Battery shell processing quality inspection method and system based on visual assistance

Abstract

The invention relates to the technical field of image processing, in particular to a battery shell processing quality inspection method and system based on vision assistance. The method comprises the following steps: firstly, detecting a welding seam and a corresponding welding crack in a shell image; crack extensibility is determined according to diffusion conditions of welding cracks in different directions; the texture uniformity and the texture interval of the weld joint are analyzed, and the heat input degree of the weld joint is determined; based on the abnormal heat input degree, the weld joints are clustered, and weld joint clusters are obtained; according to the heat input degree and the crack extensibility of each welding seam cluster, the affected degree of the welding crack is determined; determining a shell defect coefficient according to the deformation condition of the battery shell, the position of the welding crack and the sharp condition; and comparing the crack extensibility with the shell defect coefficient, determining a welding risk coefficient, and determining the processing fault type of the battery shell in combination with the affected degree. According to the invention, whether the defects of the battery shell are caused by processing can be judged.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a method and system for quality inspection of battery case processing based on visual assistance. Background Art

[0002] Battery cases are usually made of metal materials, and the outer shape of battery cases is complex and requires high processing precision; various defects may occur during the production process of battery cases, such as cracks, scratches, shape deviations, etc.; if these defects are not detected in time, they may affect the performance and safety of the battery and even cause battery failure or safety accidents. For example, cracks or poor sealing of the battery case may cause electrolyte leakage. For lithium-ion batteries, the electrolyte is flammable and explosive, and contact with air or moisture may cause short circuits, thermal runaway, and even fire and explosion; when structural defects such as deformation and damage occur, the mechanical strength of the outer shell will be reduced, and when subjected to impact, extrusion or vibration, the internal battery core may be damaged, resulting in internal short circuits or thermal runaway. Traditional quality inspection methods for battery cases mainly rely on manual inspection and simple automated equipment, and these methods have deficiencies in terms of detection efficiency, accuracy, and consistency. Especially in large-scale production, manual inspection is often easily affected by factors such as fatigue and negligence. And through simple automated equipment, it is impossible to determine whether the defects of the battery case are caused by weld processing or the battery case itself has defects. Summary of the Invention

[0003] In order to solve the technical problem of difficultly determining whether the defects of the battery case are caused by weld processing, the purpose of the present invention is to provide a method and system for quality inspection of battery case processing based on visual assistance, and the specific technical solutions adopted are as follows: In the first aspect, an embodiment of the present invention provides a method for quality inspection of battery case processing based on visual assistance, and the method includes: Obtain the outer shell image of the battery case after processing; detect the welds and corresponding welding cracks in the outer shell image; Determine the crack extensibility of the welding crack according to the diffusion situation of the welding crack in different directions; Analyze the texture uniformity and texture interval of the weld to determine the heat input degree corresponding to the weld; based on the abnormal heat input degree, cluster the welds to obtain weld clusters; Determine the affected degree of the welding crack according to the heat input degree and crack extensibility of each weld cluster; Determine the outer shell defect coefficient according to the deformation situation of the battery case, the position and sharpness of the welding crack; compare the crack extensibility and the outer shell defect coefficient to determine the welding risk coefficient; Combine the affected degree and the welding risk coefficient to determine the processing fault type of the battery case.

[0004] Further, determining the crack extensibility of the welding crack according to the diffusion conditions of the welding crack in different directions includes: Determining the extension direction line of the welding crack; dividing the welding crack into multiple pixel point clusters based on the vertical direction of the extension direction line of the welding crack; For each pixel point cluster, calculating the distance between each pixel point in the pixel point cluster and the extension direction line; calculating the ratio between the maximum distance and the information entropy of the distance as the diffusion coefficient of the pixel point cluster; Determining the crack extensibility of the welding crack on the battery housing according to the diffusion coefficients between different pixel point clusters.

[0005] Further, analyzing the texture uniformity and texture interval of the weld seam to determine the heat input degree corresponding to the weld seam includes: Performing edge detection on the weld seam to obtain the edges on the weld seam and recording them as weld seam textures; for any weld seam texture in the weld seam, obtaining the distance between the weld seam texture and the weld seam textures on both sides as the weld seam texture distance; Performing uniform analysis on the weld seam textures to obtain the texture uniform density of the weld seam; Combining the texture uniform density, the weld seam texture distance, and the color of the weld seam texture to determine the heat input degree corresponding to the weld seam.

[0006] Further, the method for obtaining the abnormal heat input degree is: Calculating the absolute value of the difference between the currently detected heat input degree and the normal heat input degree, and when the absolute value of the difference is greater than a preset normal difference threshold, determining that the currently detected heat input degree is an abnormal heat input degree.

[0007] Further, determining the affected degree of the welding crack according to the heat input degree and crack extensibility of each weld cluster includes: Obtaining the information entropy and the mean value of the crack extensibility of all welding cracks during the processing of the battery housing, and obtaining the mean value of the heat input degree in each weld cluster; adjusting the information entropy of the crack extensibility according to the difference between the mean value of the heat input degree and the mean value of the crack extensibility, and performing normalization processing on the adjusted value to obtain the affected degree of the welding crack.

