Tab welding detection method and device, storage medium and program product

Abstract

The invention discloses a tab welding detection method and device, a storage medium and a program product. The tab welding detection method comprises the steps that welding height data of a to-be-detected battery tab is obtained; acquiring a welding image of the battery tab under the condition that the welding height data is judged to meet a preset abnormal condition based on the target statistical characteristics; the target statistical characteristics are statistical data of welding height data of a plurality of normal battery tabs; and judging whether the battery tab has a welding missing problem or not according to the welding image. Abnormality judgment is performed on the welding height data of the battery tab by using the statistical characteristics of a plurality of normal welding height data, so that the situation that the accuracy is reduced by using threshold judgment is avoided. In consideration of the situation that the welding height is reduced due to tab folding, and the numerical value of the welding height can be influenced by the abrasion of an adapter sheet, a foil material, a welding head and the like, comprehensive analysis and judgment are carried out by combining the welding height data and the welding image, so that the over-killing problem caused by judgment by using a single data source is avoided, and the detection accuracy is improved.

Description

Technical Field

[0001] The present application relates to the technical field of batteries, and particularly to a method and device for detecting ear welding, a storage medium, and a program product. Background Art

[0002] Due to process and equipment reasons, there will be certain defects in the production process of lithium batteries, and various detection means are required to detect the defects to improve the battery yield. For example, there is a risk of missed welding in the ultrasonic welding of battery ears and adapter plates. Especially during the transportation of batteries, the lower ears directly contact the tray, and frequent folding leads to missed welding, which is difficult to monitor. Currently, there is no effective detection means. If it flows to subsequent processes, it will cause batch disassembly and material loss. Moreover, missed welding may make the connection between the ear and the adapter plate unstable, resulting in a decline in battery performance and even posing a safety risk.

[0003] The above statements are only used to provide background technical information related to the present application, and do not necessarily constitute prior art. Summary of the Invention

[0004] In view of the above problems, the present application proposes a method and device for detecting ear welding, a storage medium, and a program product to solve the problem of battery anomalies caused by missed welding in the welding of battery ears in the prior art.

[0005] In a first aspect of the present application, a method for detecting ear welding is proposed. The method includes: Obtaining first welding height data of a battery ear to be detected; When it is determined that the first welding height data meets a preset abnormal condition based on a target statistical feature, obtaining a welding image of the battery ear; the target statistical feature is statistical data of second welding height data of multiple normal battery ears; Determining whether there is a problem of missed welding of the battery ear according to the welding image.

[0006] In the technical solution of the embodiment of the present application, by using the statistical feature of the welding height data of multiple normal battery ears to determine the abnormality of the welding height data of the battery ear to be detected, the use of threshold determination is avoided to reduce the detection accuracy. And considering that in addition to the folding of the ear causing the welding height to become smaller, the adapter plate, foil thickness, and wear of the welding head will also affect the value of the welding height. Therefore, through comprehensive analysis and determination by combining the welding height data and the welding image of the battery ear, the overkill problem caused by using a single data source for determination is avoided, thereby improving the accuracy of ear welding detection.

[0007] In some embodiments, the determining that the first welding height data is abnormal based on the target statistical feature includes: Determine multiple metrics for the first welding height data according to the target statistical features; when each of the metrics meets its corresponding abnormal determination condition, determine that the first welding height data is abnormal.

[0008] In this embodiment, by using the target statistical features, metrics in multiple dimensions of the first welding height data are calculated, and thus abnormal determination is performed by combining the metrics in multiple dimensions. When the metrics in multiple dimensions all meet their respective abnormal determination conditions, it is considered that there is an abnormality in the first welding height data of the battery tab, thereby reducing the overkill problem in tab welding detection.

[0009] In some embodiments, the multiple metrics at least include Mahalanobis distance and percentile; the Mahalanobis distance is obtained by using the covariance matrix and the mean value in the first welding height data and the target statistical features, the percentile is one of the features in the target statistical features, and the percentile is used to represent the second welding height data at the corresponding position of a preset percentage in the second welding height data of multiple normal battery tabs; the abnormal determination condition corresponding to the Mahalanobis distance is that the Mahalanobis distance exceeds a preset value; the abnormal determination condition corresponding to the percentile is that the first welding height data is less than the percentile.

