controller also receives a second identification signal from the second position sensor. When the second identification signal indicates that the weld to be inspected has been identified, the main controller controls each camera to turn off. The above-described solution of the present invention frees up human resources, significantly improves production efficiency, and effectively avoids errors in manual weld positioning.

[0004] However, in the automotive manufacturing industry, automotive parts are usually quite complex. The different parts that need to be welded have different functions and require different structural strengths. The criteria for judging whether the welds in different parts are qualified are also different. For parts that require strong structural strength, the welds need to be strictly judged to ensure the quality of the finished product. For parts that require weak structural strength, the welds do not need to be strictly judged to ensure work efficiency. Therefore, a visual weld inspection system for automotive parts welding is needed that can balance work efficiency and finished product quality. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides a visual weld inspection system for automotive parts welding, which balances operational efficiency and finished product quality.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A visual weld inspection system for automotive parts welding that balances operational efficiency and finished product quality includes a scanning sensor, a control module, a recognition module, and a database. The scanning sensor is used to scan the surface of each weld seam on the part and upload the scanning data to the control module.

[0008] The identification module is used to identify the type of the part to be inspected and report to the control module. The control module identifies each weld according to the part type. The control module has pre-input the structural strength requirements of each part and adjusts the judgment criteria of each weld according to the structural strength requirements of each part. After receiving the scanning data of several welds, the control module judges whether several welds are qualified according to the adjusted judgment criteria.

[0009] As a preferred embodiment of the present invention, it also includes a database for storing the structural strength requirements of various parts.

[0010] As a preferred technical solution of the present invention, the control module is pre-inputting structural strength requirements X1, X2, ..., Xn for each part and standard structural strength requirement X0. When the control module determines whether the weld with serial number n is qualified, it reduces the standard range by a factor of A1, where A1 = Xn / X0 × c + d, and c and d are pre-input constants.

[0011] As a preferred embodiment of the present invention, it further includes an input panel, which is used to input the values ​​of X1, X2, ..., Xn and X0, and to display the judgment results of whether several welds are qualified.

[0012] As a preferred technical solution of the present invention, it further includes an ambient light detection module, which is used to detect the ambient light around the scanning sensor and upload it to the control module. The control module determines whether the ambient light exceeds a threshold and narrows the standard range when it exceeds the threshold.

[0013] As a preferred embodiment of the present invention, the ambient light detection module is used to detect the ambient light around the scanning sensor and upload the light intensity data L to the control module. The control module reduces the standard range by a factor of A2, where A2 = L / L0 × e, and L0 and e are pre-input constants.

[0014] As a preferred technical solution of the present invention, it also includes a reflectivity sensor, which is used to detect the reflectivity of the scanning area of ​​the scanning sensor and upload it to the control module. The control module determines whether the reflectivity exceeds a threshold and expands the standard range when the reflectivity exceeds the threshold.

[0015] As a preferred technical solution of the present invention, the reflectivity sensor is used to detect the reflectivity F of the scanning area of ​​the scanning sensor and upload it to the control module. The standard range is expanded by A3 times, where A3 = F0 / F×y, and F0 and y are pre-input constants.

[0016] As a preferred technical solution of the present invention, the control module adjusts the judgment standard of the scanning sensor to A1×A2 / A3 times the standard range. The control module is used to determine whether A1×A2 / A3 is less than a preset threshold, and executes a manual review procedure when the judgment result is yes.

[0017] As a preferred embodiment of the present invention, the input panel is used to display the coupling relationship curve between the Xn / X0 ratio and ambient light L and reflectivity F corresponding to each weld location. The curve is automatically generated based on the statistical regularity of historical qualified weld data of the part.

[0018] The beneficial effects of this invention are as follows:

[0019] (1) The identification module is used to identify the type of the part to be inspected and report to the control module. The control module identifies each weld according to the part type. The control module has pre-input the structural strength requirements of each part and adjusts the judgment criteria of each weld according to the structural strength requirements of each part. After receiving the scanning data of several welds, the control module judges whether several welds are qualified according to the adjusted judgment criteria. This ensures that for parts that require strong structural strength, the welds are strictly judged to ensure the quality of the finished product. For parts that require weak structural strength, there is no need to strictly judge the welds to ensure work efficiency.