[0008] Further, determining the housing defect coefficient according to the deformation condition of the battery housing, the position and sharpness of the welding crack includes: Determining the distance degree of each welding crack according to the distance between the centroid of each welding crack on the battery housing and the weld edge; Determine the shell deformation coefficient of the battery shell corresponding to each welding crack according to the change of the upper edge of the battery shell plates on both sides corresponding to each welding crack; Obtain the sharpness of the crack according to the fluctuation between the edge pixel points of the welding crack; Combine the degree of distance, the shell deformation coefficient and the sharpness of the crack to determine the shell defect coefficient of each welding crack; wherein, both the shell deformation coefficient and the sharpness are positively correlated with the shell defect coefficient, and the degree of distance is negatively correlated with the shell defect coefficient.

[0009] Further, comparing the crack extensibility and the shell defect coefficient to determine the welding risk coefficient includes: Perform curve fitting on the shell defect coefficients of each welding crack of the battery shell to obtain a first fitting curve; perform curve fitting on the crack extensibility of each welding crack of the battery shell to obtain a second fitting curve; calculate the mean square error between the first fitting curve and the second fitting curve as the welding risk coefficient.

[0010] Further, combining the degree of influence and the welding risk coefficient to determine the processing fault type of the battery shell includes: Correct the degree of influence through the welding risk coefficient, and use the corrected result value as the processing fault degree of the battery shell; When the processing fault degree is greater than or equal to the preset fault threshold, it is determined that there is a problem with the welding process of the battery shell; When the processing fault degree is less than the preset fault threshold, it is determined that there is a problem with the ex-factory quality of the battery shell.

[0011] Further, detecting the welding cracks in the shell image includes: Learn the normal weld seam features through an autoencoder, and identify the welding cracks in the shell image through the reconstruction error.

[0012] In a second aspect, a vision-assisted quality inspection system for battery shell processing is provided, and the system includes the following modules: An initial detection module for obtaining the shell image after the battery shell is processed; detecting the weld seams and corresponding welding cracks in the shell image; An extension analysis module for determining the crack extensibility of the welding crack according to the diffusion of the welding crack in different directions; A heat input analysis module for analyzing the texture uniformity and texture interval of the weld seam to determine the heat input degree corresponding to the weld seam; clustering the weld seams based on the abnormal heat input degree to obtain weld seam clusters; An influence analysis module for determining the degree of influence of the welding crack according to the heat input degree and crack extensibility of each weld seam cluster; A processing risk determination module, configured to determine a housing defect coefficient according to the deformation condition of the battery housing, the position and sharpness of the welding crack; compare the crack extensibility and the housing defect coefficient to determine the welding risk coefficient; A processing fault determination module, configured to determine the type of processing fault of the battery housing by combining the degree of influence and the welding risk coefficient.

[0013] In a third aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor. An executable code is stored in the memory. When the processor executes the executable code, the implementation is as in each possible implementation embodiment of the first aspect.

[0014] In a fourth aspect, an embodiment of the present invention provides a computer program product, which includes: computer program code. When the computer program code runs on a computer, the computer is enabled to execute the method in the above first aspect or any possible implementation manner of the first aspect.

[0015] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed in a computer, the computer is enabled to execute each possible implementation embodiment of the first aspect.

[0016] The embodiments of the present invention at least have the following beneficial effects: In the embodiments of the present invention, by identifying the battery housing image, the battery image with cracks therein is obtained. According to the morphology of the cracks in the weld of the battery housing, the crack extensibility of the cracks is obtained, and through the texture of the weld, the influence of the welding settings on the cracks is analyzed during the welding process to obtain the degree of influence of the welding cracks; further, according to the position, sharpness of the cracks and the deformation of the housing in the welding processing result of the battery housing, the quality of the battery housing is evaluated to obtain the welding risk coefficient caused by welding cracks. Finally, through the welding risk coefficient corresponding to the welding process during processing, the degree of influence on the weld is corrected, and the finally obtained processing fault degree of the battery housing is more accurate compared with the traditional battery housing processing quality inspection method, and based on the type of processing fault, it can be judged whether the defect of the battery housing is caused by weld processing. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 The method flow chart of the visual - assisted battery housing processing quality inspection method provided by an embodiment of the present invention; Figure 2 The schematic diagram of the housing image after battery housing processing provided by an embodiment of the present invention; Figure 3 The partial schematic diagram of the welding crack provided by an embodiment of the present invention; Figure 4 The schematic diagram of the extension direction line of the welding crack and the pixel point clustering provided by an embodiment of the present invention; Figure 5 The schematic diagram of the weld seam with a higher heat input provided by an embodiment of the present invention; Figure 6 The schematic diagram of the weld seam with a normal heat input provided by an embodiment of the present invention; Figure 7 The schematic diagram of the weld seam with a lower heat input provided by an embodiment of the present invention; Figure 8 The system block diagram of the visual - assisted battery housing processing quality inspection system provided by an embodiment of the present invention. Detailed implementation manners

[0019] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation manners, structures, features and their effects of a visual - assisted battery housing processing quality inspection method and system according to the present invention.

