Methods, apparatus and cells for preparing battery cells

Abstract

This application relates to a method, apparatus, and battery cell for manufacturing a single battery cell. The method includes: winding a positive electrode sheet, a negative electrode sheet, and a separator to form a battery cell assembly; pre-pressing and shaping the battery cell assembly; conveying the shaped battery cell assembly to a visual inspection station and positioning the battery cell assembly at the visual inspection station; acquiring multiple perspective images of the battery cell assembly; inspecting the separator state of the battery cell assembly based on the multiple perspective images to determine whether the battery cell assembly has separator wrinkling defects, obtaining the inspection result of the battery cell assembly; if the inspection result shows that the separator state of the battery cell assembly meets the preset state, the battery cell assembly is determined to be a good product, and the battery cell assembly continues to be manufactured along a preset good product flow path; if the inspection result shows that the separator state of the battery cell assembly does not meet the preset state, the battery cell assembly is determined to be a defective product, and the battery cell assembly is picked up and sent to the defective product processing path. This improves the accuracy of the inspection.

Description

Technical Field

[0001] This application relates to the field of battery cell technology, and in particular to a method, apparatus and battery cell for preparing a single battery cell. Background Technology

[0002] In lithium batteries, the separator acts as a barrier between the positive and negative electrodes, preventing short circuits caused by contact between them. After the winding machine completes the winding process, the battery cell remains in a loose state. Factors such as static electricity in the separator, internal airflow disturbances, separator vibration caused by the grippers releasing the cell, and the start / stop of the logistics line affecting the inner separator ring can all cause the separator to fold inwards, leading to contact between the positive and negative electrodes and posing a safety hazard during battery cell use. Therefore, detecting whether the separator is wrinkled during the battery cell manufacturing process is crucial.

[0003] In traditional technologies, during the cell manufacturing process, it is usually determined whether the separator is wrinkled by manual visual inspection or by high-potential test (Hipot) inspection.

[0004] However, traditional cell manufacturing methods suffer from low accuracy in detecting wrinkles in the separator. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, and battery cell for preparing a battery cell that can improve the accuracy of separator wrinkle detection, in order to address the above-mentioned technical problems.

[0006] In a first aspect, this application provides a method for preparing a battery cell, comprising:

[0007] The positive electrode sheet, negative electrode sheet, and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0008] The finalized battery cell assembly is transported to the visual inspection station and positioned at the visual inspection station.

[0009] Acquire multiple perspective images of the battery cell assembly;

[0010] Based on the multiple perspective images, the state of the separator of the battery cell assembly is detected to determine whether there is a separator wrinkling defect in the battery cell assembly, and the detection result of the battery cell assembly is obtained.

[0011] If the detection result indicates that the separator condition of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product, and the battery cell assembly continues to be manufactured along the preset good product flow path; if the detection result indicates that the separator condition of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product, and the battery cell assembly is picked up and sent to the defective product processing path.

[0012] In one embodiment, the visual inspection station is equipped with a photoelectric sensor and at least one positioning block; positioning the battery cell assembly at the visual inspection station includes:

[0013] The photoelectric sensor detects that the battery cell assembly has arrived at the visual inspection station and generates a trigger signal to be sent to the controller.

[0014] In response to the trigger signal, the controller drives the at least one positioning block to define the position of the battery cell assembly.

[0015] In one embodiment, acquiring multiple viewpoint images of the battery cell assembly includes:

[0016] In response to the trigger signal, the controller sends a photo-taking command to the at least two cameras to capture images of at least one side of the battery cell assembly from different angles; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0017] In one embodiment, the step of detecting the separator state of the battery cell assembly based on the multiple perspective images, determining whether the battery cell assembly has separator wrinkling defects, and obtaining the detection result of the battery cell assembly includes:

[0018] The multiple viewpoint images are combined to obtain a composite image of the battery cell assembly;

[0019] Edge detection is performed on the synthesized image to obtain a binary image of the edge of the battery cell assembly;

[0020] The state of the separator of the battery cell assembly is detected based on the edge binary image to determine whether there is a separator wrinkling defect in the battery cell assembly, and the detection result of the battery cell assembly is obtained.

[0021] In one embodiment, the step of detecting the separator state of the battery cell assembly based on the edge binary image to determine whether the battery cell assembly has separator wrinkling defects and obtaining the detection result of the battery cell assembly includes:

[0022] Based on the edge binary image, determine the edge length and the number of separator layers of the battery cell assembly;

[0023] When both the edge length and the number of membrane layers meet the preset good product conditions, the detection result of the battery cell assembly is determined based on the synthesized image and the preset detection model.

[0024] In one embodiment, the method further includes:

[0025] When the edge length is greater than a length threshold and the number of membrane layers is greater than a layer number threshold, the battery cell assembly is determined to meet the good product condition.

[0026] In one embodiment, the synthesized image includes a tab-side image and a non-tab-side image; the step of performing edge detection on the synthesized image to obtain a binary edge image of the battery cell assembly includes:

[0027] The electrode region in the electrode side image is determined based on the difference in pixel grayscale values ​​in the electrode side image;

[0028] The electrode side image is segmented according to the electrode region to obtain the segmented electrode side image;

[0029] Edge detection is performed on the cut electrode side image and the non-electrode side image to obtain the edge binary image of the battery cell assembly.

[0030] In one embodiment, the process of constructing the detection model includes:

[0031] Control the at least two cameras to take pictures of the battery cell assembly with defective separators, and acquire multiple sets of images of defective separators;

[0032] Image synthesis processing is performed on the multiple sets of images of diaphragm defects to obtain a training dataset;

[0033] The initial deep learning model is trained based on the training dataset to obtain the detection model.

[0034] Secondly, this application also provides an apparatus for manufacturing a single battery cell, the apparatus comprising a controller and a computer-readable storage medium, the controller comprising a memory and a processor, the computer-readable storage medium storing a computer program, wherein the computer program in the computer-readable storage medium, when executed by the processor of the controller, implements the steps of the method described in the first aspect; the apparatus further comprises:

[0035] The winding and shaping module is used to wind the positive electrode sheet, the negative electrode sheet and the separator to form a battery cell assembly, and to pre-press and shape the battery cell.

[0036] A conveying module is used to convey the shaped battery cell assembly to the visual inspection station and position the battery cell assembly at the visual inspection station.

[0037] An acquisition module is used to acquire multiple perspective images of the battery cell assembly;

[0038] The detection module is used to detect the membrane state of the battery cell assembly based on the multiple perspective images, determine whether the battery cell assembly has membrane wrinkling defects, and obtain the detection result of the battery cell assembly.

