Welding method and welding system based on image recognition

Abstract

The application relates to the technical field of battery product welding, in particular to a welding method and a welding system based on image recognition, which comprises the following steps: acquiring first pole column point data of each pole column, first marker point data and second marker point data of two marker points; converting the first pole column point data into second pole column point data according to the actual conversion relationship determined according to the second marker point data and the first marker point data; acquiring first basic addressing data range for each pole column point; and comparing and judging whether the second pole column point data exceeds the first basic addressing data range, whether the distance difference between the first marker point data and the second marker point data exceeds a first distance threshold value in sequence; if both are not, welding is carried out based on the second pole column point data. According to the first basic addressing data and the distance of the marker points in the welding station and the addressing station, the application can effectively identify the abnormality of camera grabbing and axial movement and data transmission, comprehensively identify data conversion deviation, and improve the foolproof effect.

Description

Technical Field

[0001] This application relates to the technical field of battery product welding, and in particular to a welding method and welding system based on image recognition. Background Technology

[0002] In the battery manufacturing process, multiple cells are typically assembled together and connected to each other via busbars. Currently, in the field of laser welding, visual addressing technology can effectively improve the accuracy of welding points. Compared to mechanical positioning and blind welding, visual addressing technology can reduce the dynamic deviation of unstable welding point positions on different modules during the welding process, thereby improving the overall accuracy and precision of the welding process.

[0003] The entire welding process can be divided into two parts based on functionality: addressing and welding. Traditional welding combines these two parts, requiring each weld point to be photographed and addressed before welding. This takes up a lot of equipment operating time and greatly reduces equipment production efficiency.

[0004] The relevant documents disclose a laser welding visual positioning guidance method, which divides the entire welding process into two stations: the welding station and the addressing station. The steps include: product module positioning; acquiring pole addressing data; servo-driven camera taking pictures of marked points to locate coordinates; and converting pole addressing data into pole coordinates in the welding camera coordinate system. This enables welding to be performed simultaneously with addressing, reducing equipment running time and improving equipment production efficiency.

[0005] Regarding the aforementioned technologies, there is a problem that deviations can occur during the conversion of addressing data into welding data, and these deviations can have an irreversible impact on the final welding point. Summary of the Invention

[0006] In order to effectively identify camera capture and axial movement, data transmission anomalies, and comprehensive identification of data conversion deviations by using the first basic addressing data and combining the distance judgment of the marker point between the welding station and the addressing station, and to improve the error prevention effect, the purpose of this application is to provide a welding method based on image recognition.

[0007] The image recognition-based welding method provided in this application adopts the following technical solution:

[0008] Firstly, the image recognition-based welding method provided in this application adopts the following technical solution:

[0009] A welding method based on image recognition includes the following steps:

[0010] Obtain the addressing data of the battery product to be welded in the addressing station coordinate system, wherein the addressing data includes the first pole point data of each pole of the battery product to be welded and the first marker point data of two marker points;

[0011] Obtain the data of the second marker point of the two marker points of the battery product to be welded in the coordinate system of the welding station;

[0012] The addressing data is called to the welding station to obtain the first pole point data and the first marker point data.

[0013] Based on the second marker point data and the first marker point data called, the actual transformation relationship between the addressing station coordinate system and the welding station coordinate system is determined, and the first polar point data called is converted into the second polar point data in the welding station coordinate system based on the actual transformation relationship.

[0014] Obtain the first basic addressing data range of each pole point in the welding coordinate system, and determine whether the data of the second pole point exceeds the corresponding first basic addressing data range. The first basic addressing data range is generated based on product design and process requirements.

[0015] If so, perform the first corrective welding;

[0016] If not, obtain the first distance between the two markers based on the first marker data, obtain the second distance between the two markers based on the second marker data, compare the first distance and the second distance to obtain the distance difference, and determine whether the distance difference exceeds a preset first distance threshold.

[0017] If so, perform a second correction weld;

[0018] If not, welding is performed on the product to be welded based on the second pole point data.

[0019] By adopting the above technical solution, the welding operation is divided into two parts, one completed at the addressing station and the other at the welding station, based on functionality. This achieves addressing and welding simultaneously, reducing equipment uptime and improving production efficiency. For each pole point, a first basic addressing data range is pre-recorded. This first basic addressing data is obtained based on the design specifications and process requirements of the corresponding batch of products. The system then determines whether the second pole point data exceeds the corresponding first basic addressing data range. For cases exceeding the range, a first correction welding is performed. This effectively identifies and judges anomalies caused by camera misalignment at the addressing station, data transmission errors, etc., improving data accuracy. Furthermore, by comparing the first distance length and the second distance length to determine whether they exceed a preset first distance threshold, a second correction welding is performed for cases exceeding the range. This effectively identifies anomalies caused by camera misalignment at the addressing station and welding station, camera axial movement deviation, etc. Based on the normal data transmission and addressing station camera judgments, this combined judgment improves the overall data accuracy.

[0020] Optionally, after determining that the distance difference does not exceed the first distance threshold, the process includes:

[0021] Obtain the first basic welding range of the battery product to be welded in the welding coordinate system, wherein the first basic welding range is generated based on product design and process requirements;

[0022] Determine whether the data of the second pole point exceeds the welding range of the first base;

[0023] If so, perform a third correction weld;

[0024] If not, welding is performed on the product to be welded based on the second pole point data.

