A lithium ion battery pole piece uniformity detection method, system, product and storage medium

Abstract

A lithium ion battery pole piece uniformity detection method, system, product and storage medium, relate to the field of analyzing materials by means of determining the physical properties of materials. The method comprises: controlling a mechanical grinding device to remove the surface layer material of the pole piece to expose the interface to be measured; using a distance measuring sensor to scan the interface to obtain height data and calculate a height distribution matrix; moving the detection probe to the target position, and synchronously adjusting the vertical height of the probe according to the height value in the matrix; controlling the spectrometer to collect spectral signal data; extracting the characteristic peak intensity values of the active material and the binder, and obtaining the composition distribution value of the target position by calculating the ratio of the two. The technical scheme provided in the application improves the accuracy of lithium ion battery pole piece uniformity detection.

Description

Technical Field

[0001] This application relates to the field of materials analysis by measuring the physical properties of materials, and more particularly to a method, system, product, and storage medium for detecting the uniformity of lithium-ion battery electrode sheets. Background Technology

[0002] Currently, lithium-ion batteries are widely used in new energy vehicles and energy storage systems due to their high energy density and long cycle life. The manufacturing quality of the battery electrodes, especially the uniformity of the distribution of active materials, binders, and other components within the coating, determines the battery's electrochemical performance and safety.

[0003] In related technologies, physical layering or cross-sectional sample preparation combined with spectral analysis is commonly used to analyze the compositional distribution of electrodes. The surface material of the electrode is removed by mechanical scraping, cutting, or grinding to expose interfaces at specific depths within the electrode. Using detection equipment and specific scanning parameters, the exposed interfaces are scanned point-by-point or area-by-area. By reading the signal intensity of characteristic peaks of chemical composition at different locations, the spatial distribution of the material within the electrode can be inferred.

[0004] However, since the electrode is a composite material composed of hard active material particles and soft polymer binders, the mechanical removal process results in an exposed interface that is microscopically rough and uneven. In related technologies, when using spectroscopic equipment to scan this non-ideal plane, because the detection probe is usually maintained at a fixed focal plane height, the fluctuations in the microstructure cause the laser focus point to frequently deviate from the actual sample surface (defocusing) or produce irregular light scattering. This causes the acquired signal intensity to fluctuate due to differences in physical morphology, easily misinterpreting morphological noise as changes in component content, thus making it difficult to accurately reflect the true uniformity of the electrode. Summary of the Invention

[0005] This application provides a method, system, product, and storage medium for detecting the uniformity of lithium-ion battery electrodes, which can improve the accuracy of lithium-ion battery electrode uniformity detection.

[0006] The first aspect of this application provides a method for detecting the uniformity of lithium-ion battery electrodes, the method comprising:

[0007] The mechanical grinding device is controlled to remove a layer of surface material from the lithium-ion battery electrode according to preset depth parameters, exposing the interface to be tested. The ranging sensor is controlled to perform a grid-like scan of the interface to be tested, acquiring the height coordinate data of multiple discrete measurement points on the interface. Based on the height coordinate data, the height distribution matrix corresponding to the interface to be tested is calculated. The detection probe is controlled to move to the target test position, and the vertical height of the detection probe relative to the interface to be tested is adjusted synchronously according to the height value corresponding to the target test position in the height distribution matrix. The spectrometer is controlled to collect spectral signal data at the target test position. The first characteristic peak intensity value of the active material and the second characteristic peak intensity value of the binder are extracted from the spectral signal data, and the ratio of the second characteristic peak intensity value to the first characteristic peak intensity value is calculated to obtain the component distribution value at the target test position.

[0008] In the above embodiments, a height distribution matrix of the interface under test is constructed using a ranging sensor, and the detection probe is controlled to adjust its vertical height synchronously in real time according to this matrix. This compensates for microscopic surface undulations caused by mechanical grinding, ensuring that the laser focus of the spectrometer falls on the material surface at every test position. This eliminates optical signal intensity fluctuations caused by changes in focal length, allowing the extracted characteristic peak intensity ratio to accurately reflect the molar ratio of active material to binder. This eliminates morphological interference and improves the accuracy of lithium-ion battery electrode composition uniformity detection.

