Solar cell visual positioning method and laser scribing method

Abstract

The application provides a solar cell visual positioning method and a laser scribing method. The method comprises the following steps: obtaining a first image and a second image of a cell; determining two side edges of the cell in an x direction according to the first image of the cell, so as to determine the positioning coordinates of the cell in the x direction; determining a candidate region according to the second image of the cell; obtaining the definition scores of all fine grids in the candidate region; determining k target fine grids according to the definition scores of the fine grids in the candidate region; and determining the positioning coordinates of the cell in a y direction according to the k target fine grids and the two side edges of the cell in the x direction. The obtained positioning coordinates of the cell have high precision and stability, and can effectively overcome the influence of elastic warping caused by placing the cell on a suction platform on the positioning precision of the cell. Laser scribing is performed according to the obtained positioning coordinates of the cell, so that the scribing track is accurately aligned with the grid gap, and the risk of hidden cracking and grid breaking is significantly reduced.

Description

Technical Field

[0001] This application belongs to the field of photovoltaic cell manufacturing technology, specifically, it relates to a visual positioning method and a laser scribing method for solar cells. Background Technology

[0002] In the laser scribing process of solar cells, a vision-guided positioning system is required to achieve high-precision scribing. Currently, a combination of backlight imaging and edge detection algorithms is commonly used for cell positioning. Specifically, the cell is placed on a movable adsorption platform, with a large-area uniform backlight positioned below the platform and a high-resolution industrial camera positioned above it. Under the illumination of the backlight, the camera captures images of the cell, and the cell outline appears as a high-contrast shadow in the image. Image processing is used to extract the edges of each side of the cell, fit straight lines, calculate geometric coordinates, and send the coordinates to a host computer. The host computer then drives the linear motor controlling the movement of the adsorption platform and the laser to complete the scribing of the cell.

[0003] like Figure 1 As shown, the adsorption platform consists of only six elongated adsorption platforms, with a certain distance between adjacent platforms. Adsorption holes are provided on the side of each platform that contacts the solar cell. After the solar cell is placed on the platform, the adsorption holes on each platform adsorb and fix it to the platform. For ultra-thin solar cells with a thickness reduced to below 130 μm, the areas not in contact with the adsorption platforms are prone to elastic warping, especially along the width direction (y-direction) of the adsorption platform. The edges of the solar cell located on both sides of the platform (denoted as the first edge and the second edge, respectively) are more susceptible to elastic warping due to being suspended. In actual production, without external force intervention, the average vertical distance between the highest point of the solar cell warping and the ideal horizontal plane can reach 128 μm. This deformation can cause local defocusing, shadow breaks, or positional shifts at the first and second edges in backlight imaging.

[0004] In existing methods for positioning solar cells, the y-coordinate needs to be determined by the midpoint between the first and second edges. When the first and second edges cannot be stably detected due to warping, the positioning error of the solar cell in the y-direction reaches ±0.15 mm, which is far higher than the ±0.05 mm tolerance required by the laser scribing process. This error will directly cause the laser scribing trajectory to deviate from the grid line gap, leading to grid breakage, microcracks, or loss of electrical performance, thereby affecting the yield. Summary of the Invention

[0005] In view of this, this application provides a visual positioning method and a laser scribing method for solar cells to solve the above problems.

[0006] A visual positioning method for solar cells, comprising:

[0007] Acquire a first image and a second image of the battery cell; wherein the first image is an outline image of the battery cell, and the second image is an image showing a plurality of fine grids extending along the x-direction, where the x-direction is the transport direction of the battery cell;

[0008] Based on the first image of the battery cell, determine the two sides of the battery cell in the x direction;

[0009] Based on the two sides of the battery cell in the x-direction, determine the positioning coordinates of the battery cell in the x-direction;

[0010] Based on the second image of the battery cell, a candidate region is determined; along the y-direction, the candidate region is located in the middle region of the battery cell, and the y-direction is a direction perpendicular to the x-direction in the same plane;

[0011] Obtain the sharpness score of all fine grids in the candidate region;

[0012] Based on the sharpness score of each fine grid in the candidate region, k target fine grids are determined;

[0013] Based on the k target fine grids and the two sides of the solar cell in the x direction, the positioning coordinates of the solar cell in the y direction are determined.

[0014] Preferably, acquiring the first and second images of the battery cell includes:

[0015] The first surface of the battery cell is illuminated by a first light source, and the middle region of the second surface of the battery cell along the y direction is illuminated by a second light source.

[0016] A visual imaging mechanism is used to acquire images of the battery cell from the second surface of the battery cell, thereby obtaining a first image and a second image of the battery cell;

[0017] The first surface and the second surface are two opposing surfaces of the battery cell, and the second surface of the battery cell has a plurality of fine grids extending along the x-direction.