[0025] By adopting the above technical solution, the first basic welding range of the entire battery product to be welded is obtained on a macroscopic scale, and compared with the converted second pole point data. Based on the early microscopic local and data source-based error prevention, the later macroscopic overall and data conversion-based error prevention is added to further improve the welding accuracy.

[0026] Optionally, after determining that the distance difference does not exceed a preset first distance threshold, and before obtaining the first basic welding range of the battery product to be welded in the welding coordinate system, the method further includes:

[0027] Obtain the first marker point data of one of the marker points, and transform it to the welding coordinate system based on the theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first marker point transformation data;

[0028] Obtain the third distance length between the first marker point transformation data and the second marker point data of the given marker point;

[0029] Determine whether the length of the third distance exceeds a preset first length threshold;

[0030] If so, perform the fourth correction welding;

[0031] If not, obtain the first basic welding range for the battery product to be welded.

[0032] By adopting the above technical solution, the first marker point data of one of the marker points is obtained and converted using the inherent theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first marker point conversion data. The positional relationship between the first marker point conversion data and the second marker point data actually obtained by the camera is determined. If the first length threshold is exceeded, a fourth correction welding is performed. This can effectively identify and judge anomalies caused by camera deviation, data transmission, etc., from the point, thereby improving data accuracy.

[0033] Optionally, the first correction welding, the second correction welding, the third correction welding, and the fourth correction welding all include:

[0034] Pre-welding is performed using a weak laser based on the data from the second pole point;

[0035] Based on the pre-welding results, determine whether the deviation of each welding point from the center point of the corresponding pole exceeds the set welding deviation threshold.

[0036] If so, welding is performed at the welding station using a self-service welding method;

[0037] If not, welding is performed on the product to be welded based on the second pole point data.

[0038] By adopting the above technical solution, since some special products may indeed have relatively large welding errors during the assembly or stacking process, exceeding the threshold set according to the normal correction range, after pre-welding with a weak laser, it is combined with the judgment of whether the welding deviation threshold is exceeded, and the judgment is made again according to the actual situation, thereby reducing the probability of misjudgment.

[0039] Optionally, the first correction welding, prior to the pre-welding using a weak laser based on the second pole point data, further includes:

[0040] Determine whether the first pole point data being called exceeds the corresponding second basic addressing data range, wherein the second basic addressing data range is generated based on the first basic addressing data range;

[0041] If so, welding is performed at the welding station using a self-service welding method;

[0042] If not, pre-welding is performed using a weak laser based on the data from the second pole point.

[0043] By adopting the above technical solution, a second basic addressing data range is set. The second basic addressing data range is quantitatively expanded based on the first basic addressing data range and practical experience. When performing the first correction welding, a comparison and judgment are made again before pre-welding. While ensuring the error prevention effect, the welding process is comprehensively simplified.

[0044] Optionally, the second correction welding, prior to pre-welding using a weak laser based on the second pole point data, also includes:

[0045] Determine whether the distance difference exceeds a preset second distance threshold, wherein the second distance threshold is generated based on a first distance threshold;

[0046] If so, welding is performed at the welding station using a self-service welding method;

[0047] If not, pre-welding is performed using a weak laser based on the data from the second pole point.

[0048] By adopting the above technical solution, a second distance threshold is set. The second distance threshold is quantitatively expanded based on the first distance threshold and combined with practical experience. The comparison and judgment are made again before pre-welding. While ensuring the error prevention effect, the welding process is comprehensively simplified.

[0049] Optionally, the third correction welding, prior to pre-welding using a weak laser based on the second pole point data, also includes:

[0050] Determine whether the data of the second pole point exceeds the corresponding second basic welding range, wherein the second basic welding range is generated based on the first basic welding range;

[0051] If so, welding is performed at the welding station using a self-service welding method;

[0052] If not, pre-welding is performed using a weak laser based on the data from the second pole point.

[0053] By adopting the above technical solution, a second basic welding range is set. The second basic welding range is generated quantitatively based on the first basic welding range and combined with practical experience. The range is compared and judged again before pre-welding. While ensuring the error prevention effect, the welding process is comprehensively simplified.

[0054] Optionally, the fourth correction welding, before pre-welding using a weak laser based on the second pole point data, also includes:

[0055] Determine whether the length of the third distance exceeds a preset second length threshold, wherein the second length threshold is generated based on the first length threshold;

[0056] If so, welding is performed at the welding station using a self-service welding method;

[0057] If not, pre-welding is performed using a weak laser based on the data from the second pole point.

[0058] By adopting the above technical solution, a second length threshold is set. The second length threshold is generated based on the first length threshold and combined with practical experience. The two thresholds are compared and judged again before pre-welding. This ensures the error prevention effect while comprehensively simplifying the welding process.

[0059] Optionally, the actual transformation relationship between the addressing station coordinate system and the welding station coordinate system is determined based on the second marker point data and the retrieved first marker point data, including:

[0060] The first marker point data is transferred to the welding station coordinate system based on the theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first reference position coordinates of the marker point in the welding station.

[0061] Obtain the angle between the line connecting the points corresponding to the first reference position coordinates and the line connecting the points corresponding to the second marker data, and get the offset angle;

[0062] After rotating and offsetting the line connecting the points corresponding to the first reference position coordinates by the offset angle, the second reference position coordinates are obtained.

