Pole piece thickness control method and device

Abstract

The invention provides a pole piece thickness control method and device, and the method comprises the steps: inputting a flatness scanning image of a pole piece into a preset appearance detection model, and obtaining a flatness feature parameter of the pole piece; calculating rolling parameters according to the flatness characteristic parameters; and controlling rolling equipment to roll the pole piece according to the rolling parameter set.

Description

Technical Field

[0001] This application relates to the field of lithium battery technology, and in particular to a method and apparatus for controlling electrode thickness. Background Technology

[0002] In the manufacturing process of lithium battery electrodes, the rolling process is a crucial step in controlling the thickness and density of the electrodes. Traditional methods for controlling electrode thickness mainly rely on manual experience and fixed parameter settings, which cannot be dynamically adjusted according to the actual flatness of the electrode surface. Traditional control methods also face the problem of inconsistent thickness in the transverse and longitudinal directions. When the thickness of the electrode is inconsistent from side to side, it is difficult to make precise pressure adjustments, and electrode rebound also affects the thickness control effect. Summary of the Invention

[0003] This application provides a method and apparatus for controlling electrode thickness, used to improve the thickness accuracy of electrode sheets during the rolling process.

[0004] In a first aspect, embodiments of this application provide a method for controlling electrode thickness, the method comprising: The flatness scan image of the electrode is input into a preset appearance detection model to obtain the flatness feature parameters of the electrode. Calculate the roll pressing parameters based on the flatness characteristic parameters; The rolling equipment is controlled to perform rolling processing on the electrode sheet according to the rolling parameter set.

[0005] In some embodiments, inputting the flatness scan image of the electrode into a preset appearance detection model to obtain the surface feature parameters of the electrode includes: The flatness scan image is preprocessed and then input into a preset appearance detection model for feature recognition to obtain flatness feature parameters. The flatness characteristic parameters include: edge thickness deviation value, warping coefficient, transverse thickness index, and longitudinal thickness change rate.

[0006] In some embodiments, calculating the rolling parameters based on the surface feature parameters includes: The pressure difference value on the opposite side is calculated based on the edge thickness deviation value. When the edge thickness exceeds the preset threshold of the center thickness, the rolling pressure in the corresponding area is reduced. The roll pressing temperature parameter is determined based on the warpage coefficient. When the warpage exceeds the critical value, the roll pressing temperature is increased and the hot roll mode is activated. The rolling speed parameters are calculated based on the transverse thickness index and the longitudinal thickness change rate. When the thickness inconsistency is large, the rolling speed is reduced. The rolling parameters are generated by combining the pressure difference value on the opposite side, the rolling temperature parameter, and the rolling speed parameter.

[0007] In some embodiments, controlling the rolling equipment to roll the electrode sheet according to the rolling parameter set includes: Adjust the roller pressure on the left and right sides of the roller press according to the pressure difference value on the opposite side, and apply a smaller pressure to the thick edge area to balance the transverse compaction density. The hot roller heating system is controlled according to the roller pressing temperature parameters to heat the warping risk area to reduce the inward and outward concavity phenomenon; The operating speed of the rolling equipment is controlled according to the rolling speed parameters, and low-speed rolling is used in areas with large thickness variations to prevent electrode rebound. The rolling stress changes are monitored in real time during the rolling process, and the subsequent rolling parameters are automatically adjusted when the thickness deviates from the target value.

[0008] In some embodiments, the step of preprocessing the flatness scan image and then inputting it into a preset appearance detection model for feature recognition to obtain flatness feature parameters includes: The flatness scan image is subjected to grayscale conversion and noise filtering to obtain a preprocessed image; The preprocessed image is input into a pre-defined appearance detection model that includes convolutional layers, pooling layers, and fully connected layers. The pooling layer is used to reduce the dimensionality of the convolutional features, and the ReLU activation function is used to enhance the nonlinear expressive power of the model. The pooled feature map is input into the fully connected layer for feature fusion, and the softmax classifier is used to identify thick edge regions, warped regions and regions with uneven thickness. Based on the recognition results, the geometric and statistical characteristics of each region are calculated, and the edge thickness deviation value, warping coefficient, lateral thickness index, and longitudinal thickness change rate are output.

