Analysis method of pole piece section particle distribution condition
By combining image acquisition, segmentation, and feature extraction methods with K-means and DBSCAN algorithms, the shortcomings of existing technologies in silicon particle distribution assessment are addressed, enabling accurate assessment and intuitive display of silicon particle distribution, which is applicable to the production of silicon-carbon anode materials.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-25
- Publication Date
- 2026-04-07
AI Technical Summary
In existing technologies, the use of standard deviation to describe the uniformity of silicon particle distribution on the electrode cross-section is insufficient, as it cannot accurately reflect the spatial distribution pattern of silicon particles and is sensitive to outliers.
Image acquisition, image segmentation, and feature extraction methods, combined with K-means and DBSCAN clustering algorithms, were used to obtain the geometric features of silicon particles, and the uniformity of silicon particle distribution was evaluated by chi-square test and coefficient of variation.
It provides statistical information on the number and spatial location of silicon particles, improving the accuracy and comprehensiveness of distribution assessment. It can intuitively display the distribution of silicon particles and is suitable for product process optimization by silicon-carbon anode material manufacturers.
Smart Images
Figure CN121810709A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for analyzing the particle distribution on the cross-section of an electrode. Background Technology
[0002] The general steps of the existing method for quantifying the uniformity of particle distribution in the cross section of an electrode are as follows: first, obtain a scanning electron microscope (SEM) image of the porous electrode material; second, divide the SEM image into N equal parts and determine the number of silicon particles in each part; and finally, determine the uniformity of silicon material dispersion in the porous electrode based on the number of silicon particles in each part.
[0003] However, using standard deviation to quantitatively describe the uniformity of silicon distribution has some drawbacks. First, standard deviation can only provide the degree to which data points deviate from the average value, but it cannot reflect the spatial distribution pattern of data points. Second, standard deviation is more sensitive to outliers. In the cross-section of the electrode, if there are a small number of silicon particle aggregation areas, these outliers may significantly affect the calculation results of standard deviation. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing technology that only uses standard deviation to describe uniformity, and to provide an analytical method for the particle distribution of the electrode cross section that can provide not only statistical information on the number of silicon particles, but also spatial position information between silicon particles and copper foil, so as to more accurately reflect the actual distribution of silicon particles.
[0005] The present invention solves the above-mentioned technical problems through the following technical solution:
[0006] This invention provides a method for analyzing the particle distribution in the cross-section of an electrode sheet, which includes the following steps:
[0007] S1. Image acquisition: Acquire a microscopic image of the cross-section of the electrode, wherein the microscopic image contains image information of the current collector and active material particles;
[0008] S2. Image segmentation: Cluster analysis is performed on the pixels in the microscopic image obtained in step S1. The microscopic image is segmented in the first stage to separate the current collector region and the active material particle region. The active material particle region is then segmented in the second stage to separate each active material particle.
[0009] S3. Feature Extraction: Extract the geometric features of each active substance particle obtained in step S2 to obtain the dataset;
[0010] S4. Statistical analysis: Perform statistical analysis on the dataset to obtain the particle distribution of the electrode cross-section.
[0011] In step S1, the microscopic image can be obtained by conventional microscopic imaging equipment in the art. Preferably, the microscopic imaging equipment is SEM or FIB-SEM.
[0012] In step S1, the magnification of the microscopic image can be 200 to 1000 times, preferably 400 times.
[0013] In some preferred embodiments, step S1 further includes grayscale processing of the microscopic image.
[0014] In step S1, the current collector may include conventional current collector materials in the art, preferably one or more of copper foil, aluminum foil and nickel foil, such as copper foil.
[0015] In step S1, the active material particles can be conventional active material particles in the art, preferably including silicon particles and / or graphite particles, such as silicon particles.
[0016] In step S2, the first segmentation can be performed using the K-means clustering algorithm, and the second segmentation can be performed using the DBSCAN clustering algorithm.
[0017] In step S2, the K value of the K-means algorithm can be 2 to 6, preferably 3 to 5, for example 5.
[0018] In some preferred embodiments of the present invention, the number of random seeds in the K-means algorithm is 0.
[0019] In step S2, the radius of the DBSCAN algorithm can be 1 to 10, preferably 3 to 10, for example 10.
[0020] In step S2, the minimum number of points in the DBSCAN algorithm can be 3 to 18, preferably 13 to 18, for example 16.
[0021] In step S2, the number of active substance particles in the active substance particle region can be 50 to 200, preferably 80 to 120, for example 100.