[0018] Preferably, determining the positioning coordinates of the battery cell in the x-direction based on its two sides in the x-direction includes:

[0019] Linear fitting is performed on the two sides of the battery cell in the x direction to obtain lines L1 and L2;

[0020] Obtain the coordinates x1 of the midpoint of line L1 in the x-direction and the coordinates x2 of the midpoint of line L2 in the x-direction;

[0021] The average value of coordinates x1 and x2 is used as the positioning coordinate of the battery cell in the x-direction.

[0022] Preferably, obtaining the sharpness score of all fine grids in the candidate region includes:

[0023] Extract all fine gates in the candidate region to obtain a set of fine gates;

[0024] Obtain multiple sub-indices for each fine grid in the fine grid set;

[0025] The sharpness score of all fine grids in the candidate region is obtained by weighted fusion of multiple sub-indices for each fine grid.

[0026] Preferably, the step of extracting all fine grids within the candidate region to obtain a set of fine grids includes:

[0027] The candidate regions are preprocessed to enhance the features of each fine gate within the candidate regions;

[0028] Linear fitting is performed on each fine grating in the candidate region to obtain the linear equation of each fine grating in the candidate region, and the linear equation of each fine grating is used as the set of fine gratings.

[0029] Preferably, the preprocessing of the candidate region to enhance the features of each fine gate in the candidate region includes:

[0030] The candidate region is then subjected to Gaussian filtering for noise reduction.

[0031] Enhance the contrast of the candidate regions;

[0032] Morphological post-processing is performed on the candidate regions, with opening operations removing isolated noise points and closing operations connecting broken gate segments.

[0033] Preferably, obtaining multiple sub-indices for each fine gate in the fine gate set includes:

[0034] Calculate the average gradient magnitude of the region where each fine grating is located;

[0035] Calculate the effective consecutive pixel length of each fine gate;

[0036] Calculate the linear fitting residuals for each fine grid;

[0037] The average gradient magnitude, effective continuous pixel length, and linear fitting residual of each fine gate are used as sub-indices for each fine gate.

[0038] Preferably, the k target fine grids are the k fine grids with the highest sharpness scores in the candidate region, where k = 2 to 5.

[0039] Preferably, determining the positioning coordinates of the battery cell in the y-direction based on the k target fine grids and the two sides of the battery cell in the x-direction includes:

[0040] Obtain the first intersection point P between each target fine grid and the two sides of the solar cell in the x-direction. left (x) left y left ) and the second intersection point P right (x) right y right );

[0041] Set the y-coordinate of the first intersection point left The y-coordinate of the second intersection point right The average value is used as the candidate y value for the target fine gate;

[0042] The candidate y values ​​of each target fine grid are fused to obtain the positioning coordinates of the battery cell in the y direction.

[0043] Preferably, the step of fusing the y-candidate values ​​of each of the target fine grids to obtain the positioning coordinates of the battery cell in the y-direction includes:

[0044] Obtain the mean and standard deviation of the y-candidate values ​​for each of the target fine gates;

[0045] Based on the mean and standard deviation of the y-candidate values ​​of each target fine gate, retain the y-candidate values ​​that meet the preset conditions;

[0046] Based on the retained y-candidate values ​​and the corresponding sharpness scores of the target grating, a weighted fusion is performed to obtain the weighted average y-coordinate Y. final And use it as the positioning coordinate of the battery cell in the y direction.

[0047] In another aspect, the present invention provides a laser scribing method for solar cells, which uses the aforementioned visual positioning method for solar cells to obtain the positioning coordinates of the solar cells;

[0048] Based on the positioning coordinates of the battery cell in the x direction, the adsorption platform carrying the battery cell is driven to move in the x direction;

[0049] Based on the positioning coordinates of the battery cell in the y-direction, the position of the laser beam emitted by the laser on the battery cell is adjusted, and the laser cuts the battery cell along a preset cutting path; wherein, the preset cutting path is the direction of fine grid extension.

[0050] The beneficial effects of this application are as follows: First and second images of the solar cell are obtained. Based on the first image, two sides of the solar cell in the x-direction are determined, thereby determining the positioning coordinates of the solar cell in the x-direction. Based on the second image, candidate regions are determined, and then the sharpness scores of all fine grids in the candidate regions are obtained. Based on the sharpness scores of each fine grid in the candidate regions, k target fine grids are determined. Based on the k target fine grids and the two sides of the solar cell in the x-direction, the positioning coordinates of the solar cell in the y-direction are determined. The obtained positioning coordinates of the solar cell in both the x and y directions have high accuracy and stability, effectively overcoming the influence of elastic warping caused by the solar cell being placed on the adsorption platform on the positioning accuracy of the solar cell. Attached Figure Description

[0051] 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 or prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0052] The structures, proportions, sizes, etc., shown in the accompanying drawings are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the implementation conditions of this application. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size should still fall within the scope of the technical content disclosed in this application, provided that they do not affect the effects and purposes that this application can produce.