[0009] In conjunction with some embodiments of the first aspect, in some embodiments, controlling the spectrometer to acquire spectral signal data at the target test location specifically includes:

[0010] Based on the height distribution matrix, the local surface curvature of the target test location is calculated; it is determined whether the local surface curvature is concave or convex; when the local surface curvature is concave, the target test location is determined to be a thermally sensitive area, and a low-power excitation command is generated; when the local surface curvature is convex, the target test location is determined to be a light scattering enhanced area, and a high-power excitation command is generated; the spectrometer is controlled to emit laser light of corresponding power to the target test location according to the low-power or high-power excitation command and to collect the original spectral signal; the laser power value corresponding to the original spectral signal is obtained, and the intensity of the original spectral signal is normalized using the laser power value to obtain the spectral signal data.

[0011] In the above embodiments, the laser excitation power is dynamically adjusted based on the local surface curvature. Power is reduced in the concave thermosensitive region to avoid thermal damage and fluorescence interference, while power is increased in the convex scattering region to compensate for signal loss caused by light scattering. Combined with subsequent power normalization processing, this not only ensures the signal acquisition quality under complex microscopic surfaces but also eliminates the dimensional differences in intensity introduced by different excitation conditions, ensuring that the spectral data truly reflects the intrinsic properties of the material, thereby improving the accuracy of lithium-ion battery electrode uniformity detection.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, the control spectrometer emits laser light of corresponding power toward the target test location and acquires the original spectral signal according to a low-power excitation command or a high-power excitation command, specifically including:

[0013] Using the neighborhood height data centered on the target test location in the height distribution matrix, the surface gradient modulus of the target test location is calculated. If the surface gradient modulus is greater than a preset scattering threshold, an extended integration time command is generated. If the surface gradient modulus is less than or equal to the scattering threshold, a standard integration time command is generated. The spectrometer is controlled to acquire the original spectral signal according to the determined low-power excitation command or high-power excitation command, and based on the exposure time corresponding to the extended integration time command or standard integration time command. The exposure time value corresponding to the original spectral signal is obtained, and the intensity of the original spectral signal is time-normalized using the exposure time value.

[0014] In the above embodiments, the surface gradient modulus is used to identify large tilt angle regions and adaptively extend the exposure time, physically compensating for the deviation of the scattered light cone and signal attenuation caused by surface tilt. Combined with time normalization processing, the correlation between surface geometry and spectral response intensity is decoupled, eliminating signal artifacts caused by uneven roughness. This ensures that test data at different microscopic angles have a unified dimensional benchmark, thereby improving the accuracy and reliability of the electrode composition distribution characterization.

[0015] In conjunction with some embodiments of the first aspect, in some embodiments, a height distribution matrix corresponding to the interface to be measured is calculated based on height coordinate data, specifically including:

[0016] An original surface topography model is constructed using height coordinate data, and the local spatial curvature of each measuring point in the original surface topography model is calculated. Measuring points with local spatial curvature greater than a preset peak threshold are identified as adhesive stringing feature points. The height coordinate data of the adhesive stringing feature points is replaced and corrected using the average height of non-feature points in the neighborhood around the feature points to obtain corrected height coordinate data. A height distribution matrix is ​​generated based on the corrected height coordinate data.

[0017] In the above embodiments, the drawing peaks caused by the extension of the binder are identified by the local spatial curvature, and their height is corrected based on the neighborhood statistics. This eliminates the interference of physical burrs on the morphology model, restores the real electrode substrate layer, and ensures that the subsequent detection probe can penetrate the surface noise and focus on the real distribution plane of the active material and binder. This avoids the sampling deviation caused by focusing on the drawing tip, thereby improving the authenticity and accuracy of the micro-component characterization of the electrode.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the mechanical grinding device is controlled to remove a layer of surface material from the lithium-ion battery electrode according to preset depth parameters, exposing the interface to be tested, specifically including:

[0019] The thickness sensor in the control system measures the initial thickness of the lithium-ion battery electrode at the current grinding position; based on the initial thickness value and the preset depth parameters, the target thickness threshold for grinding termination is calculated; the mechanical grinding device is controlled to start rotating grinding, and the negative pressure adsorption module in the system is simultaneously turned on to collect dust at the first adsorption power; during the grinding process, the current thickness value fed back by the thickness sensor is obtained in real time, and when the current thickness value reaches the target thickness threshold, the mechanical grinding device is controlled to stop rotating and lift; the negative pressure adsorption module is controlled to switch the adsorption power to a second adsorption power higher than the first adsorption power, and is turned off after maintaining the preset cleaning time.

[0020] In the above embodiments, by combining real-time thickness closed-loop control with two-stage variable frequency negative pressure adsorption, the problem of dust residue interference is solved while ensuring accurate grinding depth. In particular, the high-power cleaning after grinding removes loose particles and debris attached to the surface, preventing them from obstructing or scattering noise to the spectral signal. A clean and representative test interface is constructed, ensuring that the signals subsequently acquired originate from the electrode body rather than processing waste, thereby improving the reliability and signal-to-noise ratio of the electrode uniformity detection data.