[0020] In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0021] Among them, in the description of the embodiments of the present invention, unless otherwise specified, " / " means "or". For example, A / B can represent A or B. The "and / or" in the text is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present invention, "a plurality of" means two or more than two.

[0022] Hereinafter, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as implying or indicating relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include one or more of such features.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs.

[0024] The embodiments of the present invention will be described below with reference to the accompanying drawings. As is known to those of ordinary skill in the art, with the development of technology and the emergence of new scenarios, the technical solutions provided by the embodiments of the present invention are equally applicable to similar technical problems.

[0025] The specific solutions of the battery housing processing quality inspection method and system based on visual assistance provided by the present invention will be specifically described below with reference to the accompanying drawings.

[0026] Please refer to Figure 1 , which shows a flowchart of the steps of a battery housing processing quality inspection method based on visual assistance provided by an embodiment of the present invention. The method includes the following steps: Step S100, obtain the housing image after the battery housing is processed; detect the welds and corresponding welding cracks in the housing image.

[0027] New energy batteries are usually composed of multiple battery cells through welding or other connection methods to form a battery module. The batteries within each module are connected in parallel or in series to achieve the required voltage and capacity. Usually, in order to ensure the stability and safety of the battery pack, multiple battery cells or battery modules are fixed in a sealed housing.

[0028] When encapsulating the battery module, first, the battery cells or modules need to be connected and fixed. After fixation, heat dissipation materials or a heat dissipation system need to be added, and a battery management system needs to be installed. After installation, a metal housing is used to encapsulate the battery. During encapsulation, it needs to be fixed by means such as screws, buckles, or welding. After encapsulation, a sealant is used at the seams of the housing to ensure that the inside of the battery module is not affected by the external environment such as water, dust, and oil.

[0029] Therefore, in order to ensure the welding quality at the weld before sealing, the quality of the welds of the battery housing after welding is inspected. An industrial camera with an appropriate resolution is selected according to the size of the battery. And in order to avoid reflection or shadow and the texture of the housing surface, uniform illumination is also required for the inspection scene.

[0030] Use an industrial camera to obtain the housing image after the battery housing is processed. The housing image is a grayscale image. Please refer to Figure 2 , Figure 2 which is a schematic diagram of the housing image after the battery housing is processed.

[0031] After obtaining the image of the processed and welded housing, the histogram equalization processing can be first performed on the housing image to enhance the contrast of the crack area. The Canny operator can be used to detect the edges in the housing image, and the Sobel operator can also be used to detect the edges in the housing image. After determining the edges, the closed areas can be extracted by using the region generation algorithm or the contour analysis algorithm, and then the weld area can be located.

[0032] Then, the normal weld features are learned through the autoencoder, and the welding cracks in the housing image are identified through the reconstruction error.

[0033] It should be noted that the method for locating the weld area from the closed area can be to identify the weld area through the trained neural network, which will not be elaborated here.

[0034] Please refer to Figure 3 , Figure 3 for the partial schematic diagram of the welding crack.

[0035] For the housing image in which no welding crack is identified, no subsequent processing operation is performed. The embodiment of the present invention only judges the battery housing in which a welding crack is detected in the housing image, and judges whether the welding crack is caused by the defect of the battery housing itself or by improper processing during the welding process through subsequent steps.

[0036] Step S200, determine the crack extensibility of the welding crack according to the diffusion condition of the welding crack in different directions.

[0037] In the image of the battery housing weld with cracks obtained by the above operations, some cracks may appear due to welding process problems during welding, and some cracks may appear due to housing substrate problems; in order to adjust the subsequent processing quality, it is necessary to analyze the generation conditions of the welding cracks.

[0038] First, determine the extension direction line of the welding crack. Specifically: obtain the edge pixel points of the welding crack in the housing image, and use the line segment connecting the two edge pixel points with the largest distance as the extension direction line of the crack connected domain, and this extension direction line represents the extension direction of the welding crack.

[0039] Based on the vertical direction of the extension direction of the welding crack, the welding crack is divided into multiple pixel clusters. Specifically: according to the perpendicular line of the extension direction line, the welding crack is divided into multiple pixel clusters perpendicular to the extension direction line, where each line segment formed by each pixel cluster is perpendicular to the extension direction line.

[0040] Please refer to Figure 4 , Figure 4Schematic diagram of the extension direction line of the welding crack and pixel clustering Figure 4 The closed area in Figure 4 is the welding crack, and the long line segment a in Figure 4 is the extension direction line of the welding crack, and the short line segment b in Figure 4 is the perpendicular line to the extension direction line of the welding crack; the pixel points belonging to the short line segment b and within the boxed part in

[0041] are pixel points belonging to the same cluster. For each pixel cluster, calculate the distance between each pixel point in the pixel cluster and the extension direction line; calculate the ratio between the maximum distance and the information entropy of the distance as the diffusion coefficient of the pixel cluster.

