Truck steel structure quality prediction method and system based on girder welding result

Abstract

The invention discloses a truck steel structure quality prediction method and system based on a girder welding result, and relates to the technical field of truck steel structure welding quality prediction. The invention provides a truck steel structure quality prediction method based on a girder welding result, and the method comprises the steps: obtaining the welding spot positions of a welding seam in order according to a welding seam image of a girder welding position; according to the positions of the welding spots, the interval distribution characteristics of the adjacent welding spots are obtained; the first welding quality of the weld joint is evaluated according to the interval distribution characteristics; according to the welding spot positions, obtaining at least one welding seam image block containing a preset number of welding spots, and performing welding seam defect identification on the welding seam image block to obtain defect types of the welding seam image; acquiring defect degrees of all the weld joint blocks according to the defect types; evaluating the second welding quality of the weld joint according to the defect degree; and predicting the truck steel structure quality according to the first welding quality and the second welding quality.

Description

Technical Field

[0001] The present application relates to the technical field of truck steel structure welding quality prediction, and in particular to a truck steel structure quality prediction method and system based on beam welding results. Background Art

[0002] In modern industrial production, trucks are important logistics transportation tools, and the quality of their steel structures is directly related to the safety, durability and transportation efficiency of the vehicles. In particular, the beam part of the truck, as a key structure that bears the weight of the vehicle body and cargo load, its welding quality is even more important. However, due to the presence of multiple variables in the welding process, such as welding parameters, material properties, operating techniques, etc., the quality of the beam weld is often difficult to accurately evaluate through a single method.

[0003] Traditional truck beam welding quality inspection usually relies on manual visual inspection or simple measuring tools. These methods are not only inefficient, but also greatly affected by human factors, making it difficult to ensure the objectivity and accuracy of the inspection results. Summary of the invention

[0004] The main purpose of this application is to provide a truck steel structure quality prediction method and system based on beam welding results, aiming to solve the technical problem that traditional truck beam welding quality inspection usually relies on manual visual inspection or simple measuring tools, which are not only inefficient but also difficult to detect.

[0005] To achieve the above objectives, in a first aspect, the present application provides a method for predicting the quality of a truck steel structure based on beam welding results, comprising:

[0006] According to the weld image of the beam welding position, the weld point positions of the weld are obtained in order;

[0007] According to the positions of the solder joints, obtaining the spacing distribution characteristics of adjacent solder joints;

[0008] evaluating a first welding quality of the weld according to the spacing distribution characteristic;

[0009] According to the positions of the weld spots, obtaining at least one weld image block including a preset number of weld spots, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image;

[0010] According to the defect category, the defect degree of all weld seam blocks is obtained;

[0011] According to the defect degree, evaluating the second welding quality of the weld;

[0012] The quality of the steel structure of the truck is predicted according to the first welding quality and the second welding quality.

[0013] Optionally, the step of obtaining the spacing distribution characteristics of adjacent solder joints according to the solder joint positions includes:

[0014] According to the arrangement order of the welding points on the weld, two adjacent welding points are paired to obtain paired welding points, and the Euclidean distance between the center points of each group of paired welding points is used as the welding point spacing;

[0015] The average weld point spacing of the weld is obtained according to the weld point distances of all paired weld points, and the weld point spacing of each group is compared with the average weld point spacing to obtain the spacing distribution characteristics of the paired weld points, wherein the spacing distribution characteristics characterize the weld point uniformity of the weld.

[0016] Optionally, the step of obtaining an average weld point spacing of a weld according to the weld point distances of all paired weld points, and comparing each group of weld point spacings with the average weld point spacing to obtain the spacing distribution characteristics of the paired weld points includes:

[0017] According to the solder joint distances of all paired solder joints, the maximum solder joint spacing, the minimum solder joint spacing and the average solder joint spacing are obtained;

[0018] Get the maximum solder point spacing, the minimum solder point spacing and the average solder point spacing, and get the normal solder point spacing range;

[0019] Compare the spacing between each group of solder joints with the normal spacing between solder joints, obtain abnormal paired solder joints, and send abnormal information.

[0020] Optionally, the step of evaluating the first welding quality of the weld according to the spacing distribution characteristic comprises:

[0021] Formula for assessing the first weld quality of a weld:

[0022]

[0023] Among them, d i It represents the distance between the i-th and i+1-th welding points. There are n welding points in total. D represents the average welding gap between adjacent welding points. K is the coefficient.

[0024] Optionally, the step of acquiring at least one weld image block including a preset number of welds according to the weld position, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image includes:

[0025] Process the weld image to obtain a grayscale image, calculate the grayscale co-occurrence matrix of the grayscale image, and provide texture features;

[0026] A threshold is set according to grayscale value or texture feature, the grayscale image is segmented into foreground and background, and edge information of defects in the foreground is detected by an edge detection operator;

[0027] The defect category of the weld image is obtained according to the edge information.