[0039] The determination module is used to determine that the battery cell assembly is a good product when the detection result shows that the separator state of the battery cell assembly meets the preset state, and the battery cell assembly continues to be manufactured along the preset good product flow path; when the detection result shows that the separator state of the battery cell assembly does not meet the preset state, the battery cell assembly is determined to be an abnormal product, and the battery cell assembly is picked up and sent to the abnormal product processing path.

[0040] Thirdly, this application also provides a battery cell, which is prepared using the battery cell preparation method described in the first aspect.

[0041] Fourthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0042] The positive electrode sheet, negative electrode sheet, and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0043] The finalized battery cell assembly is transported to the visual inspection station and positioned at the visual inspection station.

[0044] Acquire multiple perspective images of the battery cell assembly;

[0045] Based on the multiple perspective images, the state of the separator of the battery cell assembly is detected to determine whether there is a separator wrinkling defect in the battery cell assembly, and the detection result of the battery cell assembly is obtained.

[0046] If the detection result indicates that the separator condition of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product, and the battery cell assembly continues to be manufactured along the preset good product flow path; if the detection result indicates that the separator condition of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product, and the battery cell assembly is picked up and sent to the defective product processing path.

[0047] The aforementioned battery cell preparation method, apparatus, and battery cell involve winding a positive electrode sheet, a negative electrode sheet, and a separator to form a battery cell assembly; pre-pressing and shaping the battery cell assembly; conveying the shaped battery cell assembly to a visual inspection station and positioning it there; acquiring multiple perspective images of the battery cell assembly; inspecting the separator state based on these images to determine if there are separator wrinkling defects, and obtaining the inspection result; if the inspection result indicates that the separator state of the battery cell assembly meets the preset state, the battery cell assembly is determined to be a good product and continues subsequent preparation along the preset good product flow path; if the inspection result indicates that the separator state of the battery cell assembly does not meet the preset state, the battery cell assembly is determined to be an abnormal product and is picked up and sent to the abnormal product processing path. It enables online, non-contact detection of membrane wrinkling defects, replacing inefficient manual visual inspection, improving detection accuracy, and raising detection efficiency to match production line efficiency, ensuring that defective products are intercepted in real time, thereby improving the overall safety and consistency of battery cells. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a schematic diagram of the traditional battery cell preparation method;

[0050] Figure 2 This is an application environment diagram of the battery cell preparation method in one embodiment;

[0051] Figure 3 This is a schematic flowchart of a method for preparing a single battery cell in one embodiment;

[0052] Figure 4 This is the normal state of the separator in the battery cell assembly;

[0053] Figure 5 Schematic diagram of diaphragm wrinkling after pre-pressing and shaping of battery cell assembly;

[0054] Figure 6 A schematic diagram showing the wrinkled separator after the battery cell assembly is disassembled;

[0055] Figure 7 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0056] Figure 8 This is a schematic diagram of the tab side of the battery cell assembly;

[0057] Figure 9 This is a schematic diagram of the non-tab side of the battery cell assembly;

[0058] Figure 10 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0059] Figure 11 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0060] Figure 12 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0061] Figure 13 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0062] Figure 14 This is a schematic flowchart of a method for preparing a single battery cell in another embodiment;

[0063] Figure 15 This is a structural block diagram of a battery cell preparation apparatus in one embodiment;

[0064] Figure 16 This is an internal structural diagram of a computer device in one embodiment.

[0065] Explanation of reference numerals in the attached figures:

[0066] 1. Logistics line; 2. Light source; 3. Image acquisition unit on the tab side;

[0067] 4. Positioning block on the tab side; 5. Positioning block on the first electrode side;

[0068] 6. Positioning block on the second motor side;

[0069] 7. Positioning block on the non-electrode side; 8. Image acquisition unit on the non-electrode side;

[0070] 9. Photoelectric sensor; 10. Battery cell assembly. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0072] like Figure 1As shown, in the traditional method of preparing a battery cell, the positive electrode sheet, negative electrode sheet and separator are wound to form a cell assembly. After the cell assembly is pre-pressed and shaped and then pre-pressed again, the condition of the separator is inspected by manual visual inspection. Manual visual inspection can observe whether the edge of the separator is continuous. However, the cell separator has many layers, small gaps between layers, and the production line is fast. It is also strongly related to the skills and working conditions of the employees, resulting in low detection rate and accuracy of manual visual inspection.

[0073] Based on this, this application provides a method for preparing a battery cell that can improve the accuracy of separator wrinkle detection.

[0074] The method for preparing a battery cell provided in this application embodiment can be applied to, for example... Figure 2 The application environment is shown below. In this environment, 1 is the material flow line, 2 is the light source, 3 is the image acquisition unit on the tab side, 4 is the positioning block on the tab side, 5 is the positioning block on the first electrode side, 6 is the positioning block on the second electrode side, 7 is the positioning block on the non-tab side, 8 is the image acquisition unit on the non-tab side, 9 is the photoelectric sensor, and 10 is the battery cell assembly. The tab side and the non-tab side are positioned opposite each other, and the first motor side and the second motor side are positioned opposite each other. After processing the positive electrode, negative electrode, and separator to obtain the shaped battery cell assembly 10, the controller acquires multiple perspective images of the battery cell assembly 10 from the image acquisition unit 3 on the tab side and the image acquisition unit 8 on the non-tab side. Based on these multiple perspective images, the controller detects the separator state of the battery cell assembly 10, determines whether there are separator wrinkling defects, and obtains the detection result of the battery cell assembly.

[0075] In one embodiment, such as Figure 3 As shown, a method for preparing a single battery cell is provided, comprising:

[0076] S201 involves winding the positive electrode sheet, negative electrode sheet, and separator to form a battery cell assembly, and then pre-pressing and shaping the battery cell assembly.

[0077] In this embodiment, the battery cell preparation apparatus includes modules such as a winding and shaping module, a conveying module, a vision inspection station, an image acquisition unit, and a controller.

[0078] In this embodiment, the cut positive electrode sheet, negative electrode sheet, and two layers of separator are arranged and fixed on the winding needle of the winding device in the stacking order of "positive electrode sheet, separator, negative electrode sheet, separator". Further, the winding needle rotates and winds according to a preset program, tightly winding the multiple layers together to form a cylindrical or square bare battery cell assembly with a preset number of turns and layer structure. After the winding operation is completed, the winding needle can perform a flipping or transferring action to remove the initially formed battery cell assembly from the winding area, and attach finishing tape to the outermost layer of the battery cell to fix the winding structure and prevent it from loosening.