[0010] In this embodiment, the Mahalanobis distance of the first welding height data is calculated by using the covariance matrix and the mean value in the target statistical features. The Mahalanobis distance can characterize the degree to which the first welding height data deviates from the distribution of the normal data set. Therefore, when the Mahalanobis distance exceeds the preset value, it indicates that the first welding height data deviates from the normal range. Considering that there are two possibilities for the reason of abnormal Mahalanobis distance, that is, the welding height data deviates from the lower limit of the normal range and the upper limit of the normal range, and missed welding will only cause the welding height data to deviate downward. Therefore, on the basis of judging the abnormal Mahalanobis distance, by comparing the percentile in the target statistical features with the first welding height data, the percentile can characterize the data distribution. If the first welding height data is smaller than the percentile, it indicates that the first welding height data is abnormal and deviates downward, so as to more accurately realize the determination of abnormal welding height data and avoid the overkill problem.

[0011] In some embodiments, the method further includes: Set a counter; increment the count value of the counter by 1 for each battery tab detection completed; when the count value of the counter reaches a preset quantity, obtain the second welding height data of normal battery tabs from the battery tabs whose detection has been completed, reset the counter, and continue to execute the step of incrementing the count value of the counter by 1 for each battery tab detection completed; update the target statistical features by using the obtained second welding height data.

[0012] In this embodiment, considering that the industrial production environment is dynamically changing, for example, the wear of the welding head and the welding seat will affect various parameters in the ultrasonic welding process. By automatically adjusting the target statistical features using the welding height data of normal battery tabs after detecting a certain number of battery tabs each time, the dynamic changes in the production environment can be reflected in real time, thereby maintaining the detection accuracy and reliability.

[0013] In some embodiments, the updating of the target statistical features using the obtained second welding height data includes: Determining the covariance matrix and the mean value of the obtained second welding height data; respectively replacing the covariance matrix and the mean value in the target statistical features with the determined covariance matrix and mean value; sorting the obtained second welding height data in ascending order; determining the target welding height data at the position corresponding to the preset percentage using the sorted second welding height data; and replacing the percentile in the target statistical features with the target welding height data.

[0014] In this embodiment, by calculating the covariance matrix, the mean value, and the percentile at the preset percentage position of the newly obtained second welding height data to update the target statistical features, the accuracy of subsequent calculations of multi-dimensional index values (such as Mahalanobis distance) can be ensured.

[0015] In some embodiments, the determining whether there is a problem of missed welding of the battery tab according to the welding image includes: Determining a gray level threshold for distinguishing the foreground and the background based on the welding image; converting the welding image into a binary image using the gray level threshold; and inputting the binary image into a trained detection model so that the detection model determines whether there is a problem of missed welding of the battery tab according to the binary image.

[0016] In this embodiment, considering that the acquisition of the welding image is easily affected by light interference, especially when the tab material is metal foil, which has a reflective property. Therefore, by determining a dynamic gray level threshold to perform binary processing on the welding image, while accurately distinguishing the foreground and the background, the data processing amount of the detection model is reduced, so that the trained detection model can perform anomaly detection on a relatively small binary image, improving the detection efficiency.

[0017] In some embodiments, the determining a gray level threshold for distinguishing the foreground and the background based on the welding image includes: Determining the gray level histogram of the welding image; and using a preset between-class variance relation and the gray level histogram to solve for the threshold that maximizes the between-class variance in the preset between-class variance relation as the gray level threshold for distinguishing the foreground and the background.

[0018] In this embodiment, the optimal grayscale threshold for binarization processing is obtained by using the method of maximizing the variance between classes, so as to be applicable to images under different lighting conditions.

[0019] In some embodiments, the first welding height data at least includes the height before welding and the height after welding. The height before welding is the thickness of the battery tab before welding, and the height after welding is the thickness of the battery tab after welding.

[0020] In this embodiment, considering that there is a certain correlation between the height before welding and the height after welding, the fluctuation of one variable will be masked by the fluctuation of the other variable, and there will be missed detections when analyzing the height before welding and the height after welding separately. Therefore, during the detection process, the height before welding and the height after welding of the battery tab are obtained for subsequent abnormal determination, so as to avoid the influence of the correlation between the height before welding and the height after welding on the detection accuracy.

[0021] The second aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the method described in the first aspect are implemented.

[0022] The third aspect of the present application provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by the processor, the steps of the method described in the first aspect are implemented.

[0023] An embodiment of the fourth aspect of the present application provides a computer program product, including a computer program, and the computer program is executed by a processor to implement the method described in the first aspect above.