[0063] Obtain the offset between the second reference position coordinates and the second marker point data, where the offset angle and offset amount constitute the actual conversion relationship.

[0064] By adopting the above technical solution, the coordinates of the first reference position of the marker point in the welding station are determined based on the theoretical conversion relationship, and the offset angle is obtained by comparing the corresponding line of the second marker point data. Based on the offset angle, the offset amount is determined, and the actual conversion relationship is obtained, thereby reducing the deviation generated in the process of converting addressing data into welding data.

[0065] Secondly, the image recognition-based welding system provided in this application adopts the following technical solution:

[0066] The first acquisition module is used to acquire the addressing data of the battery product to be welded in the addressing station coordinate system and the second marker point data of the two marker points of the battery product to be welded in the welding station coordinate system. The addressing data includes the first pole point data of each pole of the battery product to be welded and the first marker point data of the two marker points.

[0067] The calling module is used to call the addressing data to the welding station to obtain the first pole point data and the first marker point data to be called.

[0068] The conversion module is used to determine the actual conversion relationship between the addressing station coordinate system and the welding station coordinate system based on the second marker point data and the first marker point data called, and to convert the first polar point data called into the second polar point data in the welding station coordinate system based on the actual conversion relationship.

[0069] The second acquisition module is used to acquire the first basic addressing data range of each pole point in the welding coordinate system.

[0070] The first judgment module is used to determine whether the second pole point data exceeds the corresponding first basic addressing data range, wherein the first basic addressing data range is generated based on product design and process requirements;

[0071] The first welding module is used to perform a first correction welding when the second pole point data exceeds the corresponding first basic addressing data range;

[0072] The third acquisition module is used to obtain the first distance between two marker points based on the first marker point data called, and to obtain the second distance between two marker points based on the second marker point data, when the second pole point data does not exceed the corresponding first basic addressing data range, and to compare the first distance length and the second distance length to obtain the distance difference.

[0073] The second judgment module is used to determine whether the distance difference exceeds a preset first distance threshold.

[0074] The second welding module is used to perform a second correction welding when the distance difference exceeds a preset first distance threshold; and to weld the product to be welded based on the second pole point data when the distance difference does not exceed the preset first distance threshold.

[0075] By adopting the above technical solutions, abnormalities caused by issues such as camera misalignment, data transmission errors, and camera axial movement deviation in the addressing station camera and welding station camera can be effectively identified and judged, thereby improving welding accuracy.

[0076] In summary, this application includes at least one of the following beneficial technical effects:

[0077] 1. The welding operation is broken down into two separate operations: one at the addressing station and the other at the welding station. This enables addressing and welding, reduces equipment uptime, and improves equipment production efficiency.

[0078] 2. Record the first basic addressing data range in advance for each pole point, and determine whether the data of the second pole point exceeds the corresponding first basic addressing data range. Perform the first correction welding for cases that exceed the range. This effectively identifies and judges anomalies caused by camera misalignment, data transmission errors, etc., and improves data accuracy.

[0079] Based on the normal data transmission and addressing station camera judgment, further, based on the distance difference between a pair of marker points at the addressing station and the welding station, anomalies caused by the welding station camera's grasp deviation, camera axial movement deviation, etc. are effectively identified, improving the overall data accuracy; further, based on the conversion data of one of the marker points (first marker point conversion data) and the second marker point data, conversion judgment is added at the point, and mistaken error prevention is performed again, further improving data accuracy;

[0080] 3. Based on the initial micro-level local data analysis and the emphasis on data source error prevention to improve overall data accuracy, obtain the first basic welding range for the entire battery product to be welded, and compare it with the converted second pole point data to increase the later macro-level overall data analysis and the emphasis on data conversion error prevention, thereby further improving welding accuracy.

[0081] 4. During the correction welding process, the probability of misjudgment is reduced by combining pre-welding judgment; furthermore, a comparison judgment is made again before pre-welding to ensure the error prevention effect while comprehensively simplifying the welding process. Attached Figure Description

[0082] Figure 1 This is a schematic flowchart of the welding method based on image recognition described in one embodiment of this application;

[0083] Figure 2 This is a schematic diagram of the first correction welding process described in one embodiment of this application;

[0084] Figure 3 This is a schematic diagram of the first correction welding process described in one embodiment of this application;

[0085] Figure 4 This is a schematic flowchart of the welding method based on image recognition described in one embodiment of this application. Detailed Implementation

[0086] The following is in conjunction with the appendix Figure 1 -Appendix Figure 4 This application will be described in further detail below.

[0087] This application discloses a welding method based on image recognition, which can be executed by a processor of a computer device. In implementation, the computer device can be any suitable device with data processing capabilities. It is understood that in industrial production, the computer device can refer to a host computer, such as a server, laptop, tablet, desktop computer, or smartphone, or a slave computer, such as an industrial control computer or a programmable logic controller (PLC). In some embodiments, the computer device may include a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the program to implement the coordinate transformation method.

[0088] The implementation of the image recognition-based welding method will be described in further detail below with reference to Examples 1-4.

[0089] Example 1

[0090] A welding method based on image recognition, referring to Figure 1 This includes the following steps:

[0091] S101. Obtain the addressing data of the battery product to be welded in the addressing station coordinate system. The addressing data includes the first pole point data of each pole and the first marker point data of the two marker points.