[0079] In this embodiment, the wound battery cell assembly is fed into a pre-pressing and shaping process. The battery cell assembly is placed between a pre-pressing mold or two flat pressure plates. The pressure plates close under the control of a preset program, applying a calculated and set pressure to the battery cell assembly. This pressure reduces the gap between the internal electrodes and the separator, improves interlayer adhesion, and eliminates most of the air and loose structure introduced during winding. This results in the battery cell assembly achieving initial mechanical stability and a consistent thickness without damaging the internal structure. The battery cell assembly is held under constant pressure for a period of time to ensure the shaping effect.

[0080] S202, the finalized battery cell assembly is transported to the visual inspection station and positioned there.

[0081] In this embodiment, after the pre-pressing and shaping of the battery cell assembly is completed, the battery cell assembly is transported to the visual inspection station via a conveying device such as a logistics line. When the battery cell assembly reaches the preset position of the inspection station, the positioning sensing device is triggered. Furthermore, the positioning mechanism on the inspection station constrains and aligns the battery cell assembly from multiple directions to ensure that its position is fixed and its posture is consistent during subsequent image acquisition.

[0082] S203, acquire multiple perspective images of the battery cell assembly.

[0083] In this embodiment, the visual inspection station is equipped with multiple image acquisition units. These units simultaneously or sequentially capture images of the area to be inspected in the battery cell assembly from different preset angles to overcome blind spots or obstructions that may exist from a single perspective, thereby obtaining multi-view data that comprehensively reflects the edge state of the separator of the battery cell assembly. The multiple image acquisition units can be divided into multiple groups, with each group containing multiple image acquisition units installed at different angles but irradiated at the same location, i.e., irradiating the same position on the battery cell.

[0084] For example, Figure 4 This is the normal state of the separator in the battery cell assembly. Figure 5 A schematic diagram showing the wrinkling of the separator after pre-pressing and shaping of the battery cell assembly. Figure 6 This is a schematic diagram showing the wrinkled diaphragm after the battery cell assembly has been disassembled.

[0085] For example, when there are two image acquisition units, the distance between the image acquisition unit and the battery cell assembly can be determined according to Equation 1:

[0086] (Equation 1)

[0087] Where Z is the distance between the image acquisition unit and the battery cell assembly; f is the focal length of the image acquisition unit, which is a fixed parameter; B is the baseline distance between the two image acquisition units; and d is the pixel difference of the same feature point in the images acquired by the two image acquisition units, which can be calculated using the feature point matching method. When the coordinates of image acquisition unit 1 are (X... c1 ,Y c1 Z c1 The coordinates of image acquisition unit 2 are (X... c2 ,Y c2 Z c2 When the baseline distance is ), .

[0088] For example, the angle between the optical axes of the two image acquisition units can be determined according to Equation 2:

[0089] (Equation 2)

[0090] For example, when the measurement accuracy of the image acquisition unit is ±0.05mm, the measurement area length of the battery cell assembly is 130mm, the camera resolution is 5 megapixels (2448×2048), and the pixel size is 3.45μm, we can choose f as 25mm, Z as [400mm, 500mm], B as [100mm, 150mm] (i.e., 1 / 4 to 1 / 3 of Z), and θ as [14°, 21°]. The image field of view can be expressed as (pixel size × image width × Z) / f. Based on these parameters, we can determine that the image field of view acquired by the two image acquisition units is approximately 152mm, which meets the 130mm detection requirement, and determine the angle between the optical axes of the two image acquisition units. .

[0091] S204. Based on images from multiple perspectives, the state of the separator in the battery cell assembly is detected to determine whether there are separator wrinkling defects in the battery cell assembly, and the detection result of the battery cell assembly is obtained.

[0092] In this embodiment, after acquiring images from multiple perspectives, the acquired images are preprocessed and fused to construct a complete visual image for analysis. Further, the state of the separator in the battery cell assembly is detected, analyzing whether the edge morphology of the separator is continuous and smooth. Feature analysis can be performed on the fused complete visual image, and it can be compared with a preset normal separator state benchmark to determine whether the battery cell assembly has a wrinkling defect where the separator bends inward, thereby outputting the detection result of the battery cell assembly.

[0093] S205, if the test result shows that the separator of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product and the battery cell assembly continues to be processed along the preset good product flow path; if the test result shows that the separator of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be an abnormal product and the battery cell assembly is picked up and sent to the abnormal product processing path.

[0094] In this embodiment, if the detection result shows that the separator condition of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product. The battery cell assembly is then released along the preset good product flow path and flows to the next production process, such as casing or liquid injection. If the detection result shows that the separator condition does not meet the preset condition, i.e., there is a separator wrinkling defect, the battery cell assembly is determined to be a defective product. The defective battery cell assembly is removed from the main transfer path by a sorting component and transferred to a defective product processing path, such as a rework line or waste collection device, thereby preventing defective products from flowing into subsequent processes. The sorting component can be an ejector rod or a diverter plate.

[0095] In the above-mentioned method for preparing a single battery cell, the positive electrode sheet, negative electrode sheet, and separator are wound to form a battery cell assembly, which is then pre-pressed and shaped. The shaped battery cell assembly is then transported to a visual inspection station and positioned there. Multiple images of the battery cell assembly from various perspectives are acquired. Based on these images, the condition of the separator is inspected to determine if wrinkling defects exist, resulting in an inspection result. If the separator condition conforms to a preset condition, the battery cell assembly is considered a good product and continues its subsequent preparation along a preset good product flow path. If the separator condition does not conform to the preset condition, the battery cell assembly is considered an abnormal product and is moved to the abnormal product processing path. This method achieves online, non-contact inspection of separator wrinkling defects, replacing inefficient manual visual inspection, improving inspection accuracy, and increasing inspection efficiency to match production line efficiency. This ensures that defective products are intercepted in real time, thereby improving the overall safety and consistency of the battery cells.

[0096] In one embodiment, one implementation of S202 above is provided, wherein the visual inspection station is equipped with a photoelectric sensor and at least one positioning block, such as... Figure 7 As shown, the above-mentioned "positioning of the battery cell assembly at the visual inspection station" includes:

[0097] S301, the photoelectric sensor detects that the battery cell assembly has arrived at the visual inspection station and generates a trigger signal to be sent to the controller.