[0024] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are hereinafter specifically described. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. And in all the drawings, the same reference numerals are used to represent the same components. In the drawings: Figure 1 It is a schematic diagram of flattening the tab after folding in the prior art; Figure 2 It is a schematic diagram of image acquisition of tab folding in the prior art; Figure 3Schematic diagram of pre-welding height and post-welding height shown according to an exemplary embodiment; Figure 4 Schematic diagram of image acquisition of ear wrinkling in the prior art; Figure 5 Flowchart of a method for detecting ear welding shown according to an exemplary embodiment; Figure 6 Online detection flowchart of ear welding shown according to an exemplary embodiment; Figure 7 Schematic diagram of detection results of a large number of battery cells shown according to an exemplary embodiment; Figure 8 Schematic diagram of the hardware structure of an electronic device shown according to an exemplary embodiment; Figure 9 Schematic diagram of the structure of a storage medium shown according to an exemplary embodiment. Detailed implementation manners

[0026] Embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present application more clearly, so they are only examples and cannot be used to limit the protection scope of the present application.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion.

[0028] In the description of the embodiments of this application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of this application, "a plurality of" means more than two unless otherwise specifically defined.

[0029] Referring to "embodiments" herein means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of this application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0030] In the description of the embodiments of the present application, the term "and / or" is merely a relational description of associated objects, indicating three possible relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. Additionally, in this text, the character " / " generally indicates an "or" relationship between the associated objects before and after.

[0031] In the description of the embodiments of the present application, the term "plurality" refers to two or more (including two). Similarly, "multiple groups" refers to two or more groups (including two groups), and "multiple pieces" refers to two or more pieces (including two pieces).

[0032] Currently, regarding the problem of missed welding caused by the lower layer of the battery tab being folded during the ultrasonic welding process, there are roughly three solutions: The first one adopts a physical method: as Figure 1 shown, a flattening wire is placed above the tray. When transferring the battery, when the battery is placed on the tray from directly above, the lower layer of the folded tab is flattened by the flattening wire.

[0033] The second one adopts a vision method: as Figure 2 shown, after the tab welding is completed, a CCD (Charge Coupled Device) installed below can take pictures of the tab area, and the image algorithm is used to determine whether the tab is folded.

[0034] The third one adopts a data analysis method: The ultrasonic welding process can be divided into the welding head pressing the tab and then performing high-speed vibration and downward pressure. By monitoring the vertical height between the welding head and the welding seat, this vertical height can include the height before welding and the height after welding. As Figure 3 shown, the height before welding refers to the thickness of the tab plus the adapter before welding, and the height after welding refers to the thickness of the tab plus the adapter after welding). When the tab is folded, both the height before welding and the height after welding will become smaller due to the lack of the tab in the welding area, so whether the tab is missed welded is identified by comparing the height before welding or the height after welding with the threshold range.

[0035] For the first physical method mentioned above, to ensure that the flattening wire can play a flattening role, it is necessary to ensure that the end of the folded tab does not exceed the position of the flattening wire, and the crease position of the fold needs to exceed the position of the flattening wire. These limiting conditions result in this method only being able to solve some of the missed welding problems, and there is also a risk of scratching the tab by the flattening wire; for the second vision method mentioned above, the CCD image is easily interfered by light. Especially, the metal foil of the tab has the characteristic of reflecting light, which is likely to cause misjudgment, and the image algorithm has a relatively low ability to distinguish between the phenomenon of the tab being wrinkled and the tab being folded. As Figure 4 shown, the CCD image of the wrinkled tab and the above Figure 2The CCD images of the tab folding shown are almost similar and difficult to distinguish. For the data analysis adopted in the third case, in addition to the tab folding causing the welding height to become smaller, the wear of the adapter plate, foil, and welding head will also affect the post-welding height value. And when the foil is thin, the folding of a few layers has little impact on the height data. Therefore, it is very difficult to judge the abnormality of the height data.

[0036] Based on this, some embodiments of the present application propose a tab welding detection scheme. By using the target statistical features of the welding height data of multiple normal battery tabs, the welding height data of the battery tabs to be detected are determined for abnormality, avoiding the reduction of data determination accuracy by using threshold determination. And considering that in addition to the tab folding causing the welding height to become smaller, the adapter plate, foil thickness, welding head wear, etc. will all affect the value of the welding height. Therefore, through comprehensive analysis and determination by combining the welding height data and welding images of the battery tabs, the overkill problem caused by using a single data source for determination is avoided, thereby improving the accuracy of tab welding detection.

