Welding method for automatically searching battery cell pole based on CAD (computer-aided design) import

Abstract

The invention discloses a CAD (computer-aided design) import-based welding method for automatically searching a battery cell pole column, which comprises the following steps of: acquiring the battery cell pole column and a MARK point according to a CAD drawing of a product, calculating to obtain a relative theoretical position of the battery cell pole column and the MARK point, and finishing preparation work; obtaining actual position coordinates of the MARK point, and obtaining theoretical shooting coordinates of the battery cell pole according to the actual position coordinates and the relative theoretical position; operating the robot to shoot an image at the theoretical shooting coordinate, and acquiring the offset of the actual position coordinate of the battery cell pole relative to the theoretical position coordinate by using a multi-template matching method; judging whether the offset is within a preset threshold range or not; if yes, adding the theoretical shooting coordinate of the battery cell pole and the offset, and marking as a qualified welding position; and if not, marking the theoretical shooting coordinate of the battery cell pole as a failure welding position. The invention relates to the technical field of automatic welding, and solves the technical problems of insufficient precision and low efficiency in a traditional welding method.

Description

Technical Field

[0001] The present invention belongs to the field of automatic welding, involves CAD data processing technology, and specifically is an automatic method for finding the welding position of battery cell poles based on CAD import. Background Art

[0002] The battery cell pole is a core component of energy storage devices such as lithium-ion batteries, undertaking the key functions of current input and output. Its welding quality directly affects the internal resistance, thermal stability, and long-term reliability of the battery. In the rapidly developing battery industry, the production of battery cells tends to be automated and precise, and the pole welding needs to meet the requirements of high precision, high efficiency, and strong consistency.

[0003] However, traditional methods for positioning the welding of battery cell poles mainly rely on manual operation or simple mechanical positioning. In terms of positioning accuracy, traditional methods have problems of insufficient accuracy and poor repeatability. Since manual operation mainly relies on experience and visual judgment of the pole position, it is difficult to achieve high-precision positioning; while simple mechanical positioning is limited by its own structure and technology and also cannot meet the requirements of precise positioning, resulting in excessive deviation of the welding position, affecting the internal current conduction of the battery cell, reducing battery performance, and may also cause problems such as virtual soldering and de-soldering during long-term use, seriously damaging the reliability of the battery cell and shortening its service life. In terms of production efficiency, the manual adjustment and calibration process is also time-consuming and laborious. In the current situation of large-scale production of battery cell manufacturing, this low-efficiency operation process hinders the improvement of production capacity and cannot meet the rapid and efficient production requirements. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; for this purpose, the present invention proposes an automatic method for finding the welding position of battery cell poles based on CAD import, which is used to solve the problems of insufficient accuracy, low production efficiency, and poor repeatability existing in the existing methods for positioning the welding of battery cell poles.

[0005] To achieve the above object, the present invention provides an automatic method for finding the welding position of battery cell poles based on CAD import, including:

[0006] S1, calculating the relative theoretical positions of the battery cell poles and several MARK points according to the CAD drawing of the product, and completing the preparatory work according to the CAD drawing of the product;

[0007] S2, obtaining the actual position coordinates of several MARK points, and obtaining the theoretical shooting coordinates of the battery cell poles according to the actual position coordinates and the relative theoretical positions;

[0008] S3, operating the robot to take several images at the theoretical shooting coordinates, and using the multi-template matching method to obtain the offset of the actual position coordinates of the battery cell poles relative to the theoretical position coordinates according to the several images;

[0009] S4. Determine whether the offset is within a preset threshold range. If so, add the theoretical shooting coordinates of the battery cell pole to the offset and mark it as a qualified welding position. If not, mark the theoretical shooting coordinates of the battery cell pole as a failed welding position.

[0010] Through the above technical steps, the present invention realizes the automation and precision of the welding of the battery cell pole, solves the problems of insufficient precision and low efficiency in traditional welding methods, and significantly improves the welding quality and production efficiency.

[0011] Further, the calculating the relative theoretical positions of the battery cell pole and several MARK points according to the CAD drawing of the product includes:

[0012] S11-1. Convert the CAD drawing of the product into a dxf file, and perform lightweight processing on the dxf file according to the attribute data of the battery cell pole and the MARK points to obtain a lightweight file.