[0098] In this embodiment, when the battery cell assembly is transported from the logistics line into the visual inspection station, the photoelectric sensor installed at the station entrance is triggered by a change in the light intensity detection value. Preferably, when the battery cell assembly has not passed the photoelectric sensor, the detection value can be calibrated to 0; when the photoelectric sensor is completely blocked, the detection value is x; the photoelectric sensor threshold can be set to x / 2; when the battery cell assembly passes the photoelectric sensor, the photoelectric sensor value changes; when the detection value is greater than x / 2, it is determined that the battery cell has passed.

[0099] Preferably, the photoelectric sensor can be a through-beam photoelectric sensor, with the transmitting and receiving ends of the photoelectric sensor respectively positioned on opposite sides of the transport path, forming a detection light curtain. During transport, the battery cell assembly enters and blocks this light curtain, causing a change in the light intensity detected by the receiving end. The internal circuitry of the photoelectric sensor converts this light intensity change into an electrical signal and generates a high-level trigger signal. This trigger signal is immediately sent to the controller via a fieldbus or I / O module.

[0100] S302, in response to a trigger signal, the controller drives at least one positioning block to define the position of the battery cell assembly.

[0101] In this embodiment, after receiving a trigger signal from the photoelectric sensor, the controller sends a command to the drive component that controls the movement of the positioning block. The drive component then pushes at least one positioning block from a standby position toward the battery cell assembly on the logistics line. Preferably, the drive component is any one of a pneumatic solenoid valve, a servo motor driver, or a cylinder.

[0102] Preferably, the controller can be a programmable logic controller (PLC).

[0103] Preferably, the positioning block can move collaboratively from multiple directions, such as the sides and ends of the battery cell assembly, to contact the reference surface of the battery cell assembly and apply a preset clamping force, thereby completely restricting the degrees of freedom of the battery cell assembly and stably stopping and fixing it at the preset image origin of the vision inspection station. After positioning is completed, the controller sends a ready signal to the vision system to start the image acquisition process.

[0104] In the above-mentioned embodiments, a trigger signal is generated by real-time detection of the photoelectric sensor, and the controller immediately drives the positioning block to perform the action, realizing the rapid completion of the battery cell assembly from dynamic transportation to static positioning. This replaces manual placement and alignment, ensuring the consistency of the spatial position and attitude of the battery cell assembly during each test, and improving the consistency and accuracy of the testing process.

[0105] In one embodiment, an implementation of the above-mentioned S203 is provided, wherein the above-mentioned "acquiring multiple perspective images of the battery cell assembly" includes: the controller responding to a trigger signal sends a photo-taking command to at least two cameras, and the at least two cameras acquire images of at least one side end face of the battery cell assembly from different angles; the at least one side is at least one side of the non-tab side and the tab side of the battery cell assembly.

[0106] In this embodiment, the tab side and the non-tab side are opposite sides, the first electrode side and the second electrode side are two sides that intersect with the tab side, and the first electrode side and the second electrode side are opposite sides.

[0107] In this embodiment, the controller synchronously sends a trigger signal as a shooting instruction to at least two pre-configured cameras. The at least two cameras that receive the instruction start exposure and take pictures at the same time. When multiple cameras correspond to the same side end face, multiple original images of the same detection side end face can be acquired at one time.

[0108] Preferably, at least two cameras are mounted on a preset bracket, and the optical axes of the at least two cameras are aligned with the end face of the battery cell assembly to be tested at preset different angles. The end face to be tested can be the non-tab side or the tab side. The at least two cameras can be aligned with the same end face or with different end faces.

[0109] Preferably, the camera can be a charge-coupled device (CCD) image sensor camera. CCD cameras can be configured in pairs, with each CCD detection assembly acquiring images of the same side face. For example, each CCD detection assembly may include two CCD cameras, a light source, and a corresponding signal transmission system. The light source can be a ring light source, and its angle is adjustable. The optimal light source angle can reduce shadows and improve image contrast, thus better detecting defects on the battery cell surface. In this embodiment, the angle range between the light source and the battery cell is (20°, 30°). The two CCD cameras are installed at different angles but at the same illumination position to facilitate image synthesis. Based on the PLC signal, both cameras simultaneously trigger image capture. The CCD camera uses an automatic zoom camera. Before capturing the image, an ultrasonic generator emits a signal to the surface of the battery cell under test, automatically adjusting the focal length according to the distance between the battery cell and the camera.

[0110] For example, Figure 8 This is a schematic diagram of the tab side of the battery cell assembly. Figure 9 This is a schematic diagram of the non-tab side of the battery cell assembly.

[0111] In the above-mentioned embodiments, by simultaneously imaging the same detection surface from different angles, the problems of single-view visual blind spots and image distortion caused by the softness and warping of the cell separator itself and the obstruction of the tab structure can be effectively overcome, reducing the risk of missed defects due to missing information and improving the accuracy of defect detection.

[0112] In one embodiment, one implementation of S204 above is provided, such as... Figure 10 As shown, the above-mentioned "detecting the separator state of the battery cell assembly based on multiple perspective images, determining whether the battery cell assembly has separator wrinkling defects, and obtaining the detection results of the battery cell assembly" includes:

[0113] S401 performs image synthesis processing on images from multiple perspectives to obtain a composite image of the battery cell assembly.

[0114] In this embodiment, image synthesis processing is performed on multiple viewpoint images. This synthesis process may include steps such as feature point matching, image geometric correction and alignment, and pixel-level fusion, thereby generating a synthesized image that displays the entire detected end face. Preferably, a corresponding synthesized image can be generated for each detected end face.

[0115] For example, multiple viewpoint image data can be loaded and read first, and an overlapping region search step can be performed. Algorithms such as normalized cross-correlation can be used to calculate the optimal matching region between two viewpoint images, determine the spatial offset between the two viewpoint images, and evaluate the confidence level of the match. Then, based on the calculated offset, one image is subjected to geometric corrections such as rotation or scaling to spatially align it with the other image, eliminating geometric differences caused by different shooting angles. Finally, a progressive fade-in / fade-out fusion algorithm is used to smoothly merge the two aligned images. This algorithm uses a weighted average method for transition in the overlapping regions of the images; for example, the left image has a higher weight closer to the left image, and the right image has a higher weight closer to the right image, thus achieving seamless stitching. After fusion, a composite image containing complete information from multiple viewpoints is output.

[0116] S402, perform edge detection on the synthesized image to obtain a binary image of the edge of the battery cell assembly.

[0117] In this embodiment, a preset image algorithm is used to analyze the regions of pixel grayscale variation in the synthesized image. Through a preset gradient threshold and the image algorithm, the edge lines of the battery cell module separator in the synthesized image are identified, resulting in a binary image of the battery cell module's edge. In this application scenario, the primary target is the edge lines of the battery cell module separator.