[0021] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the ratio of the intensity value of the second characteristic peak to the intensity value of the first characteristic peak to obtain the component distribution value of the target test location, the method further includes:

[0022] If the current grinding depth does not reach the preset total electrode thickness threshold, the preset depth parameter is reduced to obtain a new depth parameter. The new depth parameter is then used to control the mechanical grinding device to remove the surface material of the next gradient until a new composition distribution numerical sequence at multiple different depths is obtained at the target test location. Based on the new composition distribution numerical sequence, a longitudinal composition distribution model of the target test location is constructed, and the rate of change of the composition ratio between adjacent depths is calculated. Depth intervals where the rate of change of the composition ratio is greater than the preset smearing effect threshold are marked as abnormal gradient regions. The composition distribution values ​​within the abnormal gradient regions are then weighted and corrected to generate an evaluation index for the longitudinal uniformity of the electrode.

[0023] In the above embodiments, a longitudinal distribution model is constructed using multi-depth sequences, and smear artifacts caused by tool dragging are identified based on the rate of composition change. This strategy, by weighted correction of abnormal gradient regions, removes interlayer interference noise introduced by machining, restoring false compositional mutations to the true material distribution trend. This not only solves the problem of longitudinal resolution distortion in traditional grinding inspection but also ensures that the evaluation indicators truly reflect the electrochemical active structure inside the electrode, thereby improving the accuracy and reliability of the three-dimensional uniformity characterization of the electrode.

[0024] In conjunction with some embodiments of the first aspect, in some embodiments, controlling the spectrometer to acquire spectral signal data at the target test location specifically includes:

[0025] The spectrometer is controlled to perform a short-exposure pre-scan of the target test position to acquire the verification spectral signal; the total integrated intensity of the full spectrum of the verification spectral signal is calculated; when the total integrated intensity of the full spectrum is less than the preset void anomaly threshold, the detection probe is controlled to perform micro-step displacement within the preset neighborhood of the target test position until the acquired verification spectral signal meets the void anomaly threshold; the current probe position is locked, and the spectrometer is controlled to acquire spectral signal data for extracting characteristic peak intensity values.

[0026] In the above embodiments, a short-exposure pre-scanning mechanism is used to identify and avoid microscopic pore defects on the electrode surface in real time. When signal attenuation caused by laser light falling into pores is detected, a micro-stepping optimization strategy is triggered to ensure that the detection probe always locks onto valid material entities for sampling. This eliminates invalid data entry caused by porous structures at the source, ensures a strict correspondence between spectral signals and material composition, thereby eliminating the interference of structural noise on statistical results and improving the effectiveness and representativeness of electrode uniformity detection data.

[0027] In a second aspect, embodiments of this application provide a lithium-ion battery electrode uniformity detection 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 lithium-ion battery electrode uniformity detection system to perform the method described in the first aspect and any possible implementation thereof.

[0028] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a lithium-ion battery electrode uniformity detection system, cause the lithium-ion battery electrode uniformity detection system to perform the method described in the first aspect and any possible implementation thereof.

[0029] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a lithium-ion battery electrode uniformity detection system, cause the lithium-ion battery electrode uniformity detection system to perform the method described in the first aspect and any possible implementation thereof.

[0030] Understandably, the lithium-ion battery electrode uniformity detection 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 lithium-ion battery electrode uniformity detection method 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.

[0031] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0032] 1. This application utilizes a ranging sensor to construct a height distribution matrix of the interface under test, and controls the detection probe to perform real-time vertical height synchronous adjustment according to this matrix. This compensates for microscopic surface undulations caused by mechanical grinding, ensuring that the laser focus of the spectrometer falls on the material surface at every test position. It eliminates optical signal intensity fluctuations caused by focal length changes, allowing the extracted characteristic peak intensity ratio to accurately reflect the molar ratio of active material to binder, thereby eliminating morphological interference and improving the accuracy of lithium-ion battery electrode composition uniformity detection.

[0033] 2. This application dynamically adjusts the laser excitation power based on the local surface curvature. Power is reduced in the concave thermosensitive region to avoid thermal damage and fluorescence interference, while power is increased in the convex scattering region to compensate for signal loss caused by light scattering. Combined with subsequent power normalization processing, this not only ensures the signal acquisition quality under complex microscopic surfaces but also eliminates the dimensional differences in intensity introduced by different excitation conditions, ensuring that the spectral data truly reflects the intrinsic properties of the material, thereby improving the accuracy of lithium-ion battery electrode uniformity detection.