[0013] S11-2. Use an object detection algorithm to automatically detect the battery cell pole and the MARK points in the lightweight file to obtain the theoretical position coordinates of the battery cell pole and the coordinates of several MARK points.

[0014] S11-3. According to the theoretical position coordinates of the battery cell pole and the coordinates of several MARK points, calculate the relative distance between the battery cell pole and each MARK point to obtain the relative theoretical positions of the battery cell pole and several MARK points.

[0015] Further, the preparation work further includes:

[0016] S12-1. Establish a robot tool coordinate system and make the robot tool coordinate system parallel to the XY of the copper nozzle three-axis.

[0017] S12-2. Use the nine-point calibration method to unify the camera center and the robot tool coordinate system, and calculate the pixel equivalent using a calibration board. The calibration board is placed on the product to be welded for calibration tasks.

[0018] S12-3. Use a laser range finder to obtain the relative height between the robot and the welding surface, and set the spatial position parameters of the copper nozzle three-axis according to the CAD drawing and the relative height.

[0019] Further, the calculating the pixel equivalent using the calibration board includes:

[0020] Divide the calibration board into N×M rectangular areas with a side length of L to obtain several rectangular frames.

[0021] Take pictures of the calibration board using a camera, and use a machine learning algorithm to obtain the upper left corner coordinates (X1, Y1) and the lower right corner coordinates (X2, Y2) of a single rectangular frame on the captured image;

[0022] Subtract the horizontal and vertical coordinates of the lower right corner coordinates (X2, Y2) from the horizontal and vertical coordinates of the upper left corner coordinates (X1, Y1) respectively to obtain the length A1 and width B1 of the rectangular frame;

[0023] Calculate the pixel equivalent E according to the formula E = (A / A1 + B / B1) / 2; where A and B respectively represent the length and width of the calibration board, and A = N×L, B = M×L.

[0024] Further, the obtaining of the relative height between the robot and the welding surface by using the laser range sensor includes:

[0025] Adjust the galvanometer center and the ranging center of the robot, calculate the relative position coordinates of the galvanometer center and the ranging center of the robot with respect to the camera center when they are in fixed positions, and obtain the galvanometer-vision position and the ranging-vision position;

[0026] Fix the galvanometer-vision position and the ranging-vision position, and adjust the robot to emit light at a preset rated power on the metal plate of the product to be welded;

[0027] Use the laser range sensor to measure the distance to the calibration board to obtain the relative height DZ between the robot and the welding surface 1 .

[0028] Further, the setting of the spatial position parameters of the copper nozzle on the three axes according to the CAD drawing and the relative height includes:

[0029] Operate the robot to move to several MAKD points in the lightweight file, and record the coordinates of the moved robot to obtain the photographing positions of several MARK points;

[0030] Align the center of the copper nozzle with the centers of several MARK points in sequence, and record and fix the positions of the X and Y axes of the copper nozzle;

[0031] Adjust the Z axis of the copper nozzle to completely press the bus bar of the battery cell pole column, and trigger the laser range sensor to obtain the current Z axis data DZ;

[0032] Judge whether the difference between DZ and the relative height DZ 1 is less than the preset height deviation; if yes, record and fix the Z axis position of the copper nozzle on the three axes; if not, jump to S2-1 to recalibrate the galvanometer-vision position and the ranging-vision position.

[0033] Further, the obtaining method of the preset height deviation includes:

[0034] Operate the robot to move to the welding position of the battery cell pole, adjust the laser ranging spot to align with the pole surface, measure and record the current Z-axis data DZ 2 ;

[0035] According to the formula ΔZ = DZ 2 - DZ 1 Calculate the preset height deviation and input ΔZ into the robot control parameters.

[0036] Further, obtaining the actual position coordinates of several MARK points includes:

[0037] The robot moves to the photographing position of the several MARK points to take pictures and obtains several MARK point images;

[0038] Use the template matching method to detect the point coordinates in several MARK point images and obtain the actual positions of several MARK points where i represents the index of the MARK point.

