A method and system for testing the thermal shrinkage of a battery separator

Abstract

The application relates to a method and system for testing the thermal shrinkage of a battery diaphragm, and relates to the field of testing or analyzing materials by measuring the chemical or physical properties of the materials, and the method comprises the following steps: collecting a backlight image of a diaphragm sample before heating, extracting contour coordinates to calculate the area before heating; obtaining a backlight image, a top light image and a side light image of the sample after heating, which are respectively used for determining an edge point set, identifying a wrinkle area and obtaining a three-dimensional depth feature; dividing the edge points into flat edge sections and wrinkle edge sections based on the gray scale variance of the top light image; accumulating the Euclidean distance of adjacent points of the flat edge sections to obtain a flat measurement length, and calculating a wrinkle correction length of the wrinkle edge sections according to the three-dimensional depth feature and a flattening correction coefficient; splicing the lengths of the sections in a spatial sequence to obtain a real edge length after heating, and calculating the area after heating; and calculating the thermal shrinkage according to the area before and after heating. By implementing the method, the calculation deviation of the thermal shrinkage of the battery diaphragm can be reduced.

Description

Technical Field

[0001] This application relates to the field of testing or analyzing materials by measuring their chemical or physical properties, and more particularly to a method and system for testing the thermal shrinkage rate of a battery separator. Background Technology

[0002] The separator in a lithium-ion battery acts as a physical barrier between the positive and negative electrodes, preventing short circuits. The separator thickness is typically less than 10 micrometers. Under high temperatures, it undergoes thermal shrinkage. Excessive thermal shrinkage can cause the positive and negative electrodes to come into contact, leading to a short circuit and a safety hazard. Therefore, accurately testing the thermal shrinkage rate of the battery separator at different temperatures is crucial for ensuring battery safety.

[0003] Currently, the thermal shrinkage rate of battery separators is tested using a combination of oven heating and dimensional measurement. First, the length and width of the separator sample are measured to calculate the initial area. Then, the separator is clamped and fixed with a glass plate and placed in an oven for heating for several hours. After heating, it is removed and cooled, and the dimensions are measured again to calculate the area after heating. Finally, the thermal shrinkage rate is calculated based on the change in area.

[0004] However, after the diaphragm is heated to a high temperature, irregular wrinkles and warping will appear at the edges, making it difficult to determine a uniform benchmark during measurement, and resulting in poor repeatability of multiple measurements on the same sample. There is a difference between the actual unfolded length of the wrinkled area and the projected length in the plane, while conventional measurements can only obtain the projected length, leading to deviations in the calculation of the heat shrinkage rate. Summary of the Invention

[0005] This application provides a method and system for testing the thermal shrinkage rate of battery separators, which can reduce the calculation deviation of the thermal shrinkage rate of battery separators.

[0006] In a first aspect, this application provides a method for testing the thermal shrinkage rate of a battery separator, applied to a battery separator thermal shrinkage rate testing system. The method includes: acquiring a backlight image of the separator sample before heating; extracting initial contour coordinates from high-contrast edges in the backlight image; calculating the area before heating based on the initial contour coordinates; sequentially acquiring a backlight image, a top-light image, and a side-light image of the heated separator sample, wherein the backlight image is used to determine the set of edge points of the overall contour after heating, the top-light image is used to identify wrinkled regions in the edge point set, and the side-light image is used to obtain the three-dimensional depth features of the wrinkled regions; dividing the edge point set into smooth edge segments and wrinkled edge segments based on the local grayscale variance of the top-light image, wherein the smooth edge segments... The grayscale variance is below a preset threshold and the edge direction is continuous. The grayscale variance of the wrinkled edge segment is above a preset threshold and there are periodic changes in brightness. For the smooth edge segment, the Euclidean distance between adjacent edge points is calculated and accumulated to obtain the smooth measurement length. For the wrinkled edge segment, the wrinkle height parameter is calculated based on the three-dimensional depth features. The wrinkle correction length is calculated based on the projected measurement length of the wrinkled edge segment and the flattening correction coefficient based on the wrinkle height parameter. The smooth measurement length of each smooth edge segment and the wrinkle correction length of each wrinkled edge segment are spliced ​​together according to their spatial position in the overall contour to obtain the true side length of the four sides after heating. The area after heating is calculated based on the true side length. The thermal shrinkage rate is calculated based on the area before heating and the area after heating.

[0007] In the above embodiments, the system acquires complete morphological information of the diaphragm sample through multi-source collaborative imaging. Backlighting provides clear contour boundaries, top-lighting reveals surface wrinkle texture features, and side-lighting supplements depth dimension data. Based on local grayscale variance feature values, the system automatically identifies flat and wrinkled areas. A three-dimensional depth correction coefficient is introduced to the wrinkled edge segments to compensate for projection measurement errors. The corrected edge segment lengths are then stitched together in spatial order to calculate the true side length, thereby accurately obtaining the actual area after heating and improving the accuracy of thermal shrinkage rate measurement.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of dividing the set of edge points into smooth edge segments and wrinkled edge segments based on the local gray-level variance of the top-lit image specifically includes: for each edge point in the set of edge points, extracting the gray-level value within a preset window range centered on the edge point in the top-lit image, and calculating the gray-level variance of all pixels within the preset window range as a local gray-level variance feature value; comparing the local gray-level variance feature value with a preset gray-level threshold, and merging the edge point sequence with a local gray-level variance feature value lower than the preset gray-level threshold and spatially continuous into a smooth edge segment; performing a frequency domain transformation of the gray-level value sequence on the region corresponding to the edge points with a local gray-level variance feature value not lower than the preset gray-level threshold, identifying the edge point sequence with periodic brightness changes, and merging the edge point sequence into a wrinkled edge segment.

[0009] In the above embodiments, the system extracts the gray-level variance within a preset window range as a local feature value for each edge point, and initially filters out smooth and uneven regions by comparing it with a preset gray-level threshold. For regions with high gray-level variance, further frequency domain transformation analysis is performed to identify wrinkled edge segments with periodic brightness variations, effectively distinguishing random noise interference from genuine wrinkled structures. This hierarchical discrimination mechanism ensures the accuracy and robustness of edge classification, providing a reliable basis for subsequent differentiated measurement strategies.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of calculating the fold height parameter based on the three-dimensional depth features of the fold edge segment specifically includes: extracting the side view contour line corresponding to the fold edge segment from the side-lit image; performing peak and trough detection on the side view contour line to identify the local highest point as the fold peak point and the local lowest point as the fold trough point; calculating the vertical distance between each fold peak point and the adjacent fold trough point as the height value of a single fold; statistically analyzing the height values ​​of all folds within the fold edge segment; calculating the average and maximum values ​​of the height values; and combining the average and maximum values ​​as the fold height parameter.

[0011] In the above embodiment, the system extracts the side-view contour lines of the fold edge segments from the side-lit image, locates the highest and lowest points of each fold using a peak-valley detection algorithm, and calculates the vertical distance to obtain the height value of a single fold. The average of all fold height values ​​reflects the overall degree of fluctuation, while the maximum value represents extreme deformation. The combination of these two values ​​forms a comprehensive fold height parameter. This parameter accurately quantifies the three-dimensional characteristics of the folds, providing a reliable geometric basis for calculating the flattening correction coefficient.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, the step of calculating the fold correction length based on the projected measurement length of the fold edge segment and the flattening correction coefficient based on the fold height parameter specifically includes: accumulating the Euclidean distances between adjacent edge points in the fold edge segment to obtain the projected measurement length of the fold edge segment; calculating the fold fluctuation amplitude index based on the average and maximum values ​​in the fold height parameter, and determining the flattening correction coefficient based on the fold fluctuation amplitude index; and multiplying the projected measurement length by the flattening correction coefficient to obtain the fold correction length of the fold edge segment.