[0009] Secondly, embodiments of this application provide an electrode thickness control device, which includes: The feature extraction module is used to input the flatness scan image of the electrode into a preset appearance detection model to obtain the flatness feature parameters of the electrode. The parameter calculation module is used to calculate the roll pressing parameters based on the flatness characteristic parameters; The equipment control module is used to control the rolling equipment to perform rolling processing on the electrode sheet according to the rolling parameter set.

[0010] Thirdly, embodiments of this application provide a manufacturing apparatus, the manufacturing apparatus including a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and, when executing the computer program, implement the electrode thickness control method as described in any of the embodiments of this application.

[0011] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to implement the electrode thickness control method as described in any of the embodiments of this application.

[0012] This application provides a method for controlling electrode thickness. The method includes: inputting a flatness scan image of the electrode into a preset appearance inspection model to obtain flatness feature parameters; calculating rolling parameters based on the flatness feature parameters; and controlling a rolling equipment to perform rolling processing on the electrode according to the rolling parameter set. In this method, by inputting the flatness scan image of the electrode into a preset appearance inspection model to obtain flatness feature parameters, the thick edge areas, warping deformation, and uneven thickness distribution on the electrode surface can be accurately identified. The rolling parameters calculated based on these feature parameters can be differentiated for different areas, effectively solving the problem of uneven thickness caused by traditional fixed parameter settings. During the rolling process, automatically adjusting the left and right rolling pressure, rolling temperature, and rolling speed according to the flatness characteristics can significantly reduce electrode warping, eliminate lateral and longitudinal thickness differences, and prevent electrode rebound, thereby greatly improving the accuracy and consistency of electrode thickness control and providing high-quality electrode products for subsequent slitting and winding processes. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A schematic flowchart illustrating an electrode thickness control method provided in an embodiment of this application; Figure 2 This is a schematic block diagram of an electrode thickness control device provided in an embodiment of this application. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described below with reference to the accompanying drawings.

[0016] The terms "first" and "second," etc., used in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0017] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0018] It should be understood that in this application, "at least one (item)" means one or more, "more than one" means two or more, "at least two (items)" means two or three or more, and "and / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0019] Please see Figure 1 , Figure 1 This is a schematic flowchart illustrating an electrode thickness control method provided in an embodiment of this application. Figure 1 As shown, the specific steps of this electrode thickness control method include: S101-S103.

[0020] S101. Input the flatness scan image of the electrode into a preset appearance detection model to obtain the flatness feature parameters of the electrode. For example, a laser scanning device scans the electrode surface line by line to acquire a flatness scan image containing height information. This image records the thickness distribution at each point on the electrode surface in grayscale values. The original scan image is preprocessed, including noise filtering, grayscale normalization, and size normalization, to ensure image quality meets the model input requirements. The preprocessed image is then fed into a deep learning-based appearance detection model. This model extracts image features through a multi-layer convolutional network to identify abnormal regions and thickness variation patterns on the electrode surface. The model automatically detects thick edges, calculating the edge thickness deviation by comparing the thickness difference between the edge and center regions. Simultaneously, it analyzes the curvature of the electrode surface, identifying deformation features such as concavity and convexity, and generating a warping coefficient. Statistical analysis is performed on the thickness distribution in the transverse direction of the electrode, calculating the ratio of the standard deviation to the mean of the thickness on both sides to obtain the transverse thickness index. The thickness variation trend at different locations along the longitudinal direction of the electrode is measured, and the longitudinal thickness variation rate is obtained through gradient calculation. These four key parameters constitute a complete set of flatness feature parameters, providing a quantitative data foundation for subsequent roll forming parameter calculations.

[0021] In one specific embodiment, laser scanning of a batch of electrodes revealed that the edge thickness was 8 micrometers higher than the center thickness. The model identified this as a typical thick-edge phenomenon, and the calculated edge thickness deviation value was 1.2. The electrode surface exhibited slight wavy deformation, with a warping coefficient of 0.3. The thickness difference between the left and right sides was small, with a lateral thickness index of 0.15, while the longitudinal thickness change was gradual, with a longitudinal thickness variation rate of 0.08. These flatness characteristic parameters reflect the overall quality of the electrode.