[0039] Further, obtaining the theoretical shooting coordinates of the battery cell pole according to the actual position coordinates and the relative theoretical position includes:

[0040] According to the formula Calculate the theoretical shooting coordinates (X j , Y j ) of the battery cell pole; where represents the theoretical position coordinates of the battery cell pole.

[0041] Further, using the multi-template matching method to obtain the offset of the actual position coordinates of the battery cell pole relative to the theoretical position coordinates according to several images includes:

[0042] Collect several photos of the pole area to obtain a module library containing several templates;

[0043] Use the multi-template matching algorithm to calculate the similarity between several templates and a single image, and screen to obtain the template coordinates (X, Y) with the highest similarity;

[0044] Multiply the template coordinates (X, Y) with the highest similarity by the pixel equivalent to obtain the battery cell pole offset (EX, EY).

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] Combined with the vision system through CAD drawing import, precise positioning of the cell pole is achieved; the automated process reduces manual intervention and significantly improves the welding efficiency; and for the same product, only one preparation work is required, and subsequent tasks can be completed automatically; accurate distance measurement and positioning ensure the consistency of welding quality; the method is applicable to the welding of cell poles of different models and specifications and has strong versatility. Brief Description of the Drawings

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a schematic flow chart of an automated method for finding and welding cell poles based on CAD import provided by the present invention. Detailed Embodiments

[0049] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0050] Please refer to Figure 1 , the embodiments of the present invention provide an automated method for finding and welding cell poles based on CAD import, including:

[0051] S1. Obtain the cell pole and several MARK points according to the CAD drawing of the product, calculate the relative theoretical positions of the cell pole and the several MARK points, and complete the preparation work according to the CAD drawing of the product;

[0052] S2. Obtain the actual position coordinates of the several MARK points, and obtain the theoretical shooting coordinates of the cell pole according to the actual position coordinates and the relative theoretical positions;

[0053] S3. Operate the robot to take several images at the theoretical shooting coordinates, and use the multi-template matching method to obtain the offset of the actual position coordinates of the cell pole relative to the theoretical position coordinates according to the several images;

[0054] S4. Determine whether the offset is within the preset threshold range; if yes, add the theoretical shooting coordinates of the cell pole and the offset, and mark it as a qualified welding position; if not, mark the theoretical shooting coordinates of the cell pole as a failed welding position.

[0055] Specifically, the present invention mainly includes two parts: preparatory work and formal work. The preparatory work is used to construct an accurate coordinate system and obtain key parameters, providing a basic support for precise positioning and welding in the subsequent formal work; while the formal work is based on the data obtained from the preparatory work and the established coordinate system to achieve the automatic positioning and welding of the cell pole columns.

[0056] In one implementation, in step S1, the preparatory work is completed according to the CAD drawing of the product, including:

[0057] The preparatory work is an important basis for the automatic method of finding the cell pole column welding based on CAD import, and its operation process is as follows:

[0058] 1. Process the CAD drawing;

[0059] Format conversion and lightweight processing: Convert the CAD drawing of the product into a dxf file, and perform lightweight processing on the dxf file according to the attribute data of the cell pole column and MARK points, only retaining the information of the pole column and MARK points to generate a lightweight file, so as to accelerate the inference speed of subsequent data processing, improve the overall work efficiency, and reduce the burden of software data processing;

[0060] Detection and calculation: Use the object detection algorithm to automatically detect the cell pole column and MARK points in the lightweight file to obtain the theoretical position coordinates of the cell pole column and the coordinates of several MARK points;

[0061] Then, calculate the relative distance between the cell pole column and each MARK point according to these coordinates, so as to obtain the relative theoretical position of the cell pole column and several MARK points, providing key data support for subsequent positioning.