[0053] Figure 1 This is a schematic diagram of the battery cell provided in this application placed on the adsorption platform;

[0054] Figure 2 Another schematic diagram showing the placement of the battery cell provided in this application on the adsorption platform;

[0055] Figure 3 A flowchart illustrating a visual positioning method for solar cells provided in this application;

[0056] Figure 4 A schematic diagram of a process for obtaining a first image and a second image of a battery cell provided in this application;

[0057] Figure 5 This application provides a flowchart illustrating how to determine the positioning coordinates of a battery cell in the x-direction based on its two sides in the x-direction.

[0058] Figure 6A schematic diagram of a process for obtaining the sharpness score of all fine grids in a candidate region, provided for this application;

[0059] Figure 7 A schematic diagram of a process for extracting all fine grids within a candidate region to obtain a set of fine grids, as provided in this application;

[0060] Figure 8 This application provides a flowchart illustrating how to determine the positioning coordinates of a solar cell in the y direction based on k target grids and two side edges of the solar cell in the x direction.

[0061] Figure 9 This application provides a schematic diagram of a process for fusing candidate y values ​​of a target fine gate. Detailed Implementation

[0062] The embodiments of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0063] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] In the typical laser scribing process for solar cells (hereinafter referred to as cells), the cells are adsorbed and fixed onto an adsorption platform. The adsorption platform is driven by a driving mechanism (such as a cylinder or other mechanism that drives objects to move linearly) to move back and forth in the x-direction (the direction of cell transport). The adsorption platform carries the cells to a vision positioning station, where a large-area uniform backlight located below the adsorption platform illuminates the cells. A vision imaging mechanism (such as a camera) located above the adsorption platform captures an image of the cells. Then, the edges of each side of the cells are extracted from the image, and straight lines are fitted to each side edge. The positioning coordinates of the cells are obtained based on the straight line equations of each side edge.

[0065] like Figure 1-2 As shown, the adsorption platform includes multiple elongated adsorption platforms 1 spaced apart along the y-direction. The length direction of the adsorption platform 1 is parallel to the x-direction. The side of the adsorption platform 1 that contacts the battery cell 2 is provided with several adsorption holes, through which the battery cell 2 is adsorbed and fixed onto the adsorption platform 1. The y-direction is a direction perpendicular to the x-direction in the same plane.

[0066] For ultrathin solar cells with a thickness reduced to below 130 μm, the areas not in contact with the adsorption mesa are highly susceptible to elastic warping, especially along the y-direction. The edges of the solar cells located on both sides of the adsorption mesa (denoted as the first edge 21 and the second edge 22, respectively) are more prone to elastic warping due to their suspension (see reference). Figure 2 ).

[0067] Because the positioning coordinates of the solar cell in the y-direction need to be determined by the midpoint between the first and second edges, when the first and second edges cannot be stably detected due to warping, the positioning error of the solar cell in the y-direction reaches ±0.15 mm, which is far higher than the ±0.05 mm tolerance required by the laser scribing process. This error will directly cause the laser scribing trajectory to deviate from the grid line gap, leading to grid breakage, microcracks, or loss of electrical performance, thereby affecting the yield.

[0068] Therefore, this application proposes a visual positioning method for solar cells, referring to... Figure 3 ,include:

[0069] S1: Obtain the first and second images of the battery cell.

[0070] The first image is a contour image of the battery cell, and the second image is an image showing several fine grids extending along the x-direction, which is the transport direction of the battery cell.

[0071] S2: Based on the first image of the battery cell, determine the two sides of the battery cell in the x direction.

[0072] like Figure 1 As shown, the two sides of the battery cell 2 in the y direction are the first edge 21 and the second edge 22, respectively, and the two sides of the battery cell 2 in the x direction are the third edge 23 and the fourth edge 24, respectively. The second surface of the battery cell has a plurality of fine grids 25 extending along the x direction.

[0073] Based on the first image of the battery cell, which is a contour image of the battery cell, the third and fourth edges of the battery cell in the first image can be extracted.

[0074] For example, a subpixel edge detection algorithm can be used to extract the third and fourth edges of the battery cells in the first image.

[0075] S3: Determine the positioning coordinates of the battery cell in the x-direction based on the two sides of the battery cell in the x-direction.

[0076] The adsorption platforms supporting the solar cells are spaced apart along the y-direction. After the solar cells are placed on the adsorption platforms, the first and second edges of the solar cells are prone to elastic warping, but the third and fourth edges are less affected. Therefore, the positioning coordinates of the solar cells in the x-direction can be determined based on the third and fourth edges of the solar cells in the x-direction.

[0077] S4: Determine candidate regions based on the second image of the battery cell.