[0020] (2) By setting the ambient light detection module to scan the ambient light around the sensor and upload it to the control module, the control module determines whether the ambient light exceeds the threshold. When the ambient light is strong, it will interfere with the laser-based detection results, thus narrowing the standard range and reducing the probability of judging unqualified welds as qualified due to the influence of ambient light, thereby improving the accuracy of weld detection.

[0021] (3) By setting a reflectivity sensor to detect the reflectivity of the scanning area of ​​the scanning sensor and uploading it to the control module, the control module determines whether the reflectivity exceeds the threshold. When the reflectivity exceeds the threshold, the standard range is expanded. When the reflectivity is weak and will interfere with the laser-based detection results, the standard range is reduced, reducing the probability of judging unqualified welds as qualified due to the influence of reflectivity, and further improving the accuracy of weld detection. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the overall system architecture of the present invention;

[0024] Figure 2 This is a schematic diagram illustrating the working principle of the scanning sensor of the present invention;

[0025] Figure 3 This is a schematic diagram of the adjustment process of the weld judgment criteria by the control module of the present invention;

[0026] Figure 4 This is a schematic diagram illustrating the function of the ambient light detection module and the reflectivity sensor of the present invention.

[0027] Figure 5 This is a schematic diagram illustrating the working process of the ambient light detection module and reflectivity sensor of the present invention. Detailed Implementation

[0028] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0029] refer to Figures 1-5 A visual weld inspection system for welding automotive parts includes a scanning sensor, a control module, a recognition module, and a database. The scanning sensor is used to scan the surface of each weld seam of the part and upload the scanning data to the control module.

[0030] like Figure 1 As shown, the core components of a visual weld inspection system for automotive parts welding include a scanning sensor, a control module, a recognition module, a database, an input panel, an ambient light detection module, and a reflectivity sensor. The scanning sensor includes at least a camera and a laser. The components are interconnected via data lines or signal lines to form a complete inspection system.

[0031] like Figure 2 As shown, the working principle of the scanning sensor is as follows: a laser emits light, which passes through a circular lens to form a line laser, which is then projected onto the surface of the weld to be inspected. A camera is installed at a certain angle to the laser to capture the reflected light from the line laser on the weld surface. Different heights on the weld surface cause the position of the reflected light in the camera's field of view to change, thereby enabling the measurement of the three-dimensional contour.

[0032] Specifically, the scanning sensor includes a camera and a laser, both of which are electrically connected to the control module. Depending on the actual application scenario, the scanning sensor, including the camera and laser, is optimized for specific processor speeds and welding conditions. Meanwhile, the control module also includes an image processing module. The sensor unit works with the image processing module. In this mode, the system uses previously established inspection criteria to evaluate the welding quality.

[0033] The measurement principle using a laser is as follows: The light source used in the measurement includes a laser source with a circular lens. In this embodiment, the laser source is a line laser. The camera is at a certain angle to the light source. A line of light illuminates the scanned object in the camera's field of view. This line of light will be displaced and distorted, propelling the object to different heights along the line. The angle between the camera and the light source, i.e., the triangulation angle, determines the degree of displacement and distortion of the light in the height direction. The larger the angle, the greater the magnification. Since the triangulation angle is known, the three-dimensional points on the object can be inferred from the position of the light, thereby inferring the object's outline. At the same time, the control module pre-inputs a standard size range for the outline, i.e., the standard range. Then, the size of the object's outline is compared with the standard size. Based on whether the detected weld outline size falls within the standard range, it is determined whether it is qualified. When the weld outline size falls within the standard range, it is determined to be qualified; otherwise, it is unqualified.