[0022] S102. Calculate the roll pressing parameters based on the flatness characteristic parameters; For example, the acquired flatness characteristic parameters are input into the rolling parameter calculation algorithm, which determines the optimal combination of rolling parameters based on a preset mapping relationship. When the edge thickness deviation exceeds a set threshold, the algorithm calculates the pressure difference on the opposite side and compensates for the thickness difference by reducing the rolling pressure corresponding to the thick edge area, ensuring that the compaction density of each area of ​​the electrode tends to be consistent after rolling. According to the magnitude of the warpage coefficient, the algorithm determines the appropriate rolling temperature parameter. When the coefficient exceeds the critical value, the rolling temperature is automatically increased and the hot roller heating function is activated to reduce the elastic deformation tendency of the material using the thermal effect. Combining the transverse thickness index and the longitudinal thickness change rate, the algorithm calculates the most suitable rolling speed parameter. When the thickness inconsistency is large, the rolling speed is reduced to ensure sufficient compaction and prevent the electrode from rebounding due to excessive speed. The algorithm also considers the mutual influence between different parameters and performs coupled optimization calculations to ensure that the parameters coordinate to achieve the best rolling effect. The calculated pressure difference on the opposite side, rolling temperature parameter, and rolling speed parameter are combined to form a complete set of rolling parameters.

[0023] In one specific embodiment, based on the aforementioned flatness characteristic parameters, the algorithm calculates that the left-side roller pressure should be 0.5 MPa lower than the right-side pressure to compensate for the thick edge effect. Since the warpage coefficient is 0.3, which is at a moderate level, the roller temperature is set to 85°C and the hot roller function is enabled. Considering the relatively small thickness index and rate of change, the roller speed is set to 2.5 m / min, ensuring both compaction effect and avoiding excessive stress concentration.

[0024] S103. Control the rolling equipment to perform rolling processing on the electrode sheet according to the rolling parameter set.

[0025] For example, the rolling mill receives the calculated set of rolling parameters and begins to execute a differentiated rolling control strategy. Based on the pressure difference between the two sides, the hydraulic cylinder pressures of the left and right rollers are adjusted independently using precision pressure control valves, ensuring that the thicker edge area experiences less rolling pressure while the central area maintains a normal pressure level. The hot roller heating device starts the heating program according to the set rolling temperature parameters, precisely controlling the roller surface temperature at the target temperature value through built-in resistance heating elements, providing appropriate preheating for the electrode sheet to pass through. The drive motor of the rolling mill adjusts its speed according to the calculated rolling speed parameters, ensuring that the electrode sheet passes through the roller gap at a constant speed through frequency conversion speed control. During the rolling process, a real-time monitoring device continuously detects the thickness change and stress distribution of the electrode sheet. When a deviation from the target thickness value is detected, the control device immediately adjusts the rolling parameters for subsequent areas, achieving dynamic optimization control. The entire rolling process remains continuous and stable, avoiding damage to the electrode sheet caused by sudden parameter changes, ensuring that the thickness uniformity and surface flatness of the rolled electrode sheet meet the process requirements.

[0026] In one specific embodiment, after the electrode enters the rolling area, the pressure of the left hydraulic cylinder is adjusted to 15 MPa, and the pressure of the right cylinder is set to 15.5 MPa, creating a pressure difference of 0.5 MPa. The temperature of the hot roller is stabilized at 85°C, and the electrode passes through at a uniform speed of 2.5 m / min. Monitoring shows that the thickness deviation of the electrode after rolling is controlled within ±2 micrometers, the warping phenomenon is significantly improved, and the consistency of thickness in both the transverse and longitudinal directions is enhanced.

[0027] To more clearly illustrate the technical solution of this application, the technical solution of this application will be described below through specific embodiments. It should be noted that the specific embodiments are used to expand the description of the technical solution of this application, and are not intended to limit this application.