[0037] The battery tabs in the tab welding detection scheme disclosed in the embodiments of the present application may, but are not limited to, the tabs of a single battery cell. The single battery cell may be a secondary battery, and a secondary battery refers to a single battery cell that can be activated by charging after discharging and can continue to be used. The single battery cell may be a lithium-ion battery, sodium-ion battery, sodium-lithium-ion battery, lithium-metal battery, sodium-metal battery, lithium-sulfur battery, magnesium-ion battery, nickel-metal hydride battery, nickel-cadmium battery, lead-acid battery, etc., and the embodiments of the present application do not limit this.

[0038] To enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application.

[0039] Figure 5 As shown in the flowchart of a tab welding detection method according to an exemplary embodiment, the tab welding detection method includes the following steps: Step 501: Obtain the first welding height data of the battery tabs to be detected; Step 502: When it is determined that the first welding height data is abnormal based on the target statistical features, obtain the welding image of the battery tabs. The target statistical features are the statistical data of the second welding height data of multiple normal battery tabs; Step 503: Determine whether there is a problem of missed welding of the battery tabs according to the welding image.

[0040] Both the first welding height data and the second welding height data are welding data monitored during the ultrasonic welding of battery tabs. The first welding height data can be regarded as the height data of the tab to be detected, and the second welding height data can be regarded as the height data corresponding to the battery tabs with normal detection results. Among them, a normal detection result means that there is no problem of missed welding in the battery tabs.

[0041] Exemplarily, the first welding height data includes two dimensions: the height before welding and the height after welding. There is an obvious correlation between the height before welding and the height after welding. The fluctuation of one variable will be masked by the fluctuation of the other variable. Analyzing the height before welding and the height after welding separately may result in missed detections. Therefore, by obtaining the first welding height data including the height before welding and the height after welding for anomaly determination, the influence of the correlation between the height before welding and the height after welding on the detection accuracy can be avoided.

[0042] The target statistical feature is used to characterize the statistical distribution of the dataset composed of the welding height data of the battery tabs with normal detection results.

[0043] The welding image can be understood as the image of the tab area after ultrasonic welding is completed. Specifically, it refers to the image collected by the CCD on the side of the battery tab facing away from the welding head, and can also be called the bottom CCD image.

[0044] It should be noted that the execution order of the above steps 502 and 503 is not specifically limited in this application. It is also possible to first perform the determination of the welding image, and on the basis of determining an anomaly, then perform the anomaly determination of the first welding height data.

[0045] So far, the above Figure 5 The tab welding detection process shown is completed. By using the statistical features of the welding height data of multiple normal battery tabs, the anomaly determination of the welding height data of the battery tab to be detected is performed, avoiding the reduction of detection accuracy by using threshold determination. And considering that in addition to the fact that the flipping of the tab will cause the welding height to become smaller, the adapter plate, foil thickness, welding head wear, etc. will all affect the value of the welding height. Therefore, by combining the welding height data and the welding image of the battery tab for comprehensive analysis and determination, the problem of overkill caused by using a single data source for determination is avoided, thereby improving the accuracy of tab welding detection.

[0046] In some embodiments, the process of determining that the first welding height data is abnormal based on the target statistical feature in the above step 502 may include: According to the target statistical feature, multiple indicators of the first welding height data are determined. When each indicator meets its corresponding anomaly determination condition, it is determined that the first welding height data is abnormal.

[0047] Multiple metrics can be understood as criteria for determining the abnormality of the first welding height data in different dimensions.

[0048] The abnormality determination condition refers to the condition that the metrics need to meet when indicating the abnormality of the first welding height data. Different metrics have different abnormality determination conditions.

[0049] In this embodiment, by using the target statistical features, multiple-dimensional metrics of the first welding height data are calculated, and then abnormality determination is performed by combining multiple-dimensional metrics. When the metrics in multiple dimensions all meet their respective abnormality determination conditions, it is considered that the first welding height data of the battery tab is abnormal, thus avoiding the overkill problem in tab welding detection.

[0050] In some embodiments, the multiple metrics described above may include Mahalanobis distance and percentile.