[0078] This embodiment selects a region from the second image of the battery cell as a candidate region for subsequent steps. Typically, the region with the clearest image in the second image of the battery cell is selected as the candidate region. In this embodiment, the middle region of the battery cell is selected along the y-direction as the candidate region to facilitate subsequent extraction of the fine grid and determination of the battery cell's positioning coordinates in the y-direction. Further details will be provided later.

[0079] S5: Obtain the sharpness score of all fine grids in the candidate region.

[0080] After determining the candidate regions, the sharpness score of each fine grid line in the candidate regions is calculated.

[0081] S6: Determine k target fine grids based on the sharpness score of each fine grid in the candidate region.

[0082] After obtaining the sharpness scores of each fine grid in the candidate region, select k fine grids with the highest sharpness scores and use them as target fine grids. Where k = 2~5.

[0083] In this embodiment, only 2 to 5 fine grids with the highest clarity scores need to be retained, and all remaining fine grids in the candidate area are removed, which improves the efficiency of subsequent processing and avoids the use of blurry or broken fine grids that affect the positioning accuracy of the solar cell in the y direction.

[0084] S7: Determine the positioning coordinates of the battery cell in the y direction based on the k target grids and the two sides of the battery cell in the x direction.

[0085] After selecting k target grids from the candidate region, the positioning coordinates of the solar cell in the y direction are calculated by combining the third and fourth edges of the solar cell.

[0086] This application embodiment acquires a first image and a second image of the battery cell. Based on the first image, two sides of the battery cell in the x-direction are determined, and the positioning coordinates of the battery cell in the x-direction are determined based on these two sides. Based on the second image, candidate regions are determined, and the sharpness scores of all fine grids in the candidate regions are obtained. Based on the sharpness scores of each fine grid in the candidate regions, k target fine grids are determined. Based on the k target fine grids and the two sides of the battery cell in the x-direction, the positioning coordinates of the battery cell in the y-direction are determined. The obtained positioning coordinates of the battery cell in both the x- and y-directions have high accuracy and stability, effectively overcoming the influence of elastic warping caused by the battery cell being placed on the adsorption platform on the positioning accuracy of the battery cell. When the first and second edges of the battery cell cannot be effectively detected due to warping, this application embodiment can obtain stable and accurate positioning coordinates of the battery cell in the y-direction through the k target fine grids in the middle region of the battery cell and the third and fourth edges of the battery cell. The positioning error of the obtained positioning coordinates of the battery cell can be controlled within ±0.02 mm.

[0087] In addition, selecting k fine grids with the highest clarity scores from the candidate area as target fine grids and eliminating blurry or broken fine grids can address the interference of industrial field problems such as partial occlusion of fine grids, stains, or uneven lighting on cell positioning, demonstrating high robustness and adaptability.

[0088] The solar cell visual positioning method of this embodiment can complete the entire algorithm execution process within 250 ms, which can adapt to the current production line's requirements for production cycle time and meet the needs of high-speed cycle time.

[0089] Figure 4 The diagram illustrates a process for acquiring a first and second image of a battery cell, specifically including:

[0090] S11: The first surface of the battery cell is illuminated by a first light source, and the middle area of ​​the second surface of the battery cell along the y direction is illuminated by a second light source.

[0091] The first surface and the second surface are two opposing surfaces of the battery cell. For example, the first surface can be the side of the battery cell that contacts the adsorption platform, i.e., the first surface is the lower surface of the battery cell, and the first light source can be positioned below the adsorption platform, illuminating the first surface of the battery cell from below the adsorption platform upwards. The second surface can be the side of the battery cell that does not contact the adsorption platform, i.e., the second surface is the upper surface of the battery cell, and the second light source can be positioned above the adsorption platform, illuminating the second surface of the battery cell from above the adsorption platform downwards.

[0092] In a preferred embodiment, the second light source is positioned at an angle of 30-60° above the second surface of the battery cell and illuminates the middle region of the second surface of the battery cell along the y-direction.

[0093] S12: Use a vision imaging mechanism to acquire images of the battery cell from the second surface of the battery cell, and obtain a first image and a second image of the battery cell.

[0094] In this embodiment, the visual imaging mechanism and the second light source are located on the same side of the battery cell. Specifically, both the visual imaging mechanism and the second light source are positioned above the second surface of the battery cell.

[0095] With the first light source illuminating the first surface of the solar cell, the visual imaging mechanism can obtain an outline image of the solar cell from above the second surface. Understandably, in order for the visual imaging mechanism to obtain a clear outline image of the solar cell, the illumination range of the first light source should cover the entire first surface of the solar cell.

[0096] like Figure 1 As shown, the second surface of the battery cell 2 in this embodiment has a plurality of fine grids 25 extending in the x-direction and parallel to each other. When the second surface of the battery cell is illuminated by a second light source, the second image obtained by the visual imaging mechanism shows the plurality of fine grids extending in the x-direction.