[0022] In step S3, the method for extracting the geometric features may include:
[0023] The centroid position of the active substance particle is obtained, and the geometric characteristics of the active substance particle are calculated based on the centroid position.
[0024] In step S3, the geometric feature may include the distance d of the active material particles, where the distance d refers to the distance from the active material particles to the boundary between the current collector region and the active material region.
[0025] In step S3, the geometric feature may include the radius r of the active material particles.
[0026] In step S3, the geometric features may include the perimeter l of the active material particles.
[0027] In step S3, the geometric features may include the area S of the active material particles.
[0028] In step S3, the method for extracting the distance d may include:
[0029] (1) Obtain the vertical coordinate y2 of the centroid of the active material particle and the vertical coordinate y1 of the junction between the current collector region and the active material region, respectively;
[0030] (2) Calculate the distance d: d = min(y2 - y1);
[0031] (3) Repeat steps (1) and (2) to obtain the distance dataset.
[0032] In step S3, the method for extracting the radius r may include:
[0033] (1) Obtain the coordinates (x, y) of the centroid of the active substance particles respectively. c ,y c The number of pixels m at the boundary between the current collector region and the active material region, and the coordinate (x, y) of the j-th pixel at the boundary between the current collector region and the active material region. j ,y j );
[0034] (2) Calculate the radius r:
[0035] ;
[0036] (3) Repeat steps (1) and (2) to obtain the radius dataset.
[0037] In step S4, the statistical analysis method may include:
[0038] Based on the dataset, plot the probability distribution function curve of the active substance particles;
[0039] The probability distribution function curve is preferably a cumulative distribution function curve and / or a probability density function curve.
[0040] In some preferred embodiments of the present invention, the geometric feature extracted in step S3 is distance d, resulting in a distance dataset. The method for statistical analysis of the distance dataset in step S4 includes the following steps:
[0041] (1) Plot the cumulative distribution function curve based on the distance dataset, denoted as the actual cumulative distribution function curve, and plot the reference cumulative distribution function curve with uniform distribution.
[0042] The reference cumulative distribution function curve is plotted using the following function expression:
[0043] ;
[0044] The x m The maximum value of the data in the dataset;
[0045] (2) The uniformity of particle distribution on the electrode cross section is evaluated by comparing and analyzing the actual cumulative distribution function curve and the reference cumulative distribution function curve.
[0046] In some more preferred embodiments of the present invention, the comparative analysis includes the following methods ① and / or ②:
[0047] ① Observe the degree of difference between the actual cumulative distribution function curve and the reference cumulative distribution function curve;
[0048] ② Perform a chi-square test on the actual cumulative distribution function curve and the reference cumulative distribution function curve, and calculate the chi-square statistic.
[0049] The chi-square test is performed using the following formula:
[0050] ;
[0051] Among them, the This represents the chi-square statistic, where A represents the number of groups participating in the statistic, and i represents the index number of the i-th group currently being calculated; O i E represents the measured frequency belonging to the i-th group in the dataset. i This represents the theoretical frequency of the i-th group under the reference cumulative distribution function distribution curve;
[0052] In some more preferred embodiments of the present invention, the method for statistical analysis of the distance dataset in step S4 further includes: calculating the coefficient of variation of the distance dataset.
[0053] The coefficient of variation test is performed using the following formula:
[0054] ;
[0055] Wherein, CV is the coefficient of variation of the dataset, and s is the standard deviation of the dataset. The average value of the dataset is calculated using the following formula:
[0056] ;
[0057] ;
[0058] Where n is the number of data points in the dataset.
[0059] i is the sequence number of the data in the dataset.
[0060] x i The data value with index i in the dataset.
[0061] The average value of the data in the dataset.
[0062] s is the standard deviation of the data in the dataset.
[0063] In some preferred embodiments of the present invention, the geometric feature extracted in step S3 is the radius r, resulting in a radius dataset. The method for statistical analysis of the radius dataset in step S4 includes the following steps:
[0064] (1) Draw a frequency distribution histogram based on the radius dataset and plot the probability density function curve;
[0065] (2) Analyze the plotted probability density function curve to evaluate the particle size distribution in the electrode section.
[0066] In some particularly preferred embodiments of the present invention, the method for analyzing the particles in the cross-section of the electrode includes the following steps:
[0067] S1. Image acquisition: Acquire a grayscale microscopic image of the electrode cross section. The microscopic image contains image information of the current collector and active material particles. The microscopic image is obtained by FIB-SEM and the magnification of the microscopic image is 200~1000 times.