[0051] Among them, the Mahalanobis distance is a statistical method for measuring the distance between a multi-dimensional data point and the distribution of a data set, which takes into account the correlation between each dimension in the data set. As mentioned above, the first welding height data includes two dimensions: the height before welding and the height after welding. A relatively large Mahalanobis distance may indicate abnormal changes in the first welding height data (height before welding and height after welding), such as welding parameter deviations from the normal range caused by wear of the welding head or welding seat. Therefore, the larger the Mahalanobis distance, the greater the degree to which the data point deviates from the data set distribution, and it may be an abnormal point.

[0052] It can be seen that the abnormality determination condition corresponding to the Mahalanobis distance is: the Mahalanobis distance exceeds a preset value. This preset value can be set according to actual experience. For example, the preset value of the Mahalanobis distance is set to 5.

[0053] The Mahalanobis distance is obtained by using the first welding height data, the covariance matrix, and the mean value in the target statistical features. The covariance matrix describes the linear relationship between each dimension in multi-dimensional data and reflects the correlation between data dimensions. In the calculation of the Mahalanobis distance, the covariance matrix is used to standardize the data and eliminate the influence of the correlation between dimensions. The mean value is the sum of all data points in the data set divided by the number of data points, which reflects the central tendency of the data. In welding quality detection, the mean value can represent the typical values of the height before welding and the height after welding. The calculation formula of the Mahalanobis distance is as follows:

[0054] In the above formula, , represents the first welding height data of the battery tab to be detected, is the height before welding and is the height after welding; , represents the mean value, that is, the mean vector; represents the covariance matrix, , the calculation formulas for the elements in the covariance matrix are as follows:

[0055]

[0056]

[0057] Among them, represents the pre-welding height of the i-th battery tab among n normal battery tabs, represents the post-welding height of the i-th battery tab among n normal battery tabs.

[0058] The percentile is one of the features in the target statistical features. The percentile represents the second welding height data at the position corresponding to the preset percentage in the second welding height data of multiple normal battery tabs, and it reflects the distribution of the data. In welding quality inspection, the percentile can be used to determine whether the first welding height data is too small or too large. Considering that there are two possibilities for the cause of the abnormal Mahalanobis distance, that is, the welding height data deviates from the lower limit of the normal range and the upper limit of the normal range, and missed welding only causes the first welding height data to be on the lower side. Therefore, on the basis of judging the abnormal Mahalanobis distance, the percentile is continued to be used to determine whether the first welding height data is too small.

[0059] It can be seen from this that the abnormal determination condition corresponding to the percentile is: the first welding height data is less than the percentile.

[0060] It can be understood that the first welding height data includes the pre-welding height and the post-welding height. Therefore, the pre-welding height is compared with its corresponding percentile, and the post-welding height is compared with its corresponding percentile.

[0061] In this embodiment, the Mahalanobis distance of the first welding height data is calculated by using the covariance matrix and the average value in the target statistical features. The Mahalanobis distance can characterize the degree to which the first welding height data deviates from the distribution of the normal data set. Therefore, when the Mahalanobis distance exceeds the preset value, it indicates that the first welding height data deviates from the normal range. Considering that there are two possibilities for the cause of the abnormal Mahalanobis distance, that is, the welding height data deviates from the lower limit of the normal range and the upper limit of the normal range, and missed welding only causes the welding height data to be on the lower side. Therefore, on the basis of judging the abnormal Mahalanobis distance, by comparing the percentile in the target statistical features with the first welding height data, the percentile can characterize the distribution of the data. If the first welding height data is smaller than the percentile, it indicates that the first welding height data is abnormal and on the lower side, so as to more accurately realize the determination of the abnormal welding height data and avoid the overkill problem.

[0062] It should be noted that in addition to using Mahalanobis distance and percentile to determine anomalies, other metrics such as Euclidean distance, clustering analysis, and isolation forest can also be used to determine the anomalies in the first welding height data. It is understandable that different metrics need to be designed with appropriate calculation methods according to actual requirements, and the target statistical features may be expanded. For example, when using Euclidean distance to measure the deviation degree of the first welding height data, the Euclidean distance between the first welding height data and the average value of the data set can be calculated.

[0063] In some embodiments, during the process of battery tab detection, it may further include: Setting a counter; incrementing the count value of the counter by 1 for each completed battery tab detection. When the count value of the counter reaches a preset quantity, obtaining the second welding height data of normal battery tabs from the completed battery tabs, resetting the counter, and continuing to execute the step of incrementing the count value of the counter by 1 for each completed battery tab detection, so as to update the target statistical features using the obtained second welding height data.