[0118] It should be noted that in the binary image of the edge, the boundary of the separator of the battery cell assembly is a white line, while the non-edge area is black. The edge of a normal separator appears as a set of continuous, smooth, and parallel white lines. If there is a separator wrinkling defect, the white edge lines will appear broken, concave inward, or exhibit abnormal bending and forking.

[0119] Preferably, the image algorithm can be an edge detection operator such as Canny or Sobel.

[0120] S403, the state of the separator of the battery cell assembly is detected based on the edge binary image to determine whether there is a separator wrinkling defect in the battery cell assembly, and the detection result of the battery cell assembly is obtained.

[0121] In this embodiment, the membrane condition of the battery cell assembly is detected based on the generated edge binary image. A normal membrane edge should be a set of basically parallel, continuous and smooth lines. When the membrane has a wrinkling defect, the membrane edge will have local breakage, inward indentation or irregular additional curved lines. By analyzing the edge binary image, the features of the membrane edge are extracted, and the features of the membrane edge are compared with the preset good product standard to determine whether the battery cell assembly has a membrane wrinkling defect, thereby outputting the detection result.

[0122] In the above-mentioned embodiments, the limitations of single-view field of view and geometric distortion are overcome by image synthesis processing, generating a synthetic image that can completely and realistically reflect the morphology of the entire diaphragm. Furthermore, the complex image information is abstracted into a binary line graph that clearly represents the geometric contour of the diaphragm through edge detection, ensuring the comprehensiveness and high accuracy of the detection results and improving the efficiency and robustness of defect detection.

[0123] In one embodiment, one implementation of S403 described above is provided, such as... Figure 11 As shown, the above-mentioned "detecting the separator state of the battery cell assembly based on the edge binary image, determining whether the battery cell assembly has separator wrinkling defects, and obtaining the detection result of the battery cell assembly" includes:

[0124] S501, determine the edge length and number of separator layers of the battery cell assembly based on the edge binary image.

[0125] In this embodiment of the application, the edge length of the battery cell assembly is obtained by calculating the sum of the geometric lengths of the connected regions formed by all white pixels marked as edges in the edge binary image; and the edge lines in the edge binary image are identified and counted to determine the number of lines that conform to the edge characteristics of the separator, thereby determining the number of separator layers in the battery cell assembly.

[0126] S502, when the edge length and the number of separator layers both meet the preset good product conditions, the detection result of the battery cell assembly is determined based on the synthesized image and the preset detection model.

[0127] In this embodiment, when the edge length is within a preset length range and the number of membrane layers is within a preset number of layers, the synthesized image is used as input to a pre-trained detection model. The detection model identifies whether a membrane wrinkling defect exists in the synthesized image based on pre-trained logic. When the detection model outputs that no membrane defect exists, the battery cell assembly is determined to be a good product; when the detection model outputs that a membrane defect exists, the battery cell assembly is determined to be a defective product.

[0128] Preferably, when the edge length is greater than a length threshold and the number of separator layers is greater than a layer number threshold, the battery cell assembly is determined to meet the good product conditions. That is, when the edge length is greater than the length threshold and the number of separator layers is greater than the layer number threshold, the battery cell assembly is initially determined to meet the good product conditions, and the synthesized image is input into the detection model. When the edge length is not greater than the length threshold and the number of separator layers is not greater than the layer number threshold, the battery cell assembly is determined to be a defective product. The length threshold and layer number threshold are related to the type of battery cell assembly.

[0129] In the above-mentioned application embodiments, by integrating the deterministic rules based on image processing with the recognition capabilities of deep learning models, the efficiency and accuracy are optimized. This not only enables rapid identification of defective products and improves detection efficiency, but also allows for further identification based on initial qualified products, thereby improving detection accuracy.

[0130] In one embodiment, an implementation of S402 described above is provided, wherein the synthesized image includes a tabular image and a non-tabular image, such as... Figure 12 As shown, the above-mentioned "performing edge detection on the synthesized image to obtain a binary image of the edge of the battery cell assembly" includes:

[0131] S601, determine the electrode region in the electrode side image based on the pixel grayscale value difference in the electrode side image.

[0132] In this embodiment, since the tab is typically a metal conductive electrode sheet, its surface material and reflective properties differ from those of the diaphragm region, resulting in distinct pixel grayscale value characteristics in the image. Therefore, by analyzing the pixel grayscale value differences between adjacent regions in the synthesized image, a connected region with uniform grayscale distribution and forming a high-contrast boundary with the surrounding diaphragm region can be identified—that is, the boundary of the tab region—thereby determining the tab region in the tab-side image.

[0133] S602, the electrode side image is segmented according to the electrode region to obtain the segmented electrode side image.

[0134] In this embodiment, the tabular region is excluded from the original tabular side image. For example, the pixels within the tabular region can be modified to a uniform background color, or the tabular region can be marked so that it can be ignored in subsequent analysis. Further, the segmented tabular side image is obtained.

[0135] S603 performs edge detection on the cut tab-side image and non-tab-side image to obtain a binary image of the edge of the battery cell assembly.

[0136] In this embodiment of the application, the same edge detection algorithm is applied to the cut tabular side image and the non-tabular side image to identify and extract the boundary lines of all diaphragms in the two images respectively, so as to obtain the edge binary image corresponding to the tabular side image and the edge binary image corresponding to the non-tabular side image.

[0137] In the above-mentioned embodiments, the tab region is first identified and accurately located, and then the interference of the tab region is removed by image segmentation. This eliminates the loss or misjudgment of edge information that may be caused by tab occlusion, reduces the risk of missed detection due to local information loss, and improves the accuracy and reliability of diaphragm defect detection.

[0138] In one embodiment, such as Figure 13 As shown, the construction process of the above detection model includes:

[0139] S701 controls at least two cameras to take pictures of the battery cell assembly with defective separators, and acquires multiple sets of images of the defective separators.

[0140] Among them, battery cell components with defective separators can be battery cell components with separator wrinkles of different locations and degrees that are artificially induced.

[0141] In this embodiment of the application, battery cell assemblies with defective separators are placed sequentially at the visual inspection station, and at least two cameras are controlled to take pictures of each battery cell assembly with defective separators according to the same procedure as normal inspection, thereby obtaining multiple sets of images of defective separators. Each set of images includes multiple perspective images of the same battery cell assembly with defective separators taken from different angles.