[0092] In some embodiments, the battery product to be welded may include a battery module comprising multiple battery cells, each battery cell having a positive terminal and a negative terminal, collectively referred to as terminals in this embodiment. For example, a battery cell may be a battery cell. In some embodiments, the battery product to be welded may include a battery pack composed of multiple battery modules.

[0093] In practice, those skilled in the art can set marking points on the battery product to be welded according to the actual situation. The location of the marking points can be any point on the battery product to be welded except for the terminal post. The number of marking points is set according to the actual situation. In one embodiment, the marking points are set at intervals along the circumference of the battery product to be welded, and multiple marking points tend to be evenly spaced under the condition of satisfying the setting. In this embodiment, the surface to be welded of the battery product to be welded is rectangular. Two marking points are set in the battery product to be welded, which are defined as the main feature point and the auxiliary feature point, respectively. The two marking points are located at opposite corners of the battery product to be welded.

[0094] The battery products to be welded first flow into the addressing station, which is set before the welding station. The addressing station is used to carry the battery products to be welded and is equipped with a first addressing mechanism. The first addressing mechanism may include a camera A driven by a servo mechanism.

[0095] At the addressing station, all target points (including all pole points and all marker points) are located by servo positioning (the servo mechanism drives the camera A to position). Then, each target point is photographed one by one. During implementation, for each target point, the first addressing mechanism moves the camera above the target point by servo positioning, so that the target point is as close to the center of the camera as possible, thereby increasing the field of view of the camera and improving the positioning effect. Then, a photo of the corresponding target point is taken.

[0096] Based on the photo, obtain the coordinate data of each polar point and marker point in the pixel coordinate system. The pixel coordinate system is a uov two-dimensional rectangular coordinate system. The origin o can be set at the upper left corner of the photo, and the u-axis and v-axis are parallel to the two sides of the corresponding image plane of the image.

[0097] The camera A's shooting position is determined by the movement path of the servo mechanism. The coordinate data of the pole point and the marker point in the pixel coordinate system are unified into the addressing station coordinate system in conjunction with the shooting position of the camera A to obtain the addressing data. The addressing data includes the first pole point data of each pole and the first marker point data of the two marker points. The addressing station coordinate system can be a unified coordinate system constructed for each physical position in the addressing station.

[0098] The obtained addressing data (the first pole point data of each pole and the first marker point data of the two marker points) is stored in the database of the host computer for use by the welding station upon request.

[0099] S102. Obtain the data of the second marker point in the welding station coordinate system of the two marker points of the battery product to be welded.

[0100] During implementation, the battery product to be welded flows from the addressing station into the welding station. The welding station is used to carry the battery product to be welded and is equipped with a second addressing mechanism. The second addressing mechanism may include a camera B driven by a servo mechanism.

[0101] At the welding station, two marker points are positioned by servo (the servo mechanism drives the camera B to position), and then the two marker points are photographed one by one to obtain the corresponding photos of the marker points.

[0102] Based on the photo, obtain the coordinate data of each marker point in the pixel coordinate system. Determine the shooting position of camera B through the movement path of the servo mechanism. Combine the shooting position of camera B with the coordinate data of the marker point in the pixel coordinate system to unify it into the welding station coordinate system. That is, obtain the actual position of the marker point in the welding station coordinate system (second marker point data). The second marker point data can be used to perform initial positioning of the entire battery product to be welded in the welding station coordinate system.

[0103] S103. Call the addressing data of the corresponding battery product to be welded to the welding station. The addressing data to be called includes the first pole point data of each pole and the first marker point data of the two marker points, thus obtaining the first pole point data and the first marker point data to be called.

[0104] In this embodiment, the call is to obtain addressing data at the welding station, even if the addressing data serves the welding operation at the welding station;

[0105] S104. Based on the first marker point data (coordinate data of the marker point on the addressing station) and the second marker point data on the welding station, determine the actual transformation relationship between the addressing station coordinate system and the welding station coordinate system; here, the points on the addressing station coordinate system can be correspondingly transferred to the welding station coordinate system through the actual transformation relationship.

[0106] In one embodiment, determining the actual transformation relationship includes:

[0107] The first marker point data is transferred to the welding station coordinate system based on the theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first reference position coordinates of the marker point at the welding station. The theoretical transformation relationship is the inherent transformation relationship between the addressing station coordinate system and the welding station coordinate system, that is, the transformation relationship between the two when setting the addressing station coordinate system and the welding station coordinate system.

[0108] Obtain the angle between the line connecting the two points corresponding to the first reference position coordinates and the line connecting the two points corresponding to the second marker data to get the offset angle; here, the line connecting the two points corresponding to the first reference position coordinates is defined as the first line, and the line connecting the two points corresponding to the second marker data is defined as the second line. Both the first line and the second line are set based on the welding station coordinate system, and the angle between their intersection is the offset angle.

[0109] After rotating and offsetting the line connecting the points corresponding to the first reference position coordinates by an offset angle, the second reference position coordinates are obtained; at this time, the rotated first line is parallel to the second line and their extension directions overlap.