[0028] In some embodiments, the flatness scan image of the electrode is input into a preset appearance detection model to obtain the surface feature parameters of the electrode, including: preprocessing the flatness scan image and then inputting it into the preset appearance detection model for feature recognition to obtain flatness feature parameters; wherein, the flatness feature parameters include: edge thickness deviation value, warping coefficient, transverse thickness index and longitudinal thickness change rate.

[0029] For example, a laser scanning device continuously scans the electrode surface to acquire a flatness scan image containing thickness distribution information. This image uses pixel grayscale values ​​to represent the height information of each point on the electrode. The original scanned image undergoes preprocessing operations, including grayscale conversion (converting the color image to a single-channel grayscale image), removing random noise generated during scanning using a Gaussian filter, and standardizing the image size to ensure a consistent input format. The preprocessed image is then fed into a pre-set appearance detection model for feature recognition. This model automatically extracts the geometric and texture features of the electrode surface using a deep learning algorithm. The model analyzes the thickness difference between the edge and center regions of the electrode, calculating the edge thickness deviation value by measuring the deviation between the maximum edge thickness and the average center thickness. This value reflects the severity of the thick edge phenomenon. Simultaneously, the model detects the bending deformation of the electrode surface, analyzing the distribution and degree of deformation features such as concavity and convexity, and generating a warping degree coefficient characterizing the overall warping level. Statistical analysis is performed on the thickness distribution in the lateral direction of the electrode, calculating the ratio of the variance to the mean of the thickness on both sides to obtain a lateral thickness index, which quantifies the consistency level of the lateral thickness. The thickness values ​​at multiple consecutive points along the longitudinal direction of the electrode are measured. By calculating the thickness gradient between adjacent points, the longitudinal thickness change rate is obtained, reflecting the thickness fluctuation of the electrode in the production direction. For example, if a batch of electrodes is found to have an edge thickness that is 12 micrometers thicker than the center, the model identifies this as a severe edge thickening phenomenon and calculates the edge thickness deviation value. At the same time, a slight wavy deformation feature is detected, generating a corresponding warping coefficient. The thickness analysis in the transverse and longitudinal directions produces corresponding exponential and rate-of-change values, respectively.

[0030] In some embodiments, calculating roll forming parameters based on surface feature parameters includes: calculating the pressure difference value on the opposite side based on the edge thickness deviation value; determining the roll forming temperature parameter based on the warpage coefficient, wherein the roll forming temperature is increased and a hot roll mode is enabled when the warpage coefficient exceeds a critical value; calculating the roll forming speed parameter based on the transverse thickness index and the longitudinal thickness change rate; and combining the pressure difference value on the opposite side, the roll forming temperature parameter, and the roll forming speed parameter to generate the roll forming parameters.

[0031] For example, the acquired surface feature parameters are input into a roll forming parameter calculation algorithm, which determines the optimal roll forming control strategy based on a preset parameter mapping relationship. When the edge thickness deviation exceeds a preset threshold, the algorithm automatically calculates the pressure difference on the opposite side and compensates for the thickness unevenness by applying a smaller roll forming pressure to the thicker edge area, ensuring that the compaction density of each area tends to be consistent after roll forming. The algorithm analyzes the magnitude of the warpage coefficient, and when the coefficient exceeds a set critical value, it automatically determines that the roll forming temperature needs to be increased and the hot roll mode needs to be activated, using the heating effect to soften the material and reduce its tendency for elastic deformation. Combining the values ​​of the transverse thickness index and the longitudinal thickness change rate, the algorithm calculates the most suitable roll forming speed parameter. When the thickness inconsistency is large, the roll forming speed is reduced to ensure sufficient compaction and prevent electrode rebound caused by excessive speed. The algorithm also considers the mutual coupling effect between various parameters and performs comprehensive optimization calculations to ensure the coordination of parameter combinations. The calculated pressure difference on the opposite side, roll forming temperature parameters, and roll forming speed parameters are combined according to a standard format to generate a complete roll forming parameter set, providing accurate instruction data for subsequent equipment control. For example, based on the detected thick edge phenomenon, the algorithm calculates that the rolling pressure on the left side should be reduced by a certain value compared with the right side to compensate for the thickness difference. It sets an appropriate rolling temperature according to the degree of warping and enables the hot roller function, and determines an appropriate rolling speed parameter considering the thickness consistency.