[0013] In the above embodiment, the system first accumulates the Euclidean distances between adjacent edge points in the fold edge segment to obtain the projected measurement length, which reflects the projected size of the fold on the plane. Based on the average and maximum values ​​of the fold height parameter, a fold fluctuation amplitude index is calculated, which comprehensively reflects the degree of fold undulation. A flattening correction coefficient is determined based on the fluctuation amplitude index, and the projected length is multiplied by the correction coefficient to obtain the true length of the fold after flattening, effectively compensating for the length loss of the three-dimensional fold in the two-dimensional projection.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the step of splicing the flatness measurement length of each flat edge segment and the wrinkle correction length of each wrinkled edge segment according to their spatial position order in the overall contour to obtain the true side length of the four sides after heating specifically includes: identifying the four corner points of the diaphragm sample based on the overall contour in the backlight image; dividing the set of edge points into edge point subsets corresponding to the four sides according to the four corner points; extracting the flatness measurement length of the flat edge segment and the wrinkle correction length of the wrinkled edge segment in sequence according to the spatial position order of the edge points in the edge point subset corresponding to each side; summing all the flatness measurement lengths and wrinkle correction lengths in the edge point subset corresponding to each side to obtain the true side length of each of the four sides after heating.

[0015] In the above embodiments, the system identifies four corner points based on the overall contour in the backlit image, dividing the edge point set into subsets corresponding to the four sides, ensuring the accuracy of edge segment attribution. For each edge point subset, the flat measurement length and wrinkle correction length are extracted sequentially according to spatial position, maintaining the continuity of the measurement sequence. The true side length of the edge is obtained by summing all length values ​​on the same side, avoiding measurement errors caused by directly connecting endpoints in traditional methods, and achieving accurate reconstruction of deformed boundaries.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the thermal shrinkage rate based on the area before heating and the area after heating, the method further includes: acquiring real-time temperature data of the diaphragm sample during the heating process and intermediate state images at corresponding times, the intermediate state images being used to record the deformation evolution process of the diaphragm sample from the start to the end of heating; performing contour extraction and area calculation on the intermediate state images to obtain intermediate area values ​​corresponding to multiple times, and associating the intermediate area values ​​with the corresponding real-time temperature data; calculating the stage shrinkage rate of the diaphragm sample in different temperature ranges based on the area before heating, the intermediate area values, and the area after heating, the stage shrinkage rate characterizing the thermal shrinkage characteristics of the diaphragm sample within a specific temperature range.

[0017] In the above embodiments, the system continuously acquires real-time temperature data and corresponding intermediate state images during the heating process, recording the complete deformation evolution of the diaphragm sample. Contour extraction and area calculation are performed on the intermediate state images to obtain intermediate area values ​​at multiple times, which are then correlated with the temperature data to establish an area-temperature correspondence. Based on the area data before, during, and after heating, the stage shrinkage rate in different temperature ranges is calculated, revealing the thermal shrinkage characteristics of the diaphragm material within a specific temperature range and providing more refined parameters for material performance evaluation.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the stage shrinkage rate of the diaphragm sample in different temperature ranges based on the area before heating, the intermediate area value, and the area after heating, the method further includes: performing a global wrinkle distribution analysis on the top light image of the heated diaphragm sample, calculating the proportion of the total length of the wrinkle edge segments to the total perimeter of the overall contour to obtain the wrinkle ratio; judging the heat shrinkage uniformity of the diaphragm sample based on the wrinkle ratio and the heat shrinkage rate, determining uniform shrinkage when the wrinkle ratio is lower than a preset ratio threshold, and determining non-uniform shrinkage when the wrinkle ratio is higher than the preset ratio threshold; and generating a quality evaluation result of the diaphragm sample based on the determination result of heat shrinkage uniformity and the heat shrinkage rate.

[0019] In the above embodiments, the system performs global wrinkle distribution analysis on the top-light image of the heated diaphragm sample, and calculates the wrinkle ratio parameter by statistically analyzing the proportion of the total length of the wrinkle edge segments to the total perimeter of the overall contour. The wrinkle ratio is compared with a preset ratio threshold, and the uniformity of thermal shrinkage of the diaphragm sample is determined by combining this with the thermal shrinkage rate value. A low wrinkle ratio is considered uniform shrinkage, while a high wrinkle ratio is considered uneven shrinkage. Based on the thermal shrinkage uniformity determination result and the thermal shrinkage rate, a quality evaluation result is generated, achieving a comprehensive assessment of the thermal stability performance of the diaphragm sample.

[0020] In a second aspect, embodiments of this application provide a battery separator heat shrinkage rate testing system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the battery separator heat shrinkage rate testing system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a battery separator heat shrinkage rate testing system, cause the battery separator heat shrinkage rate testing system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a battery separator heat shrinkage rate testing system, cause the battery separator heat shrinkage rate testing system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the battery separator thermal shrinkage rate testing system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. This application utilizes three light sources—backlight, top light, and side light—to collaboratively image and acquire complete three-dimensional morphological information of the separator sample. Based on local grayscale variance characteristics, it automatically identifies smooth and wrinkled edge segments. For smooth areas, direct Euclidean distance accumulation is used for measurement. For wrinkled areas, a flattening correction coefficient is calculated based on three-dimensional depth characteristics to compensate for projection errors. The corrected segment lengths are then stitched together in spatial order to obtain the true side length. This effectively solves the problem of inaccurate measurement of wrinkled areas leading to large deviations in the calculation of thermal shrinkage rate in existing technologies, thereby achieving high-precision measurement of the thermal shrinkage rate of battery separators.

[0026] 2. This application extracts the gray-level variance within a preset window range as a local feature value for each point in the edge point set, and initially filters the region type by comparing it with a preset gray-level threshold. For regions with high gray-level variance, further frequency domain transformation analysis is performed to identify folded edge segments with periodic brightness variations, distinguishing between random noise and real folded structures. This effectively solves the problem of accurately distinguishing between flat and folded regions in existing technologies, thereby achieving automated and accurate classification of edge types.

[0027] 3. This application extracts the side-view contour lines of the fold edge segments from side-lit images, locates the highest and lowest points of each fold using a peak-valley detection algorithm, and calculates the vertical distance to obtain the height value of a single fold. The average and maximum values ​​of all fold heights are combined to form a comprehensive fold height parameter, accurately quantifying the three-dimensional features of the folds. This effectively solves the problem of not being able to obtain fold depth information in existing technologies, thereby achieving accurate three-dimensional reconstruction of the fold morphology. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a method for testing the thermal shrinkage rate of a battery separator in an embodiment of this application.

[0029] Figure 2 This is another flowchart illustrating the method for testing the thermal shrinkage rate of the battery separator in this application embodiment;

[0030] Figure 3 This is a schematic diagram of the physical device structure of a battery separator thermal shrinkage rate testing system in the embodiments of this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] To facilitate understanding, the application scenarios of the embodiments of this application are described below.

[0034] The separator in a lithium-ion battery is a critical safety component positioned between the positive and negative electrodes, and its thickness is typically less than 10 micrometers. During battery use, especially at high temperatures, the separator undergoes thermal shrinkage. Excessive thermal shrinkage can cause separator deformation, leading to direct contact between the positive and negative electrodes and a short circuit, potentially resulting in serious safety incidents such as battery overheating, fire, or even explosion. Therefore, accurately testing the thermal shrinkage rate of the battery separator under different temperature conditions is crucial for ensuring the safety performance of lithium-ion batteries.