[0034] 3. This application utilizes surface gradient modulus to identify large tilt angle regions and adaptively extends the exposure time, physically compensating for the deviation of the scattered light cone and signal attenuation caused by surface tilt. Combined with time normalization processing, the correlation between surface geometry and spectral response intensity is decoupled, eliminating signal artifacts caused by uneven roughness. This ensures that test data at different microscopic angles have a unified dimensional benchmark, thereby improving the accuracy and reliability of the electrode composition distribution characterization. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating a method for detecting the uniformity of lithium-ion battery electrode sheets in an embodiment of this application.

[0036] Figure 2 This is another flowchart illustrating the method for detecting the uniformity of lithium-ion battery electrode sheets in this application.

[0037] Figure 3 This is an exemplary hardware structure diagram of a lithium-ion battery electrode uniformity detection system in an embodiment of this application. Detailed Implementation

[0038] 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 and appended claims of this application, the singular expressions “a,” “an,” “the,” “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 and includes any or all possible combinations of one or more of the listed items.

[0039] 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.

[0040] In related technologies, to analyze the internal composition distribution of lithium-ion battery electrodes, surface materials are typically removed through mechanical grinding or physical delamination to expose the interface to be tested at a specific depth. This interface is then scanned point-by-point using spectroscopic equipment, and the spatial distribution of the material is inferred by reading the signal intensity of characteristic peaks of the chemical composition. However, since electrodes are composite materials composed of hard active material particles and soft polymer binders, the mechanical removal process often results in a microscopically uneven and rough interface. In these technologies, the detection probe is typically maintained at a fixed focal plane height for scanning. This microscopic morphology fluctuation can cause the laser focus point to frequently deviate from the actual sample surface (i.e., defocusing) or produce irregular light scattering. Therefore, the acquired signal intensity will fluctuate nonlinearly due to differences in physical morphology, making it easy to misinterpret morphological noise as changes in component content, thus failing to accurately reflect the true uniformity of the electrode.

[0041] This application proposes a method for detecting the uniformity of lithium-ion battery electrode sheets. The method first controls a mechanical grinding device to remove the surface layer of the electrode sheet to expose the interface to be tested. A ranging sensor then performs a gridded scan of this interface, constructing a height distribution matrix reflecting microscopic undulations. During the detection process, the detection probe is controlled to adjust its vertical height relative to the interface in real time according to the height values ​​at corresponding positions in this height distribution matrix. This allows for the acquisition of spectral signals and the calculation of the ratio of characteristic peaks of the active material to those of the binder. This scheme effectively compensates for surface roughness caused by mechanical grinding through a dynamic focusing mechanism, ensuring that the laser focus accurately falls on the material surface at each test position. It eliminates optical signal fluctuations caused by changes in focal length, thereby eliminating morphological interference and improving the accuracy of electrode composition uniformity detection.

[0042] Figure 1 This is a flowchart illustrating the method for detecting the uniformity of lithium-ion battery electrode sheets in the embodiments of this application, including the following steps:

[0043] S101. Control the mechanical grinding device to remove a layer of surface material from the lithium-ion battery electrode according to the preset depth parameters, exposing the interface to be tested.

[0044] Among them, mechanical grinding equipment refers to mechatronic equipment used for precision removal and polishing of material surfaces, such as high-precision rotary grinding and polishing machines or CNC milling platforms; lithium-ion battery electrode refers to a composite electrode material consisting of a current collector (such as copper foil or aluminum foil) and an active material layer coated thereon (containing active particles, conductive agents and binders); preset depth parameter refers to a pre-set material thickness value that needs to be removed downward from the original surface of the electrode. This value is usually based on the total coating thickness of the electrode and the required longitudinal detection resolution (such as detection every 5 micrometers), and is derived by experimental calibration or by those skilled in the art based on the electrode design specifications; the interface to be tested refers to the internal cross-section of the electrode exposed after the surface material is removed, which is used for subsequent spectral detection.

[0045] Specifically, firstly, based on preset depth parameters (e.g., removing 10μm), the feed mechanism of the mechanical grinding device (such as a Z-axis stepper motor) is controlled to lower the grinding head to the designated position. Then, the grinding power source is activated, utilizing the grinding media (such as sandpaper, diamond grinding discs, or milling cutters) to generate relative motion with the electrode surface. Under physical friction or cutting action, the composite material on the electrode surface is mechanically peeled off, crushed, and removed.