[0028] Optionally, the step of obtaining the defect levels of all weld seam images according to the defect categories includes:

[0029] Set the window parameters for each defect category. When the defect category is pore and spatter defects, the calculation formula for the defect degree is:

[0030]

[0031] Among them, w i represents the weight of the i-th defect; I(x, y) represents the gray value of the pixel at position (x, y) in the weld image; μ represents the mean gray value of the weld image; K i (x, y) represents the Gaussian kernel function value corresponding to the i-th defect centered at (x, y).

[0032] Optionally, the step of obtaining the defect levels of all weld seam images according to the defect categories includes:

[0033] Set the window parameters for each defect category. When the defect category is under-welding and defect, the calculation formula for the defect degree is:

[0034]

[0035] Among them, w i represents the weight of the i-th defect, ΔI i (x, y) represents the gray value difference of the i-th defect at the (x, y) position, ΔI weld (x, y) represents the gray value difference of the weld area at the position (x, y), and γ represents a constant.

[0036] Optionally, the step of evaluating the second welding quality of the weld according to the defect degree includes:

[0037] According to the defect degree of all weld blocks, weighted summary is performed according to defect type to evaluate the second welding quality of the weld.

[0038] Optionally, the step of predicting the quality of the steel structure of the truck according to the first welding quality and the second welding quality includes:

[0039] A weighted calculation is performed according to the first welding quality and the second welding quality to obtain a comprehensive score, and the quality of the truck steel structure is characterized by the comprehensive score.

[0040] In a second aspect, the present application provides a truck steel structure quality prediction system based on beam welding results, comprising:

[0041] A welding point position acquisition module, which is configured to acquire welding point positions of the welds arranged in order according to the weld image of the beam welding position;

[0042] A spacing distribution characteristic acquisition module, which is configured to acquire the spacing distribution characteristics of adjacent solder joints according to the solder joint positions;

[0043] A first welding quality assessment module, configured to assess a first welding quality of the weld according to the spacing distribution characteristic;

[0044] A defect category acquisition module, configured to acquire at least one weld image block including a preset number of welds according to the weld position, perform weld defect recognition on the weld image block, and obtain a defect category of the weld image;

[0045] A defect degree acquisition module, which is configured to acquire the defect degrees of all weld seam blocks according to the defect category;

[0046] A second welding quality acquisition module, configured to evaluate a second welding quality of the weld according to the defect degree;

[0047] A truck steel structure quality prediction module is configured to predict the truck steel structure quality according to the first welding quality and the second welding quality.

[0048] Beneficial effects that this application can achieve:

[0049] The present application proposes a truck steel structure quality prediction method and system based on beam welding results, which provides a comprehensive evaluation of weld quality by comprehensively considering the weld point location, spacing distribution characteristics, and the type and degree of weld defects. Compared with traditional single feature analysis, this multi-dimensional analysis method can more accurately reflect the actual quality status of the weld, thereby improving the accuracy of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of a truck steel structure quality prediction method based on beam welding results in this application;

[0051] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0052] 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.

[0053] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0054] In the present invention, unless otherwise clearly specified and limited, the terms "connection", "fixation", etc. should be understood in a broad sense. For example, "fixation" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0055] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the meaning of "and / or" appearing in the full text includes three parallel schemes. Taking "A and / or B" as an example, it includes scheme A, or scheme B, or a scheme that satisfies both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in the field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0056] Example 1

[0057] Reference Figure 1 The present application embodiment provides a truck steel structure quality prediction method based on beam welding results, comprising the following steps:

[0058] S10. According to the weld image of the beam welding position, the weld point positions of the weld are obtained in order.

[0059] Optionally, the beam is welded to the truck bed to form the truck steel structure, and several welding points form the weld. Select appropriate industrial cameras and lenses to ensure that they can meet the resolution and clarity requirements of weld image acquisition. According to the welding process and drawings, determine the specific position and direction of the weld. Use a robotic arm, guide rail or other positioning device to accurately move the industrial camera above the weld to ensure that the camera lens maintains an appropriate distance and angle with the weld. After acquiring the weld image, the weld image is preprocessed, including denoising, contrast enhancement, distortion correction, etc., to ensure that the image quality meets the requirements of subsequent processing. Use image processing algorithms (such as edge detection, template matching, etc.) to detect the weld position in the preprocessed weld image. According to the actual position of the weld on the weld, the detected welds are sorted to form a list of weld positions arranged in order.

[0060] S20: Obtaining spacing distribution characteristics of adjacent solder joints according to the solder joint positions.