[0035] Traditional methods for testing the thermal shrinkage rate of battery separators involve combining oven heating with dimensional measurement. The specific procedure is as follows: First, the length and width of the separator sample are measured using calipers or other measuring tools to calculate the initial area; then, the separator sample is clamped and fixed with a glass plate and placed in an oven at a set temperature for several hours; after heating, the sample is removed and cooled to room temperature, and the dimensions are measured again to calculate the area after heating; finally, the thermal shrinkage rate is calculated based on the area change rate.

[0036] However, this traditional method suffers from significant measurement errors. After being heated to high temperatures, the edges of the diaphragm sample often exhibit irregular wrinkles, warping, and wavy deformations. These three-dimensional deformations make it difficult to establish a unified measurement benchmark, resulting in poor repeatability of multiple measurements of the same sample. More importantly, there is a significant difference between the actual unfolded length of the wrinkled area and its projected length on the plane. Conventional two-dimensional measurement methods can only obtain the projected length and cannot reflect the true length after the wrinkles are flattened. This directly leads to a systematic deviation in the calculation results of the thermal shrinkage rate.

[0037] To address the aforementioned technical issues, this application proposes a method for testing the thermal shrinkage rate of battery separators based on multi-source collaborative imaging and three-dimensional depth correction. This method obtains clear contour boundaries through backlight imaging, identifies surface wrinkle texture features using top-light imaging, and supplements depth dimension data with side-light imaging, achieving complete three-dimensional morphological reconstruction of deformed separator samples. During measurement, the system automatically identifies flat and wrinkled areas based on local gray-level variance features. Flat edge segments are measured using direct Euclidean distance accumulation, while wrinkled edge segments are compensated for projection measurement errors by introducing a flattening correction coefficient based on three-dimensional depth features. Finally, the corrected lengths of each edge segment are spliced ​​together in spatial order to calculate the true side length, thereby significantly improving the accuracy and repeatability of thermal shrinkage rate measurement.

[0038] To facilitate understanding, the method provided in this implementation will be described in detail below, using the above scenario as an example. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a method for testing the thermal shrinkage rate of a battery separator in an embodiment of this application.

[0039] S101. Acquire a backlight image of the diaphragm sample before heating, extract the initial contour coordinates from the high-contrast edges in the backlight image, and calculate the area before heating based on the initial contour coordinates.

[0040] The backlit image refers to an image formed by placing a light source behind the diaphragm sample, allowing light to pass through from behind, creating a high-contrast boundary between the sample outline and the background. High-contrast edges indicate a significant grayscale difference between the sample edges and the background area in the image. Initial outline coordinates represent the positions of points on the sample's edges before heating in the image coordinate system, stored as pixel coordinates. The area before heating refers to the actual area of ​​the sample before heat treatment, serving as a baseline parameter for calculating the thermal shrinkage rate.

[0041] Specifically, the system activates the backlight source and places the sample flat on a transparent stage. Parallel light emitted from the backlight penetrates the sample and is received by the camera, forming an image where the sample area is dark and the background is bright. The system performs grayscale conversion and noise filtering preprocessing on the image, and applies Canny or Sobel edge detection algorithms to extract areas with drastic grayscale gradient changes, obtaining a sequence of edge point pixel coordinates. The system checks the continuity of the coordinate sequence and connects breakpoints to form a closed contour. Based on the camera calibration parameters, the pixel coordinates are converted to physical coordinates, and the area of ​​the polygon is calculated using the shoelace formula. The contour coordinate points are connected sequentially, and the triangles formed by adjacent coordinate points are summed to obtain the area before heating.

[0042] In some embodiments, backlight image acquisition and contour extraction can be achieved in various ways. Optionally, the system uses an LED array as the backlight source, adjusts the current to control the uniformity of illumination, uses an industrial camera to capture images, binarizes the images to separate the sample from the background, applies morphological closing operations to fill internal holes, and extracts edge coordinates using a contour tracking algorithm. Optionally, the system uses a laser backlight source in conjunction with a linear scan camera for scanning imaging. The linear scan camera moves vertically to acquire transmitted light intensity line by line and stitches the images together. Adaptive threshold segmentation is applied to the images to determine edges, the least squares method is used to fit edge curves, and the corner coordinates are determined and the area is calculated based on the intersection of the fitted curves. It is understood that other optical imaging methods and image processing algorithms can also be used, and are not limited here.

[0043] S102. Sequentially acquire backlight image, top light image and side light image of the heated diaphragm sample. The backlight image is used to determine the set of edge points of the overall contour after heating. The top light image is used to identify the wrinkled area in the set of edge points. The side light image is used to acquire the three-dimensional depth features of the wrinkled area.

[0044] The backlit image is used to determine the set of edge points of the overall contour after heating, and the deformed outer contour is extracted through contrast differences. The top-lit image is a reflected light image captured by a camera from the same direction, where the light source shines vertically from directly above the sample; it can reveal surface texture details. The side-lit image represents an image captured by a camera when the light source shines from the side of the sample at a specific angle; the contrast of light and dark in the side-lit image can highlight changes in surface height. The wrinkled region refers to the wavy, uneven area formed at the edge of the sample after heating; its actual unfolded length is greater than the planar projection length. The three-dimensional depth feature is used to represent the height offset of each point in the wrinkled region relative to the reference plane, including peak height, trough depth, and undulation period.

[0045] Specifically, after the sample has been heated and cooled, the system sequentially switches light sources to acquire multimodal images. First, the backlight source is activated to acquire backlight images, and edge detection is applied to extract the set of edge points of the deformed contour. Next, the top ring LED light source is activated to vertically illuminate the sample surface and capture top-lit images. Due to the change in the surface normal vector, the intensity of reflected light in the wrinkled areas changes periodically, appearing as alternating bright and dark stripes in the image. The system performs texture analysis on the image to identify the distribution location of the wrinkles. Then, the side linear light source is activated to illuminate the sample edges at a 45-degree grazing angle. The raised parts of the wrinkles appear as bright areas, while the recessed parts form shadows. The camera captures side-lit images, and the peak and trough positions are determined by the boundaries of the bright and dark areas. The distance between them is measured to obtain the wrinkle height and periodic characteristics. The system spatially registers the three types of image information to establish the correspondence between the edge point set, wrinkle distribution, and depth features.

[0046] In some embodiments, multimodal image acquisition and feature extraction can be achieved in various ways. Optionally, the system uses a fixed mechanical structure to install three light sources, and a programmable controller controls the switching of the light sources sequentially. After each switch, a delay is set to wait for stabilization. The camera uses a fixed focal length to ensure a consistent field of view. By identifying marker points, the coordinate transformation relationship between images is established, and the features of the top-lit and side-lit images are mapped to the set of edge points of the backlit image. Optionally, the system uses a rotatable light source support, and a robotic arm automatically moves the light source to a designated position while the sample remains stationary. After each adjustment, the camera posture is fine-tuned using visual servoing, and a feature point matching algorithm is applied to the image to calculate the affine transformation matrix and unify it to the same coordinate system. It is understood that other light source layouts and image registration methods can also be used, and are not limited here.

[0047] S103. Based on the local gray-level variance of the top light image, the set of edge points is divided into smooth edge segments and wrinkled edge segments. The gray-level variance of the smooth edge segment is lower than a preset threshold and the edge direction is continuous. The gray-level variance of the wrinkled edge segment is higher than a preset threshold and there are periodic changes in brightness.