[0032] In some embodiments, controlling the rolling equipment to roll the electrode sheet according to the rolling parameter set includes: adjusting the pressure of the rollers on the left and right sides of the rolling equipment according to the pressure difference value on both sides, applying a smaller pressure to the thick edge area to balance the transverse compaction density; controlling the hot roller heating system according to the rolling temperature parameter; controlling the running speed of the rolling equipment according to the rolling speed parameter; and detecting the rolling stress in real time during the rolling process, and automatically adjusting the subsequent rolling parameters when the detected rolling stress deviates from the preset stress value.

[0033] For example, the rolling equipment receives the calculated set of rolling parameters and begins to execute a precise differentiated rolling control strategy. Based on the pressure difference between the two sides, the hydraulic drive devices of the left and right rollers are adjusted separately. Independent pressure control on both sides is achieved through precision pressure regulating valves, applying less pressure to the thicker areas to balance the lateral compaction density and ensure lateral consistency of the electrode thickness after rolling. The operating state of the hot roller heating device is controlled according to the calculated rolling temperature parameters. The built-in resistance heating element precisely controls the roller surface temperature to the target value, preheating areas at risk of warping to reduce the probability of concavity and convexity. The running speed of the rolling equipment is precisely controlled according to the rolling speed parameters. A frequency converter ensures the electrode passes through the roller gap at a constant speed. A low-speed rolling strategy is used for areas with large thickness variations to prevent electrode rebound. Throughout the rolling process, the changes in rolling stress are monitored in real time. Stress sensors continuously collect stress data at various points on the electrode. When the rolling stress deviates from the preset stress value, the control device immediately adjusts the rolling parameters for subsequent areas to achieve dynamic optimization control. For example, when the electrode passes through the rolling zone, the hydraulic cylinders on the left and right sides are set to different pressure levels according to the difference value, the temperature of the hot roller is kept within the calculated range, the equipment operating speed is strictly executed according to the calculation results, and the monitoring device provides real-time feedback on the rolling effect and triggers necessary parameter adjustments.

[0034] In some embodiments, the flatness scan image is preprocessed and then input into a preset appearance detection model for feature recognition to obtain flatness feature parameters. This includes: performing grayscale processing, median filtering for noise reduction, and histogram equalization on the flatness scan image to obtain a preprocessed image; inputting the preprocessed image into the preset appearance detection model to construct a grayscale co-occurrence matrix based on the preprocessed image, and performing convolutional feature extraction on the preprocessed image to obtain a convolutional feature matrix; extracting directional feature parameters based on the texture feature matrix in four directions with increasing angles; fusing the convolutional feature matrix and directional feature parameters based on a pooling layer to obtain fused features; performing feature learning and classification on the fused features to obtain region classification results, which include: identifying thick edge regions, warped regions, and regions with uneven thickness; calculating edge thickness deviation values, warping coefficients, lateral thickness indices, and longitudinal thickness change rates based on the region classification results; and generating flatness feature parameters based on the edge thickness deviation values, warping coefficients, lateral thickness indices, and longitudinal thickness change rates.