[0061] Optionally, the solder joint position list is traversed to calculate the spacing between adjacent solder joints. The calculated spacing is statistically analyzed, such as calculating the average spacing, the standard deviation of the spacing, the distribution form of the spacing, etc., to obtain the spacing distribution characteristics.

[0062] S30. Evaluate a first welding quality of the weld according to the spacing distribution characteristic.

[0063] Optionally, based on industry experience or experimental data, formulate evaluation standards for spacing distribution characteristics, such as the allowable range of spacing, the maximum value of standard deviation, etc. Compare the actual spacing distribution characteristics with the evaluation standards to evaluate the first welding quality of the weld, such as judging whether the weld points of the weld are uniform, whether there is excessive spacing fluctuation, etc. By analyzing the position of the weld points, the first welding quality of the weld is characterized by the uniformity of the weld points. The reasons for uneven weld points may be unstable welding speed, unstable movement of the welding gun, etc. The uniformity of the weld points can be further used to guide the setting of welding parameters, etc.

[0064] S40, acquiring at least one weld image block including a preset number of welds according to the weld point positions, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image.

[0065] Optionally, according to the weld spot positions and the preset tile size, a weld tile containing a preset number of weld spots is extracted from the weld image. Defect recognition is performed on the weld tile using a machine learning algorithm or image processing technology (such as convolutional neural network, morphological processing, etc.) to identify defect categories such as pores, spatter, under-welding, and weld nodules.

[0066] S50: Obtain the defect levels of all weld seam blocks according to the defect categories.

[0067] Optionally, the weld block is further analyzed. For different defect types and defect locations, the defect degree of the weld is further subdivided to accurately analyze the defect degree.

[0068] S60: Evaluate the second welding quality of the weld according to the defect degree.

[0069] Optionally, according to the defect type and degree, an evaluation standard for the second welding quality of the weld is formulated, such as the maximum allowable degree of defects, the maximum number of defects, etc. The actual defect degree is compared with the evaluation standard to evaluate the second welding quality of the weld.

[0070] S70. Predict the quality of the truck steel structure according to the first welding quality and the second welding quality.

[0071] Optionally, the first welding quality and the second welding quality are comprehensively considered to comprehensively evaluate the overall quality of the weld. According to the overall quality of the weld, combined with the stress analysis and fatigue life prediction of the truck steel structure, the overall quality of the truck steel structure is predicted, such as whether the truck steel structure meets the design requirements and whether there are safety hazards.

[0072] In this embodiment, by comprehensively considering the consistency of the weld spacing and the condition of the weld defects, the weld quality can be evaluated more comprehensively, thereby improving the accuracy of the truck steel structure quality prediction. Timely quality prediction helps to discover potential problems in the production process, guide production adjustments, reduce defective rates, optimize production processes, and reduce costs. Accurate truck steel structure quality prediction can ensure the reliability and safety of the vehicle structure, reduce the risk of safety accidents caused by structural problems, and provide strong support for the manufacturing and quality control of truck steel structures.

[0073] Example 2

[0074] Based on Example 1, this example provides a method for predicting the quality of a truck steel structure based on beam welding results, comprising the following steps:

[0075] S10. According to the weld image of the beam welding position, the weld point positions of the weld are obtained in order.

[0076] S20: Obtaining spacing distribution characteristics of adjacent solder joints according to the solder joint positions.

[0077] Optionally, the step of obtaining the spacing distribution characteristics of adjacent solder joints according to the solder joint positions includes:

[0078] S201. Pair two adjacent welds according to the arrangement order of the welds on the weld to obtain paired welds, and use the Euclidean distance between the center points of each group of paired welds as the weld spacing.

[0079] Specifically, the positions of all welds are accurately identified in the weld image through image processing techniques (such as edge detection, template matching or deep learning algorithms), and the welds are sorted according to the welding order or the spatial position relationship in the image. According to the arrangement order of the welds, two adjacent welds are grouped into a pair to form paired welds. Ensure that each pair of welds is directly adjacent to each other on the weld without omission or duplication. For each group of paired welds, the Euclidean distance between their center points, that is, the weld spacing, is calculated. The Euclidean distance is the most direct and commonly used distance measurement method between two points, which can accurately reflect the actual interval between welds.

[0080] Through solder point pairing and spacing calculation, the distribution of solder points on the weld can be quantified, providing basic data for subsequent spacing distribution characteristics analysis. Accurate solder point identification and sorting are the key to ensuring the correctness of spacing calculation, which helps reduce errors and misjudgments.

[0081] S202, obtaining an average weld point spacing of the weld according to the weld point distances of all paired weld points, comparing each group of weld point spacing with the average weld point spacing, and obtaining spacing distribution characteristics of the paired weld points, wherein the spacing distribution characteristics characterize the weld point uniformity of the weld.