[0048] Local gray-level variance represents the dispersion of pixel gray-level values ​​relative to the average value within the neighborhood of an edge point, reflecting the complexity of local texture. A smooth edge segment refers to an area on the sample edge without obvious wrinkles, where the surface normal vector is essentially perpendicular to the sample plane, exhibiting a uniform gray-level distribution in the top-lit image. A wrinkled edge segment represents an area with wrinkles and deformations on the edge, exhibiting periodic undulations on the surface. The preset threshold is a critical value set based on sample statistical data to distinguish between smooth and wrinkled areas. Continuous edge direction means that the tangent direction of adjacent edge points changes smoothly without abrupt changes. Periodic brightness variations indicate that the gray-level values ​​in the wrinkled area exhibit a regular alternating distribution of high and low values ​​along the edge direction.

[0049] Specifically, the system extracts grayscale values ​​within a preset window (e.g., 7×7 pixels) for each edge point, calculates the average grayscale value of pixels within the window, then calculates the squared difference between each pixel and the average value, sums these differences, and divides the sum by the total number of pixels to obtain the local grayscale variance feature value. This value is compared with a preset grayscale threshold, typically between 50 and 150, determined statistically using standard samples. When the grayscale variance is below the threshold and the tangent direction change between adjacent points is less than 15 degrees, the continuous edge point sequence is marked as a smooth edge segment. For edge points with grayscale variance above the threshold, a sequence of 20 to 50 points is extracted, and a one-dimensional Fourier transform analysis is performed on the frequency domain energy distribution. If a significant energy peak exists at a specific frequency, indicating periodic brightness changes, the continuous point sequence with the same dominant frequency is marked as a wrinkled edge segment.

[0050] In some embodiments, edge point classification can be achieved in multiple ways. Optionally, the system uses a sliding window to traverse edge points, extracts the gray-level standard deviation within a circular neighborhood radius of 5 pixels, establishes a gray-level variance curve and performs smoothing filtering, sets dual thresholds for segmentation, classifies points below the lower threshold as flat, and classifies points above the higher threshold as wrinkled, classifies intermediate values ​​according to the categories of adjacent points, and performs connectivity analysis on edge points in wrinkled regions to merge them into segments. Optionally, the system constructs a training dataset, extracts multi-dimensional feature vectors of edge points including gray-level variance, gradient magnitude, curvature, texture entropy, etc., manually labels the categories, trains a support vector machine classifier, the classifier automatically determines the category based on the feature vectors, applies morphological operations to remove isolated erroneous points, and connects edge points of the same category into segments. It is understood that other feature extraction and classification algorithms can also be used, and are not limited here.

[0051] In some embodiments, this step specifically includes:

[0052] For each edge point in the edge point set, the gray value within a preset window range centered on the edge point is extracted in the top light image. The gray variance of all pixels within the preset window range is calculated as the local gray variance feature value. The local gray variance feature value is compared with a preset gray threshold. The edge point sequence with a local gray variance feature value lower than the preset gray threshold and spatially continuous is merged into a smooth edge segment. For the region corresponding to the edge point with a local gray variance feature value not lower than the preset gray threshold, the frequency domain transformation of the gray value sequence is performed to identify the edge point sequence with periodic brightness changes. The edge point sequence is then merged into a wrinkled edge segment.

[0053] The local gray-level variance represents the dispersion of the gray-level values ​​of neighboring pixels around an edge point relative to the average gray-level; a larger value indicates a more complex texture. A smooth edge segment refers to a region on the sample edge with a flat surface and no obvious undulations, exhibiting a uniform gray-level distribution in a top-lit image. A wrinkled edge segment represents a region with wavy, undulating deformations on the edge, where surface undulations cause periodic changes in gray-level. The preset window range represents a rectangular neighborhood region centered on the edge point, used to extract local gray-level information. The preset threshold is a critical value determined based on material properties and experimental statistics. Periodic brightness variations indicate that gray-level values ​​along the edge direction exhibit a regular alternation of high and low values, manifested as energy peaks at specific frequencies in the frequency domain. For example, a smooth region has a gray-level variance of 30, while a wrinkled region has 120; setting a threshold of 80 can effectively distinguish them.

[0054] The system iterates through the set of edge points, extracting a 7×7 or 9×9 pixel window centered on each edge point in the top-lit image. It calculates the average grayscale value of all pixels within the window, then squares the difference between each pixel and the average value, sums these values, and divides the sum by the total number of pixels to obtain the local grayscale variance feature value. The system compares this feature value with a preset grayscale threshold, typically between 50 and 150. When the grayscale variance of an edge point is below the threshold, it checks whether the grayscale variance of adjacent edge points is also below the threshold, and whether the angle of change of the tangent direction of adjacent points is less than 15 degrees. Sequences of consecutive edge points meeting these conditions are merged into a smooth edge segment. For edge points with grayscale variances above the threshold, an edge sequence containing 20 to 50 points centered on that point is extracted. The grayscale values ​​of each point are read to construct a grayscale value sequence, which is then converted into a frequency domain signal using a one-dimensional fast Fourier transform. The system analyzes the frequency domain amplitude spectrum, searching for amplitude peaks within a period of 0.1 to 0.5 pixels. If a peak exceeds twice the average amplitude, periodic brightness variations are identified. The system merges edge point sequences with the same dominant frequency and spatial continuity into folded edge segments, thus completing the classification.

[0055] S104. Calculate the Euclidean distance between adjacent edge points for the flat edge segment and sum them to obtain the flat measurement length. For the folded edge segment, calculate the fold height parameter based on the three-dimensional depth feature. Calculate the fold correction length based on the projected measurement length of the folded edge segment and the flattening correction coefficient based on the fold height parameter.

[0056] In this context, Euclidean distance represents the straight-line distance between two points in a two-dimensional plane, calculated as the square root of the sum of the squares of the differences between the horizontal and vertical coordinates. The flattened measurement length represents the actual length of the flattened edge segment; since there is no three-dimensional deformation, the projected length is the true length. The fold height parameter describes the fold geometry, including the maximum height of the crest, the average height difference between the crest and trough, and the undulation period. The projected measurement length represents the projected length of the fold edge segment on the horizontal plane, which is less than the actual length after flattening. The flattening correction factor is a coefficient that converts the projected length to the true unfolded length; it is greater than 1 and depends on the fold height and shape. The fold correction length represents the true length of the flattened fold edge segment, obtained by multiplying the projected length by the correction factor.

[0057] Specifically, for the smoothed edge segment, the system extracts the coordinates (x1, y1) and (x2, y2) of adjacent edge points in sequence and calculates the Euclidean distance. The flattened measurement length is obtained by summing the distances between all adjacent points. For the folded edge segment, the side view contour line is extracted from the side-lit image. The peak and trough points are identified using first-order derivative zero-point detection. The vertical distance between each peak and the adjacent trough is calculated as the fold height value. The average value h_avg and the maximum value h_max of all height values ​​are statistically calculated to form the fold height parameter. The projected length L_proj is obtained by summing the Euclidean distances between adjacent points of the folded edge segment. The fluctuation amplitude index A=(h_avg+h_max) / 2 is calculated based on the fold height parameter. The flattening correction coefficient is determined using the formula k=1+α×(A / L_proj), where α is the fitting coefficient with a value ranging from 0.5 to 2.0, obtained through statistical regression of known samples. Finally, the fold correction length L_corrected=k×L_proj is calculated.