[0068] S2. Image segmentation: Cluster analysis is performed on the pixels in the microscopic image obtained in step S1. The first segmentation is performed on the microscopic image using the K-means algorithm to segment the current collector region and the active material particle region. The second segmentation is performed on the active material particle region using the DBSCAN algorithm to extract each active material particle.
[0069] S3. Feature Extraction: Extract geometric features for each active substance particle obtained in step S2: Based on the position of the active substance particle in the microscopic image, obtain its centroid position, and calculate based on the centroid position to obtain the geometric features of the active substance particle, thereby obtaining the dataset.
[0070] The dataset includes distance data and radius data. The distance data d is obtained through the following steps: d = min(y2 - y1), where y2 is the vertical axis coordinate of the centroid of the active material particle, and y2 is the vertical axis coordinate of the boundary between the current collector region and the active material region.
[0071] Wherein, the radius data It is obtained through the following steps:
[0072] ;
[0073] Where m is the number of pixels at the particle boundary, (x c ,y c (x) represents the coordinates of the particle's centroid. j ,y j () represents the coordinates of the j-th pixel on the particle boundary;
[0074] S4. Based on the dataset, plot the cumulative probability density distribution function curve of the active particles and the reference cumulative distribution function curve, and then perform a chi-square test on the cumulative probability density distribution function curve and the reference cumulative distribution function curve.
[0075] The reference cumulative distribution function curve is plotted using the following function expression:
[0076] ;
[0077] The d m The maximum value of distance data d in the dataset;
[0078] The chi-square test is performed using the following formula:
[0079] ;
[0080] Among them, the This represents the chi-square statistic, where A represents the number of groups participating in the statistic, and i represents the index number of the i-th group currently being calculated; O i E represents the measured frequency belonging to the i-th group in the dataset. i This represents the theoretical frequency of the i-th group under the reference cumulative distribution function distribution curve;
[0081] The coefficient of variation (CV) was also calculated for the dataset. Finish;
[0082] Where s is the standard deviation of the dataset, and is the mean of the dataset, calculated according to the following formula:
[0083] ;
[0084] ;
[0085] Where n is the number of data points in the dataset;
[0086] i is the index of the distance data in the dataset;
[0087] d i The distance data value with index i in the dataset;
[0088] Let be the mean of the distance data in the dataset, and s be the standard deviation of the distance data in the dataset.
[0089] The positive and progressive effects of this invention are as follows:
[0090] 1. Precise image segmentation technology: Combining K-means clustering and DBSCAN clustering algorithms, it can accurately segment silicon particles and copper foil from complex electrode cross-section images, providing a reliable data foundation for subsequent distribution analysis.
[0091] 2. Multi-dimensional quantitative evaluation system: In addition to using the chi-square statistic to measure the uniformity of distribution, the coefficient of variation is also introduced as a supplementary indicator to comprehensively evaluate the distribution of silicon particles from different perspectives, making the evaluation results more accurate and comprehensive.
[0092] 3. Intuitive Visual Analysis: By plotting CDF curves, complex distribution data is displayed in an intuitive graphical form, making it easy for technicians to quickly understand and judge the uniformity of silicon particle distribution, specifically whether it is enriched near the copper foil or randomly enriched.
[0093] 4. High operability: The entire quantification method is based on common image processing and statistical analysis techniques, which are easy to implement and apply. It does not require complex equipment and high cost investment, and can be widely used in the product process optimization of silicon-carbon anode material manufacturers.
[0094] 5. This method uses the chi-square between the sample distribution function and the assumed uniform distribution function as the first criterion, which can better quantitatively describe its uniformity;
[0095] 6. This solution can provide spatial location information between active material particles and current collectors, thus reflecting the distribution of active material particles more comprehensively. Attached Figure Description
[0096] Figure 1 This is a FIB-SEM image of the electrode cross-section obtained in Example 1.
[0097] Figure 2 This is a clustering result obtained using the K-means algorithm in Example 1.
[0098] Figure 3 This is a clustering result obtained using the DBSCAN algorithm in Example 1.
[0099] Figure 4 The cumulative distribution function curve is obtained from the distance data of the centroid of silicon particles to the copper foil for each sample in Example 1.
[0100] Figure 5 The graph shows the chi-square statistics and coefficient of variation results for each sample in Example 1.