[0097] The solar cell is placed on the adsorption platform. Along the y-direction, the central region of the solar cell is least affected by elastic warping. Therefore, in this embodiment, the second light source illuminates the central region of the second surface of the solar cell. More preferably, the second light source illuminates the central region of the second surface of the solar cell at an angle of 30-60° from above, so that the central region of the second image obtained by the visual imaging mechanism is the clearest. Then, the selected candidate region is located in the central region of the solar cell for subsequent extraction of the fine grid.

[0098] In a preferred embodiment, to improve image acquisition efficiency, typically with a first light source illuminating the first surface of the battery cell and a second light source illuminating the middle region of the second surface of the battery cell along the y-direction, the visual imaging mechanism acquires an image of the battery cell from above the second surface, obtaining a first image and a second image of the battery cell. It is understood that the visual imaging mechanism acquires an image that includes both the outline image of the battery cell and an image showing the battery cell with several fine grids extending along the x-direction. Specifically, the outline of this image is used as the outline image of the battery cell, i.e., as the first image of the battery cell, and the area of ​​this image excluding the outline (i.e., the image showing several fine grids extending along the x-direction) is used as the second image of the battery cell.

[0099] Since the second light source only illuminates the middle area of ​​the second surface of the solar cell along the y-direction, the second light source does not affect the outline imaging of the solar cell. That is, the image acquired by the vision imaging mechanism can present a clear outline of the solar cell while making the imaging of the middle area of ​​the second surface of the solar cell along the y-direction clear.

[0100] The second light source in this embodiment can be a point light source, a strip LED array, a coaxial light source, or a ring light source, etc. This embodiment does not make specific limitations, as long as the reflective features of each fine grid on the second surface of the battery cell can be highlighted under the illumination of the second light source, so that each fine grid is clearly imaged and the outline imaging of the battery cell is not affected.

[0101] Figure 5 The diagram illustrates a process for determining the positioning coordinates of a solar cell in the x-direction based on its two sides in the x-direction, specifically including:

[0102] S31: Perform linear fitting on the two sides of the battery cell in the x direction to obtain straight lines L1 and L2.

[0103] After extracting the third and fourth edges of the battery cell from the first image, straight lines are fitted to the third and fourth edges respectively to obtain straight lines L1 and L2.

[0104] S32: Get the coordinates x1 of the midpoint of line L1 in the x-direction and x2 of the midpoint of line L2 in the x-direction.

[0105] This embodiment can use existing image processing algorithms to determine the coordinates (x1, y1) of the midpoint of line L1 and the coordinates (x2, y2) of the midpoint of line L2.

[0106] S33: The average value of coordinates x1 and x2 is used as the positioning coordinate of the battery cell in the x direction.

[0107] Since the third and fourth edges of the solar cell are less affected by elastic warping, the solar cell's positioning coordinates in the x-direction can be obtained by fitting a straight line to the third and fourth edges and calculating the average of the x-coordinates of the midpoints of the straight lines L1 and L2.

[0108] Figure 6 A flowchart illustrating a process for obtaining sharpness scores for all fine rasteres in a candidate region is shown, specifically including:

[0109] S51: Extract all fine grids in the candidate region to obtain a set of fine grids.

[0110] For example, each fine grid detected in the candidate region can be fitted with a straight line to obtain the corresponding straight line parameters, thereby forming a set of fine grids. Alternatively, other image processing methods capable of identifying parallel linear structures in an image can be used to obtain the set of fine grids, such as template matching, frequency domain analysis, Hough transform, least squares fitting, or semantic segmentation based on convolutional neural networks (CNNs).

[0111] In a preferred embodiment, this example obtains a set of fine grids by performing linear fitting on each fine grid in the candidate region.

[0112] Figure 7 This diagram illustrates a process for extracting all fine gates within a candidate region to obtain a set of fine gates, specifically including:

[0113] S511: Preprocess the candidate region to enhance the features of each fine gate in the candidate region.

[0114] In one implementation, the candidate region is preprocessed by performing Gaussian filtering to denoise the candidate region, then enhancing the contrast of the candidate region, and finally performing morphological postprocessing on the candidate region, including opening operations to remove isolated noise points and closing operations to connect broken gate segments, thereby enhancing the features of each fine gate in the candidate region so as to obtain a set of fine gates in the subsequent process.

[0115] When performing Gaussian filtering to denoise the candidate region, a 3*3 Gaussian kernel can be used to suppress noise.

[0116] To enhance the contrast of the candidate region, adaptive histogram equalization can be used to enhance the grayscale difference between the fine grid and the background in the candidate region.

[0117] S512: Perform linear fitting on each fine grid in the candidate region to obtain the linear equation of each fine grid in the candidate region, and use the linear equation of each fine grid as the set of fine grids.