[0058] In some embodiments, wrinkle length correction can be achieved in several ways. Optionally, the system constructs stereo vision using backlight and sidelight images, calculates the three-dimensional coordinates of each point in the wrinkle region through triangulation, connects the coordinate points into a three-dimensional curve according to the spatial direction, calculates and accumulates the three-dimensional Euclidean distance between adjacent points on the curve, and directly obtains the true unfolded length as the wrinkle correction length. Optionally, the system decomposes the wrinkle into multiple small segments based on the peak and trough positions, assuming each segment is a triangle or arc. The actual arc length or hypotenuse length is calculated using geometric relationships based on the projected length and height parameters. For example, the approximate actual length of a triangle is √(projected length² + height²). The wrinkle correction length is obtained by accumulating the lengths of all small segments. It is understood that other geometric modeling and calculation strategies can also be used, and are not limited here.

[0059] In some embodiments, this step specifically includes:

[0060] The side-view contour line corresponding to the fold edge segment is extracted from the side-lit image. Peak and trough detection is performed on the side-view contour line to identify the local highest point as the fold peak point and the local lowest point as the fold trough point. The vertical distance between each fold peak point and the adjacent fold trough point is calculated as the height value of a single fold. The height values ​​of all folds in the fold edge segment are counted, and the average and maximum height values ​​are calculated. The average and maximum values ​​are combined as the fold height parameter. The Euclidean distances between adjacent edge points in the fold edge segment are accumulated to obtain the projected measurement length of the fold edge segment. The fold fluctuation amplitude index is calculated based on the average and maximum values ​​in the fold height parameter. The flattening correction coefficient is determined based on the fold fluctuation amplitude index. The projected measurement length is multiplied by the flattening correction coefficient to obtain the fold correction length of the fold edge segment.

[0061] In this context, Euclidean distance represents the straight-line distance between two points in a two-dimensional plane, calculated as the square root of the sum of the squares of the differences between the horizontal and vertical coordinates. The flattened measurement length represents the actual length of the flattened edge segment; since there is no three-dimensional deformation, the projected length is the true length. The three-dimensional depth feature represents the height offset of each point in the folded region relative to the reference plane. The fold height parameter describes the fold geometry, including the maximum height of the crest and the average height difference between the crest and trough. The side view contour line refers to the side profile curve of the folded edge segment extracted from the side-lit image, exhibiting an undulating wave shape. Crest and trough detection refers to the process of identifying local extreme points on the curve; the crest is the local highest point, and the trough is the local lowest point. The projected measurement length represents the projected length of the folded edge segment on the horizontal plane, which is less than the actual length after flattening. The flattening correction coefficient is a coefficient that converts the projected length to the true unfolded length; it is greater than 1 and depends on the fold height and shape. The fold fluctuation amplitude index quantifies the degree of fold undulation, calculated from the average height and maximum height. For example, if the projected length of a fold segment is 50 mm, the fluctuation range index is 2 mm, and the correction coefficient is 1.08, then the corrected length of the fold is 54 mm.

[0062] For smooth edge segments, the system extracts the coordinates (x1, y1) and (x2, y2) of adjacent edge points in contour order and calculates the Euclidean distance. The flattened measurement length is obtained by summing the distances between all adjacent points in the segment. For the folded edge segment, the system locates the corresponding region in the side-lit image and extracts the pixel coordinate sequence of the side-view contour line. The system calculates the first derivative of the contour line; points where the derivative is zero are extreme points. The system identifies peak points where the derivative changes from positive to negative and trough points where the derivative changes from negative to positive by judging the change in the sign of the derivative. The system calculates the vertical coordinate difference between each peak point and its left and right adjacent trough points, and takes the smaller value as the height value of the fold. All height values ​​are counted, and the average h_avg and maximum h_max are calculated and combined to form the fold height parameter. The system sums the Euclidean distances between adjacent edge points of the folded edge segment to obtain the projected measurement length L_proj. The fold fluctuation amplitude index A=(h_avg+h_max) / 2 is calculated, and the flattening correction coefficient is determined using the empirical formula k=1+α×(A / L_proj). α is the fitting coefficient, usually ranging from 0.5 to 2.0, obtained through statistical regression of known samples. Finally, calculate the wrinkle correction length L_corrected = k × L_proj, which represents the actual side length after the wrinkle is flattened.

[0063] S105. The flatness measurement length of each flat edge segment and the fold correction length of each fold edge segment are spliced ​​together in the order of their spatial positions in the overall outline to obtain the true side lengths of the four sides after heating, and the area after heating is calculated based on the true side lengths.

[0064] The spatial order refers to the actual arrangement of each edge segment on the overall contour of the sample, arranged sequentially according to the continuous direction of the contour. "Splicing" means connecting the measured length values ​​of each segment end-to-end in spatial order and adding them together. The true side length refers to the actual length of each edge after the sample is heated and flattened, including the increase in length after the wrinkles are unfolded. The area after heating represents the actual area value of the sample after thermal shrinkage and deformation, calculated based on the true side length.

[0065] Specifically, the system identifies the start and end positions of each edge segment within the overall contour based on the arrangement order of the edge point set. Following a clockwise or counter-clockwise direction along the contour, the system extracts the length of each edge segment sequentially; for smooth edge segments, it extracts the measured length for smoothness; for wrinkled edge segments, it extracts the corrected length for wrinkles. For rectangular samples, the system identifies the four corner points and divides the contour into four sides (top, bottom, left, and right). Each side may contain multiple edge segments. The system accumulates the lengths of all edge segments on each side to obtain the true side lengths for each of the four sides. After obtaining the four true side lengths, the system multiplies the lengths of two adjacent sides to calculate the area after heating.

[0066] In some embodiments, edge segment splicing and area calculation can be implemented in various ways. Optionally, the system establishes an edge segment index table, recording the start and end coordinates, length, and type of each edge segment, sorting them by the starting point position, traversing the list and accumulating the length sequentially, marking the detection of a corner point as the end of one edge and the start of the next, taking the average of the lengths of the two opposite sides as the effective side length, and calculating the product of the effective side lengths in the two directions to obtain the area. Optionally, the system identifies four corner points by detecting local maxima of edge curvature, divides the contour into four segments based on the corner points, traverses the edge segments contained in each segment, selects the corresponding length according to the type, and accumulates it to obtain the true side length. If the sample is close to a rectangle, the length and width are directly multiplied; if the shape is irregular, the quadrilateral area formula is used in combination with the diagonal length for calculation. It is understood that other edge segment organization and area calculation methods can also be used, which are not limited here.

[0067] In some embodiments, this step specifically includes:

[0068] Based on the overall contour in the backlit image, the four corner points of the diaphragm sample are identified. The edge point set is divided into edge point subsets corresponding to the four sides according to the four corner points. For each edge point subset corresponding to the side, the flat measurement length of the flat edge segment and the wrinkle correction length of the wrinkled edge segment are extracted in sequence according to the spatial position of the edge points. All flat measurement lengths and wrinkle correction lengths in the edge point subset corresponding to each side are summed to obtain the true side length of each of the four sides after heating.

[0069] In this context, spatial position order refers to the actual arrangement order of each edge segment on the overall contour of the sample, arranged sequentially according to the continuous direction of the contour. "Splicing" refers to concatenating and summing the measured length values ​​of each segment in spatial order. "True side length" refers to the actual length of each edge after the sample is heated and flattened, including the added length after the wrinkles are unfolded. "Corner point" refers to a feature point on the contour where the curvature changes significantly; for a rectangular sample, this refers to the four corner positions. "Edge point subset" represents the set of all edge points belonging to the same edge, separated by corner points. "Heated area" represents the actual area value of the sample after thermal shrinkage and deformation, calculated based on the true side length. For example, if the coordinates of the four corner points of the sample are (10, 10), (110, 10), (110, 80), and (10, 80), the contour is divided into four sides: top, bottom, left, and right. The top side contains three flat segments and two wrinkled segments; summing the lengths of each segment yields the true side length.