[0082] Specifically, the average solder joint spacing of the weld is calculated based on the solder joint spacing of all paired solder joints. This can be achieved by adding up all solder joint spacings and dividing by the number of paired solder joints. The solder joint spacing of each group is compared with the average solder joint spacing to analyze the discrete degree, fluctuation range and other characteristics of the solder joint spacing. These characteristics can reflect the uniformity of the solder joints of the weld, that is, whether the solder joint spacing is consistent, whether there is abnormal spacing, etc. Based on the comparison results, the distribution characteristics of the solder joint spacing are quantitatively characterized, such as calculating statistical indicators such as standard deviation and coefficient of variation, or drawing a spacing distribution diagram to intuitively display the distribution of solder joints.

[0083] Through the analysis of the spacing distribution characteristics, potential problems such as uneven distribution of weld points in the weld, too large or too small spacing, etc. can be discovered in time, providing a strong basis for production adjustments. From the perspective of the weld point position alone, the more uniform the weld point distribution, the higher the welding quality. Accurate spacing distribution characteristics analysis helps to optimize welding process parameters, improve weld quality and production efficiency, and reduce costs and defective rates.

[0084] Optionally, the step of obtaining an average weld point spacing of a weld according to the weld point distances of all paired weld points, and comparing each group of weld point spacings with the average weld point spacing to obtain the spacing distribution characteristics of the paired weld points includes:

[0085] S2021. Obtain a maximum solder point spacing, a minimum solder point spacing, and an average solder point spacing based on the solder point distances of all paired solder points.

[0086] Specifically, on the basis of calculating the distance between the weld points of all paired weld points, the maximum, minimum and average values ​​of these distances are further counted. The maximum distance between weld points indicates the maximum distance between adjacent weld points in the weld. The minimum distance between weld points indicates the minimum distance between adjacent weld points in the weld. The average distance between weld points is the sum of all the distances between weld points divided by the number of paired weld points, which is used to reflect the overall level of the distance between weld points. By counting the maximum, minimum and average values ​​of the distance between weld points, we can preliminarily understand the overall distribution of weld points in the weld, laying the foundation for the subsequent analysis of the distance distribution characteristics.

[0087] S2022. Obtain a maximum solder point spacing, a minimum solder point spacing, and an average solder point spacing, and obtain a normal solder point spacing range.

[0088] Specifically, based on the statistically obtained maximum solder joint spacing, minimum solder joint spacing, and average solder joint spacing, combined with production process requirements and welding standards, a reasonable normal solder joint spacing interval is determined. This interval can be adjusted according to actual conditions to ensure that it can include most normal solder joint spacings and effectively identify abnormal solder joint spacings. Determining the normal solder joint spacing interval is the basis for identifying abnormal solder joint spacings, which helps to accurately distinguish normal and abnormal solder joint pairings in subsequent steps.

[0089] S2023, comparing the spacing between each group of solder joints with the normal spacing between solder joints, obtaining abnormal paired solder joints, and sending abnormal information.

[0090] Specifically, the spacing between each group of welds is compared with the normal spacing between welds to determine whether the spacing between each group of welds falls within the interval. If the spacing between a group of welds is not within the normal interval, the group of welds is marked as abnormal paired welds. For the identified abnormal paired welds, the system should send abnormal information, such as notifying relevant personnel to handle it through interface display, sound alarm or sending email. By identifying abnormal paired welds, possible quality problems in the weld can be discovered in time, such as uneven distribution of welds and welding deviation. Sending abnormal information helps to promptly remind relevant personnel to take measures to intervene and avoid potential quality risks and safety hazards.

[0091] S30. Evaluate a first welding quality of the weld according to the spacing distribution characteristic.

[0092] Optionally, the step of evaluating the first welding quality of the weld according to the spacing distribution characteristic includes:

[0093] Formula for assessing the first weld quality of a weld:

[0094]

[0095] Among them, d iIt represents the distance between the i-th and i+1-th welding points. There are n welding points in total. D represents the average welding gap between adjacent welding points. K is the coefficient.

[0096] Specifically, the denominator in the above formula calculates the standard deviation (or square root of mean square error) of the difference between the distance between the welds and the average weld gap, which reflects the degree of discreteness of the distance between the welds. When the distance between the welds is closer to the average weld gap, the smaller the standard deviation, the smaller the denominator, and thus the larger the evaluation value S, indicating that the welds of the weld are more evenly distributed and the welding quality is higher. The coefficient K is used to adjust the numerical range of the evaluation result to make it more in line with the needs of practical applications. This formula provides a quantitative method to evaluate the uniformity of the weld distribution of the weld, thereby indirectly reflecting the welding quality. It is easy to operate by simply calculating the distance between the welds and the average weld gap and then substituting them into the formula to obtain the evaluation result. Compared with visual observation or simple measurement alone, this formula can more accurately reflect the degree of discreteness of the weld distribution, thereby improving the accuracy of welding quality evaluation.