[0042] The information entropy of the distance is the information entropy of the multiple distances corresponding to the extension direction line of each pixel point belonging to the same pixel cluster. The maximum distance is also the maximum distance between each pixel point in the same pixel cluster and the extension direction line.

[0043] Arrange the pixel clusters according to their positions on the extension direction line to determine the next pixel cluster for each pixel cluster. For example, in Figure 4 the pixel clusters are arranged from top to bottom according to their positions on the extension direction line.

[0044] Furthermore, determine the crack extensibility of the welding crack on the battery housing according to the diffusion coefficients between different pixel clusters.

[0045] In the embodiments of the present invention, the crack extensibility is determined by combining the length of the extension direction line and the difference in the diffusion coefficients corresponding to adjacent pixel clusters. Among them, the coefficient determined by the difference in the diffusion coefficients corresponding to adjacent pixel clusters is the overall diffusion coefficient. Among them, both the length of the extension direction line and the overall diffusion coefficient are positively correlated with the crack extensibility.

[0046] Because, for the length of the extension direction line, the greater the length, the greater the extensibility of the crack, and the more likely it is that quality problems will occur in the corresponding area of the crack. And the overall diffusion coefficient is determined by the difference in the diffusion coefficients corresponding to adjacent pixel clusters. The greater the diffusion coefficients corresponding to adjacent pixel clusters, it indicates that the diffusion coefficient of the pixel clusters of the crack is continuously increasing. Therefore, it indicates that the crack is continuously spreading outwards, so its extensibility is higher.

[0047] In some embodiments, the normalized value of the product of the length of the extension direction line and the overall diffusion coefficient is used as the crack extensibility.

[0048] In some embodiments, the method for obtaining the overall diffusion coefficient is: the average value of the differences in the diffusion coefficients corresponding to all adjacent pixel clusters.

[0049] The stronger the extensibility of the crack in the outer shell image, the greater the tendency for the crack to extend outward. When the tendency for the crack to extend outward is greater, the extension of the crack may cause damage to the integrity of the outer shell, thereby leading to the exposure of the internal components of the battery and posing a safety hazard.

[0050] Since the reason for the generation of the weld seam is due to some problems in the welding process, it is necessary to obtain the influence of the welding process on the crack based on the crack image on the weld seam in the battery outer shell.

[0051] Step S300: Analyze the texture uniformity and texture interval of the weld seam to determine the heat input level corresponding to the weld seam; based on the abnormal heat input level, cluster the weld seams to obtain weld seam clusters.

[0052] The settings of various parameters during the welding process will affect the total heat transferred to the workpiece during welding, thereby affecting the formation of the weld seam, the size of the weld pool, the cooling rate, and the flow of the metal, and directly affecting the shape, thickness, and regularity of the surface texture of the weld seam.

[0053] When the heat input is excessive or too little, it will cause the welding metal to melt excessively or the cooling to be uneven, thereby generating cracks.

[0054] Therefore, in order to analyze the influence of parameter settings during the welding process on the welding cracks of the battery outer shell, it is first necessary to judge the abnormal situation of the welding process parameters based on the cracks; Perform edge detection on the weld seam to obtain the edges on the weld seam, which are recorded as weld seam textures.

[0055] When the heat input is excessive, it will cause a larger weld pool and a slower cooling rate during the processing, thereby possibly forming coarser, larger, and uneven textures; while when the heat input is low, it will cause a smaller weld pool and a faster cooling rate during the processing, thereby possibly forming finer and uniform textures, but due to the relatively fine texture, there may be a discontinuous situation.

[0056] Please refer to Figure 5 , Figure 5 which is a schematic diagram of the weld seam with a higher heat input; please refer to Figure 6 , Figure 6 which is a schematic diagram of the weld seam with a normal heat input; please refer to Figure 7 , Figure 7 which is a schematic diagram of the weld seam with a lower heat input.

[0057] Figure 5 The weld seam in Figure 7 has coarser textures and poorer uniformity between the textures due to the higher heat input; Figure 6The weld seam in [it] has a more appropriate heat input. At this time, the weld seam texture will be uniform and have an appropriate depth.

[0058] Perform a uniform analysis on the weld seam texture to obtain the texture uniform density of the weld seam.

[0059] First, perform a similarity analysis on the weld seam texture to obtain the texture similarity of the weld seam. The higher the texture similarity, the more appropriate the heat input during the current welding is characterized.

[0060] In some embodiments, for each weld seam texture, construct a weld seam texture gray value sequence based on the gray values of the pixel points in each weld seam texture. Use the DTW algorithm to calculate the DTW distance between the weld seam texture gray value sequences of any two weld seam textures, and perform a negative correlation mapping on the DTW distance to obtain a single texture similarity; in a certain order, calculate the mean value of the single texture similarities between adjacent two weld seam textures respectively to obtain the texture similarity of the weld seam. Among them, in a certain order, it can be in the order from top to bottom, and can also be in the order from bottom to top. If the weld seam is vertical, it can also be in the order from left to right. In short, calculate the texture similarity of adjacent weld seam textures according to a certain arrangement order.