[0070] The system reads the overall contour edge point set extracted from the backlit image and performs corner detection on the contour. Corner detection employs curvature analysis to calculate the local curvature at each edge point. Curvature is the rate of change of the angle formed by three adjacent points. Points with an absolute curvature value exceeding a threshold (e.g., greater than 30 degrees) are identified as candidate corner points. The system filters the candidate corner points, retaining the four points with the largest curvature as the final corner points, labeled as corner point 1 to corner point 4 according to their spatial position. Based on the positions of the four corner points, the system divides the edge point set into four subsets: the edge points between corner point 1 and corner point 2 are assigned to the subset of the first edge, the edge points between corner point 2 and corner point 3 are assigned to the subset of the second edge, and so on, forming subsets of edge points corresponding to the four edges. For each subset of edges, the system traverses the edge points in order on the contour, identifying whether each edge point belongs to a flat or wrinkled edge segment type, and extracts the length value of that segment sequentially. For flat edge segments, the flat measurement length is extracted; for wrinkled edge segments, the wrinkle correction length is extracted. The system sums the lengths of all edge segments of the edge to obtain the true side length. The same operation is performed on each of the four edges to obtain the true side lengths L1, L2, L3, and L4 after heating. For a rectangular sample, the system calculates the length and width by taking the average of two opposite sides, and the area is (L1+L3) / 2×(L2+L4) / 2, thus obtaining the area after heating.

[0071] S106. Calculate the thermal shrinkage rate based on the area before heating and the area after heating.

[0072] The thermal shrinkage rate refers to the percentage reduction in area of ​​a sample after heat treatment, used to quantitatively evaluate the thermal stability of a material. The area before heating represents the original area of ​​the sample before heat treatment, serving as the calculation baseline. The area after heating represents the actual area of ​​the sample after heat treatment, taking into account wrinkle correction. The thermal shrinkage rate is calculated as the ratio of the area change to the area before heating, expressed as a percentage.

[0073] Specifically, the system reads the area before heating (A_before) and the area after heating (A_after). It calculates the area change ΔA = A_before - A_after; a positive value indicates sample shrinkage. The system calculates the heat shrinkage rate using the formula S = (ΔA / A_before) × 100%, outputting it as a percentage. The system checks the validity of the results, determining if the heat shrinkage rate is within a reasonable range. If it is negative or exceeds a preset upper limit, a warning is issued indicating potential measurement error. The system saves the heat shrinkage rate, sample information, and test parameters to the database and generates a test report.

[0074] In some embodiments, the calculation and processing of heat shrinkage rate can be achieved in various ways. Optionally, the system performs repeated tests on the same sample, calculates the heat shrinkage rate for each test, performs statistical analysis to calculate the average, standard deviation, and coefficient of variation. When the standard deviation is less than a preset threshold, the average is used as the final result; otherwise, outlier data is removed and the calculation is repeated. Optionally, the system queries a database for a reference range of heat shrinkage rate based on the sample material type and heating conditions, compares the test results with the reference range to determine whether the sample is qualified, calculates the shrinkage rates in the longitudinal and transverse directions corresponding to the length and width directions, and compares the shrinkage differences in the two directions to analyze material anisotropy, providing data support for process optimization. It is understood that other statistical analysis methods and evaluation standards can also be used to accurately determine the heat shrinkage rate, which is not limited here.

[0075] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is another flowchart illustrating the method for testing the thermal shrinkage rate of the battery separator in this application embodiment.

[0076] S201. Calculate the thermal shrinkage rate based on the area before heating and the area after heating.

[0077] The area before heating refers to the original area of ​​the sample before heat treatment, serving as the calculation basis. The area after heating refers to the actual area of ​​the sample after heat treatment, taking into account wrinkle correction. The thermal shrinkage rate represents the percentage reduction in area of ​​the sample after heat treatment, reflecting the thermal stability of the material. The calculation formula is the ratio of the area change to the area before heating, converted to a percentage. For example, if the area of ​​the sample before heating is 100 square millimeters and the area after heating is 95 square millimeters, then the thermal shrinkage rate is 5%.

[0078] The system reads two stored values: the area before heating (A_before) and the area after heating (A_after). First, it calculates the area change ΔA using the formula ΔA = A_before - A_after. A positive value indicates shrinkage, while a negative value indicates expansion. Then, it calculates the shrinkage rate S according to the definition of thermal shrinkage using the formula S = (ΔA / A_before) × 100%, and outputs the result as a percentage. The system performs a range check on the calculation results to determine if the thermal shrinkage rate is within the material's theoretical range. If the value is abnormal, such as exceeding 50% or being negative, it is marked as abnormal data and recorded. The system stores the thermal shrinkage rate value along with parameters such as the sample number, heating temperature, and heating time in the database as a core indicator for quality assessment.

[0079] S202. Obtain real-time temperature data and corresponding intermediate state images of the diaphragm sample during the heating process. The intermediate state images are used to record the deformation evolution process of the diaphragm sample from the start of heating to the end of heating.

[0080] Real-time temperature data refers to the sequence of temperature measurements collected chronologically during the heating process, recording the temperature change trajectory of the sample. Intermediate state images refer to images of the sample taken at multiple time points during the heating process, capturing the deformation process of the sample from its initial state to its final state. The deformation evolution process represents the dynamic process of the sample's size and shape gradually changing with increasing temperature, including stages such as the onset of shrinkage, accelerated shrinkage, and shrinkage stabilization. For example, images are taken at 60℃, 80℃, and 100℃ to record the morphology of the sample at different temperatures.

[0081] After the heating device is activated, the system collects the sample surface temperature in real time using a contact temperature sensor or infrared thermometer. The sampling frequency is set to once per second or once every 2 seconds, and the temperature data is recorded synchronously with the timestamp. A backlit imaging system is installed inside the heating device. The camera triggers an image at fixed time intervals, such as every 10 seconds or every 5°C increase in temperature, to acquire intermediate state images of the sample. The system establishes a temperature-image correspondence table, binding and storing each intermediate state image with the temperature value at the time of capture. The intermediate state images use the same light source configuration and shooting parameters as the images before heating to ensure comparability of images at different times. The system plots the collected real-time temperature data into a temperature-time curve and arranges the intermediate state image sequence in chronological order to form a continuous image sequence reflecting the deformation evolution process.

[0082] S203. Extract the contour and calculate the area of ​​the intermediate state image to obtain the intermediate area values ​​corresponding to multiple time points, and associate the intermediate area values ​​with the corresponding real-time temperature data.

[0083] Contour extraction refers to identifying and extracting the set of coordinate points of the sample edge from an intermediate state image. The intermediate area value represents the actual area of ​​the sample at a certain moment during the heating process. Association refers to establishing a one-to-one correspondence between the intermediate area value and the corresponding temperature data, forming a temperature-area data pair. For example, if the extracted contour at a temperature of 70℃ calculates to an area of ​​98 square millimeters, then this intermediate area value is associated with and stored at the 70℃ temperature.