[0110] The offset between the second reference position coordinates and the second marker point data is obtained, where the offset angle and the offset amount constitute the actual transformation relationship. Here, the second reference coordinates are moved based on the offset and overlapped with the corresponding second marker point data. The first reference position coordinates are corrected to the second marker point data through the actual transformation relationship.

[0111] After obtaining the actual conversion relationship, based on the actual conversion relationship and the first pole point data of each pole of the battery product to be welded, the second pole point data of the pole of the battery product to be welded in the welding station coordinate system is determined, thereby obtaining the set of pole coordinates of multiple poles of the battery product to be welded in the welding station coordinate system, effectively reducing conversion deviation.

[0112] S105. Obtain the first basic addressing data range of each terminal point of the battery product to be welded in the welding coordinate system, wherein the basic addressing data range of each terminal point is set based on product design and process requirements.

[0113] In one embodiment, when designing the battery product to be welded, basic addressing data of each terminal point in the welding coordinate system is obtained. Further, based on process requirements and considering the acceptable error range (which is set by those skilled in the art), the first basic addressing data range for each terminal point is determined by adjusting the basic addressing data according to this acceptable error range. Specifically, the acceptable error range can be determined using the following two methods:

[0114] Method 1: When there is a corresponding cell stacking process standard for the battery product to be welded, the cell stacking process standard shall be used to determine the process. For example, if the standard limits the deviation of cell stacking to no more than 5mm, then 5mm shall be taken as the acceptable error range.

[0115] Method 2: When there is no corresponding cell stacking process standard for the battery product to be welded, since the automatic stacking of the equipment is uniform and consistent, multiple sets of normal module addressing points are used to compare the differences simultaneously, and the maximum and minimum differences are compared to determine the acceptable error range; for example, as shown in Table 1 below, the deviation of 4 sets of normal module addressing points in the welding coordinate system, where the module is composed of the battery product to be welded.

[0116] Table 1

[0117]

[0118] According to Table 1, the acceptable error range along the X-axis can be set to 3.006 mm, and the acceptable error range along the Y-axis can be set to 2.7345 mm.

[0119] S106. Among all pole points, for each pole point, determine whether the data of the second pole point exceeds the range of the corresponding first basic addressing data.

[0120] If (at least one pole point does not meet the above judgment requirements, that is, the data of the second pole point exceeds the corresponding first basic addressing data range), perform the first correction welding;

[0121] If not, proceed to step S107;

[0122] In one embodiment, reference is made to Figure 2 The first correction welding includes:

[0123] Weak laser is used for pre-welding based on the second pole point data; here, weak laser refers to laser with low power density or low energy radiation. In the welding field, it can be 10%-30% of the power used by the laser. For example, for a laser with a power of 4000W, the power of its weak laser is set to 400W-1200W. In this embodiment, it can be set to 400W.

[0124] Based on the pre-welding results, determine whether the deviation of each welding point from the corresponding pole center point exceeds the set off-center welding process standard. If yes, activate the self-welding mode; otherwise, weld the product to be welded based on the second pole point data. Here, calculate the distance between the center point of each pole and the welding point to obtain the welding deviation, and determine whether the welding deviation exceeds the set off-center welding process standard.

[0125] In another embodiment, reference Figure 3 Before pre-welding, the first correction welding also includes: determining whether the data of the second pole point exceeds the corresponding second basic addressing data range. If so, the self-photographing and self-welding mode is started for welding, that is, at the welding station, each pole point is addressed, photographed, and welded; if not, pre-welding is performed. The second basic addressing data range is generated based on the first basic addressing data range. Specifically, for each pole point, the second basic addressing data range covers the first basic addressing data range. That is, relative to the first basic addressing data range, an adaptive expansion adjustment is made based on the actual situation. In one embodiment, the second basic addressing data range is proportionally expanded by 2 times relative to the first basic addressing data range. For example, when the standard limits the deviation of cell stacking to no more than 5mm, then 5mm is taken as the acceptable error range when the first basic addressing data range is determined, and 10mm is taken as the acceptable error range when the second basic addressing data range is determined.

[0126] S107. Obtain the first distance between the two marker points based on the data of the first marker point called from the two marker points;

[0127] The second distance between the two marker points is obtained based on the second marker point data of the two marker points;

[0128] By comparing the first distance length and the second distance length, the distance difference (absolute value) is obtained. It is then determined whether the distance difference exceeds a set first distance threshold. In one embodiment, the first distance threshold can be set as follows: Define two marker points M1 and M2. Using the line connecting the two marker points M1 and M2 (product design standard) as the hypotenuse (length C), construct a right triangle. The first right-angled side is formed corresponding to marker point M1, and the second right-angled side is formed corresponding to marker point M2. Determine the maximum deviation of each marker point according to the maximum deviation process standard. Set the maximum deviation of marker point M1 as x and the maximum deviation of marker point M2 as y. Extend the first right-angled side by x to obtain an extended marker point M1'. Extend the second right-angled side by y to obtain an extended marker point M2'. Connect the two extended marker points M1' and M2' to determine the length C'. Therefore, C'-C is set as the first distance threshold. In this embodiment, the first distance threshold can be 1 mm.

[0129] If yes, perform a second correction weld; otherwise, weld the product to be welded based on the second pole point data.