[0061] Denote the average width of the weld seam as the weld seam width, and for any weld seam texture in the weld seam, obtain the distance between the weld seam texture and the weld seam textures on both sides as the weld seam texture distance. Obtain the difference between the minimum value and the maximum value in the weld seam texture distances in the weld seam as the maximum difference of the weld seam texture distances.

[0062] For each weld seam texture, calculate the ratio of the weld seam texture distance of each weld seam texture to the weld seam width, and find the mean value of the ratios of all weld seam textures. Denote this mean value as the weld seam width uniformity value; Take the product of the weld seam width uniformity value, the variance of the weld seam texture distance, and the texture similarity as the texture uniform density of the weld seam.

[0063] Combine the texture uniform density, the weld seam texture distance, and the color of the weld seam texture to determine the heat input degree corresponding to the weld seam.

[0064] In the embodiments of the present invention, combine the texture uniform density, the maximum difference of the weld seam texture distance, and the average gradient value of the weld seam texture to determine the heat input degree corresponding to the weld seam. Among them, the color of the weld seam texture is characterized by the average gradient value of the gray values of the weld seam texture.

[0065] In some embodiments, the calculation formula for the heat input degree RS is: ; where j is the texture uniform density, is the maximum difference of the weld seam texture distance; is the average gradient value of the weld texture; exp is the exponential function with the natural constant as the base.

[0066] Among them, The larger the value of, the shallower and more uniform the texture on the surface weld, and the texture in the weld may be interrupted. Therefore, the surface weld part is more likely to be caused by a low heat input.

[0067] Regarding the heat input degree during the welding process of the battery case, either too large or too small heat input degree may cause problems during the welding process of the battery case. From the historical database, obtain the median of the heat input degree of the weld corresponding to the battery case image without welding cracks as the normal heat input degree.

[0068] After determining the heat input amount in the welding process of the currently detected battery case, compare the difference between the currently detected heat input degree and the normal heat input degree to obtain the heat input abnormal amount in the welding process of the tram shell. More specifically: calculate the absolute value of the difference between the currently detected heat input degree and the normal heat input degree. When the absolute value of this difference is greater than the preset normal difference threshold, determine that the currently detected heat input degree is the heat input abnormal amount, that is, an abnormal heat input degree. In the embodiment of the present invention, the value of the preset normal difference threshold can be 0.2, and in other embodiments, the threshold can also be adjusted by the implementer according to the actual situation.

[0069] Cluster the heat input abnormal amounts in the welding process of the battery case for the same battery case to obtain multiple abnormal amount clusters. According to the abnormal amount clusters, divide the welding cracks belonging to the same abnormal amount cluster into the same cluster to obtain weld clusters. The heat input abnormal amounts of the welding cracks during the processing of the battery cases belonging to the same weld cluster are relatively close. Among them, each abnormal amount cluster corresponds to a weld cluster. It can also be understood that the welding cracks are divided based on the heat input abnormal amount as the data basis for clustering.

[0070] Step S400, determine the affected degree of the welding crack according to the heat input degree and crack extensibility of each weld cluster.

[0071] Obtain the information entropy and mean value of the crack extensibility of the welding cracks in the battery case processing in each weld cluster; Obtain the mean value of the heat input degree in each weld cluster; According to the difference between the mean value of the heat input degree and the mean value of the crack extensibility, adjust the information entropy of the crack extensibility, and perform normalization processing on the adjusted value to obtain the affected degree of the welding crack.

[0072] The smaller the difference between the mean of the heat input degree and the mean of the crack extensibility, the higher the synchronization between the crack extensibility in the cluster and the abnormal amount of welding heat input, indicating that the higher the abnormal heat input, the more likely it is to have cracks with strong extensibility after welding.

[0073] In some embodiments, the influence degree YX of the welding cracks in the cluster is: ; where th is the hyperbolic function; N is the number of clusters; is the mean value of the abnormal heat input of the nth weld cluster; is the mean value of the crack extensibility of all welding cracks; is the information entropy of the crack extensibility of all welding cracks; is a non-zero constant between 0 and 1.

[0074] Among them, in the calculation formula of the influence degree, adding is to prevent when the result of

[0075] is 0, it will affect the subsequent calculation results.

[0076] Step S500, determine the shell defect coefficient according to the deformation condition of the battery shell, the position and sharpness of the welding crack; compare the crack extensibility and the shell defect coefficient to determine the welding risk coefficient.

[0077] Quality problems in the battery shell material itself will cause welding cracks in the battery shell during welding, and even exacerbate the welding cracks caused by welding parameter problems; therefore, it is necessary to analyze the battery shell material according to the performance of the welds on the battery shell.