[0046] To ensure that the exposed interface to be tested has good flatness for subsequent optical inspection, the grinding process is usually accompanied by precise pressure and speed control to ensure that the removal process is uniform and carried out layer by layer, and finally a new physical plane, i.e. the interface to be tested, is formed at a predetermined depth.

[0047] In some embodiments, grinding control of this step can be achieved in a variety of ways:

[0048] Optionally, CNC milling can be used: First, a laser thickness gauge is used to determine the zero point position of the current surface of the electrode; second, a high-speed rotating diamond milling cutter is controlled to move on the surface of the electrode according to a grating path, while the Z-axis indentation is set to a preset depth parameter; finally, the milling area is swept by high-pressure airflow to complete the material removal.

[0049] It is understandable that other methods can also be used, such as laser ablation or ion beam etching, which are not limited here.

[0050] In some embodiments, where a high binder content in the electrode leads to material softening and easy smearing, a cryogenic assisted grinding mechanism can be introduced. Specifically, because soft binders easily soften and coat the surface of hard particles under the heat of grinding at room temperature, resulting in an blurred interface, while controlling the operation of the mechanical grinding device, liquid nitrogen spray or a semiconductor cooling module is controlled to locally cool the grinding area, lowering the binder temperature on the electrode surface below the glass transition temperature, making it brittle. Grinding in this state allows the binder to be removed through brittle fracture rather than plastic deformation and smearing, thereby exposing the clearly defined interface between the active material particles and the binder distribution.

[0051] In other embodiments, when there are tolerance fluctuations in the macroscopic thickness of the electrode and grinding dust is prone to secondary pollution, an adaptive grinding and graded negative pressure cleaning strategy based on thickness closed-loop feedback can be adopted to achieve precise control of grinding depth and thorough removal of interface residues, thereby improving the cleanliness and layer accuracy of the interface to be tested.

[0052] Specifically, firstly, the high-precision thickness sensor integrated on the grinding platform (such as a contact dial indicator or a non-contact laser thickness gauge) is activated to measure the initial thickness of the electrode position to be ground, obtaining the initial thickness value T. initial This eliminates the interference of manufacturing tolerances of the electrode itself (such as thicker coating at the edges and thinner coating in the middle) on the absolute depth determination. Next, the processor reads the preset depth parameter D. set (This parameter is usually set by process engineers based on the average particle size of the electrode active material particles and the required longitudinal resolution, after experimental verification, such as setting it to 5μm or 10μm), and the target thickness threshold T for grinding termination is calculated. target =T initial -D setAfter establishing the target, the mechanical grinding device is started, and the grinding disc begins to rotate and descend. Simultaneously, the negative pressure adsorption module is activated at its first adsorption power (e.g., 50% of the rated power). This first adsorption power is set to generate a moderate laminar negative pressure, which effectively removes dust generated during grinding, preventing it from spreading into the equipment, while avoiding excessive suction that could cause the electrode to vibrate or deform, affecting the grinding smoothness. During grinding, the thickness sensor continuously monitors the current thickness of the electrode at a high-frequency sampling rate and feeds it back to the controller. The controller constantly compares the current thickness with the target thickness threshold. When the two are equal or the difference enters the allowable error range, a stop command is immediately issued, controlling the grinding device to stop rotating and quickly raising the cutting head, thus achieving immediate stopping upon reaching the target thickness and precisely ensuring the uniformity of the removed thickness. At the moment grinding stops, although large dust particles have been removed, fine binder powder may be electrostatically adsorbed onto the newly exposed interface. At this point, the negative pressure adsorption module instantly switches its power to the second adsorption power (e.g., 100% or higher of the rated power). This high-intensity pulsed negative pressure airflow generates powerful shearing force, overcoming the van der Waals forces and electrostatic forces between fine dust and the interface, thus removing stubborn residual dust. This high-power state is maintained for a preset cleaning time (e.g., 3-5 seconds, determined based on dust settling rate experiments) before being shut off, ultimately exposing the interface.

[0053] By adopting the above technical solution, on the one hand, the blind grinding error caused by uneven thickness of electrode raw materials is overcome by calculating the closed-loop logic of grinding control, ensuring that each removal is a layer of equal thickness relative to the current surface; on the other hand, by using graded negative pressure control with low suction during grinding and high suction after shutdown, the processing stability is guaranteed, the problem of secondary dust adsorption is solved, false component interference is avoided, and the signal-to-noise ratio of spectral detection is improved.