[0034] Meanwhile, the control module pre-inputs possible weld shapes, fillet welds, or other weld shapes and possible welding processes, such as metal inert gas welding (MIG) / metal active gas welding (MAG) welds, laser welds, steel-aluminum dissimilar material welds, etc., as well as weld length, weld height, holes, porosity, pore spots, edge notches, weld location, weld volume, weld length, protrusions, depressions, weld thickness, symmetry, unfilled end craters, and others;

[0035] In this embodiment, the detection speed is 6-50 m / min, the scanning speed (beam frequency) is 345-2805 Hz, the resolution is 10 micrometers, the laser device laser class (compliant with EN 60825-1:2008 standard) is Class 3B (corresponding to sensor model 103490) or Class 2 (corresponding to sensor model 106719); the laser output power is ≤40mW; the safe distance (NOHD) is 1400mm.

[0036] In the automotive manufacturing industry, automotive parts are usually quite complex. The different parts that need to be welded have different functions and require different structural strengths. The criteria for judging whether the welds in different parts are qualified are different. For parts that require strong structural strength, the welds are strictly judged to ensure the quality of the finished product. For parts that require weak structural strength, there is no need to strictly judge the welds to ensure work efficiency.

[0037] Therefore, the identification module is used to identify the type of the part to be inspected and report to the control module. The control module identifies each weld according to the part type. The control module has pre-input the structural strength requirements of each part and adjusts the judgment criteria of each weld according to the structural strength requirements of each part. After receiving the scan data of several welds, the control module judges whether several welds are qualified according to the adjusted judgment criteria.

[0038] like Figure 3 As shown, the control module adjusts the weld judgment criteria based on the part type, weld location, and structural strength requirements. First, the identification module identifies the part type and reports it to the control module; then, the control module identifies each weld according to the part type and adjusts the weld judgment criteria according to the pre-input structural strength requirements for each part; finally, the control module performs a pass / fail judgment on the weld scanning data based on the adjusted judgment criteria.

[0039] Specifically, after reading each part, the control module assigns a serial number to each potential weld area. The serial number is a single-point incrementing natural number 1, 2, ..., n. At the same time, the control module adjusts the passing order of different parts of the current car part with the scanning sensor, thereby adjusting the passing order of different welds with the scanning sensor. By adjusting the passing order of different parts of the current car part with the scanning sensor, the control module ensures that the welds pass through the scanning sensor in ascending order according to the weld numbers in the database.

[0040] At this point, the first weld scan data received by the control module is weld number one, the second weld scan data is weld number two, and so on.

[0041] The control module can then know which weld number the current scan data is. When the control module judges whether a certain weld number meets the standard, it simultaneously calls the structural strength requirements of the corresponding position of this number. In turn, when judging welds in different locations, it simultaneously adjusts the judgment standard of the weld.

[0042] Specifically, the control module is pre-inputted with the structural strength requirements X1, X2, ..., Xn for each part and the standard structural strength requirement X0. When the control module judges whether the weld with the number n is qualified, it reduces the standard range by a factor of A1, where A1 = Xn / X0 × c + d, and c and d are pre-input constants.

[0043] When the structural strength requirement Xn of a certain part is large, it means that even a small deviation in the weld shape has a high probability of causing a decrease in structural strength. It is necessary to narrow the standard range. At this time, the value of A1=Xn / X0×c+d is large, the reduction factor is large, and the weld inspection result is more likely to fall outside the judgment standard and be judged as unqualified. Therefore, for parts that require strong structural strength, the weld should be strictly judged.

[0044] When the structural strength requirement Xn of a certain part is small, the weld shape deviation has little impact on the shape. There is no need to narrow the judgment standard too much, so as not to affect production efficiency. At this time, the value of A1=Xn / X0×c+d is small, the reduction factor is small, and the probability that the weld inspection result will fall outside the judgment standard and be judged as unqualified is small. Therefore, the judgment standard for welds can be relaxed for parts with weaker required structural strength.

[0045] The identification module identifies the type of the part to be inspected and reports it to the control module. The control module identifies each weld according to the part type. The control module has pre-input the structural strength requirements of each part and adjusts the judgment criteria of each weld according to the structural strength requirements of each part. After receiving the scanning data of several welds, the control module judges whether several welds are qualified according to the adjusted judgment criteria. This allows for strict judgment of welds in parts that require strong structural strength to ensure the quality of finished products, while no strict judgment is required for parts that require weak structural strength, thus ensuring work efficiency.

[0046] To facilitate the storage of data on the types of parts and the structural strength requirements of each part, a database is also included, which is used to store the structural strength requirements of each part.