[0078] When there are quality problems in the battery shell material, the welds of the battery shell will have brittle fractures due to the defects of the material itself, and due to the material itself, the cracks will occur in the fusion zone, and the fusion zone is generally located at the edge of the weld. And because the brittle fracture caused by the material reason is relatively sharp, the shell defect coefficient of the battery shell can be determined through the deformation condition of the battery shell and the sharpness of the welding crack.

[0079] First, determine the distance degree of each welding crack based on the distance between the centroid of each welding crack on the battery housing and the edge of the weld seam. Specifically: Obtain the minimum distance between the centroid of each welding crack on the battery housing and the edge of the weld seam as the distance degree of each welding crack. When the welding crack is closer to the edge of the weld seam, that is, the welding crack is not close to the middle of the weld seam but closer to the edge of the weld seam, it indicates that the crack is more likely caused by the battery housing.

[0080] Secondly, determine the housing deformation coefficient of each welding crack according to the change of the upper edge of the battery housing plates on both sides corresponding to each welding crack; among them, the battery housing plates on both sides corresponding to the welding crack include: the edge of the battery housing plate with the same ordinate as the welding crack, and the edge of the battery housing plate with the same abscissa as the welding crack. Calculate the deformation coefficient for the edges of the battery housing plates obtained in the two cases of the same abscissa and the same ordinate respectively, and select the maximum deformation coefficient corresponding to the same abscissa and the same ordinate as the deformation coefficient of the welding crack. Specifically, the method for obtaining the deformation coefficient is: for the edges of the battery housing plates on both sides with the same ordinate as the welding crack, calculate the average value of the differences in the abscissa of any one side of the welding crack as the first deformation coefficient; calculate the average value of the differences in the abscissa of the other side of the welding crack as the second deformation coefficient; take the average value of the first deformation coefficient and the second deformation coefficient as the deformation coefficient of the edges of the battery housing plates on both sides with the same ordinate as the welding crack.

[0081] Then, obtain the sharpness of the crack according to the fluctuation of the edge pixel points of the welding crack. Specifically: Obtain the edge chain code on the edge of the welding crack, and obtain the sharpness according to the average value of the differences at the corresponding positions between adjacent chain code values.

[0082] Finally, determine the housing defect coefficient of the battery housing by combining the distance degree, the housing deformation coefficient of the battery housing, and the sharpness of the crack, where the housing deformation coefficient and the sharpness are both positively correlated with the housing defect coefficient, and the distance degree is negatively correlated with the housing defect coefficient.

[0083] In some embodiments, the calculation formula for the housing defect coefficient ZL of the welding crack is: ; where x is the sharpness of the welding crack; r is the housing deformation coefficient of the battery housing corresponding to the welding crack; c is the distance degree of the welding crack; th is the hyperbolic tangent function.

[0084] In the calculation formula of the housing defect coefficient, the larger the value, the sharper and closer to the fusion zone the crack on the battery housing is and the more obvious the housing deformation is, so it indicates that there are more quality problems with the battery housing.

[0085] Curve fitting is performed on the shell defect coefficient of each welding crack of the battery shell to obtain a first fitting curve; curve fitting is performed on the crack extensibility of each welding crack of the battery shell to obtain a second fitting curve; the mean square error between the first fitting curve and the second fitting curve is calculated, and this mean square error is denoted as the welding risk coefficient; the larger the mean square error, the greater the difference between the current two fitting curves, the smaller the probability that the crack is caused by shell defects, the greater the probability that the crack is caused by welding, and the higher the corresponding welding risk coefficient; conversely, the smaller the mean square error, the smaller the difference between the current two fitting curves, and the lower the corresponding welding risk coefficient.

[0086] Step S600, combining the degree of influence and the welding risk coefficient, determine the processing failure type of the battery shell.

[0087] When the influence of welding on the welding crack is more serious, the corresponding degree of influence is greater, indicating that the welding processing problem of the battery shell is greater; and the larger the value of the welding risk coefficient, the smaller the probability that the crack is caused by the battery shell, and the greater the probability that the crack appears due to welding processing. It can also be understood that the welding risk coefficient is obtained by analyzing the morphology of the battery and is used as a coefficient for correcting the degree of influence. The degree of influence is corrected by the welding risk coefficient to ensure that the evaluation result of the welding processing problem will not be too high.

[0088] In some embodiments, the product value of the degree of influence and the welding risk coefficient is used as the processing failure degree of the battery shell.

[0089] As a preferred embodiment of the present invention, after obtaining the processing failure degree of the battery shell, compare the size of the processing failure degree and the preset failure threshold. When the processing failure degree is greater than or equal to the preset failure threshold, it is determined that there is a problem with the welding processing of the battery shell. At this time, it is necessary to detect the setting of various welding parameters and the welding equipment. When the processing failure degree is less than the preset failure threshold, it is determined that there is a factory quality problem with the battery shell. At this time, the battery shell needs to be returned to the factory, and if necessary, the raw materials of the battery shell can also be inspected for quality.