[0130] In one embodiment, the second correction weld includes:

[0131] Pre-welding is performed using a weak laser based on the data from the second pole point;

[0132] Based on the pre-welding results, determine whether the deviation of each welding point from the center point of the corresponding electrode exceeds the set welding process standard. If yes, start the self-welding mode. If no, it means that there is indeed a large welding error in the assembly or stacking process of the battery product, which exceeds the first distance threshold set according to normal conditions. Weld the product to be welded based on the second electrode point data.

[0133] In another embodiment, before performing pre-welding, the second correction welding further includes: determining whether the distance difference exceeds a set second distance threshold; if so, initiating a self-welding mode for welding; if not, performing a pre-welding operation. The second distance threshold is generated based on the first distance threshold and is adaptively increased based on the actual situation. In one embodiment, the second distance threshold is proportionally increased by 1.4-1.8 times relative to the first distance threshold. In this embodiment, it is proportionally increased by 1.6 times, that is, the second distance threshold is set to 1.6 mm.

[0134] The implementation effect of Embodiment 1 of this application is as follows: Based on functionality, the welding operation is divided into two parts to be completed at the addressing station and the welding station respectively. At the addressing station, the coordinate information of each pole point and two marker points is obtained by taking pictures. Then, the coordinate information of the two marker points is obtained by taking pictures at the welding station. During the error prevention process, the range of the first basic addressing data is recorded in advance, and it is determined whether the data of the second pole point exceeds the corresponding range of the first basic addressing data. For cases that exceed the range, the first correction welding is performed. This effectively identifies and judges anomalies caused by camera misalignment at the addressing station, data transmission errors, etc., and improves data accuracy. Furthermore, by comparing the first distance length and the second distance length corresponding to the marker points, the anomalies caused by camera misalignment at the addressing station and the welding station, camera axial movement deviation, etc. are effectively identified. On the basis of normal data transmission and the judgment of the addressing station camera, the overall data accuracy is improved.

[0135] Example 2:

[0136] Embodiment 2 of this application discloses a preparation method for a welding method based on image recognition, referring to... Figure 4 Steps S201-S207 correspond one-to-one with steps S101-S107 in Embodiment 1. The difference is that in step S207, after determining that the distance difference does not exceed the set first distance threshold, refer to Figure 4 ,include:

[0137] S208. Obtain the first basic welding range of the battery product to be welded, wherein the first basic welding range is set based on the product design and process requirements.

[0138] In one embodiment, the method for determining the first basic welding range is to obtain the basic welding data of the outer ring pole points on the welding surface of the product when designing the battery product to be welded, and further determine the first basic welding range of the battery product by comprehensively considering the acceptable error range based on the process requirements.

[0139] In another embodiment, the method for determining the first basic welding range is to use the positions of the four corners of the teaching module, and the range enclosed by the four points constitutes the welding range of the welding module.

[0140] S209. For the battery product to be welded, determine whether the data of the second electrode point exceeds the corresponding first basic welding range.

[0141] If yes, perform the third correction welding; otherwise, weld the product to be welded based on the second pole point data.

[0142] In one embodiment, the third correction welding includes: pre-welding with a weak laser based on the second pole point data; determining whether the deviation of each welding point from the center point of the corresponding pole exceeds the set off-center welding process standard based on the pre-welding result; if so, starting the self-welding mode; if not, welding the product to be welded based on the second pole point data.

[0143] In another embodiment, the third correction welding, before pre-welding, further includes: determining whether the data of the second pole point exceeds the corresponding second basic welding range; if so, starting the self-photographing and self-welding mode for welding, that is, at the welding station, addressing and photographing each pole point for welding; if not, performing pre-welding, wherein the second basic welding range is generated based on the first basic welding range, and the second basic welding range covers the first basic welding range, that is, relative to the first basic welding range, it is adaptively expanded and adjusted based on the actual situation. In one embodiment, the second basic welding range is proportionally expanded by 2 times relative to the first basic welding range.

[0144] The implementation effect of Embodiment 2 of this application is as follows: Based on the early stage of micro-local and data source error prevention to improve the overall data accuracy, at the welding station, the first basic welding range of the entire battery product to be welded is obtained and compared with the converted second pole point data, which increases the later macro-overall and data conversion error prevention, and further improves the welding accuracy.

[0145] Example 3:

[0146] Embodiment 3 of this application discloses a preparation method for a welding method based on image recognition. Steps S301-S307 correspond one-to-one with steps S101-S107 of Embodiment 1. The difference is that in step S307, after determining that the distance difference does not exceed the set first distance threshold, refer to Figure 4 ,include:

[0147] S308. Obtain the third distance length between the first and second marker data of one of the marker points;

[0148] S309. Determine whether the length of the third distance exceeds the set first length threshold. In one embodiment, the first length threshold may be 0.5 mm.

[0149] If yes, perform the fourth correction welding; if not, weld the product to be welded based on the second pole point data.

[0150] In one embodiment, the fourth correction welding includes: pre-welding with a weak laser based on the second pole point data; determining whether the deviation of each welding point from the center point of the corresponding pole exceeds the set off-center welding process standard based on the pre-welding result; if so, starting the self-welding mode; if not, welding the product to be welded based on the second pole point data.

[0151] In another embodiment, the fourth correction welding further includes determining whether the distance difference exceeds the set second length threshold before pre-welding. If yes, the self-welding mode is started for welding; if no, pre-welding is performed. The second length threshold is generated based on the first length threshold and is adaptively increased based on the actual situation. In one embodiment, the second distance threshold is proportionally increased by 2 times relative to the first distance threshold.