[0047] To facilitate the input of values ​​for X1, X2, ..., Xn and X0, an input panel is also included. The input panel is used to input the values ​​for X1, X2, ..., Xn and X0, and to display the judgment results of whether several welds are qualified.

[0048] In the above process, the judgment of whether the threshold is exceeded is affected by ambient light. When the ambient light is strong, it will interfere with the laser-based detection results, causing unqualified welds to be judged as qualified. In order to reduce the probability of unqualified welds being judged as qualified due to the influence of ambient light, an ambient light detection module is also included. The ambient light detection module is used to detect the ambient light around the scanning sensor and upload it to the control module. The control module judges whether the ambient light exceeds the threshold and narrows the standard range when it exceeds the threshold.

[0049] like Figure 4-5As shown, the ambient light detection module and reflectivity sensor play different roles in the visual weld inspection system: The ambient light detection module detects the intensity of ambient light around the scanning sensor and uploads the data to the control module. This allows for narrowing the judgment criteria and improving detection accuracy when the ambient light is too strong. The reflectivity sensor detects the reflectivity of the scanning area and also uploads the data to the control module. This allows for narrowing the judgment criteria and improving the detection standard when the reflectivity is high, and widening the judgment criteria and reducing the false positive rate when the reflectivity is low.

[0050] Specifically, the ambient light detection module is used to detect the ambient light around the scanning sensor and upload the light intensity data L to the control module. The control module reduces the standard range by a factor of A2, where A2 = L / L0 × e, and L0 and e are pre-input constants.

[0051] When the ambient light is strong, it will interfere with the laser-based detection results. There is a probability that the unqualified weld will be judged as qualified due to the influence of the ambient light. It is necessary to tighten the judgment standard. At this time, the value of A2=L / L0×e is large, thus completing the tightening of the judgment standard when the light intensity is strong.

[0052] When the ambient light is weak, the interference with the laser-based detection results is small, and the probability of judging unqualified welds as qualified is low. There is no need to tighten the judgment criteria. At this time, the value of A2 = L / L0 × e is small, so the judgment criteria are not tightened to a large extent when the light intensity is strong.

[0053] By setting up an ambient light detection module to collect the ambient light around the scanning sensor and upload it to the control module, the control module can determine whether the ambient light exceeds the threshold. When the ambient light is strong and interferes with the laser-based detection results, the standard range is narrowed, reducing the probability of judging unqualified welds as qualified due to the influence of ambient light, thus improving the detection accuracy.

[0054] In the above process, when the reflectivity of the material where the weld is located is high, more light information will be reflected to the camera. At this time, the reflected light received by the camera can represent the actual surface shape to a greater extent. At this time, the probability of the detection result based on laser is small, and it is necessary to narrow the standard range and improve the judgment standard. For this purpose, a reflectivity sensor is also included. The reflectivity sensor is used to detect the reflectivity of the scanning area of ​​the scanning sensor and upload it to the control module. The control module determines whether the reflectivity exceeds the threshold and narrows the standard range when the reflectivity exceeds the threshold.

[0055] Specifically, the reflectivity sensor is used to detect the reflectivity F of the scanning area of ​​the scanning sensor and upload it to the control module. The control module expands the standard range by A3 times, where A3 = F0 / F×y, and F0 and y are pre-input constants.

[0056] When the reflectivity is strong, it means that the probability of the laser-based detection result being deviated is small, and the standard range needs to be increased. At this time, the value of A3 = F0 / F×y is small, less than 1. When the control module expands the standard range by A3 times, it completes the reduction of the standard range when the reflectivity is strong, and the judgment standard is improved.

[0057] When the reflectivity is weak, it means that the probability of the laser-based detection result being deviated is high, and the standard range needs to be expanded to prevent inaccurate detection. At this time, the value of A3 = F0 / F × y is large, greater than 1. When the control module expands the standard range by A3 times, the expansion of the standard range is completed when the reflectivity is strong.