[0062] 2. Construct a coordinate system;

[0063] First, establish a robot tool coordinate system and make it parallel to the XY of the copper nozzle three-axis, ensuring that during the operation of the robot, the movement of the copper nozzle has a clear direction and position reference in a specific coordinate system, ensuring the accuracy of the welding operation;

[0064] Next, adopt the nine-point calibration method, use the calibration plate placed on the product to be welded as the calibration tool to unify the camera center and the robot tool coordinate system, and calculate the pixel equivalent. The specific steps are as follows:

[0065] Divide the calibration board into N×M rectangular areas with side length L to obtain a number of rectangular frames. After the camera takes a picture of the calibration board, use a machine learning algorithm to obtain the upper left coordinate (X1, Y1) and the lower right coordinate (X2, Y2) of a single rectangular frame on the captured image. Then, calculate the length A1 and width B1 of the rectangular frame, and calculate the pixel equivalent E according to the formula E = (A / A1 + B / B1) / 2 (where A = N×L, B = M×L);

[0066] The pixel equivalent is to establish a connection between the image coordinates and the actual physical coordinates, providing a conversion basis for subsequent position calculations based on the image;

[0067] Then obtain the relative height between the robot and the welding surface:

[0068] First, adjust the galvanometer center and the ranging center of the robot, calculate the relative position coordinates of the galvanometer center and the ranging center of the robot with respect to the camera center at a fixed position, and obtain the galvanometer-vision position and the ranging-vision position;

[0069] The specific operation steps may include:

[0070] Place a metal plate on the trolley for placing the product to be welded, and only adjust the robot's posture in the horizontal and vertical directions so that the galvanometer is at the preset focal length position directly above the metal plate, and record the robot coordinate values (mainly the X and Y values) at this position;

[0071] Establish a temporary cross pattern at the center point of the galvanometer. Use this pattern, and the galvanometer center emits low-power (20% of the rated power) laser to draw a visually recognizable cross pattern on the metal plate. Only adjust the robot's posture in the horizontal and vertical directions so that the cross pattern on the metal plate is at the center of the camera. At this time, record the robot position at this point. Subtract the X-axis and Y-axis positions of the robot twice, and the relative position coordinates of the galvanometer center with respect to the visual center can be obtained (DX 1 , DY 1 ).

[0072] Similarly, only adjust the robot's posture in the horizontal and vertical directions so that the cross pattern on the metal plate is at the center of the camera, and record the robot position at this point; and, only adjust the robot's posture in the horizontal and vertical directions so that the laser ranging spot hits the center of the cross pattern, and record the robot position at this point; subtract the X-axis and Y-axis positions of the robot twice, and the relative position coordinates of the ranging center with respect to the visual center can be obtained (DX 2 , DY 2 );

[0073] After fixing these two positions, adjust the robot to emit light at the preset rated power on the metal plate of the product to be welded (in this embodiment, 20% of the rated power is the weak power and is used as the preset rated power for light emission), and then use the laser range finder sensor to measure the distance to the calibration plate to obtain the relative height DZ of the robot from the welding surface 1 , and record the Z-axis data of the robot and define it to the robot axis variable Z 1 to ensure that the copper nozzle and the welding surface maintain an appropriate distance during welding;

[0074] Next, operate the robot to move to the welding position of the cell pole, adjust the laser range finder spot to align with the pole surface, measure and record the current Z-axis data DZ 2 , calculate the preset height deviation according to the formula ΔZ = DZ2 - DZ1, and enter it into the robot control parameters;

[0075] 3. Set the spatial position parameters of the three axes of the copper nozzle;

[0076] First, operate the robot to move to several MARK points in the lightweight file, record the coordinates of the moved robot, and obtain the photographing positions of several MARK points;

[0077] Align the center of the copper nozzle with the centers of several MARK points in sequence, record and fix the positions of the three axes X and Y of the copper nozzle;

[0078] Then adjust the Z-axis of the three axes of the copper nozzle so that it completely presses the bus bar of the cell pole, and trigger the laser range finder sensor to obtain the current Z-axis data DZ;

[0079] By judging whether the difference between DZ and the relative height DZ 1 is less than the preset height deviation to determine the final position of the Z-axis of the three axes of the copper nozzle. If the difference is less than the preset height deviation, record and fix the Z-axis position of the three axes of the copper nozzle; if not, re-calibrate the vibration-vision position and the measurement-vision position.

[0080] It should be noted that only one preparation work is required for products of the same model, and there is no need to repeat it multiple times.

[0081] In one implementation manner, the acquisition method of the actual position coordinates of the MARK points and the theoretical photographing coordinates of the cell poles in step S2 may include:

[0082] The robot moves to the photographing positions of several MARK points to take pictures and obtains several MARK point images;

[0083] Use the template matching method to detect the point coordinates in several MARK point images to obtain the actual positions of several MARK points where i represents the index of the MARK point;

[0084] According to the formula the theoretical shooting coordinates (X j , Y j ) of the cell terminal are calculated.