[0097] S40, acquiring at least one weld image block including a preset number of welds according to the weld point positions, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image.

[0098] Optionally, the step of acquiring at least one weld image block including a preset number of welds according to the weld position, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image includes:

[0099] S401, processing the weld image block, obtaining a grayscale image, calculating the grayscale co-occurrence matrix of the grayscale image, and providing texture features.

[0100] Specifically, first, the weld image is converted from the original color image to a grayscale image. This is to simplify subsequent processing while retaining the main structural information of the image. Use an appropriate grayscale algorithm, such as the weighted average method, to ensure that the grayscale image can accurately reflect the color information of the original image. Gray-Level Co-occurrence Matrix (GLCM) is a method to describe image texture. Appropriate distance and direction parameters are selected to calculate the GLCM to capture the texture features in the weld image. Texture features such as contrast, energy, and entropy are extracted from the GLCM, which will be used for subsequent defect identification.

[0101] S402 , setting a threshold according to grayscale values ​​or texture features, segmenting the grayscale image into a foreground and a background, and detecting edge information of defects in the foreground by using an edge detection operator.

[0102] Specifically, a threshold is set based on the grayscale value or the texture features extracted previously. A threshold segmentation method (such as the Otsu algorithm) is used to segment the grayscale image into the foreground (weld area) and the background. An edge detection operator (such as the Canny operator, the Sobel operator, or the Prewitt operator) is applied to detect the edge information of the defect in the foreground. The edge detection operator can identify places in the image where the grayscale value changes dramatically, which are usually the edges of the defect.

[0103] S403: Obtain a defect category of the weld image according to the edge information.

[0104] Specifically, the features of the defects, such as shape, size, direction, etc., are extracted from the detected edge information. A machine learning algorithm (such as a support vector machine, a decision tree, or a neural network) or a template matching method is used to classify the defects according to the extracted features. Welding defects include under-welding, pores, weld nodules, and spatters. For image blocks with under-welding or weld nodules, the defect category is set to the first category, for image blocks with spatters, the defect category is set to the second category, and for image blocks with pores, the defect category is set to category three.

[0105] Porosity refers to the holes formed in the weld metal due to the inability of gas to escape. For example, in the manual arc welding process, if the shielding gas contains air or the welding materials (such as welding wire and flux) contain too much water, pores may form in the weld. Porosity may appear in the form of uniform pores, surface pores, worm-like pores, etc., which seriously affect the density and strength of the weld. Porosity reduces the effective working section of the weld and reduces the mechanical strength of the joint. In weldments that require sealing, pores may also affect the sealing of the weldment.

[0106] Spatter refers to the phenomenon that molten metal particles splash out of the weld during welding. For example, in resistance spot welding, if the welding current is too large or the electrode pressure is insufficient, the liquid molten core may break through the plastic metal ring around the weld after being compressed, causing spatter. Spatter can be divided into surface spatter and internal spatter. The former affects the surface quality of the weld, while the latter may affect the mechanical properties of the weld. Spatter not only affects the appearance quality of the weld, but also may reduce the service life of the electrode and even cause a decrease in the mechanical properties of the welded joint.

[0107] Under-welding refers to the phenomenon that the weld metal is not completely fused to the parent material or between weld layers, resulting in partial lack of fusion. For example, in butt welding, if the blunt edge of the groove is too large, the root gap is too small, or the welding heat input is insufficient, it may cause root lack of fusion. In multi-layer and multi-pass welding, if the interlayer cleaning is not thorough or the welding parameters are improper, it may also cause interlayer lack of fusion. Under-welding greatly reduces the working section of the weld, causing serious stress concentration, thereby reducing the strength of the joint. It often becomes the root cause of weld cracking, seriously affecting the safety and reliability of the welded structure.

[0108] Weld nodules refer to metal nodules formed when molten metal flows onto the unmelted parent material outside the weld during the welding process. For example, in vertical and horizontal welding, if the welding current is too large, the welding rod angle is incorrect, or the operation gesture is improper, the molten metal may flow outside the weld under its own weight, forming weld nodules. Weld nodules usually appear as raised metal lumps on the surface of the weld. Weld nodules not only affect the appearance quality of the weld, but may also hide defects such as incomplete penetration under the weld nodules. The presence of weld nodules may also cause deviations in the actual size of the weld, forming stress concentration areas, thereby reducing the strength and toughness of the welded joint.

[0109] S50: Obtain the defect levels of all weld seam blocks according to the defect categories.