[0101] Figure 6 This is a PDF curve showing the particle size distribution of each sample in Example 2. Detailed Implementation
[0102] The present invention will be further illustrated by way of embodiments below, but the present invention is not limited to the scope of the embodiments described herein.
[0103] In this embodiment of the invention, the sample is prepared in the following manner:
[0104] ①Precursor preparation:
[0105] The silicon-carbon deposition sample SA04 obtained at 1500℃ was mixed with conductive carbon black Super P and binder PAA at a mass ratio of 80:10:10 to obtain precursor one.
[0106] The silicon-carbon deposition sample SA07 obtained at 1600℃ was mixed with conductive carbon black Super P and binder PAA at a mass ratio of 80:10:10 to obtain precursor II.
[0107] The silicon-carbon deposition sample SA08 obtained at 1700℃ was mixed with conductive carbon black Super P and binder PAA at a mass ratio of 80:10:10 to obtain precursor three.
[0108] The silicon-carbon deposition sample SA14 obtained at 1800℃ was mixed with conductive carbon black Super P and binder PAA at a mass ratio of 80:10:10 to obtain precursor four.
[0109] ② Mixing: Using a high-shear mixer with a speed of 2000 rpm, mix precursor one, precursor two, precursor three and precursor four for 30 minutes each to obtain slurry one, slurry two, slurry three and slurry four.
[0110] ③ Coating: Slurry 1, slurry 2, slurry 3 and slurry 4 are uniformly coated on copper foil with a thickness of 100~150 μm, and the coating thickness is controlled at 100-120 μm, to obtain sample 1, sample 2, sample 3 and sample 4 respectively.
[0111] ④ Drying: Dry samples 1, 2, 3 and 4 in an oven at 60℃ for 12 hours to ensure that the samples are completely dry and form a uniform coating. After drying, the coating thickness of samples 1, 2, 3 and 4 is 70~90μm, and the total thickness including the copper foil is 190~230μm.
[0112] These points ①, ②, ③, and ④ confirm that sample four has the most uniform particle distribution, while sample three has the most small-sized particles.
[0113] Example 1
[0114] S1. Image Acquisition: Using a Thermo Fisher Scientific DualBeam focusing ion beam scanning electron microscope, 3-5 images of the electrode sections of the four samples were taken. (Images are shown below.) Figure 1 The image recorded is a cross-sectional FIB-SEM image of the electrode obtained in Example 1, with a resolution of 1536×1024 pixels and a magnification of 1000 times.
[0115] S2. Image Preprocessing and Segmentation: After converting the image to grayscale, the K-means clustering algorithm is used, with a cluster label of 5 and a random seed of 0, for the first segmentation, separating the silicon particle and copper foil regions to obtain... Figure 2 , Figure 2 This is a clustering result obtained using the K-means algorithm in Example 1.
[0116] Then, the DBSCAN clustering algorithm was applied to the silicon particle region for a second segmentation, with a radius parameter of 5 pixels and a minimum number of points of 16, to further refine the segmentation into 100 silicon particles, thereby obtaining... Figure 3 , Figure 3 This is a clustering result obtained using the DBSCAN algorithm in Example 1.
[0117] S3. Feature Extraction: Obtain the ordinate y2 of the centroid of the silicon particle and the ordinate y1 of the boundary between the copper foil region and the silicon particle region. Calculate the nearest distance d from the centroid of each silicon particle to the copper foil region, where d = min(y2-y1). After traversing all silicon particles, a dataset containing 100 distance values is obtained.
[0118] S4. Statistical Analysis:
[0119] ① Plot the CDF curve based on the dataset, and then plot a uniformly distributed reference CDF curve, also known as the reference cumulative distribution function. The reference CDF curve is plotted using the following function expression:
[0120] ;
[0121] The d m This represents the maximum distance value in the dataset.
[0122] The comparison revealed a significant deviation between the actual CDF curve and the reference CDF curve, especially in the region near the copper foil.
[0123] ② Chi-square test and coefficient of variation calculation: The distance data for samples 1, 2, 3, and 4 are summarized separately, and CDF curves are plotted separately for chi-square test. The CDF curves of samples 1, 2, 3, and 4 are then plotted on the same coordinate axis to obtain the results. Figure 4 . Figure 4 The cumulative distribution function curves are obtained from the distance data of the centroid of silicon particles to the copper foil in samples one, two, three and four of Example 1.