[0118] This embodiment can determine multiple sampling points for each fine grid in the candidate region, and then use the least squares method to fit the optimal straight line G for the multiple sampling points of each fine grid. i The equation of the straight line is: y = k i x+b i .

[0119] Based on minimizing the sum of squared residuals in the vertical direction: It can be obtained , , will k i and b i Substituting the equations into the formula for the linear equation above, the linear equation of the fine grating can be obtained. By performing the above process on each fine grating in the candidate region, a set of fine gratings can be obtained.

[0120] S52: Obtain multiple sub-indices for each fine grid in the fine grid set.

[0121] In this embodiment, each fine grid line may have multiple sub-indices, including: average gradient magnitude, effective continuous pixel length, and linear fitting residual. The specific process may include:

[0122] Calculate the average gradient magnitude of the region where each fine grating is located;

[0123] Calculate the effective consecutive pixel length of each fine gate;

[0124] Calculate the linear fitting residuals for each fine grid;

[0125] The average gradient magnitude, effective continuous pixel length, and linear fitting residual of each fine gate are used as sub-indices for each fine gate.

[0126] This embodiment does not impose specific restrictions on the order of calculating the average gradient magnitude, the effective continuous pixel length, and the linear fitting residual.

[0127] In this embodiment, the average gradient magnitude of each fine grid region is calculated. This can be achieved by calculating the Sobel gradient magnitude map of the original fine grid image and then determining the fine grid fitting line G. i The region in which it is located (e.g., the fitted line G of the fine grid) i The gradient magnitude of all pixels in a strip region (within ±2 pixels) is extracted, and the average gradient magnitude of all pixels in the region is calculated to obtain the average gradient magnitude E corresponding to the fine gate. i .

[0128] The original fine grid image is a candidate region in the second image obtained under the illumination of the second light source, and this candidate region has not undergone the aforementioned preprocessing.

[0129] The average gradient magnitude can reflect the contrast between the fine grid and the background. The larger the average gradient magnitude, the sharper the edge of the fine grid and the higher the positioning reliability.

[0130] In this embodiment, the effective continuous pixel length of each fine grid can be calculated by fitting a straight line G along the fine grid. i Based on the projection direction, connected component analysis is performed on the binarized candidate region image to calculate the length L of the longest consecutive bright pixel segment. i (Unit: pixels), then normalization is performed to obtain the effective continuous pixel length of the fine grid: , where W is the width of the second image.

[0131] By calculating the effective continuous pixel length of each fine grid, it is possible to avoid using broken or partially missing fine grids as target fine grids, which would affect the positioning accuracy of the solar cell in the y direction.

[0132] In this embodiment, the linear fitting residuals of each fine grid can be calculated by recording the distance from each interior point to the fitted line G during the process of obtaining the linear equations of each fine grid by performing linear fitting on each fine grid in the candidate region. i The vertical distance is calculated, and then the distance from each interior point to the fitted line G is calculated. i The root mean square error (RMSE) of the perpendicular distance: , where N i Let d be the number of interior points. ij For the j-th point, the fitted line G i The vertical distance.

[0133] By calculating the linear fitting residuals of each fine grid, we can measure whether the grid is close to an ideal straight line. Understandably, warping or stains will increase the linear fitting residuals.

[0134] S53: Weighted fusion of multiple sub-indices for each fine grid is performed to obtain the sharpness score of all fine grids in the candidate region.

[0135] Specifically, the mean gradient magnitude, effective consecutive pixel length, and linear fitting residual corresponding to each fine grid can be normalized to the [0, 1] interval, and then the sharpness score of the corresponding fine grid can be obtained by weighted summation. The calculation formula is as follows: , where w1+w2+w3=1.

[0136] In this embodiment, during the calculation of the sharpness score for each fine grid, the linear fitting residual term is inverted (1-norm(R)). i This is to unify the scoring direction and make the indicators more positive.

[0137] Figure 8 The diagram illustrates a process for determining the positioning coordinates of a solar cell in the y-direction based on k target grids and two side edges of the cell in the x-direction. Specifically, it includes:

[0138] S71: Obtain the first intersection point P between each target grid and the two sides of the solar cell in the x-direction. left (x) left y left ) and the second intersection point P right (x) right y right ).

[0139] Specifically, the first intersection point P is obtained by finding the intersection point between each target grating and the straight line L1. left (x) left y left ), and find the second intersection point P for each target grating and the straight line L2. right (x)right y right ).

[0140] S72: Set the y-coordinate of the first intersection point... left The y-coordinate of the second intersection point right The average value is used as the candidate value of y for the target fine gate.

[0141] Specifically, the candidate values ​​Y of the target fine gate. i The calculation formula is: Y i =(y left + y right ) / 2.

[0142] S73: The candidate y values ​​of the target fine grid are fused to obtain the positioning coordinates of the cell in the y direction.