[0054] S102. Control the ranging sensor to perform a grid-like scan of the interface to be measured, and obtain the height coordinate data of multiple discrete measuring points on the interface to be measured.

[0055] Among them, a distance sensor refers to a precision sensor that can measure the distance from the sensor probe to the surface of the object being measured in a non-contact manner, such as a laser triangulation displacement sensor or a spectral confocal sensor; grid scanning refers to the sensor traversing and covering a certain area point by point or row by row according to a pre-planned orthogonal grid path of the X and Y axes; discrete measurement points refer to specific sampling positions selected at fixed intervals at grid intersections or scanning paths; height coordinate data refers to the longitudinal position information (usually Z-axis coordinate values) used to represent each discrete measurement point relative to the reference plane.

[0056] Specifically, because the interface after mechanical grinding is not an absolutely ideal plane at the microscopic scale, but has uneven roughness, it will affect the focusing of subsequent spectral detection. A precision displacement platform equipped with a ranging sensor is controlled to move on the XY plane of the interface under test according to a set scanning step size (e.g., one point every 10 μm). At each grid node, the ranging sensor emits a measurement beam (such as a laser) to illuminate the surface of the interface under test and receives the reflected light signal. By analyzing the changes in the spot position of the reflected light signal (triangulation) or the changes in the wavelength peak (confocal method), the sensor calculates the precise distance from the current point to the sensor, and combines this with the current XY coordinates to record the absolute height value (Z value) of that point, thereby obtaining the discrete measurement point height data.

[0057] In some embodiments, grid scanning can be implemented in a variety of ways:

[0058] Optionally, a point-by-point pause scanning mode is adopted: First, the displacement platform is controlled to move to the first measurement point coordinates (X1, Y1) and come to a complete stop; second, the distance sensor is triggered to perform a single measurement and read stable Z-axis data; finally, it moves to the next measurement point (X2, Y1) and repeats the above process until the entire area is covered.

[0059] Optionally, a continuous scanning mode is adopted: First, the displacement platform is controlled to move continuously along the X-axis at a constant speed; second, the ranging sensor continuously and at high speed collects height data according to a fixed time frequency or position trigger signal; finally, the collected time-series height data is synchronously matched with the position data of the displacement encoder to generate a coordinate point set.

[0060] It is understandable that other methods can also be used, such as using a structured light 3D camera to perform area array imaging to obtain height data, which is not limited here.

[0061] S103. Based on the height coordinate data, calculate the height distribution matrix corresponding to the interface to be measured.

[0062] The height distribution matrix refers to a two-dimensional data structure (such as an m×n numerical matrix), where each element corresponds to the Z-axis height value at a specific spatial location on the interface under test. Mathematically, this matrix represents the microscopic morphological surface of the interface under test.

[0063] Specifically, the raw height coordinate data may be a series of unordered (x, y, z) tuples or simply arranged in chronological order. The processor first establishes a virtual two-dimensional grid plane corresponding to the physical dimensions based on the grid settings used during scanning. Then, the height coordinate data of each discrete measurement point is mapped to the corresponding node in this virtual grid. If the original sampling point does not completely coincide with the grid node, the height value at the grid node is calculated using an interpolation algorithm (such as bilinear interpolation or nearest neighbor interpolation). Finally, a matrix H containing the height information of all grid nodes is generated, where H(i, j) represents the microscopic height at the position in the i-th row and j-th column.

[0064] In some embodiments, matrix calculations can be implemented in multiple ways:

[0065] Optionally, a direct mapping method can be used: First, initialize an all-zero matrix with dimensions consistent with the number of scan grids; second, traverse all collected discrete measurement point data; finally, directly fill the Z value of each measurement point into the matrix index position closest to the XY coordinate.

[0066] Optionally, a surface fitting interpolation method can be used: First, a least squares fitting surface or spline function surface is constructed using the data of all discrete measurement points; second, regular grid coordinate points are generated according to the preset detection resolution; finally, the height value of each regular grid point is calculated by substituting it into the fitting surface equation and filling it into the matrix.

[0067] It is understandable that other methods can also be used, such as generating a smooth height distribution matrix based on the Kriging interpolation algorithm, which is not limited here.

[0068] S104. Control the detection probe to move to the target test position, and synchronously adjust the vertical height of the detection probe relative to the interface to be tested according to the height value corresponding to the target test position in the height distribution matrix.

[0069] The detection probe refers to the optical front-end component of the spectrometer, which usually includes a microscope objective for laser focusing and signal collection; the target test position refers to the XY plane coordinate point where spectral acquisition is currently planned; and the vertical height refers to the physical distance (working distance) between the probe objective and the sample surface.