[0124] The CDF curves of each sample were compared with the reference CDF curve using a chi-square test, which was performed using the following formula:
[0125] ;
[0126] Among them, the This represents the chi-square statistic, where A represents the number of groups included in the statistic, chosen as 100.
[0127] i represents the index number of the i-th group currently being calculated; O i E represents the measured frequency belonging to the i-th group in the dataset. i This represents the theoretical frequency of the i-th group under the reference cumulative distribution function distribution curve;
[0128] The coefficient of variation was then calculated on the dataset. Finish;
[0129] Wherein, s is the standard deviation of the dataset. The average value of the dataset is calculated using the following formula:
[0130] ;
[0131] ;
[0132] Where n is the number of data points in the dataset;
[0133] i is the index of the distance data in the dataset;
[0134] d i The distance data value with index i in the dataset;
[0135] The average distance data in the dataset. The standard deviation of the distance data in the dataset;
[0136] The CDF curves obtained from the calculated distances from the centroid of silicon particles to the copper foil for samples 1, 2, 3, and 4, compared with the reference CDF curve, showed chi-square statistics of 1150, 620, 1200, and 720, and coefficients of variation of 1.02, 0.88, 1.02, and 1.04, respectively. Based on the chi-square statistics and coefficients of variation for each sample, [further details are needed]. Figure 5 , Figure 5 The graph shows the chi-square statistics and coefficient of variation results for each sample in Example 1.
[0137] Uniformity assessment: Based on the chi-square statistic and coefficient of variation, the uniformity of silicon particle distribution in different samples was determined. According to the results, the chi-square statistic ranking is as follows: Sample 4 has the second smallest chi-square statistic and the largest coefficient of variation, indicating that it is relatively uniformly distributed in a wide range. In other words, the silicon particle distribution in Sample 4 is the most uniform, which is consistent with the expected results during preparation and effectively proves the reliability of this evaluation method.
[0138] Example 2
[0139] Steps S1 and S2 are the same as in Example 1, except that:
[0140] S3. Feature Extraction: Obtain the coordinates (x, y) of the centroid of the silicon particles. c ,y c ), the number of pixels m at the boundary between the copper foil region and the silicon particle region, and the coordinates (x, y) of the j-th pixel at the boundary between the copper foil region and the silicon particle region. j ,y j ), and calculate the radius of each silicon particle. The calculation formula is as follows:
[0141] ;
[0142] Where m is the number of pixels at the particle boundary, (x c ,y c (x) represents the coordinates of the particle's centroid. j ,y j ) represents the coordinates of the j-th pixel on the particle boundary.
[0143] r of all silicon particles i Values form the radius dataset .
[0144] S4. Statistical Analysis: Based on the radius dataset Draw a frequency distribution histogram and a PDF curve, also known as a probability density function curve.
[0145] By comparing and analyzing the PDF curves of different samples, the concentration and dispersion characteristics of particle size distribution in each sample were quantitatively assessed. The PDF curves of different samples were plotted on the same coordinate axis, and the results are shown below. Figure 6 As shown, Figure 6 This is a probability density function (PDF) curve of the particle size distribution of each sample in Example 2.
[0146] Depend on Figure 6 As can be seen, the probability density curve of sample 3 has the narrowest radius distribution range and the most obvious peak, which is concentrated in the range of smaller radii. The probability density curve radius distribution ranges of samples 1, 4, and 2 gradually widen, indicating that the particle size distribution gradually disperses. Therefore, sample 3 has the most small-sized particles, which is consistent with the expected results obtained from the preparation, further proving the reliability of this evaluation method.
[0147] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of the present invention is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but all such changes and modifications fall within the scope of protection of the present invention.
Claims
1. A method for analyzing the particle distribution in a cross-section of an electrode sheet, characterized in that, It includes the following steps: S1. Image acquisition: Acquire a microscopic image of the cross-section of the electrode, wherein the microscopic image contains image information of the current collector and active material particles; S2. Image segmentation: Cluster analysis is performed on the pixels in the microscopic image obtained in step S1. The microscopic image is segmented in the first stage to separate the current collector region and the active material particle region. The active material particle region is then segmented in the second stage to separate each active material particle. S3. Feature Extraction: Extract the geometric features of each active substance particle obtained in step S2 to obtain the dataset; S4. Statistical analysis: Perform statistical analysis on the dataset to obtain the particle distribution of the electrode cross-section.