[0035] For example, during the process of grayscale conversion, median filtering for noise reduction, and histogram equalization enhancement of the flatness scan image, the computer converts the original RGB color flatness scan image into a 256-level grayscale image according to the weight formula 0.299R+0.587G+0.114B. At the same time, a 5×5 window median filter is used to effectively remove salt-and-pepper noise and random noise in the image. The histogram equalization enhancement technique expands the grayscale distribution of the image from the original concentration in the 128-180 range to the full range of 0-255, thereby obtaining a preprocessed image with enhanced contrast and reduced noise. When the preprocessed image is input into a preset appearance detection model, the computer constructs gray-level co-occurrence matrices in four directions (0°, 45°, 90°, 135°) based on the pixel gray-level values ​​of the preprocessed image. Each matrix has a size of 256×256 and records the co-occurrence frequency of pixel pairs at different gray levels under specific directions and distances. Simultaneously, multi-scale feature extraction is performed on the preprocessed image using 3×3 and 5×5 convolutional kernels to generate convolutional feature matrices containing edge, texture, and shape information. When extracting directional feature parameters in the four directions with increasing angles, the computer calculates 16 directional feature parameters based on the texture feature matrix, including energy (range 0.1-0.8), entropy (range 2.5-8.0), contrast (range 0.05-2.0), and correlation (range 0.3-0.95). In the process of fusing convolutional feature matrices and directional feature parameters using pooling layers, a combination of max pooling and average pooling is employed. The 512-dimensional convolutional feature matrix and 16-dimensional directional feature parameters are concatenated and weighted to generate a 528-dimensional fused feature vector. When performing feature learning and classification on the fused features, a deep neural network performs nonlinear mapping and classification processing, identifying regions with thick edges (15-25% of the total area), warped regions (5-15% of the total area), and regions with uneven thickness (10-30% of the total area). When calculating parameters based on the region classification results, the edge thickness deviation is obtained by measuring the difference between the thick edge region and the standard thickness (range 0.5-5.0 μm), the warping coefficient is calculated by analyzing the curvature changes in the warped region (range 0.1-2.5), and the lateral thickness index and longitudinal thickness change rate are obtained by statistically analyzing the thickness gradient changes in the uneven thickness regions in the horizontal and vertical directions (ranges 0.2-1.8 and 0.1-1.2, respectively). Based on these four key parameters, a weighted average algorithm is used to generate smoothness characteristic parameters for comprehensive evaluation.

[0036] In some embodiments, generating flatness feature parameters based on edge thickness deviation, warpage coefficient, lateral thickness index, and longitudinal thickness change rate includes: calculating the coupling interference factor between the edge thickness deviation and the warpage coefficient; when the coupling interference factor exceeds a first interference threshold, compensating and correcting the warpage coefficient to obtain a corrected warpage coefficient; calculating the thickness change interaction coefficient between the lateral thickness index and the longitudinal thickness change rate; when the thickness change interaction coefficient exceeds a second interference threshold, enabling a bidirectional compensation algorithm for the lateral thickness index and the longitudinal thickness change rate to generate a compensated thickness index; recalibrating the edge thickness deviation based on the corrected warpage coefficient and the compensated thickness index to obtain a calibrated edge deviation value; combining the calibrated edge deviation value, the corrected warpage coefficient, and the compensated thickness index to generate a calibration parameter set; and adaptively weighting the calibration parameter set to output the flatness feature parameters.

[0037] For example, when calculating the coupling interference factor between the edge thickness deviation value and the warpage coefficient, the Pearson correlation coefficient algorithm is used to analyze the linear correlation between the two parameters. When the absolute value of the correlation coefficient exceeds 0.7, it indicates a strong coupling relationship. At this time, the calculated coupling interference factor value ranges between 0.5 and 2.0. When the coupling interference factor exceeds the first interference threshold of 1.2, a compensation coefficient of 0.85 is introduced to linearly correct the warpage coefficient, resulting in a relatively stable corrected warpage coefficient. In the process of calculating the thickness change interaction coefficient between the transverse thickness index and the longitudinal thickness change rate, the nonlinear dependence between the two parameters is quantified using mutual information theory. The calculation results of the interaction coefficient are usually distributed in the range of 0.3-1.8. When the thickness change interaction coefficient exceeds the second interference threshold of 1.0, a bidirectional compensation algorithm is activated to simultaneously adjust the transverse thickness index by multiplying it by a correction factor of 0.92 and the longitudinal thickness change rate by multiplying it by a correction factor of 0.88, generating a compensated thickness index after the interaction effect has been eliminated. In the calculation of edge thickness deviation based on the corrected warp coefficient and the compensation thickness index, a ternary linear regression model is established to analyze the mathematical relationship among the three factors. The regression coefficients are solved using the least squares method, and the edge thickness deviation is re-estimated to obtain a calibrated edge deviation value with an error range controlled within ±0.3μm. The calibrated edge deviation value, the corrected warp coefficient, and the compensation thickness index are combined according to a fixed data structure to form a calibration parameter set containing three key parameters. When the calibration parameter set is adaptively weighted, the weight ratio is dynamically allocated according to the numerical stability and reliability of each parameter. The optimal weight combination is determined through an iterative optimization algorithm, and the flatness characteristic parameters are output.