[0070] Specifically, when deciding to inspect a specific location (target test location), the XY-axis displacement platform is first controlled to move the inspection probe horizontally above that location. Simultaneously, the height distribution matrix is ​​queried to retrieve the surface height value Z corresponding to the target test location. target Based on this height value and the standard focal length F of the probe objective, the absolute Z-axis position Z that the probe should be in is calculated. probe =Z target+F). Subsequently, the Z-axis brake (such as a piezoelectric ceramic motor or stepper motor) drives the detection probe to rise and fall vertically to the calculated position. This process compensates for the unevenness of the electrode surface caused by grinding roughness, ensuring that the laser focus always falls precisely on the surface of the electrode material, rather than focusing in the air or deep within the material.

[0071] In some embodiments, synchronization can be achieved in a variety of ways:

[0072] Optionally, open-loop preset position control is adopted: First, the Z value of the target point in the matrix is ​​read; second, the number of pulses required for the Z-axis motor to move is calculated; finally, pulse commands are directly sent to drive the motor to the position without real-time position feedback confirmation.

[0073] It is understandable that other methods can also be used, such as using a voice coil motor for high-frequency dynamic position following, which is not limited here.

[0074] In other embodiments, when high-frequency micro-ripples exist on the electrode surface and the scanning speed is fast, causing a lag in the Z-axis motor response, feedforward predictive compensation control can be employed. Specifically, due to the inertia and response delay of the mechanical motor, directly following the height matrix may result in a slight delay. Based on the height data of the target test position in the height distribution matrix and a series of subsequent planned paths, Kalman filtering or multinomial extrapolation algorithms are used to predict the surface height change trend at the next moment. An action command is sent to the Z-axis motor a certain preset time in advance, using the predicted value to offset the physical lag of the mechanical system, thereby achieving a more accurate dynamic focusing effect in high-speed motion detection.

[0075] S105. Control the spectrometer to collect spectral signal data at the target test location.

[0076] Among them, a spectrometer refers to an instrument used to analyze the composition of light waves, specifically a Raman spectrometer, which includes a laser, a spectroscopic system, and a detector; spectral signal data refers to the digital information recorded by the detector on the change of light intensity with wavelength (or Raman shift), usually represented as a spectrum.

[0077] Specifically, the laser inside the spectrometer is triggered to emit a monochromatic laser of a specific wavelength towards the target test location via optical fiber and a detection probe. The laser undergoes inelastic scattering (Raman scattering) with the active materials and binder molecules on the electrode surface. The scattered light carries information about the molecular vibrational energy levels, returns along its original path, and is collected by the detection probe. The grating inside the spectrometer disperses the mixed scattered light according to wavelength, projecting it onto a CCD or CMOS detector. The detector converts the optical signal into an electrical signal, which is then converted from an analog-to-digital converter (A / D) to ultimately generate spectral signal data containing a series of characteristic peaks.

[0078] In some embodiments, spectral acquisition can be achieved in a variety of ways:

[0079] Optionally, single-point static acquisition can be used: First, keep the probe and sample relatively stationary; second, open the laser shutter and start integration for a preset exposure time (e.g., 1 second); finally, close the shutter, read the charge data on the detector and output it.

[0080] It is understandable that other methods can also be used, such as Time Delay Integration (TDI) mode data acquisition, which is not limited here.

[0081] In other embodiments, when the microstructure of the electrode surface is complex and the thermal stability of different components varies greatly, an adaptive excitation and acquisition strategy based on morphology feature feedback can be adopted to maximize signal acquisition efficiency while preventing thermal damage, thereby obtaining spectral data with high signal-to-noise ratio and lateral comparability.

[0082] Specifically, the geometric information contained in the previously constructed height distribution matrix is ​​used to dynamically optimize the spectrometer's acquisition parameters point by point. First, based on the Z-axis data of the target test location and its neighborhood in the height distribution matrix, the local surface curvature of that point is calculated using a discrete differential operator (such as the Laplacian operator). The surface morphology is determined based on the sign and magnitude of the curvature: if the curvature is negative (or less than a certain threshold), it indicates a concave morphology, usually corresponding to pores rich in binder or heat traps with poor heat dissipation. In this case, it is identified as a thermally sensitive area, and a low-power excitation command is generated to avoid carbonization of the polymer material due to thermal accumulation or the generation of strong fluorescence interference; if the curvature is positive, it indicates a convex morphology, usually corresponding to hard active material particles, with good heat dissipation but strong scattering. In this case, it is identified as a light scattering enhancement area, and a high-power excitation command is generated to enhance the number of Raman scattering photons.