[0090] Please refer to Figure 8 , which shows a system block diagram of a vision-assisted quality inspection system for battery shell processing provided by another embodiment of the present invention. The system includes: An initial detection module, configured to obtain a shell image after processing the battery shell; detect the weld seams and corresponding welding cracks in the shell image; An extension analysis module, configured to determine the crack extensibility of the welding crack according to the diffusion conditions of the welding crack in different directions; A heat input analysis module for analyzing the texture uniformity and texture interval of a weld seam to determine the heat input level corresponding to the weld seam; clustering the weld seams based on the abnormal heat input level to obtain weld seam clusters; An influence analysis module for determining the affected degree of welding cracks according to the heat input level and crack extensibility of each weld seam cluster; A processing risk determination module for determining a shell defect coefficient according to the deformation condition of the battery shell, the position and sharpness of the welding crack; comparing the crack extensibility and the shell defect coefficient to determine a welding risk coefficient; A processing fault determination module for determining the type of processing fault of the battery shell by combining the affected degree and the welding risk coefficient.

[0091] Optionally, the transmission medium may be a wired link, such as but not limited to, coaxial cable, optical fiber, digital subscriber line, etc., or a wireless link, such as but not limited to, Wireless Fidelity (WIFI), Bluetooth, mobile device network, etc.

[0092] It should be noted that: for the device provided in the above embodiment, only the above division of each functional module is used for illustration. In practical applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above.

[0093] A schematic structural diagram of a computer device provided by an embodiment of the present invention. Exemplarily, the computer device includes: a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the computer device can execute any of the above-described vision-assisted battery shell processing quality inspection methods.

[0094] In addition, an embodiment of the present invention also protects a device, which may include a memory and a processor. Among them, an executable program code is stored in the memory, and the processor is used to call and execute the executable program code to execute the vision-assisted battery shell processing quality inspection method provided by an embodiment of the present invention.

[0095] An embodiment of the present invention can divide the functions of the device according to the above method examples. For example, it can correspond to each functional module, or integrate two or more functions into one processing module. The above integrated module can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is illustrative, only a logical function division, and there may be other division methods in actual implementation.

[0096] In the case where each module is divided according to each function, the device may further include a signal uploading module, a determination module, an adjustment module, etc. It should be noted that all relevant contents of each step involved in the above method embodiment can be cited in the function description of the corresponding functional module, and will not be elaborated here.

[0097] It should be understood that the device provided in the embodiments of the present invention is used to execute the above-mentioned method for visual-aided quality inspection of battery housing processing, so the same effects as the above implementation method can be achieved.

[0098] In the case of adopting integrated units, the device may include a processing module and a storage module. Among them, when the device is applied to a device, the processing module can be used to control and manage the actions of the device. The storage module can be used to support the device to execute mutual program codes, etc. Among them, the processing module can be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits described in combination with the disclosure of the present invention. The processor can also be a combination that realizes computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module can be a memory.

[0099] In addition, the device provided in the embodiments of the present invention may specifically be a chip, a component, or a module. The chip may include a connected processor and a memory; among them, the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute the above-mentioned method for visual-aided quality inspection of battery housing processing provided in the above embodiments.

[0100] The embodiments of the present invention also provide a computer-readable storage medium, in which computer program codes are stored. When the computer program codes run on a computer, the computer is enabled to execute the above-related method steps to implement the above-mentioned method for visual-aided quality inspection of battery housing processing provided in the above embodiments.

[0101] The embodiments of the present invention also provide a computer program product. When the computer program product runs on a computer, the computer is enabled to execute the above-related steps to implement the above-mentioned method for visual-aided quality inspection of battery housing processing provided in the above embodiments.

[0102] Among them, the device, computer-readable storage medium, computer program product or chip provided by the embodiments of the present invention are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be elaborated here. Through the description of the above embodiments, those skilled in the art can understand that for the convenience and conciseness of description, only the above-mentioned division of each functional module is used as an example. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In the embodiments provided by the present invention, it should be understood that the disclosed device and method can be implemented in other ways.

[0103] The device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in an electrical, mechanical or other form.

[0104] It should also be noted that in this article, the terms "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or terminal device including the said element.

[0105] It should be noted that the above-mentioned sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0106] Each embodiment in this specification is described in a progressive manner. The same or similar parts between each embodiment can be referred to each other, and the key points of each embodiment are the differences from other embodiments.

[0107] The above content is only a specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention.

Claims

1. A visually assisted battery casing processing quality inspection method, characterized in that: The method comprises the following steps: Acquire a shell image of the battery shell after processing; detect a weld and a corresponding welding crack in the shell image; Determining the crack extension of the welding crack according to the diffusion of the welding crack in different directions; Analyze the texture uniformity and texture interval of the weld to determine the heat input level corresponding to the weld; cluster the welds based on the abnormal heat input level to obtain weld clusters; Determine the impact of weld cracks based on the heat input and crack extension of each weld cluster; Determine the shell defect coefficient based on the deformation of the battery shell, the location and sharpness of the welding crack; compare the crack elongation and shell defect coefficient to determine the welding risk factor; The processing failure type of the battery casing is determined based on the degree of impact and the welding risk factor.