[0038] In some embodiments, the calibration parameter set is adaptively weighted to output flatness feature parameters, including: calculating the standardized deviation of calibration edge deviation value, correction warpage coefficient and compensation thickness index in the calibration parameter set; generating a parameter deviation evaluation table based on the standardized deviation; generating a weighting coefficient table based on the parameter deviation evaluation table; and weighting and fusing the calibration parameter set based on the weighting coefficient table to obtain the flatness feature parameters.

[0039] For example, when calculating the standardized deviation of calibration edge deviation values, corrected warpage coefficients, and compensated thickness indices in the calibration parameter set, the Z-score standardization method is used to convert each parameter into a standard normal distribution value with a mean of 0 and a standard deviation of 1. The standardized deviation is obtained by calculating the degree of deviation of each parameter from its historical mean and dividing by the standard deviation. Based on the magnitude of the standardized deviation (usually distributed in the range of -3 to +3), a parameter deviation evaluation table is generated that records the deviation status of each parameter in detail. In the process of generating the weight coefficient table based on the parameter deviation evaluation table, parameters with an absolute deviation value less than 1.0 are assigned higher weights (0.4-0.5), parameters with an absolute deviation value between 1.0 and 2.0 are assigned medium weights (0.2-0.3), and parameters with an absolute deviation value greater than 2.0 are assigned lower weights (0.1-0.15). The weight coefficient table is generated under the constraint that the sum of the weight coefficients equals 1.0. When weighting and fusing the calibration parameter set according to the weighting coefficient table, the calibration edge deviation value, the correction warping coefficient, and the compensation thickness index are multiplied by their respective weighting coefficients and then summed to obtain the flatness characteristic parameters. The flatness characteristic parameters comprehensively reflect the surface quality of the electrode.

[0040] Please see Figure 2 , Figure 2 This is a schematic block diagram of an electrode thickness control device 200 provided in an embodiment of this application. The electrode thickness control device 200 is used to execute the aforementioned electrode thickness control method. The electrode thickness control device 200 can be configured in a server.

[0041] The server can be a standalone server, a server cluster, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0042] like Figure 2 As shown, the electrode thickness control device 200 includes: a feature extraction module 201, a parameter calculation module 202, and an equipment control module 203.

[0043] The feature extraction module 201 is used to input the flatness scan image of the electrode into a preset appearance detection model to obtain the flatness feature parameters of the electrode.

[0044] The parameter calculation module 202 is used to calculate the roll pressing parameters based on the flatness characteristic parameters.

[0045] The equipment control module 203 is used to control the rolling equipment to perform rolling processing on the electrode sheets according to the rolling parameter set.

[0046] This application provides a manufacturing apparatus, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and, when executing the computer program, implement the electrode thickness control method as described in any of the embodiments of this application.

[0047] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it causes the processor to implement an electrode thickness control method as described in any of the embodiments of this application.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for controlling electrode thickness, characterized in that, The method includes: The flatness scan image of the electrode is input into a preset appearance detection model to obtain the flatness feature parameters of the electrode. Calculate the roll pressing parameters based on the flatness characteristic parameters; The rolling equipment is controlled to perform rolling processing on the electrode sheet according to the rolling parameter set.

2. The electrode thickness control method as described in claim 1, characterized in that, The step of inputting the flatness scan image of the electrode into a preset appearance detection model to obtain the surface feature parameters of the electrode includes: The flatness scan image is preprocessed and then input into a preset appearance detection model for feature recognition to obtain flatness feature parameters. The flatness characteristic parameters include: edge thickness deviation value, warping coefficient, transverse thickness index, and longitudinal thickness change rate.

3. The electrode thickness control method as described in claim 1, characterized in that, The calculation of the rolling parameters based on the surface feature parameters includes: Calculate the pressure difference on the opposite side based on the edge thickness deviation value; The rolling temperature parameters are determined based on the warpage coefficient, wherein when the warpage coefficient exceeds a critical value, the rolling temperature is increased and the hot rolling mode is activated. The roll pressing speed parameters are calculated based on the transverse thickness index and the longitudinal thickness change rate. The rolling parameters are generated by combining the pressure difference value on the opposite side, the rolling temperature parameter, and the rolling speed parameter.