[0085] In one implementation, the method for obtaining the offset of the actual position coordinates of the cell terminal relative to the theoretical position coordinates in step S3 may include:

[0086] Collect a number of photos of the terminal area to obtain a module library containing a number of templates;

[0087] Use the multi-template matching algorithm to calculate the similarity between a number of templates and a single image, and screen to obtain the template coordinates (X, Y) with the highest similarity;

[0088] Multiply the template coordinates (X, Y) with the highest similarity by the pixel equivalent to obtain the cell terminal offset (EX, EY).

[0089] Finally, in step S4, respectively determine whether the offsets (EX, EY) of the X-axis and Y-axis are within the preset threshold range; if so, add the theoretical shooting coordinates of the cell terminal and the offset, and mark it as a qualified welding position, then control the robot to weld at the qualified welding position; if not, mark the theoretical shooting coordinates of the cell terminal as a failed welding position.

[0090] Some of the data in the above formula are calculated by removing the dimension and taking their numerical values. The formula is a formula that is obtained by software simulation of a large amount of collected data and is closest to the actual situation; the preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained by simulating a large amount of data.

[0091] The working principle of the present invention:

[0092] First, convert and lightweight the CAD drawing to obtain the relative theoretical positions of the cell terminal and the MARK point. Then construct a coordinate system, unify the camera and robot tool coordinate systems and calculate the pixel equivalent, and obtain the relative height between the robot and the welding surface and the preset height deviation. Finally, determine the MARK point photographing position, adjust the three-axis position of the copper nozzle to complete the preparation work;

[0093] Then operate the robot to photograph the MARK point to obtain its actual position coordinates, and calculate the theoretical shooting coordinates of the cell terminal in combination with the relative theoretical position. Then take a photo at this coordinate, obtain the offset and determine whether it is within the threshold. If it is, obtain the qualified welding position according to the offset; if not, mark it as a failed position.

[0094] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for automatically finding battery pole welding based on CAD import, characterized in that: include: S1, calculate the relative theoretical position of the battery pole and several MARK points according to the product CAD drawing, and complete the preparation work according to the product CAD drawing; S2, obtaining the actual position coordinates of several MARK points, and obtaining the theoretical shooting coordinates of the battery cell pole according to the actual position coordinates and the relative theoretical position; S3, operating the robot to take a plurality of images at the theoretical shooting coordinates, and using a multi-template matching method to obtain the offset of the actual position coordinates of the battery cell pole relative to the theoretical position coordinates according to the plurality of images; S4, determining whether the offset is within a preset threshold range; If yes, add the theoretical shooting coordinates of the battery cell pole and the offset and mark it as a qualified welding position; If not, the theoretical shooting coordinates of the battery cell pole are marked as the failed welding position.

2. The method for automatically finding the battery pole welding based on CAD import according to claim 1 is characterized in that: The relative theoretical positions of the battery poles and several MARK points are calculated based on the CAD drawings of the product, including: S11-1, converting the CAD drawing of the product into a dxf file, and performing lightweight processing on the dxf file according to the attribute data of the battery cell pole and the MARK point to obtain a lightweight file; S11-2, using the target detection algorithm to automatically detect the battery poles and MARK points in the lightweight file, and obtain the theoretical position coordinates of the battery poles and several MARK point coordinates; S11-3, according to the theoretical position coordinates of the battery cell pole and the coordinates of several MARK points, the relative distance between the battery cell pole and each MARK point is calculated to obtain the relative theoretical positions of the battery cell pole and the several MARK points.

3. The method for automatically finding the battery pole welding based on CAD import according to claim 2 is characterized in that: The preparation work includes: S12-1, establish a robot tool coordinate system, and make the robot tool coordinate system parallel to the XY axis of the copper nozzle; S12-2, unifying the camera center and the robot tool coordinate system using a nine-point calibration method, and calculating the pixel equivalent using a calibration plate; wherein the calibration plate is placed on the product to be welded for the calibration task; S12-3, use the laser ranging sensor to obtain the relative height between the robot and the welding surface, and set the spatial position parameters of the three axes of the copper nozzle according to the CAD drawing and the relative height.