[0064] That is to say, taking the second welding height data of normal battery tabs within a specific window size as a training set, calculating the target statistical features using the second welding height data in the training set, and using the target statistical features to determine anomalies for subsequent battery tabs. When the number of detected battery tabs reaches the specific window size, obtaining the second welding height data of normal battery tabs within the window to recalculate the target statistical features for subsequent anomaly determination of battery tabs. This can keep the target statistical features in an updated state, which can reflect in real-time the changes in various parameters during the ultrasonic welding process caused by the wear of the welding head and welding seat, and avoid the decline in the generalization ability of the target statistical features due to the wear of the welding head and welding seat.

[0065] In this embodiment, considering that the industrial production environment is dynamically changing, for example, the wear of the welding head and welding seat will affect various parameters during the ultrasonic welding process. By automatically adjusting the target statistical features using the welding height data of normal battery tabs after detecting a certain number of battery tabs, the dynamic changes in the production environment can be reflected in real-time, thus maintaining the detection accuracy and reliability.

[0066] In some embodiments, as described above, when the target statistical features include covariance matrix, average value, and percentile, the process of updating the target statistical features using the obtained second welding height data may include: Determine the covariance matrix and mean value of the obtained second welding height data, use the determined covariance matrix and mean value to replace the covariance matrix and mean value in the target statistical feature respectively, sort the obtained second welding height data in ascending order, use the sorted second welding height data to determine the target welding height data at the position corresponding to the preset percentage, and use the target welding height data to replace the percentile in the target statistical feature.

[0067] The calculation methods of the covariance matrix and the mean value can be referred to the relevant descriptions in the above embodiments, and will not be elaborated here.

[0068] The preset percentage can be set according to actual experience, and the calculation formula for the position corresponding to the preset percentage is as follows: Q p =( n +1)× p Where, p is the preset percentage, and n is the number of the second welding height data, that is, the number of multiple normal battery tabs.

[0069] It should be noted here that Q p The result of Q p may be an integer or may not be an integer. If Q p the result of is an integer, directly find the target welding height data at the Q p corresponding position in the sorted second welding height data to replace the percentile in the target statistical feature. If Q p the result of is not an integer, find the target welding height data at the left and right two positions corresponding to

[0070] in the sorted second welding height data, and after linearly interpolating the target welding height data at the two positions, use the interpolation result to replace the percentile in the target statistical feature.

[0071] For example, assume that the number of the second welding height data is 1000, p =0.01, Q p =( n +1)× p= 10.01, so as to find the second welding height data at the 10th and 11th positions, and replace the percentile in the target statistical feature with the average value of these two second welding height data. That is, find the pre-welding heights at the 10th and 11th positions, replace the percentile corresponding to the pre-welding height in the target statistical feature with the average value of these two pre-welding heights, and find the post-welding heights at the 10th and 11th positions, and replace the percentile corresponding to the post-welding height in the target statistical feature with the average value of these two post-welding heights.

[0072] In this embodiment, by calculating the covariance matrix, average value, and percentile at the preset percentage position of the newly obtained second welding height data to update the target statistical feature, the accuracy of subsequent calculation of multi-dimensional index values (such as Mahalanobis distance) can be ensured.

[0073] In some embodiments, step 503 of determining whether there is a welding omission problem of the battery tab according to the welding image may include: Determine the gray threshold for distinguishing the foreground and background based on the welding image, convert the welding image into a binary image using the gray threshold, and then input the binary image into the trained detection model so that the detection model determines whether there is a welding omission problem of the battery tab according to the binary image.

[0074] The gray threshold can be understood as a dynamic threshold, and the size of this threshold is obtained according to the gray distribution of the welding image.

[0075] The binary processing formula is as follows:

[0076] Among them, is the gray threshold, that is, the optimal threshold, is the foreground pixel value, usually .

[0077] In this embodiment, considering that the acquisition of the welding image is easily affected by light interference, especially when the tab material is metal foil, which has a reflective property, a dynamic gray threshold is determined to perform binary processing on the welding image. While accurately distinguishing the foreground and background, the data processing volume of the detection model is reduced, so that the trained detection model can perform anomaly detection on a relatively small binary image, improving the detection efficiency.