[0152] Example 4

[0153] Embodiment 4 of this application discloses a preparation method of a welding method based on image recognition. Steps S401-S407 correspond one-to-one with steps S101-S107 of Embodiment 1. The difference is that in step S407, after determining that the distance difference does not exceed the set first distance threshold, steps S408-S409 are executed.

[0154] S408. Obtain the third distance length between the first and second marker data of one of the marker points;

[0155] S409. Determine whether the length of the third distance exceeds the set first length threshold.

[0156] If yes, perform the fourth correction welding; if no, proceed to step S410.

[0157] S410, Obtain the first basic welding range for the battery product to be welded;

[0158] S411. For the battery product to be welded, determine whether the data of the second electrode point exceeds the corresponding first basic welding range.

[0159] If yes, perform a third correction weld; otherwise, weld the product to be welded based on the second pole point data.

[0160] The implementation of the image recognition-based welding system will be described in further detail below with reference to Embodiment 5. Here, the implementation of the image recognition-based welding system is an exemplary structure of the software module.

[0161] Example 5

[0162] Embodiment 5 of this application discloses a welding system based on image recognition, including: a first acquisition module, a calling module, a conversion module, a second acquisition module, a first judgment module, a first welding module, a third acquisition module, a second judgment module, and a second welding module;

[0163] The first acquisition module is used to acquire the addressing data of the battery product to be welded in the addressing station coordinate system and the second marker point data of the two marker points of the battery product to be welded in the welding station coordinate system. The addressing data includes the first pole point data of each pole of the battery product to be welded and the first marker point data of the two marker points.

[0164] The calling module is used to call the addressing data to the welding station to obtain the first pole point data and the first marker point data.

[0165] The conversion module is used to determine the actual conversion relationship between the addressing station coordinate system and the welding station coordinate system based on the second marker point data and the called first marker point data, and to convert the called first pole point data into second pole point data in the welding station coordinate system based on the actual conversion relationship.

[0166] The second acquisition module is used to acquire the first basic addressing data range of each pole point in the welding coordinate system;

[0167] The judgment module is used to determine whether the second pole point data exceeds the corresponding first basic addressing data range, wherein the first basic addressing data range is generated based on product design and process requirements;

[0168] The first welding module is used to perform a first correction welding when the second pole point data exceeds the corresponding first basic addressing data range;

[0169] The third acquisition module is used to obtain the first distance between two marker points based on the first marker point data called when the second pole point data does not exceed the corresponding first basic addressing data range, and to obtain the second distance between two marker points based on the second marker point data, and to compare the first distance length and the second distance length to obtain the distance difference.

[0170] The second judgment module is used to determine whether the distance difference exceeds a preset first distance threshold.

[0171] The second welding module is used to perform a second correction welding when the distance difference exceeds a preset first distance threshold; and to weld the product to be welded based on the second pole point data when the distance difference does not exceed the preset first distance threshold.

[0172] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, all equivalent changes made to the structure, shape, and principle of this application should be covered within the scope of protection of this application.

Claims

1. A welding method based on image recognition, characterized in that, Includes the following steps: Obtain the addressing data of the battery product to be welded in the addressing station coordinate system, wherein the addressing data includes the first pole point data of each pole of the battery product to be welded and the first marker point data of two marker points; Obtain the data of the second marker point of the two marker points of the battery product to be welded in the coordinate system of the welding station; The addressing data is called to the welding station to obtain the first pole point data and the first marker point data. Based on the second marker point data and the first marker point data called, the actual transformation relationship between the addressing station coordinate system and the welding station coordinate system is determined, and the first polar point data called is converted into the second polar point data in the welding station coordinate system based on the actual transformation relationship. Obtain the first basic addressing data range of each pole point in the welding coordinate system, and determine whether the data of the second pole point exceeds the corresponding first basic addressing data range. The first basic addressing data range is generated based on product design and process requirements. If so, perform the first corrective welding; If not, obtain the first distance between the two markers based on the first marker data, obtain the second distance between the two markers based on the second marker data, compare the first distance and the second distance to obtain the absolute value of the distance difference, and determine whether the absolute value of the distance difference exceeds a preset first distance threshold. If so, perform a second correction weld; If not, weld the product to be welded based on the second pole point data; Both the first correction welding and the second correction welding include: Pre-welding is performed using a weak laser based on the data from the second pole point; Based on the pre-welding results, determine whether the deviation of each welding point from the center point of the corresponding pole exceeds the set welding deviation threshold. If so, welding is performed at the welding station using a self-service welding method; If not, weld the product to be welded based on the second pole point data; The self-photographing and self-welding process involves taking photos of each pole point at the welding station for addressing and welding.

2. The welding method according to claim 1, characterized in that, After determining that the absolute value of the distance difference does not exceed the first distance threshold, the following steps are included: Obtain the first basic welding range of the battery product to be welded in the welding coordinate system, wherein the first basic welding range is generated based on product design and process requirements; Determine whether the data of the second pole point exceeds the welding range of the first base; If so, perform a third correction weld; If not, welding is performed on the product to be welded based on the second pole point data.