[0143] This embodiment performs fusion processing on the candidate y values ​​of the target fine gate. This fusion processing can be performed using the median or weighted average, or it can be performed using RANSAC (Random Sample Consensus Algorithm) to remove outlier intersections, least squares optimization, or Kalman filtering. This embodiment does not impose any specific limitations.

[0144] For example, Figure 9 This diagram illustrates a process for fusing candidate y-values ​​of a target fine gate, specifically including:

[0145] S731: Obtain the mean and standard deviation of the y candidate values ​​for each target fine gate;

[0146] S732: Based on the mean and standard deviation of the y candidate values ​​of each target fine grid, retain the y candidate values ​​that meet the preset conditions;

[0147] S733: Based on the retained y-candidate values ​​and the corresponding sharpness scores of the target raster, a weighted fusion is performed to obtain the weighted average y-coordinate Y. final And use it as the positioning coordinate of the battery cell in the y direction.

[0148] In this embodiment, the y-candidate values ​​of each target fine gate are fused in two stages to improve stability.

[0149] First, a first-level fusion process is performed to remove outliers. Specifically, candidate values ​​Yi for each target fine gate are calculated. i The mean μY and standard deviation σY are then used to determine the condition |Y|. i Outliers with -μY|>2σY are removed, and the remaining non-outliers are retained, i.e., those satisfying |Y i Candidate values ​​for y with respect to -μY|≤2σY.

[0150] Then, weighted fusion is performed. Specifically, based on the retained y-candidate values ​​and the corresponding sharpness score S of the target raster,... i Calculate the weighted average y-coordinate Y final The calculation formula is as follows:

[0151]

[0152] Where I is the set of fine-gate indexes that have passed the validity check and have not been removed.

[0153] The calculated weighted average y-coordinate Y final This refers to the positioning coordinates of the battery cell in the y-direction.

[0154] The positioning coordinates of the solar cell in the y-direction obtained by the above method can effectively avoid the influence of solar cell warping on the positioning of the solar cell in the y-direction, and obtain accurate and stable positioning coordinates of the solar cell.

[0155] This application embodiment only requires the addition of a second light source, without the need for large-scale modification of the existing visual positioning device. The original adsorption platform, visual imaging mechanism and first light source can be retained, which has the advantages of low cost and easy integration.

[0156] It is understood that the solar cell visual positioning method of this application embodiment can be applied not only to laser scribing, but also to other photovoltaic processes that require high-precision positioning, such as the alignment of the solder ribbon and the cell during laser film opening, laser doping, or string welding.

[0157] In another embodiment of this application, a laser scribing method is provided, which uses the aforementioned visual positioning method to obtain the positioning coordinates of the battery cell.

[0158] Then, based on the positioning coordinates of the solar cell in the x-direction, the drive mechanism drives the adsorption platform carrying the solar cell to move in the x-direction, so that the solar cell passes under the laser. Simultaneously, based on the positioning coordinates of the solar cell in the y-direction, the position of the laser beam emitted by the laser on the solar cell is adjusted. That is, the laser interpolates based on the positioning coordinates of the solar cell in the y-direction, causing the laser to cut the solar cell along a preset cutting path. The preset cutting path is the direction of the fine grid extension.

[0159] By using the aforementioned positioning method, accurate and stable cell positioning coordinates are obtained. Then, based on the cell positioning coordinates, the adsorption platform moves in the x-direction, and the laser interpolates in the y-direction to ensure that the laser beam emitted by the laser cuts the cell along the pre-cutting path of the cell. This makes the laser scribing trajectory accurately aligned with the fine grid gap, significantly reducing problems such as microcracks and breakage, and improving the consistency of the cell's electrical performance.

[0160] The various embodiments in this specification are described in a progressive, parallel, or combined manner. Each embodiment focuses on its differences from other embodiments, and similar or identical parts between embodiments can be referred to interchangeably. For the apparatuses disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0161] It should be noted that, in the description of this application, the terms "upper," "lower," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. When a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be a component centrally located at the same time.

[0162] It should also be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or apparatus comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or apparatus that includes the aforementioned element.

[0163] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A visual positioning method for solar cells, characterized in that, include: Acquire a first image and a second image of the battery cell; wherein, the first image is an outline image of the battery cell, and the second image is an image showing a plurality of fine grids extending along the x-direction, where the x-direction is the transport direction of the battery cell; Based on the first image of the battery cell, determine the two sides of the battery cell in the x direction; Based on the two sides of the battery cell in the x-direction, determine the positioning coordinates of the battery cell in the x-direction; Based on the second image of the battery cell, a candidate region is determined; along the y-direction, the candidate region is located in the middle region of the battery cell, and the y-direction is a direction perpendicular to the x-direction in the same plane; Obtain the sharpness score of all fine grids in the candidate region; Based on the sharpness score of each fine grid in the candidate region, k target fine grids are determined; Based on the k target fine grids and the two sides of the solar cell in the x direction, the positioning coordinates of the solar cell in the y direction are determined.