2. The visually assisted battery shell processing quality inspection method according to claim 1 is characterized in that: Determining the crack extension of the welding crack according to the diffusion of the welding crack in different directions includes: Determine an extension direction line of the welding crack; divide the welding crack into a plurality of pixel point clusters based on a vertical direction of the extension direction line of the welding crack; For each pixel cluster, the distance between each pixel in the pixel cluster and the extension direction line is calculated; the ratio between the maximum distance and the information entropy of the distance is calculated as the diffusion coefficient of the pixel cluster; The crack extension of the welding crack on the battery casing is determined based on the diffusion coefficient between different pixel clusters.

3. The visually assisted battery shell processing quality inspection method according to claim 1 is characterized in that: The analyzing the texture uniformity and texture interval of the weld to determine the heat input level corresponding to the weld includes: Perform edge detection on the weld, obtain the edge on the weld and record it as the weld texture; for any weld texture in the weld, obtain the distance between the weld texture and the weld textures on both sides as the weld texture distance; Perform uniform analysis on the weld texture to obtain the weld texture uniform density; The degree of heat input corresponding to the weld is determined by combining the texture uniformity density, weld texture distance, and weld texture color.

4. The visually assisted battery shell processing quality inspection method according to claim 1, characterized in that: The method for obtaining the abnormal heat input degree is: The absolute value of the difference between the currently detected heat input level and the normal heat input level is calculated, and when the absolute value of the difference is greater than a preset normal difference threshold, the current heat input level is determined to be an abnormal heat input level.

5. The visually assisted battery shell processing quality inspection method according to claim 1 is characterized in that: The step of determining the degree of influence of the welding cracks according to the heat input degree and crack extension of each weld cluster comprises: The information entropy of crack extensibility of all welding cracks in the battery casing processing and the mean of crack extensibility are obtained, and the mean of the heat input degree in each weld cluster is obtained; according to the difference between the mean of the heat input degree and the mean of the crack extensibility, the information entropy of the crack extensibility is adjusted, and the adjusted value is normalized to obtain the degree of influence of the welding crack.

6. The visually assisted battery casing processing quality inspection method according to claim 1, characterized in that: The shell defect coefficient is determined according to the deformation of the battery shell, the location and sharpness of the welding crack, including: Determine the distance degree of each welding crack according to the distance between the centroid of each welding crack and the edge of the weld on the battery housing; Determine the shell deformation coefficient of the battery shell corresponding to each welding crack according to the changes in the upper edges of the battery shell plates on both sides corresponding to each welding crack; According to the fluctuation between the edge pixels of the welding crack, the sharpness of the crack is obtained; The shell defect coefficient of each welding crack is determined in combination with the distance degree, the shell deformation coefficient and the sharpness of the crack; wherein the shell deformation coefficient and the sharpness are both positively correlated with the shell defect coefficient, and the distance degree is negatively correlated with the shell defect coefficient.

7. The visually assisted battery shell processing quality inspection method according to claim 1, characterized in that: The comparison of crack extension and shell defect coefficient to determine welding risk coefficient includes: A curve fitting is performed on the shell defect coefficient of each welding crack of the battery shell to obtain a first fitting curve; a curve fitting is performed on the crack elongation of each welding crack of the battery shell to obtain a second fitting curve; and the mean square error between the first fitting curve and the second fitting curve is calculated as the welding risk coefficient.

8. The visually assisted battery shell processing quality inspection method according to claim 1, characterized in that: The combination of the degree of influence and the welding risk factor to determine the processing failure type of the battery casing includes: The affected degree is corrected by the welding risk coefficient, and the corrected result value is used as the processing failure degree of the battery shell; When the degree of processing fault is greater than or equal to the preset fault threshold, it is determined that there is a problem in the battery shell welding process; When the degree of processing fault is less than the preset fault threshold, it is determined that the battery casing has factory quality problems.

9. The visually assisted battery shell processing quality inspection method according to claim 1, characterized in that: Detecting welding cracks in the housing image includes: The normal weld features are learned by the autoencoder, and the welding cracks in the shell image are identified through the reconstruction error.

10. A visually assisted battery shell processing quality inspection system, characterized in that: The system includes the following modules: An initial detection module is used to obtain a shell image of the battery shell after processing; and detect the weld seam and the corresponding welding crack in the shell image; An extension analysis module, used to determine the crack extension of the welding crack according to the diffusion of the welding crack in different directions; Heat input analysis module, used to analyze the texture uniformity and texture spacing of the weld and determine the heat input level corresponding to the weld; Based on the abnormal heat input degree, the welds are clustered to obtain weld clusters; Impact analysis module to determine the impact of weld cracks based on the heat input and crack extension of each weld cluster; The processing risk determination module is used to determine the shell defect coefficient according to the deformation of the battery shell, the location and sharpness of the welding crack; compare the crack elongation and the shell defect coefficient to determine the welding risk coefficient; The processing fault determination module is used to determine the processing fault type of the battery casing in combination with the degree of influence and the welding risk factor.

Images