3. The welding method according to claim 2, characterized in that, After determining that the absolute value of the distance difference does not exceed a preset first distance threshold, before obtaining the first basic welding range of the battery product to be welded in the welding coordinate system, the process further includes: Obtain the first marker point data of one of the marker points, and transform it to the welding coordinate system based on the theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first marker point transformation data; Obtain the third distance length between the first and second marker data of the marker point; Determine whether the length of the third distance exceeds a preset first length threshold; If so, perform the fourth correction welding; If not, obtain the first basic welding range for the battery product to be welded.

4. The welding method according to claim 3, characterized in that, Both the third and fourth correction welding include: Pre-welding is performed using a weak laser based on the data from the second pole point; Based on the pre-welding results, determine whether the deviation of each welding point from the center point of the corresponding pole exceeds the set welding deviation threshold. If so, welding is performed at the welding station using a self-service welding method; If not, welding is performed on the product to be welded based on the second pole point data.

5. The welding method according to claim 4, characterized in that, The first correction welding, prior to the pre-welding using a weak laser based on the second pole point data, further includes: Determine whether the first pole point data being called exceeds the corresponding second basic addressing data range, wherein the second basic addressing data range is generated based on the first basic addressing data range; If so, welding is performed at the welding station using a self-service welding method; If not, pre-welding is performed using a weak laser based on the data from the second pole point.

6. The welding method according to claim 4, characterized in that, The second correction welding, prior to the pre-welding using a weak laser based on the second pole point data, further includes: Determine whether the absolute value of the distance difference exceeds a preset second distance threshold, wherein the second distance threshold is generated based on the first distance threshold; If so, welding is performed at the welding station using a self-service welding method; If not, pre-welding is performed using a weak laser based on the data from the second pole point.

7. The welding method according to claim 4, characterized in that, The third correction welding, prior to the pre-welding using a weak laser based on the second pole point data, also includes: Determine whether the data of the second pole point exceeds the corresponding second basic welding range, wherein the second basic welding range is generated based on the first basic welding range; If so, welding is performed at the welding station using a self-service welding method; If not, pre-welding is performed using a weak laser based on the data from the second pole point.

8. The welding method according to claim 4, characterized in that, The fourth correction welding, prior to the pre-welding using a weak laser based on the second pole point data, further includes: Determine whether the length of the third distance exceeds a preset second length threshold, wherein the second length threshold is generated based on the first length threshold; If so, welding is performed at the welding station using a self-service welding method; If not, pre-welding is performed using a weak laser based on the data from the second pole point.

9. The welding method according to claim 1, characterized in that, Determining the actual transformation relationship between the addressing station coordinate system and the welding station coordinate system based on the second marker point data and the invoked first marker point data includes: The first marker point data is transferred to the welding station coordinate system based on the theoretical transformation relationship between the addressing station coordinate system and the welding station coordinate system to obtain the first reference position coordinates of the marker point in the welding station. Obtain the angle between the line connecting the points corresponding to the first reference position coordinates and the line connecting the points corresponding to the second marker data, and obtain the offset angle; After rotating and offsetting the line connecting the points corresponding to the first reference position coordinates by the offset angle, the second reference position coordinates are obtained. Obtain the offset between the second reference position coordinates and the second marker point data, where the offset angle and offset amount constitute the actual conversion relationship.

10. A welding system based on image recognition, characterized in that, include: The first acquisition module is used to acquire the addressing data of the battery product to be welded in the addressing station coordinate system and the second marker point data of the two marker points of the battery product to be welded in the welding station coordinate system. The addressing data includes the first pole point data of each pole of the battery product to be welded and the first marker point data of the two marker points. The calling module is used to call the addressing data to the welding station to obtain the first pole point data and the first marker point data to be called. The conversion module is used to determine the actual conversion relationship between the addressing station coordinate system and the welding station coordinate system based on the second marker point data and the first marker point data called, and to convert the first polar point data called into the second polar point data in the welding station coordinate system based on the actual conversion relationship. The second acquisition module is used to acquire the first basic addressing data range of each pole point in the welding coordinate system; The first judgment module is used to determine whether the second pole point data exceeds the corresponding first basic addressing data range, wherein the first basic addressing data range is generated based on product design and process requirements; The first welding module is used to perform a first correction welding when the second pole point data exceeds the corresponding first basic addressing data range; The third acquisition module is used to obtain the first distance between two marker points based on the first marker point data called when the second pole point data does not exceed the corresponding first basic addressing data range, and to obtain the second distance between two marker points based on the second marker point data, and to compare the first distance length and the second distance length to obtain the absolute value of the distance difference. The second judgment module is used to determine whether the absolute value of the distance difference exceeds a preset first distance threshold. The second welding module is used to perform second correction welding when the absolute value of the distance difference exceeds a preset first distance threshold; and to weld the product to be welded based on the second pole point data when the absolute value of the distance difference does not exceed the preset first distance threshold. Both the first correction welding and the second correction welding include: Pre-welding is performed using a weak laser based on the data from the second pole point; Based on the pre-welding results, determine whether the deviation of each welding point from the center point of the corresponding pole exceeds the set welding deviation threshold. If so, welding is performed at the welding station using a self-service welding method; If not, weld the product to be welded based on the second pole point data; The self-photographing and self-welding process involves taking photos of each pole point at the welding station for addressing and welding.

Images