[0110] Optionally, the step of obtaining the defect levels of all weld seam images according to the defect categories includes:

[0111] S501. Set the window parameters of each defect category. When the defect category is pore and spatter defects, the calculation formula of the defect degree is:

[0112]

[0113] Among them, w i represents the weight of the i-th defect; I(x, y) represents the gray value of the pixel at position (x, y) in the weld image; μ represents the mean gray value of the weld image; K i (x, y) represents the Gaussian kernel function value corresponding to the i-th defect centered at (x, y).

[0114] Specifically, S i Indicates the degree of defect of type i (pores or spatter). (x,y)∈boundary |I(x,y)-μ|·K i (x,y) means the sum of all pixels (x,y) on the defect boundary. Here, boundary refers to the boundary area of ​​the defect, which can be determined by image segmentation and edge detection. |I(x,y)-μ| represents the absolute difference between the gray value of the pixel point (x,y) and the gray value mean μ of the weld block. This difference reflects the difference between the gray level of the pixel point and the overall weld. The larger the difference, the more obvious the defect may be. K i (x,y) is the Gaussian kernel function value corresponding to the i-th defect, centered at pixel (x,y). The Gaussian kernel function is a commonly used weight function whose value gradually decreases as the distance from the center increases. Here, it is used to weight the pixels on the defect boundary so that the pixels close to the defect center contribute more to the defect degree.

[0115] K iThe formula for (x,y) is as follows:

[0116]

[0117] Among them, (x c ,y c ) represents the center coordinate of the window; σ i 2 The Gaussian kernel variance representing the i-th defect is related to the window radius r i Related, usually, But the exact relationship may vary depending on the application.

[0118] According to the influence of pores and spatter defects on weld quality, the corresponding weight value w is set. i ; Sum the grayscale values ​​of all pixels in the weld image and divide it by the total number of pixels to get the grayscale mean μ. Determine the boundary area of ​​pores or spatter defects through image segmentation and edge detection. For each pixel (x, y) on the defect boundary, calculate the absolute difference between its grayscale value and the mean μ and multiply it by the corresponding Gaussian kernel function value K i (x,y). Sum the contributions of all pixels on the defect boundary and multiply them by the weight w i , and obtain the final defect degree value.

[0119] Through the above formula, we can quantitatively evaluate the severity of porosity and spatter defects, providing strong support for the evaluation and control of weld quality. At the same time, since the formula takes into account factors such as weight, grayscale difference and Gaussian kernel function, the evaluation results are more accurate and reliable.

[0120] Optionally, the step of obtaining the defect levels of all weld seam images according to the defect categories includes:

[0121] S502, setting the window parameters of each defect category. When the defect category is under-welding and weld bead defect, the calculation formula of the defect degree is:

[0122]

[0123] Among them, w i represents the weight of the i-th defect, ΔI i (x, y) represents the gray value difference of the i-th defect at the (x, y) position, ΔI weld (x, y) represents the gray value difference of the weld area at the position (x, y), and γ represents a constant.

[0124] The formula for calculating the gray value difference is:

[0125] ΔI i(x,y)=|I(x,y)-μ|·K i (x,y)

[0126] In the above formula, I(x,y) represents the grayscale value of the pixel at position (x,y) in the weld image; μ represents the mean grayscale value of the weld image; K_i(x,y) represents the Gaussian kernel function value corresponding to the i-th defect centered at (x,y).

[0127] For porosity and spatter defects, the main focus is on the grayscale value fluctuation in the boundary area; for under-welding and weld bead defects, the focus is on the grayscale value fluctuation in the entire area (including the center and the boundary).

[0128] It is necessary to add that:

[0129] For window size: set according to the typical size of the defect and the width of the weld. For under-weld and weld overhang defects, the window should be large enough to include the entire defect area, but not too large to avoid introducing too much irrelevant information.

[0130] Window shape: Rectangular or elliptical windows are usually selected to suit the shape of the weld and the distribution of defects. For welds or defects with irregular shapes, consider using windows with custom shapes.

[0131] Regarding the window position: According to the output of the defect detection algorithm, the window position is dynamically adjusted to cover the defect area. For under-weld defects, the window should mainly cover the unfused part of the weld; for weld nodule defects, the window should include the weld nodule and the surrounding weld area.

[0132] Weight w i :It is set according to the degree of influence of defects on welding quality. For example, under-weld defects may lead to reduced strength of welded joints, so they should be given a higher weight. Although weld nodules may not necessarily affect joint strength, they may affect appearance and dimensional accuracy, so they should also be given an appropriate weight.

[0133] The constant is used to adjust the effect of the gray value difference of the internal area on the total defect degree. When you need to pay more attention to the changes inside the defect, you can increase the value of γ; when you need to pay more attention to the changes in the entire defect area, you can decrease the value of γ.