[0078] In some embodiments, for the process of determining the gray threshold for distinguishing the foreground and background based on the welding image, it may include: Determine the gray histogram of the welding image, and use the preset between-class variance relationship formula and the gray histogram to solve the threshold that maximizes the between-class variance in the preset between-class variance relationship formula as the gray threshold for distinguishing the foreground and background.

[0079] The grayscale histogram is used to count the number of pixels at each grayscale level in an image. Assuming the grayscale range of the image is [0, L−1], where L is the total number of grayscale levels (usually 256), calculate the probability of each grayscale level. , is the total number of pixels in the image, i is the grayscale level (i = 0, 1, 2, …, L-1), is the number of pixels at the i-th grayscale level. The preset between-class variance relationship can be expressed as follows:

[0080] where, is the between-class variance of the foreground and background, is the threshold, is the global mean of the image, , is the cumulative probability, , is the cumulative mean, .

[0081] For each threshold , calculate the between-class variance , and then find the threshold that maximizes as the grayscale threshold for distinguishing the foreground and background.

[0082] For the technical solution given in the above embodiment, Figure 6 FIG. 45 is a flowchart of an on-line detection of tab welding according to an exemplary embodiment, including the following steps: Step 1: Collect the pre-welding height and post-welding height of the battery tab to obtain a data point, and calculate the Mahalanobis distance of the data point using the covariance matrix and mean value in the target statistical features; Step 2: Compare the Mahalanobis distance with a preset value. If the Mahalanobis distance exceeds the preset value, execute Step 3; otherwise, execute Step 5; Step 3: Compare the pre-welding height of the battery tab with the percentile of the pre-welding height in the target statistical features, and compare the post-welding height of the battery tab with the percentile of the post-welding height in the target statistical features. If the pre-welding height of the battery tab is less than the percentile of the pre-welding height and the post-welding height of the battery tab is less than the percentile of the post-welding height, execute Step 4; otherwise, execute Step 5; Step 4: Determine whether the battery tab is abnormal according to the welding image. If so, execute Step 6; otherwise, execute Step 5; Step 5: Determine that the battery tab is welded normally without missing welding; Step 6: Determine that the battery tab is missing welding and generate an interception signal.

[0083] During the actual on-line detection process, a total of 38,786 battery tabs were detected, among which 2 were actually found to be unsoldered. Using this solution, a total of 11 unsoldered battery tabs were identified. As Figure 7 shown, the horizontal axis is the number of battery tabs starting from 0, and the vertical axis is the Mahalanobis distance of the battery tabs. Figure 7 In the figure, the hollow circles are marked as unsoldered battery tabs, and the solid dots are marked as Mahalanobis distances. Among them, 2 unsoldered battery tabs were 100% identified, and the overkill rate was 0.023%. It can be seen that this solution can significantly improve the accuracy of ultrasonic welding unsoldering detection.

[0084] The descriptions of the above embodiments tend to emphasize the differences between the embodiments. Their similarities or similarities can be referred to each other. For the sake of brevity, they will not be repeated in this article.

[0085] Those skilled in the art can understand that in the above method of the specific embodiment, the writing order of each step does not mean a strict execution order and does not constitute any limitation to the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0086] The embodiment of the present application also provides an electronic device corresponding to the tab welding detection method provided in the foregoing embodiment to execute the above tab welding detection method.

[0087] Figure 8 FIG. 18 is a hardware structure diagram of an electronic device shown according to an exemplary embodiment of the present application. The electronic device may include: a communication interface 601, a processor 602, a memory 603, and a bus 604; wherein, the communication interface 601, the processor 602, and the memory 603 complete mutual communication through the bus 604. The processor 602 can execute the tab welding detection method described above by reading and executing machine-executable instructions corresponding to the control logic of the tab welding detection method in the memory 603. The specific content of this method can be found in the above embodiments and will not be repeated here.

[0088] The memory 603 mentioned in the present application can be any electronic, magnetic, optical or other physical storage device, which can contain stored information, such as executable instructions, data, etc. Specifically, the memory 603 can be RAM (Random Access Memory), flash memory, a storage drive (such as a hard disk drive), any type of storage disk (such as an optical disk, a DVD, etc.), or a similar storage medium, or a combination thereof. Through at least one communication interface 601 (which can be wired or wireless), a communication connection is realized between the system network element and at least one other network element, and the Internet, a wide area network, a local area network, a metropolitan area network, etc. can be used.

[0089] The bus 604 can be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. Among them, the memory 603 is used to store programs, and after receiving the execution instruction, the processor 602 executes the program.