[0084] The system iterates through all intermediate state images, applying an edge detection algorithm to extract the sample contour for each image. Edge detection employs the Canny operator or threshold segmentation method to obtain the pixel coordinate sequence of the contour. The system converts the pixel coordinates to physical coordinates based on camera calibration parameters and calculates the area of ​​the polygon enclosed by the contour using the shoelace formula, obtaining the intermediate area value at that moment. Since the sample typically does not produce obvious wrinkles during heating, the two-dimensional projected area can be calculated directly. The system reads the temperature value corresponding to the image from the temperature-image correspondence table, pairs and stores the intermediate area value with the temperature value, forming a temperature-area data sequence. The system sorts this sequence temporally and plots an area-temperature curve, visually displaying the trend of sample area change with temperature, providing a data foundation for analyzing the shrinkage characteristics of different temperature ranges.

[0085] S204. Based on the area before heating, the intermediate area value, and the area after heating, calculate the stage shrinkage rate of the diaphragm sample in different temperature ranges. This stage shrinkage rate characterizes the thermal shrinkage characteristics of the diaphragm sample within a specific temperature range.

[0086] The temperature range refers to several temperature intervals defined during the heating process, such as room temperature to 60℃, 60℃ to 90℃, and 90℃ to 120℃. The stage shrinkage rate represents the percentage of area shrinkage that occurs within a specific temperature range, reflecting the thermal shrinkage characteristics within that temperature range. It is calculated by dividing the difference between the initial and final areas of the sample within that range by the initial area. For example, in the 60℃ to 90℃ range, if the initial area is 98 square millimeters and the final area is 96 square millimeters, then the stage shrinkage rate is 2.04%.

[0087] The system divides the heating process into several temperature intervals based on the temperature variation range and shrinkage characteristics, using either equal temperature intervals or division according to the inflection point of the shrinkage rate change. For each temperature interval, the system extracts the area A_start corresponding to the starting temperature and the area A_end corresponding to the ending temperature from the temperature-area data sequence. The stage shrinkage rate of that interval is calculated using the formula S_stage=(A_start-A_end) / A_start×100%. The system summarizes the stage shrinkage rates of all temperature intervals to form a stage shrinkage rate distribution table, analyzing the differences in shrinkage rates between different temperature intervals. By comparing the shrinkage rate values ​​of each stage, the main shrinkage temperature range of the sample is identified. For example, if the stage shrinkage rate in the 60℃ to 90℃ range is significantly higher than other ranges, it indicates that this temperature range is the critical temperature range for sample thermal shrinkage, providing a basis for optimizing process parameters.

[0088] S205. Perform a global wrinkle distribution analysis on the top light image of the heated diaphragm sample, and calculate the proportion of the total length of the wrinkle edge segment to the total perimeter of the overall outline to obtain the wrinkle percentage.

[0089] Global wrinkle distribution analysis refers to the statistical analysis of the location and extent of wrinkles across the entire surface of the sample. The total length of wrinkled edge segments represents the sum of the lengths of all wrinkled edge segments, reflecting the overall scale of the wrinkles. The total perimeter of the overall profile refers to the complete perimeter of the sample edge, including both smooth and wrinkled edge segments. The wrinkle percentage represents the proportion of the wrinkled edge segment length in the total perimeter; a larger value indicates a more widespread wrinkle distribution. For example, if the total perimeter is 200 mm and the total length of the wrinkled edge segments is 30 mm, then the wrinkle percentage is 15%.

[0090] The system reads the top-light image of the heated sample, which has already undergone edge classification processing, with each edge segment labeled as either smooth or wrinkled. The system iterates through all edge segments, filters out those labeled as wrinkled, extracts the length value of each wrinkled edge segment, and sums them to obtain the total length L_wrinkle of the wrinkled edge segments. Simultaneously, the system sums the lengths of all edge segments, including the smoothness measurement length of smooth edge segments and the wrinkle correction length of wrinkled edge segments, to obtain the overall total perimeter L_total. The wrinkle percentage is calculated using the formula R_wrinkle = (L_wrinkle / L_total) × 100%. The system can also count the number and average length of wrinkled edge segments, analyzing the degree of wrinkle dispersion. If wrinkles are concentrated in a few locations, it indicates that the shrinkage unevenness has local characteristics; if wrinkles are scattered, it indicates that the overall shrinkage is fluctuating.

[0091] S206. Determine the thermal shrinkage uniformity of the separator sample based on the wrinkle ratio and the thermal shrinkage rate. When the wrinkle ratio is lower than the preset ratio threshold, it is determined as uniform shrinkage; when the wrinkle ratio is higher than the preset ratio threshold, it is determined as non-uniform shrinkage.

[0092] Among them, the thermal shrinkage uniformity represents the consistency of the shrinkage degree of each part of the sample. Uniform shrinkage means that the shrinkage amount of each area of the sample is basically the same, and non-uniform shrinkage means that excessive shrinkage occurs in some areas, forming wrinkles. The preset ratio threshold is a judgment critical value determined according to material standards or experimental statistics, and is used to distinguish between uniform and non-uniform shrinkage. For example, the preset ratio threshold of a certain material is 10%. When the wrinkle ratio is 8%, it is determined as uniform shrinkage; when the wrinkle ratio is 15%, it is determined as non-uniform shrinkage.

[0093] The system reads the calculated wrinkle ratio R_wrinkle and the preset ratio threshold T_threshold. The preset ratio threshold is loaded from the parameter configuration file according to the material type, and the typical value range is between 5% and 15%. The system compares the wrinkle ratio with the preset ratio threshold. If R_wrinkle < T_threshold, it is determined that the sample has uniform shrinkage, and the shrinkage uniformity identifier is marked as "uniform"; if R_wrinkle ≥ T_threshold, it is determined that the sample has non-uniform shrinkage, and the shrinkage uniformity identifier is marked as "non-uniform". The system further makes a comprehensive judgment in combination with the thermal shrinkage rate value. If the thermal shrinkage rate is too high, such as exceeding 20%, even if the wrinkle ratio is lower than the threshold, local stress concentration may occur due to excessive overall shrinkage, and the shrinkage uniformity is downgraded to "basically uniform". The judgment result is recorded together with the wrinkle ratio value and the thermal shrinkage rate value as the criterion for quality evaluation.

[0094] S207. Generate the quality evaluation result of the separator sample based on the judgment result of the thermal shrinkage uniformity and the thermal shrinkage rate.

[0095] Among them, the judgment result refers to the classification identifier of the thermal shrinkage uniformity of the sample, including uniform shrinkage or non-uniform shrinkage. The quality evaluation result represents the comprehensive evaluation of the thermal stability performance of the sample, including grades such as qualified, unqualified, or further testing required, etc. Generation means outputting the final evaluation conclusion according to multiple test indicators according to the evaluation rules. For example, when the thermal shrinkage rate is 4% and it is determined as uniform shrinkage, the evaluation result is excellent; when the thermal shrinkage rate is 18% and it is determined as non-uniform shrinkage, the evaluation result is unqualified. <0000The system reads the heat shrinkage rate (S) and the heat shrinkage uniformity assessment result, and makes a comprehensive judgment based on preset quality evaluation rules. The evaluation rules include multi-level standards. The first level is the heat shrinkage rate assessment: if S is less than 5%, it is considered excellent; 5% to 10% is good; 10% to 15% is acceptable; and greater than 15% is unacceptable. The second level is the uniformity assessment: if the assessment result is uniform shrinkage, the current level is maintained; if it is uneven shrinkage, the level is lowered by one. The system combines the two levels of assessment results to generate the final quality evaluation result. The output format includes the evaluation level, heat shrinkage rate value, wrinkle ratio value, and judgment basis explanation. The system writes the quality evaluation result into a test report, generating a complete document containing sample images, data curves, and evaluation conclusions. If the evaluation result is unacceptable, the system automatically marks the batch of samples as needing re-inspection and triggers an early warning notification to quality management personnel.