[0134] In addition, before calculating the defect degree, the image should be preprocessed, such as denoising and contrast enhancement, to improve the accuracy and reliability of defect detection. Window parameters, weights, and lengths should be optimized and adjusted according to actual conditions. Experiments or simulations can be used to compare the defect degree calculation results under different parameter settings and select the best parameter combination. The calculation results are compared and verified with the actual welding quality to ensure the accuracy and effectiveness of the defect degree calculation method.

[0135] S60: Evaluate the second welding quality of the weld according to the defect degree.

[0136] Optionally, the step of evaluating the second welding quality of the weld according to the defect degree includes:

[0137] According to the defect degree of all weld blocks, weighted summary is performed according to defect type to evaluate the second welding quality of the weld.

[0138] Specifically, first, we need to obtain the data of defect degree from all weld blocks. These data may include the degree values ​​of various defects such as pores, spatter, under-welding, and weld nodules. Different defect types have different degrees of influence on weld quality. Therefore, we need to assign a weight to each defect type based on factors such as the severity of the defect, the impact on the performance of the welded joint, and the frequency of occurrence. The size of the weight can reflect the importance of the defect to the quality of the weld. For example, for defects that seriously affect the strength and sealing of the welded joint (such as cracks, lack of fusion, etc.), a higher weight should be given. Using the weighted summation method, the degree values ​​of various defects in each weld block are summarized according to their corresponding weights. A comprehensive defect degree index is obtained, which can reflect the overall quality status of the weld. According to the summarized defect degree index, we can evaluate the second welding quality of the weld.

[0139] S70. Predict the quality of the truck steel structure according to the first welding quality and the second welding quality.

[0140] Optionally, the step of predicting the quality of the steel structure of the truck according to the first welding quality and the second welding quality includes:

[0141] A weighted calculation is performed according to the first welding quality and the second welding quality to obtain a comprehensive score, and the quality of the truck steel structure is characterized by the comprehensive score.

[0142] Specifically, the specific impact of the first welding quality and the second welding quality on the quality of the truck steel structure is deeply analyzed. This includes considering multiple aspects such as the strength, sealing, and durability of the welded joint. Based on the above analysis, reasonable weights are assigned to the first welding quality and the second welding quality. The size of the weights should reflect their respective importance to the quality of the truck steel structure. In order to eliminate the dimensional differences between different welding quality indicators, we may need to standardize the first welding quality and the second welding quality and convert them into dimensionless values. Using the assigned weights, the standardized first welding quality and the second welding quality are weighted summed to obtain a comprehensive score. The calculation formula for the comprehensive score can be expressed as:

[0143] Comprehensive score = first welding quality standardization value × first welding quality weight + second welding quality standardization value × second welding quality weight

[0144] According to historical data or industry standards, set a threshold range for the comprehensive score to judge the quality of the truck steel structure. Compare the calculated comprehensive score with the set threshold to evaluate the quality level of the truck steel structure. For example, a comprehensive score above a certain threshold may indicate excellent quality, while a score below another threshold may indicate unqualified quality. Effectively predicting the quality of truck steel structures provides strong support for production, quality control, and continuous improvement.

[0145] Example 3

[0146] Based on Example 1, this embodiment provides a truck steel structure quality prediction system based on beam welding results, including:

[0147] A welding point position acquisition module, which is configured to acquire welding point positions of the welds arranged in order according to the weld image of the beam welding position;

[0148] A spacing distribution characteristic acquisition module, which is configured to acquire the spacing distribution characteristics of adjacent solder joints according to the solder joint positions;

[0149] A first welding quality assessment module, configured to assess a first welding quality of the weld according to the spacing distribution characteristic;

[0150] A defect category acquisition module, configured to acquire at least one weld image block including a preset number of welds according to the weld position, perform weld defect recognition on the weld image block, and obtain a defect category of the weld image;

[0151] A defect degree acquisition module, which is configured to acquire the defect degrees of all weld seam blocks according to the defect category;

[0152] A second welding quality acquisition module, configured to evaluate a second welding quality of the weld according to the defect degree;

[0153] A truck steel structure quality prediction module is configured to predict the truck steel structure quality according to the first welding quality and the second welding quality.

[0154] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A truck steel structure quality prediction method based on beam welding results, characterized in that: include: According to the weld image of the beam welding position, the weld point positions of the weld are obtained in order; According to the positions of the solder joints, obtaining the spacing distribution characteristics of adjacent solder joints; evaluating a first welding quality of the weld according to the spacing distribution characteristic; According to the positions of the weld spots, obtaining at least one weld image block including a preset number of weld spots, performing weld defect recognition on the weld image block, and obtaining a defect category of the weld image; According to the defect category, the defect degree of all weld seam blocks is obtained; According to the defect degree, evaluating the second welding quality of the weld; The quality of the steel structure of the truck is predicted according to the first welding quality and the second welding quality.