[0142] S702 performs image synthesis processing on multiple sets of images of diaphragm defects to obtain a training dataset.

[0143] In this embodiment, after acquiring multiple sets of images showing diaphragm defects, referring to step S401, an image synthesis processing algorithm is used to fuse each set of images showing diaphragm defects, generating a corresponding synthesized image. The synthesized images generated from battery cell assemblies with diaphragm defects, as well as the synthesized images generated from good diaphragm products, are labeled to form a training dataset for model training.

[0144] S703 trains the initial deep learning model based on the training dataset to obtain the detection model.

[0145] In this embodiment of the application, the training dataset is input into the initial deep learning model. Through repeated iterative learning processes, the model extracts and learns the features that distinguish between good and abnormal synthetic images from the training dataset, thereby obtaining the detection model.

[0146] Preferably, the initial deep learning model can be a convolutional neural network.

[0147] In the above-mentioned embodiments, by collecting and learning diverse defect samples, the trained detection model can accurately identify complex and subtle wrinkle patterns, thereby improving the accuracy of defect detection.

[0148] In one embodiment, a complete method for preparing a single battery cell is provided, such as... Figure 14 As shown, the above method includes:

[0149] S1, the positive electrode sheet, negative electrode sheet and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0150] S2 transports the finalized battery cell assembly to the visual inspection station, and the photoelectric sensor detects the arrival of the battery cell assembly at the visual inspection station, generating a trigger signal to be sent to the controller.

[0151] S3, in response to the trigger signal, the controller drives at least one positioning block to define the position of the battery cell assembly.

[0152] S4, the controller responds to the trigger signal and sends a photo-taking command to at least two cameras to acquire images of at least one side of the battery cell assembly from different angles through the at least two cameras; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0153] S5 performs image synthesis processing on images from multiple perspectives to obtain a composite image of the battery cell assembly.

[0154] S6. Determine the electrode region in the electrode side image based on the pixel grayscale value difference in the electrode side image.

[0155] S7. The electrode side image is segmented according to the electrode region to obtain the segmented electrode side image.

[0156] S8 performs edge detection on the cut tab-side image and non-tab-side image to obtain the edge binary image of the battery cell assembly.

[0157] S9. Based on the edge binary image, determine the edge length and number of separator layers of the battery cell assembly.

[0158] S10, when the edge length is greater than the length threshold and the number of membrane layers is greater than the number of layers threshold, the cell assembly is determined to meet the good product conditions.

[0159] S11, when the edge length and the number of separator layers both meet the preset good product conditions, the detection result of the battery cell assembly is determined based on the synthesized image and the preset detection model.

[0160] S12, if the test result shows that the separator of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product and the battery cell assembly continues to be processed along the preset good product flow path; if the test result shows that the separator of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be an abnormal product and the battery cell assembly is picked up to the abnormal product processing path.

[0161] In the above-mentioned method for preparing a single battery cell, the positive electrode sheet, negative electrode sheet, and separator are wound to form a battery cell assembly, which is then pre-pressed and shaped. The shaped battery cell assembly is then transported to a visual inspection station and positioned there. Multiple images of the battery cell assembly from various perspectives are acquired. Based on these images, the condition of the separator is inspected to determine if wrinkling defects exist, resulting in an inspection result. If the separator condition conforms to a preset condition, the battery cell assembly is considered a good product and continues its subsequent preparation along a preset good product flow path. If the separator condition does not conform to the preset condition, the battery cell assembly is considered an abnormal product and is moved to the abnormal product processing path. This method achieves online, non-contact inspection of separator wrinkling defects, replacing inefficient manual visual inspection, improving inspection accuracy, and increasing inspection efficiency to match production line efficiency. This ensures that defective products are intercepted in real time, thereby improving the overall safety and consistency of the battery cells.

[0162] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0163] Based on the same inventive concept, this application also provides a battery cell preparation apparatus for implementing the above-described battery cell preparation method. The solution provided by this apparatus is similar to the solution described in the above method; therefore, the specific limitations of one or more battery cell preparation apparatus embodiments provided below can be found in the limitations of the battery cell preparation method described above, and will not be repeated here.

[0164] In one embodiment, such as Figure 15 As shown, a battery cell fabrication apparatus is provided, comprising: a winding and shaping module 10, a conveying module 11, an acquisition module 12, a detection module 13, and a determination module 14, wherein:

[0165] The winding and shaping module 10 is used to wind the positive electrode sheet, negative electrode sheet and separator to form a battery cell assembly and to pre-press and shape the battery cell.

[0166] The conveying module 11 is used to convey the shaped battery cell assembly to the vision inspection station and position the battery cell assembly at the vision inspection station.

[0167] The acquisition module 12 is used to acquire multiple perspective images of the battery cell assembly.

[0168] The detection module 13 is used to detect the state of the separator of the battery cell assembly based on multiple perspective images, determine whether there is a separator wrinkling defect in the battery cell assembly, and obtain the detection result of the battery cell assembly.

[0169] The determination module 14 is used to determine that the battery cell assembly is a good product when the detection result shows that the separator of the battery cell assembly meets the preset state, and the battery cell assembly continues to be processed along the preset good product flow path; when the detection result shows that the separator of the battery cell assembly does not meet the preset state, the battery cell assembly is determined to be an abnormal product, and the battery cell assembly is picked up to the abnormal product processing path.

[0170] In one embodiment, the visual inspection station is equipped with a photoelectric sensor and at least one positioning block; the aforementioned conveying module 11 includes: a generating unit and a driving unit, wherein:

[0171] The generation unit is used to generate a trigger signal and send it to the controller when the photoelectric sensor detects that the battery cell assembly has arrived at the vision inspection station.

[0172] A drive unit is used by the controller to drive at least one positioning block to define the position of the battery cell assembly in response to a trigger signal.

[0173] In one embodiment, the acquisition module 12 includes: an acquisition unit, configured to send a photo-taking command to at least two cameras in response to a trigger signal, and acquire images of at least one side of the battery cell assembly from different angles using the at least two cameras; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0174] In one embodiment, the detection module 13 includes: a synthesis unit, a first detection unit, and a second detection unit, wherein:

[0175] The synthesis unit is used to perform image synthesis processing on images from multiple perspectives to obtain a composite image of the battery cell assembly.

[0176] The first detection unit is used to perform edge detection on the synthesized image to obtain a binary image of the edge of the battery cell assembly.

[0177] The second detection unit is used to detect the membrane state of the battery cell assembly based on the edge binary image, determine whether there is a membrane wrinkling defect in the battery cell assembly, and obtain the detection result of the battery cell assembly.