2. The method for analyzing the particle distribution of electrode cross-sections according to claim 1, characterized in that, Step S1 satisfies one or more of the following conditions (1) to (5): (1) The microscopic image is obtained by a microscopic imaging device, preferably a SEM or a FIB-SEM; (2) The magnification of the microscopic image is 200 to 1000 times, preferably 400 times; (3) Step S1 also includes grayscale processing of the microscopic image; (4) The current collector includes one or more of copper foil, aluminum foil, and nickel foil, preferably copper foil; and, (5) The active material particles include silicon particles and / or graphite particles, preferably silicon particles.
3. The method for analyzing the particle distribution of electrode cross-sections according to claim 1, characterized in that, In step S2, the first segmentation uses the K-means clustering algorithm, and the second segmentation uses the DBSCAN clustering algorithm.
4. The method for analyzing the particle distribution of the electrode cross-section according to claim 3, characterized in that, Step S2 satisfies one or more of the following conditions (1) to (5): (1) The K value of the K-means algorithm is 2 to 6, preferably 3 to 5, for example 5; (2) The number of random seeds in the K-means algorithm is 0; (3) The radius of the DBSCAN algorithm is 1~10, preferably 3~10, for example 10; (4) The minimum number of points in the DBSCAN algorithm is 3 to 18, preferably 13 to 18, for example 16; (5) In the active substance particle region, the number of active substance particles is 50 to 200, preferably 80 to 120, for example 100.
5. The method for analyzing the particle distribution of electrode cross-sections according to claim 1, characterized in that, In step S3, the method for extracting the geometric features includes: The centroid position of the active substance particle is obtained, and the geometric characteristics of the active substance particle are calculated based on the centroid position.
6. The method for analyzing the particle distribution of electrode cross-sections according to claim 1, characterized in that, In step S3, the geometric features include one or more of the following (1) to (4): (1) The distance d of the active material particles, wherein the distance d refers to the distance from the active material particles to the boundary between the current collector area and the active material area; (2) The radius r of the active substance particles; (3) The perimeter l of the active substance particles; (4) The area S of the active substance particles.
7. The method for analyzing the particle distribution of an electrode cross-section according to claim 6, characterized in that, The method for extracting the distance d includes: (1) Obtain the vertical coordinate y2 of the centroid of the active material particle and the vertical coordinate y1 of the junction between the current collector region and the active material region, respectively; (2) Calculate the distance d: d = min(y2 - y1); (3) Repeat steps (1) and (2) to obtain the distance dataset; And / or, the method for extracting the radius r includes: (1) Obtain the coordinates (x, y) of the centroid of the active substance particles respectively. c ,y c The number of pixels m at the boundary between the current collector region and the active material region, and the coordinate (x, y) of the j-th pixel at the boundary between the current collector region and the active material region. j ,y j ); (2) Calculate the radius r: ； (3) Repeat steps (1) and (2) to obtain the radius dataset.
8. The method for analyzing the particle distribution of an electrode cross-section according to claim 1, characterized in that, In step S4, the statistical analysis method includes: Based on the dataset, plot the probability distribution function curve of the active substance particles; The probability distribution function curve is preferably a cumulative distribution function curve and / or a probability density function curve.
9. The method for analyzing the particle distribution of an electrode cross-section according to claim 8, characterized in that, The geometric feature extracted in step S3 is the distance d, resulting in a distance dataset. The method for statistical analysis of the distance dataset in step S4 includes the following steps: (1) Plot the cumulative distribution function curve based on the distance dataset, denoted as the actual cumulative distribution function curve, and plot the reference cumulative distribution function curve with uniform distribution; (2) The uniformity of particle distribution on the electrode cross-section is evaluated by comparing and analyzing the actual cumulative distribution function curve and the reference cumulative distribution function curve; Preferably, the comparative analysis includes the following methods ① and / or ②: ① Observe the degree of difference between the actual cumulative distribution function curve and the reference cumulative distribution function curve; ② Perform a chi-square test on the actual cumulative distribution function curve and the reference cumulative distribution function curve, and calculate the chi-square statistic; Preferably, the method for statistical analysis of the distance dataset in step S4 further includes: calculating the coefficient of variation of the distance dataset.
10. The method for analyzing the particle distribution of an electrode cross-section according to claim 8, characterized in that, The geometric feature extracted in step S3 is the radius r, resulting in a radius dataset. The method for statistical analysis of the radius dataset in step S4 includes the following steps: (1) Draw a frequency distribution histogram based on the radius dataset and plot the probability density function curve; (2) Analyze the plotted probability density function curve to evaluate the particle size distribution in the electrode section.