2. The truck steel structure quality prediction method based on beam welding results according to claim 1 is characterized in that: The step of obtaining the spacing distribution characteristics of adjacent solder joints according to the solder joint positions includes: According to the arrangement order of the welding points on the weld, two adjacent welding points are paired to obtain paired welding points, and the Euclidean distance between the center points of each group of paired welding points is used as the welding point spacing; The average weld point spacing of the weld is obtained according to the weld point distances of all paired weld points, and the weld point spacing of each group is compared with the average weld point spacing to obtain the spacing distribution characteristics of the paired weld points, wherein the spacing distribution characteristics characterize the weld point uniformity of the weld.

3. The truck steel structure quality prediction method based on beam welding results as claimed in claim 2, characterized in that: The step of obtaining an average weld point spacing of a weld according to the weld point distances of all paired weld points, and comparing each group of weld point spacings with the average weld point spacing to obtain the spacing distribution characteristics of the paired weld points includes: According to the solder joint distances of all paired solder joints, the maximum solder joint spacing, the minimum solder joint spacing and the average solder joint spacing are obtained; Get the maximum solder point spacing, minimum solder point spacing and average solder point spacing, and get the normal solder point spacing range; Compare the spacing between each group of solder joints with the normal spacing between solder joints, obtain abnormal paired solder joints, and send abnormal information.

4. The truck steel structure quality prediction method based on beam welding results according to claim 1 is characterized in that: The step of evaluating the first welding quality of the weld according to the spacing distribution characteristic comprises: Formula for assessing the first weld quality of a weld: Among them, d i It represents the distance between the i-th and i+1-th welding points. There are n welding points in total. D represents the average welding gap between adjacent welding points. K is the coefficient.

5. The truck steel structure quality prediction method based on beam welding results according to claim 1, characterized in that: The step of obtaining at least one weld image block including a preset number of welds according to the weld position, performing weld defect recognition on the weld image block, and obtaining the defect category of the weld image includes: Process the weld image to obtain a grayscale image, calculate the grayscale co-occurrence matrix of the grayscale image, and provide texture features; A threshold is set according to grayscale value or texture feature, the grayscale image is segmented into foreground and background, and edge information of defects in the foreground is detected by an edge detection operator; The defect category of the weld image is obtained according to the edge information.

6. The truck steel structure quality prediction method based on beam welding results according to claim 1, characterized in that: The step of obtaining the defect levels of all weld seam blocks according to the defect categories includes: Set the window parameters for each defect category. When the defect category is pore and spatter defects, the calculation formula for the defect degree is: Among them, w i represents the weight of the i-th defect; I(x, y) represents the gray value of the pixel at position (x, y) in the weld image; μ represents the mean gray value of the weld image; K i (x, y) represents the Gaussian kernel function value corresponding to the i-th defect centered at (x, y).

7. The truck steel structure quality prediction method based on beam welding results according to claim 1, characterized in that: The step of obtaining the defect levels of all weld seam blocks according to the defect categories includes: Set the window parameters for each defect category. When the defect category is under-welding and weld bead defects, the calculation formula for the defect degree is: Among them, w i represents the weight of the i-th defect, ΔI i (x, y) represents the gray value difference of the i-th defect at the (x, y) position, ΔI weld (x, y) represents the gray value difference of the weld area at the position (x, y), and γ represents a constant.

8. The truck steel structure quality prediction method based on beam welding results according to claim 1, characterized in that: The step of evaluating the second welding quality of the weld according to the defect degree comprises: According to the defect degree of all weld blocks, weighted summary is performed according to defect type to evaluate the second welding quality of the weld.

9. The truck steel structure quality prediction method based on beam welding results according to claim 1, characterized in that: The step of predicting the quality of the truck steel structure according to the first welding quality and the second welding quality comprises: A weighted calculation is performed according to the first welding quality and the second welding quality to obtain a comprehensive score, and the quality of the truck steel structure is characterized by the comprehensive score.

10. A truck steel structure quality prediction system based on beam welding results, characterized in that: include: A welding point position acquisition module, which is configured to acquire welding point positions of the welds arranged in order according to the weld image of the beam welding position; A spacing distribution characteristic acquisition module, which is configured to acquire the spacing distribution characteristics of adjacent solder joints according to the solder joint positions; A first welding quality assessment module, configured to assess a first welding quality of the weld according to the spacing distribution characteristic; A defect category acquisition module, configured to acquire at least one weld image block including a preset number of welds according to the weld position, perform weld defect recognition on the weld image block, and obtain a defect category of the weld image; A defect degree acquisition module, which is configured to acquire the defect degrees of all weld seam blocks according to the defect category; A second welding quality acquisition module, configured to evaluate a second welding quality of the weld according to the defect degree; A truck steel structure quality prediction module is configured to predict the truck steel structure quality according to the first welding quality and the second welding quality.

Images