4. The method for automatically finding the battery pole welding based on CAD import according to claim 3 is characterized in that: The method of calculating the pixel equivalent by using the calibration plate includes: The calibration plate is divided into N×M rectangular areas with a side length of L to obtain several rectangular frames; Use a camera to take a picture of the calibration plate, and use a machine learning algorithm to obtain the upper left corner coordinates (X1, Y1) and lower right corner coordinates (X2, Y2) of a single rectangular box on the captured image; Subtract the horizontal and vertical coordinates of the lower right corner coordinates (X2, Y2) from the upper left corner coordinates (X1, Y1) to obtain the length A1 and width B1 of the rectangular frame; The pixel equivalent E is calculated according to the formula E=(A / A1+B / B1) / 2, where A and B represent the length and width of the calibration plate, respectively, and A=N×L, B=M×L.

5. The method for automatically finding the battery pole welding based on CAD import according to claim 3 is characterized in that: The method of using a laser ranging sensor to obtain a relative height between the robot and the welding surface includes: Adjust the galvanometer center and the ranging center of the robot, calculate the relative position coordinates of the galvanometer center and the ranging center of the robot and the camera center when they are in a fixed position, and obtain the vibration-viewing position and the ranging-viewing position; Fix the vibration-visual position and the measurement-visual position, and adjust the robot to emit light at a preset rated power on the metal plate of the product to be welded; The laser distance sensor is used to measure the distance of the calibration plate to obtain the relative height DZ1 between the robot and the welding surface.

6. The method for automatically finding the battery pole welding based on CAD import according to claim 3 is characterized in that: The spatial position parameters of the three axes of the copper nozzle are set according to the CAD drawing and the relative height, including: Operate the robot to move to several MAKD points in the lightweight file, and record the coordinates of the robot after the movement to obtain the photographing positions of several MARK points; Align the center of the copper nozzle with the centers of several MARK points in sequence, and record and fix the X and Y positions of the three axes of the copper nozzle; Adjust the Z axis of the copper nozzle triaxial to completely press the busbar of the battery cell pole, triggering the laser ranging sensor to obtain the current Z axis data DZ; Determine whether the difference between DZ and the relative height DZ1 is less than the preset height deviation; if yes, record and fix the Z-axis position of the three axes of the copper nozzle; if no, jump to S2-1 and recalibrate the vibration-visual position and the measurement-visual position.

7. The method for automatically finding the battery pole welding based on CAD import according to claim 6 is characterized in that: The method for obtaining the preset height deviation includes: Operate the robot to move to the welding position of the battery pole, adjust the laser ranging spot to the pole surface, measure and record the current Z-axis data DZ2; The preset height deviation is calculated according to the formula ΔZ=DZ2-DZ1, and ΔZ is entered into the robot control parameters.

8. The method for automatically finding the battery pole welding based on CAD import according to claim 1 is characterized in that: The actual position coordinates of the MARK points are obtained, including: The robot moves to the photographing positions of the several MARK points to take pictures, and obtains several MARK point images; Use template matching method to detect the point coordinates of several MARK points in the image and get the actual positions of several MARK points Among them, i represents the index of the MARK point.

9. The method for automatically finding the battery pole welding based on CAD import according to claim 2 is characterized in that: Obtaining the theoretical shooting coordinates of the battery cell pole according to the actual position coordinates and the relative theoretical position includes: According to the formula Calculate the theoretical shooting coordinates (X j ,Y j );in, Indicates the theoretical position coordinates of the battery cell pole.

10. The method for automatically finding battery cell pole welding based on CAD import according to claim 3, characterized in that: The method of using a multi-template matching method to obtain the offset of the actual position coordinates of the battery cell pole relative to the theoretical position coordinates according to a plurality of images includes: Collect several pole area photos to obtain a module library containing several templates; Use the multi-template matching algorithm to calculate the similarity between several templates and a single image, and select the template coordinates (X, Y) with the highest similarity; Multiply the template coordinates (X, Y) with the highest similarity by the pixel equivalent to obtain the cell pole offset (EX, EY).

Images