2. The visual positioning method for solar cells according to claim 1, characterized in that, The process of acquiring the first and second images of the battery cell includes: The first surface of the battery cell is illuminated by a first light source, and the middle region of the second surface of the battery cell along the y direction is illuminated by a second light source. A visual imaging mechanism is used to acquire images of the battery cell from the second surface of the battery cell, thereby obtaining a first image and a second image of the battery cell; The first surface and the second surface are two opposing surfaces of the battery cell, and the second surface of the battery cell has a plurality of fine grids extending along the x-direction.

3. The visual positioning method for solar cells according to claim 1, characterized in that, Determining the positioning coordinates of the battery cell in the x-direction based on its two sides in the x-direction includes: Linear fitting is performed on the two sides of the battery cell in the x direction to obtain lines L1 and L2; Obtain the coordinates x1 of the midpoint of line L1 in the x-direction and the coordinates x2 of the midpoint of line L2 in the x-direction; The average value of coordinates x1 and x2 is used as the positioning coordinate of the battery cell in the x-direction.

4. The visual positioning method for solar cells according to any one of claims 1-3, characterized in that, The process of obtaining the sharpness score of all fine grids in the candidate region includes: Extract all fine gates in the candidate region to obtain a set of fine gates; Obtain multiple sub-indices for each fine grid in the fine grid set; The sharpness score of all fine grids in the candidate region is obtained by weighted fusion of multiple sub-indices for each fine grid.

5. The visual positioning method for solar cells according to claim 4, characterized in that, The step of extracting all fine gates within the candidate region to obtain a set of fine gates includes: The candidate regions are preprocessed to enhance the features of each fine gate within the candidate regions; Linear fitting is performed on each fine grating in the candidate region to obtain the linear equation of each fine grating in the candidate region, and the linear equation of each fine grating is used as the set of fine gratings.

6. The visual positioning method for solar cells according to claim 5, characterized in that, The preprocessing of the candidate region to enhance the features of each fine gate in the candidate region includes: The candidate region is then subjected to Gaussian filtering for noise reduction. Enhance the contrast of the candidate regions; Morphological post-processing is performed on the candidate regions, with opening operations removing isolated noise points and closing operations connecting broken gate segments.

7. The visual positioning method for solar cells according to claim 5, characterized in that, The step of obtaining multiple sub-indices for each fine gate in the fine gate set includes: Calculate the average gradient magnitude of the region where each fine grating is located; Calculate the effective consecutive pixel length of each fine gate; Calculate the linear fitting residuals for each fine grid; The average gradient magnitude, effective continuous pixel length, and linear fitting residual of each fine gate are used as sub-indices for each fine gate.

8. The visual positioning method for solar cells according to claim 1, characterized in that, The k target grids are the k grids with the highest sharpness scores in the candidate region, where k = 2 to 5.

9. The visual positioning method for solar cells according to claim 1, characterized in that, The step of determining the positioning coordinates of the battery cell in the y-direction based on the k target fine grids and the two sides of the battery cell in the x-direction includes: Obtain the first intersection point P between each target fine grid and the two sides of the solar cell in the x-direction. left (x) left y left ) and the second intersection point P right (x) right y right ); Set the y-coordinate of the first intersection point left The y-coordinate of the second intersection point right The average value is used as the candidate y value for the target fine gate; The candidate y values ​​of each target fine grid are fused to obtain the positioning coordinates of the battery cell in the y direction.

10. The visual positioning method for solar cells according to claim 9, characterized in that, The step of fusing the y-candidate values ​​of each of the target fine grids to obtain the positioning coordinates of the battery cell in the y-direction includes: Obtain the mean and standard deviation of the y-candidate values ​​for each of the target fine gates; Based on the mean and standard deviation of the y-candidate values ​​of each target fine gate, retain the y-candidate values ​​that meet the preset conditions; Based on the retained y-candidate values ​​and the corresponding sharpness scores of the target grating, a weighted fusion is performed to obtain the weighted average y-coordinate Y. final And use it as the positioning coordinate of the battery cell in the y direction.

11. A method for laser scribing solar cells, characterized in that, The positioning coordinates of the solar cell are obtained by using the visual positioning method of any one of claims 1-10. Based on the positioning coordinates of the battery cell in the x direction, the adsorption platform carrying the battery cell is driven to move in the x direction; Based on the positioning coordinates of the battery cell in the y-direction, the position of the laser beam emitted by the laser on the battery cell is adjusted, and the laser cuts the battery cell along a preset cutting path; wherein, the preset cutting path is the direction of fine grid extension.

Images