[0178] In one embodiment, the second detection result is specifically used to determine the edge length and number of separator layers of the battery cell assembly based on the edge binary image; when both the edge length and the number of separator layers meet the preset good product conditions, the detection result of the battery cell assembly is determined based on the synthesized image and the preset detection model.

[0179] In one embodiment, the second detection result is specifically used to determine that the battery cell assembly meets the good product conditions when the edge length is greater than the length threshold and the number of membrane layers is greater than the number of layers threshold.

[0180] In one embodiment, the first detection result is specifically used to determine the tab region in the tab-side image based on the difference in pixel grayscale values ​​in the tab-side image; to cut the tab-side image according to the tab region to obtain the cut tab-side image; and to perform edge detection on the cut tab-side image and the non-tab-side image to obtain the edge binary image of the battery cell assembly.

[0181] In one embodiment, the above-mentioned battery cell manufacturing apparatus further includes a training module, used to control at least two cameras to take pictures of the cell assembly with defective separators, and acquire multiple sets of images of defective separators; to perform image synthesis processing on the multiple sets of images of defective separators to obtain a training dataset; and to train an initial deep learning model based on the training dataset to obtain a detection model.

[0182] Each module in the aforementioned battery cell fabrication apparatus can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0183] In one embodiment, a controller is provided, the internal structure of which can be shown in the following diagram: Figure 16 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for manufacturing a single battery cell. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0184] Those skilled in the art will understand that Figure 16 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0185] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0186] The positive electrode sheet, negative electrode sheet and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0187] The finalized battery cell assembly is transported to the vision inspection station and positioned there.

[0188] Acquire images of the battery cell assembly from multiple perspectives;

[0189] Based on images from multiple perspectives, the condition of the separator in the battery cell assembly is detected to determine whether there are separator wrinkling defects, and the detection results of the battery cell assembly are obtained.

[0190] If the test result shows that the separator of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product and continues to be manufactured along the preset good product flow path; if the test result shows that the separator of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product and is picked up to the defective product processing path.

[0191] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0192] The photoelectric sensor detects when the battery cell assembly arrives at the visual inspection station and generates a trigger signal to be sent to the controller.

[0193] In response to a trigger signal, the controller drives at least one positioning block to define the position of the battery cell assembly.

[0194] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0195] In response to a trigger signal, the controller sends a photo-taking command to at least two cameras to capture images of at least one side of the battery cell assembly from different angles; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0196] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0197] Image synthesis processing is performed on images from multiple perspectives to obtain a composite image of the battery cell assembly;

[0198] Edge detection is performed on the synthesized image to obtain a binary image of the edge of the battery cell assembly;

[0199] The membrane condition of the battery cell assembly is detected by edge binary image to determine whether there is membrane wrinkling defect and obtain the detection result of the battery cell assembly.

[0200] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0201] Based on the binary edge image, determine the edge length and number of separator layers of the battery cell assembly;

[0202] When the edge length and the number of separator layers both meet the preset good product conditions, the detection results of the battery cell assembly are determined based on the synthesized image and the preset detection model.

[0203] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0204] When the edge length is greater than the length threshold and the number of membrane layers is greater than the number of layers threshold, the battery cell assembly is determined to meet the good product conditions.

[0205] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0206] The electrode region in the electrode side image is determined based on the difference in pixel grayscale values ​​in the electrode side image;

[0207] The polar ear side image is segmented according to the polar ear region to obtain the segmented polar ear side image;

[0208] Edge detection is performed on the cut electrode side image and the non-electrode side image to obtain the edge binary image of the battery cell assembly.

[0209] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0210] Control at least two cameras to take pictures of the battery cell assembly with defective separators, and acquire multiple sets of images of the defective separators;

[0211] Image synthesis processing was performed on multiple sets of images of diaphragm defects to obtain a training dataset;

[0212] The initial deep learning model is trained based on the training dataset to obtain the detection model.

[0213] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0214] The positive electrode sheet, negative electrode sheet and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0215] The finalized battery cell assembly is transported to the vision inspection station and positioned there.

[0216] Acquire images of the battery cell assembly from multiple perspectives;

[0217] Based on images from multiple perspectives, the condition of the separator in the battery cell assembly is detected to determine whether there are separator wrinkling defects, and the detection results of the battery cell assembly are obtained.

[0218] If the test result shows that the separator of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product and continues to be manufactured along the preset good product flow path; if the test result shows that the separator of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product and is picked up to the defective product processing path.

[0219] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0220] The photoelectric sensor detects when the battery cell assembly arrives at the visual inspection station and generates a trigger signal to be sent to the controller.

[0221] In response to a trigger signal, the controller drives at least one positioning block to define the position of the battery cell assembly.

[0222] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0223] In response to a trigger signal, the controller sends a photo-taking command to at least two cameras to capture images of at least one side of the battery cell assembly from different angles; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0224] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0225] Image synthesis processing is performed on images from multiple perspectives to obtain a composite image of the battery cell assembly;

[0226] Edge detection is performed on the synthesized image to obtain a binary image of the edge of the battery cell assembly;

[0227] The membrane condition of the battery cell assembly is detected by edge binary image to determine whether there is membrane wrinkling defect and obtain the detection result of the battery cell assembly.

[0228] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0229] Based on the binary edge image, determine the edge length and number of separator layers of the battery cell assembly;

[0230] When the edge length and the number of separator layers both meet the preset good product conditions, the detection results of the battery cell assembly are determined based on the synthesized image and the preset detection model.

[0231] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0232] When the edge length is greater than the length threshold and the number of membrane layers is greater than the number of layers threshold, the battery cell assembly is determined to meet the good product conditions.

[0233] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0234] The electrode region in the electrode side image is determined based on the difference in pixel grayscale values ​​in the electrode side image;

[0235] The polar ear side image is segmented according to the polar ear region to obtain the segmented polar ear side image;

[0236] Edge detection is performed on the cut electrode side image and the non-electrode side image to obtain the edge binary image of the battery cell assembly.

[0237] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0238] Control at least two cameras to take pictures of the battery cell assembly with defective separators, and acquire multiple sets of images of the defective separators;

[0239] Image synthesis processing was performed on multiple sets of images of diaphragm defects to obtain a training dataset;

[0240] The initial deep learning model is trained based on the training dataset to obtain the detection model.

[0241] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0242] The positive electrode sheet, negative electrode sheet and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped.

[0243] The finalized battery cell assembly is transported to the vision inspection station and positioned there.