[0090] The processor 602 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor 602 or the instructions in the form of software. The above-mentioned processor 602 can be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed by the hardware decoding processor, or executed by the combination of the hardware and software modules in the decoding processor.

[0091] The electronic device provided in the embodiments of the present application and the tab welding detection method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by it.

[0092] The embodiments of the present application also provide a computer-readable storage medium corresponding to the tab welding detection method provided in the foregoing embodiments. Please refer to Figure 9 As shown, the computer-readable storage medium shown is an optical disc 30, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it will execute the tab welding detection method provided in any of the foregoing embodiments.

[0093] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here one by one.

[0094] The computer-readable storage medium provided by the above embodiments of the present application and the tab welding detection method provided by the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0095] The embodiments of the present application also provide a computer program product corresponding to the request processing method provided by the foregoing embodiments. The computer program product includes a computer program, and the computer program is executed by a processor to implement the tab welding detection method provided by the foregoing embodiments.

[0096] The computer program product provided by the above embodiments of the present application and the tab welding detection method provided by the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0097] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include known common knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and embodiments are only to be considered as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0098] It should also be noted that the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device including the element.

[0099] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for detecting tab welding, characterized in that: The method comprises: Acquire first welding height data of the battery tab to be detected; When it is determined based on a target statistical feature that the first welding height data is abnormal, obtaining a welding image of the battery tab; the target statistical feature is statistical data of second welding height data of multiple normal battery tabs; It is determined whether the battery tab has a welding leak problem according to the welding image.

2. The method according to claim 1, characterized in that The determining, based on the target statistical feature, that the first welding height data is abnormal comprises: Determining a plurality of indicators of the first welding height data according to the target statistical characteristics; When all the indicators satisfy their corresponding abnormality determination conditions, it is determined that the first welding height data is abnormal.

3. The method according to claim 2, characterized in that The multiple indicators include at least a Mahalanobis distance and a percentile; the Mahalanobis distance is obtained by using the first welding height data, the covariance matrix in the target statistical feature and the average value; the percentile belongs to one of the features in the target statistical feature; the percentile represents the second welding height data at a position corresponding to a preset percentage in the second welding height data of the multiple normal battery tabs; The abnormality determination condition corresponding to the Mahalanobis distance is: the Mahalanobis distance exceeds a preset value; The abnormal determination condition corresponding to the percentile is: the first welding height data is less than the percentile.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Set a counter; Each time a battery tab is detected, the count value of the counter is increased by 1; When the count value of the counter reaches a preset number, obtaining the second welding height data of a normal battery pole ear from the battery pole ear that has completed the inspection, resetting the counter, and continuing to perform the step of increasing the count value of the counter by 1 each time the inspection of a battery pole ear is completed; The target statistical feature is updated using the acquired second welding height data.

5. The method according to claim 4, characterized in that The method of updating the target statistical feature by using the acquired second welding height data comprises: determining a covariance matrix and a mean value of the acquired second welding height data; Replacing the covariance matrix and the mean value in the target statistical feature with the determined covariance matrix and the mean value respectively; The acquired second welding height data are sorted in ascending order; Determine the target welding height data at the position corresponding to the preset percentage by using the sorted second welding height data; The target welding height data is used to replace the percentile in the target statistical feature.

6. The method according to claim 1, characterized in that The determining whether the battery tab has a welding leak problem according to the welding image includes: Determining a grayscale threshold for distinguishing a foreground from a background based on the welding image; Converting the welding image into a binary image using the grayscale threshold; The binary image is input into a trained detection model so that the detection model determines whether the battery tab has a welding leak problem based on the binary image.

7. The method according to claim 6, characterized in that The method of determining a grayscale threshold for distinguishing a foreground from a background based on the welding image comprises: Determining a grayscale histogram of the welding image; By using the preset inter-class variance relational expression and the grayscale histogram, a threshold value that maximizes the inter-class variance in the preset inter-class variance relational expression is solved as the grayscale threshold value for distinguishing the foreground and the background.

8. The method according to any one of claims 1-3, 6-7, characterized in that: The first welding height data at least includes a pre-welding height and a post-welding height, wherein the pre-welding height is the thickness of the battery tab before welding, and the post-welding height is the thickness of the battery tab after welding.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer program product, comprising a computer program, characterized in that The computer program is executed by a processor to implement the method according to any one of claims 1 to 8.

Images