4. The electrode thickness control method as described in claim 1, characterized in that, The step of controlling the rolling equipment to perform rolling processing on the electrode sheet according to the rolling parameter set includes: Adjust the roller pressure on the left and right sides of the roller press according to the pressure difference value on the opposite side, and apply a smaller pressure to the thick edge area to balance the transverse compaction density. The hot roller heating system is controlled according to the roller pressing temperature parameters; The operating speed of the rolling equipment is controlled according to the rolling speed parameters. The rolling stress is detected in real time during the rolling process, and the subsequent rolling parameters are automatically adjusted when the rolling stress deviates from the preset stress value.

5. The electrode thickness control method as described in claim 2, characterized in that, The step of preprocessing the flatness scan image and then inputting it into a preset appearance detection model for feature recognition to obtain flatness feature parameters includes: The flatness scan image is subjected to grayscale conversion, median filtering for noise reduction, and histogram equalization enhancement to obtain a preprocessed image; The preprocessed image is input into the preset appearance detection model, a gray-level co-occurrence matrix is ​​constructed based on the preprocessed image, and convolutional feature extraction is performed on the preprocessed image to obtain a convolutional feature matrix; In four directions with increasing angles, directional feature parameters are extracted based on the texture feature matrix. The convolutional feature matrix and the directional feature parameters are fused using a pooling layer to obtain fused features; The fused features are subjected to feature learning and classification to obtain region classification results, which include: identifying thick-edge regions, warped regions, and regions with uneven thickness. Based on the region classification results, calculate the edge thickness deviation value, warping coefficient, lateral thickness index, and longitudinal thickness change rate; Flatness characteristic parameters are generated based on the edge thickness deviation value, the warping coefficient, the transverse thickness index, and the longitudinal thickness change rate.

6. The electrode thickness control method as described in claim 2, characterized in that, The process of generating flatness feature parameters based on the edge thickness deviation value, the warping coefficient, the transverse thickness index, and the longitudinal thickness change rate includes: Calculate the coupling interference factor between the edge thickness deviation value and the warping degree coefficient. When the coupling interference factor exceeds the first interference threshold, compensate and correct the warping degree coefficient to obtain the corrected warping coefficient. Calculate the thickness change interaction coefficient between the transverse thickness index and the longitudinal thickness change rate. When the thickness change interaction coefficient exceeds the second interference threshold, enable a bidirectional compensation algorithm for the transverse thickness index and the longitudinal thickness change rate to generate a compensated thickness index. Based on the corrected warpage coefficient and the compensation thickness index, the edge thickness deviation value is recalibrated to obtain the calibrated edge deviation value. The calibrated edge deviation value, the corrected warpage coefficient, and the compensation thickness index are combined to generate a calibration parameter set. The calibration parameter set is adaptively weighted to output the flatness feature parameters.

7. The electrode thickness control method as described in claim 6, characterized in that, The step of adaptively weighting the calibration parameter set and outputting smoothness feature parameters includes: Calculate the normalized deviation of the calibration edge deviation value, the corrected warpage coefficient, and the compensation thickness index in the calibration parameter set, and generate a parameter deviation evaluation table based on the normalized deviation. Generate a weight coefficient table based on the parameter deviation evaluation table; The calibration parameter set is weighted and fused according to the weight coefficient table to obtain the flatness feature parameters.

8. An electrode thickness control device, characterized in that, The electrode thickness control device is used to perform the electrode thickness control method as described in any one of claims 1-7, and the electrode thickness control device comprises: The feature extraction module is used to input the flatness scan image of the electrode into a preset appearance detection model to obtain the flatness feature parameters of the electrode. The parameter calculation module is used to calculate the roll pressing parameters based on the flatness characteristic parameters; The equipment control module is used to control the rolling equipment to perform rolling processing on the electrode sheet according to the rolling parameter set.

Images