[0244] Acquire images of the battery cell assembly from multiple perspectives;

[0245] Based on images from multiple perspectives, the condition of the separator in the battery cell assembly is detected to determine whether there are separator wrinkling defects, and the detection results of the battery cell assembly are obtained.

[0246] If the test result shows that the separator of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product and continues to be manufactured along the preset good product flow path; if the test result shows that the separator of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product and is picked up to the defective product processing path.

[0247] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0248] The photoelectric sensor detects when the battery cell assembly arrives at the visual inspection station and generates a trigger signal to be sent to the controller.

[0249] In response to a trigger signal, the controller drives at least one positioning block to define the position of the battery cell assembly.

[0250] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0251] In response to a trigger signal, the controller sends a photo-taking command to at least two cameras to capture images of at least one side of the battery cell assembly from different angles; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

[0252] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0253] Image synthesis processing is performed on images from multiple perspectives to obtain a composite image of the battery cell assembly;

[0254] Edge detection is performed on the synthesized image to obtain a binary image of the edge of the battery cell assembly;

[0255] The membrane condition of the battery cell assembly is detected by edge binary image to determine whether there is membrane wrinkling defect and obtain the detection result of the battery cell assembly.

[0256] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0257] Based on the binary edge image, determine the edge length and number of separator layers of the battery cell assembly;

[0258] When the edge length and the number of separator layers both meet the preset good product conditions, the detection results of the battery cell assembly are determined based on the synthesized image and the preset detection model.

[0259] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0260] When the edge length is greater than the length threshold and the number of membrane layers is greater than the number of layers threshold, the battery cell assembly is determined to meet the good product conditions.

[0261] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0262] The electrode region in the electrode side image is determined based on the difference in pixel grayscale values ​​in the electrode side image;

[0263] The polar ear side image is segmented according to the polar ear region to obtain the segmented polar ear side image;

[0264] Edge detection is performed on the cut electrode side image and the non-electrode side image to obtain the edge binary image of the battery cell assembly.

[0265] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0266] Control at least two cameras to take pictures of the battery cell assembly with defective separators, and acquire multiple sets of images of the defective separators;

[0267] Image synthesis processing was performed on multiple sets of images of diaphragm defects to obtain a training dataset;

[0268] The initial deep learning model is trained based on the training dataset to obtain the detection model.

[0269] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0270] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0271] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for preparing a single battery cell, characterized in that, The method includes: The positive electrode sheet, negative electrode sheet, and separator are wound to form a battery cell assembly, and the battery cell assembly is pre-pressed and shaped. The finalized battery cell assembly is transported to the visual inspection station and positioned at the visual inspection station. Acquire multiple perspective images of the battery cell assembly; The multiple viewpoint images are combined to obtain a composite image of the battery cell assembly; Edge detection is performed on the synthesized image to obtain a binary image of the edge of the battery cell assembly; Based on the edge binary image, determine the edge length and the number of separator layers of the battery cell assembly; When both the edge length and the number of membrane layers meet the preset good product conditions, the detection result of the battery cell assembly is determined based on the synthesized image and the preset detection model. If the detection result indicates that the separator condition of the battery cell assembly meets the preset condition, the battery cell assembly is determined to be a good product, and the battery cell assembly continues to be manufactured along the preset good product flow path; if the detection result indicates that the separator condition of the battery cell assembly does not meet the preset condition, the battery cell assembly is determined to be a defective product, and the battery cell assembly is picked up and sent to the defective product processing path.

2. The method according to claim 1, characterized in that, The visual inspection station is equipped with a photoelectric sensor and at least one positioning block; positioning the battery cell assembly at the visual inspection station includes: The photoelectric sensor detects that the battery cell assembly has arrived at the visual inspection station and generates a trigger signal to be sent to the controller. In response to the trigger signal, the controller drives the at least one positioning block to define the position of the battery cell assembly.

3. The method according to claim 2, characterized in that, The acquisition of multiple perspective images of the battery cell assembly includes: In response to the trigger signal, the controller sends a photo-taking command to at least two cameras to capture images of at least one side of the battery cell assembly from different angles; the at least one side is at least one of the non-tab side and the tab side of the battery cell assembly.

4. The method according to claim 1, characterized in that, The method further includes: When the edge length is greater than a length threshold and the number of membrane layers is greater than a layer number threshold, the battery cell assembly is determined to meet the good product condition.

5. The method according to claim 1, characterized in that, The synthesized image includes a tab-side image and a non-tab-side image; the step of performing edge detection on the synthesized image to obtain a binary edge image of the battery cell assembly includes: The electrode region in the electrode side image is determined based on the difference in pixel grayscale values ​​in the electrode side image; The electrode side image is segmented according to the electrode region to obtain the segmented electrode side image; Edge detection is performed on the cut electrode side image and the non-electrode side image to obtain the edge binary image of the battery cell assembly.

6. The method according to any one of claims 1-4, characterized in that, The process of constructing the detection model includes: Control at least two cameras to take pictures of the battery cell assembly with defective separators, and acquire multiple sets of images of the defective separators; Image synthesis processing is performed on the multiple sets of images of diaphragm defects to obtain a training dataset; The initial deep learning model is trained based on the training dataset to obtain the detection model.

7. An apparatus for preparing a single battery cell, characterized in that, The apparatus includes a controller and a computer-readable storage medium, the controller including a memory and a processor, the computer-readable storage medium storing a computer program, the computer program in the computer-readable storage medium, when executed by the processor of the controller, implementing the method as described in any one of claims 1 to 6; the apparatus further includes: The winding and shaping module is used to wind the positive electrode sheet, the negative electrode sheet and the separator to form a battery cell assembly, and to pre-press and shape the battery cell. A conveying module is used to convey the shaped battery cell assembly to the visual inspection station and position the battery cell assembly at the visual inspection station. An acquisition module is used to acquire multiple perspective images of the battery cell assembly; The detection module is used to detect the membrane state of the battery cell assembly based on the multiple perspective images, determine whether the battery cell assembly has membrane wrinkling defects, and obtain the detection result of the battery cell assembly. The determination module is used to determine that the battery cell assembly is a good product when the detection result shows that the separator state of the battery cell assembly meets the preset state, and the battery cell assembly continues to be manufactured along the preset good product flow path; when the detection result shows that the separator state of the battery cell assembly does not meet the preset state, the battery cell assembly is determined to be an abnormal product, and the battery cell assembly is picked up and sent to the abnormal product processing path.

8. A single battery cell, characterized in that, The battery cell is prepared using the battery cell preparation method according to any one of claims 1 to 6.

Images