[0058] By narrowing the standard range when the reflectivity is high and widening the standard range when the reflectivity is low, the system can ensure that the reflected light received by the camera can better represent the actual surface shape. When the probability of deviation in laser-based detection results is low, the judgment standard is increased, and when the probability of deviation in laser-based detection results is high, the judgment standard is decreased, thereby reducing the false judgment rate.

[0059] Furthermore, the control module comprehensively adjusts the standard range through a composite adjustment coefficient A. While maintaining the original three adjustment parameters (structural strength, ambient light, and reflectivity), it establishes a composite algorithm to achieve multi-factor collaborative control. At the same time, a safety mechanism is set to avoid excessive automation. Specifically, the control module adjusts the judgment standard of the scanning sensor to A1×A2 / A3 times the standard range.

[0060] When the composite adjustment coefficient A1×A2 / A3 is less than the preset threshold, it means that the judgment range caused by the automatic adjustment is too small and has limited effect on production guidance. At this time, manual correction is required. Therefore, when the control module judges that A1×A2 / A3 is less than the preset threshold, it triggers the manual review mechanism. Specifically, when the manual review mechanism is triggered, the control module sends out the corresponding sound and light signals and displays the review prompt information on the input panel.

[0061] Furthermore, it also includes an input panel with a dynamic parameter configuration interface, which extends the human-computer interaction dimension, visualizes the parameter adjustment process and associates it with historical data, and can display and adjust the coupling relationship curve between the Xn / X0 ratio and ambient light L and reflectivity F for each weld location in real time. The curve is automatically generated based on the statistical rules of the historical qualified weld data of the part.

[0062] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A visual weld inspection system for automotive parts welding, characterized in that: It includes a scanning sensor, a control module, an identification module, and a database. The scanning sensor is used to scan the surface of each weld seam of the part and upload the scanning data to the control module. The identification module is used to identify the type of the part to be inspected and report to the control module. The control module identifies each weld according to the part type. The control module has pre-input the structural strength requirements of each part and adjusts the judgment criteria of each weld according to the structural strength requirements of each part. After receiving the scanning data of several welds, the control module judges whether several welds are qualified according to the adjusted judgment criteria. It also includes a database for storing the structural strength requirements of each part; the control module is pre-inputted with the structural strength requirements X1, X2, ..., Xn of each part and the standard structural strength requirement X0; when the control module judges whether the weld with the number n is qualified, it reduces the standard range by a factor of A1, where A1 = Xn / X0 × c + d, and c and d are pre-input constants; It also includes an input panel, which is used to input the values ​​of X1, X2, ..., Xn and X0, and to display the judgment results of whether several welds are qualified; It also includes an ambient light detection module, which is used to detect the ambient light around the scanning sensor and upload it to the control module. The control module determines whether the ambient light exceeds a threshold and narrows the standard range when it exceeds the threshold. The ambient light detection module is used to detect the ambient light around the scanning sensor and upload the light intensity data L to the control module. The control module reduces the standard range by a factor of A2, where A2 = L / L0 × e, and L0 and e are pre-input constants.

2. The visual weld inspection system for automotive parts welding according to claim 1, characterized in that: It also includes a reflectivity sensor, which is used to detect the reflectivity of the scanning area of ​​the scanning sensor and upload it to the control module. The control module determines whether the reflectivity exceeds a threshold and expands the standard range when the reflectivity exceeds the threshold.

3. The visual weld inspection system for automotive parts welding according to claim 2, characterized in that: The reflectivity sensor is used to detect the reflectivity F of the scanning area of ​​the scanning sensor and upload it to the control module. The control module expands the standard range by A3 times, where A3 = F0 / F × y, and F0 and y are pre-input constants.

4. The visual weld inspection system for automotive parts welding according to claim 3, characterized in that: The control module adjusts the judgment standard of the scanning sensor to A1×A2 / A3 times the standard range. The control module is used to determine whether A1×A2 / A3 is less than the preset threshold, and executes the manual review procedure when the judgment result is yes.

5. A visual weld inspection system for automotive parts welding according to claim 4, characterized in that: It also includes an input panel, which is used to display the coupling relationship curve between the Xn / X0 ratio and ambient light L and reflectivity F for each weld location. The curve is automatically generated based on the statistical regularity of historical qualified weld data of the part.

Images