[0097] The battery separator thermal shrinkage rate testing system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 3 This is a schematic diagram of a physical device structure of a battery separator thermal shrinkage rate testing system in an embodiment of this application.

[0098] It should be noted that, Figure 3 The structure of the battery separator heat shrinkage rate testing system shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0099] like Figure 3 As shown, the battery separator thermal shrinkage rate testing system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 302 or a program loaded from storage section 308 into random access memory (RAM) 303, such as performing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0100] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0101] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the various functions defined in the present invention.

[0102] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0103] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0104] Specifically, the battery separator heat shrinkage rate testing system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the battery separator heat shrinkage rate testing method provided in the above embodiment.

[0105] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the battery separator heat shrinkage rate testing system described in the above embodiments; or it may exist independently and not assembled into the battery separator heat shrinkage rate testing system. The storage medium carries one or more computer programs, which, when executed by a processor of the battery separator heat shrinkage rate testing system, cause the battery separator heat shrinkage rate testing system to implement the battery separator heat shrinkage rate testing method provided in the above embodiments.

[0106] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0107] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0108] 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. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for testing the thermal shrinkage rate of a battery separator, characterized in that, The method, applied to a battery separator thermal shrinkage rate testing system, includes: A backlight image of the diaphragm sample to be tested before heating is acquired. Initial contour coordinates are extracted from the high-contrast edges in the backlight image. The area before heating is calculated based on the initial contour coordinates. Backlight image, top light image and side light image are sequentially acquired on the heated diaphragm sample. The backlight image is used to determine the set of edge points of the overall contour after heating. The top light image is used to identify the wrinkled region in the set of edge points. The side light image is used to obtain the three-dimensional depth features of the wrinkled region. Based on the local gray-level variance of the top light image, the set of edge points is divided into smooth edge segments and wrinkled edge segments. The gray-level variance of the smooth edge segments is lower than a preset threshold and the edge direction is continuous. The gray-level variance of the wrinkled edge segments is higher than the preset threshold and there are periodic changes in brightness. The Euclidean distance between adjacent edge points is calculated for the smooth edge segment and accumulated to obtain the smooth measurement length. The fold height parameter is calculated for the fold edge segment based on the three-dimensional depth feature. The fold correction length is calculated based on the projected measurement length of the fold edge segment and the flattening correction coefficient based on the fold height parameter. The flatness measurement length of each of the flat edge segments and the fold correction length of each of the fold edge segments are spliced ​​together in the spatial order of their positions in the overall contour to obtain the true side lengths of the four sides after heating, and the area after heating is calculated based on the true side lengths. The thermal shrinkage rate is calculated based on the area before heating and the area after heating.

2. The method according to claim 1, characterized in that, The step of dividing the set of edge points into smooth edge segments and wrinkled edge segments based on the local gray-level variance of the top light image specifically includes: For each edge point in the set of edge points, extract the gray value within a preset window range centered on the edge point in the top light image, and calculate the gray variance of all pixels within the preset window range as the local gray variance feature value. The local gray-scale variance feature value is compared with a preset gray-scale threshold, and the sequence of edge points whose local gray-scale variance feature value is lower than the preset gray-scale threshold and whose spatial location is continuous is merged into the flat edge segment; For edge points whose local gray-level variance feature value is not lower than the preset gray-level threshold, a frequency domain transformation of the gray-level value sequence is performed to identify edge point sequences with periodic brightness changes, and the edge point sequences are merged into the wrinkled edge segments.

3. The method according to claim 1, characterized in that, The step of calculating the fold height parameter of the fold edge segment based on the three-dimensional depth feature specifically includes: Extract the side view contour line corresponding to the fold edge segment from the side-lit image, perform peak and valley detection on the side view contour line, and identify the local highest point as the fold peak point and the local lowest point as the fold valley point. The vertical distance between each fold crest and the adjacent fold trough is calculated as the height value of a single fold. The height values ​​of all folds within the fold edge segment are counted, and the average and maximum values ​​of the height values ​​are calculated. The average and maximum values ​​are then combined as the fold height parameter.

4. The method according to claim 1, characterized in that, The step of calculating the fold correction length based on the projected measurement length of the fold edge segment and the flattening correction coefficient based on the fold height parameter specifically includes: The Euclidean distances between adjacent edge points in the folded edge segment are accumulated to obtain the projected measurement length of the folded edge segment; The wrinkle fluctuation amplitude index is calculated based on the average and maximum values ​​of the wrinkle height parameter, and the flattening correction coefficient is determined based on the wrinkle fluctuation amplitude index. The wrinkle correction length of the wrinkle edge segment is obtained by multiplying the projected measurement length by the flattening correction coefficient.

5. The method according to claim 1, characterized in that, The step of splicing together the flatness measurement length of each of the flattened edge segments and the wrinkle correction length of each of the wrinkled edge segments according to their spatial positions in the overall contour to obtain the true side lengths of the four sides after heating specifically includes: Based on the overall contour in the backlight image, the four corner points of the diaphragm sample are identified, and the set of edge points is divided into four edge point subsets corresponding to the four sides according to the four corner points; For each edge point subset corresponding to each edge, the flat measurement length of the flat edge segment and the wrinkle correction length of the wrinkled edge segment are extracted sequentially according to the spatial position order of the edge points. The actual side length of each of the four edges after heating is obtained by summing all the flattening measurement lengths and wrinkle correction lengths in the edge point subset corresponding to each edge.

6. The method according to claim 1, characterized in that, After the step of calculating the thermal shrinkage rate based on the area before heating and the area after heating, the method further includes: The real-time temperature data and corresponding intermediate state images of the diaphragm sample during the heating process are obtained. The intermediate state images are used to record the deformation evolution process of the diaphragm sample from the start to the end of heating. Contour extraction and area calculation are performed on the intermediate state image to obtain intermediate area values ​​corresponding to multiple time points, and the intermediate area values ​​are associated with the corresponding real-time temperature data; Based on the area before heating, the intermediate area value, and the area after heating, the stage shrinkage rate of the diaphragm sample in different temperature ranges is calculated. The stage shrinkage rate characterizes the thermal shrinkage characteristics of the diaphragm sample within a specific temperature range.

7. The method according to claim 6, characterized in that, After the step of calculating the phase shrinkage rate of the diaphragm sample in different temperature ranges based on the area before heating, the intermediate area value, and the area after heating, the method further includes: A global wrinkle distribution analysis is performed on the top light image of the heated diaphragm sample, and the proportion of the total length of the wrinkle edge segments to the total perimeter of the overall outline is calculated to obtain the wrinkle ratio. The heat shrinkage uniformity of the diaphragm sample is determined based on the wrinkle ratio and the heat shrinkage rate. When the wrinkle ratio is lower than a preset ratio threshold, it is determined to be uniform shrinkage. When the wrinkle ratio is higher than the preset ratio threshold, it is determined to be uneven shrinkage. Based on the determination result of the heat shrinkage uniformity and the heat shrinkage rate, the quality evaluation result of the diaphragm sample is generated.

8. A battery separator thermal shrinkage rate testing system, characterized in that, The battery separator heat shrinkage rate testing system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the battery separator heat shrinkage rate testing system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is run on the battery separator heat shrinkage rate testing system, the battery separator heat shrinkage rate testing system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the battery separator heat shrinkage rate testing system, the battery separator heat shrinkage rate testing system performs the method as described in any one of claims 1-7.

Images