[0083] Simultaneously, the surface gradient modulus at this location (i.e., the tangent of the angle between the surface normal vector and the optical axis) is calculated to assess the degree of surface tilt. If the gradient modulus is greater than a preset scattering threshold (this threshold is typically determined through optical simulation or standard plate experiments based on the numerical aperture NA of the detection probe objective and the light collection cone angle), it indicates severe surface tilt, and the reflected light cone may deviate from the receiving aperture, leading to signal loss. Therefore, an extended integration time command is generated to compensate for photon loss by increasing the exposure time; otherwise, a standard integration time command is generated. Subsequently, the spectrometer executes the determined power and time parameters for physical acquisition. Finally, to eliminate intensity differences introduced by different excitation conditions, the actual executed laser power and exposure time values ​​are obtained, and the intensity of the acquired raw spectral signal is linearly normalized (i.e., divided by the product of power and time) to restore the intrinsic spectral intensity, which is only related to the amount of material.

[0084] By employing the above technical solution and utilizing a morphology-physical-parameter mapping mechanism, the contradiction between overexposure burns and tilting loss in rough surface detection is resolved. Low power protects the thermosensitive components, long integrals compensate for scattering losses, and normalization ensures that the data from all measurement points are on the same dimensional standard. This allows the final calculated composition distribution values ​​to truly reflect the differences in the chemical composition of the material, rather than optical artifacts caused by geometric morphology.

[0085] S106. Extract the first characteristic peak intensity value of the active substance and the second characteristic peak intensity value of the binder from the spectral signal data, and calculate the ratio of the second characteristic peak intensity value to the first characteristic peak intensity value to obtain the component distribution value at the target test location.

[0086] Among them, active material refers to the material responsible for energy storage in the electrode (such as lithium cobalt oxide, ternary materials, etc.); binder refers to the polymer material used to bond particles (such as PVDF); characteristic peak refers to the peak position in the Raman spectrum corresponding to the vibration of a specific chemical bond; intensity value refers to the peak height or peak area of ​​the characteristic peak; composition distribution value refers to the quantitative index obtained by calculation that characterizes the relative content of binder at that point.

[0087] Specifically, the acquired raw spectral signal data is first preprocessed, including baseline removal (removing fluorescence background) and smoothing / denoising. Next, based on a pre-established database, the characteristic bands of the active material (e.g., peaks near 500-600 cm⁻¹ for NCM materials) and the characteristic bands of the binder are located in the spectrum. Peak-finding algorithms are used to determine the maximum peak intensity within these bands, or the peak area is calculated by integrating the spectral curves within the bands, yielding the first characteristic peak intensity value I. active Second characteristic peak intensity value I binder Finally, perform the division operation to calculate the ratio R=I. binder / I active This ratio eliminates common-mode interference such as laser power fluctuations and minor focusing errors, and can reflect the degree of enrichment of the binder relative to the active material at this microscopic location, i.e., the component distribution value.

[0088] In some embodiments, extraction and calculation can be achieved in multiple ways:

[0089] Optionally, the peak-to-peak ratio method can be used: First, the position of the peak of the characteristic peak is determined by Gaussian fitting or Lorentz fitting; second, the ordinate of the peak of the fitted curve is read as the intensity value; finally, the ratio of the two ordinates is directly calculated.

[0090] Optionally, the integral area ratio method can be used: First, determine the wavenumbers of the left and right boundaries of the characteristic peak; second, perform numerical integration on the region enclosed by the spectral curve and the baseline within the boundary to obtain the peak area; finally, calculate the ratio of the two peak areas.

[0091] It is understandable that other methods can also be used, such as using chemometric methods (such as principal component analysis (PCA) or partial least squares method (PLS) to directly predict the component content, which is not limited here.

[0092] In the above embodiments, a height distribution matrix of the interface under test is constructed using a ranging sensor, and the detection probe is controlled to adjust its vertical height synchronously in real time according to this matrix. This compensates for microscopic surface undulations caused by mechanical grinding, ensuring that the laser focus of the spectrometer falls on the material surface at every test position. This eliminates optical signal intensity fluctuations caused by changes in focal length, allowing the extracted characteristic peak intensity ratio to accurately reflect the molar ratio of active material to binder. This eliminates morphological interference and improves the accuracy of lithium-ion battery electrode composition uniformity detection.

[0093] In other embodiments of this application, when the electrode binder content is high and has high ductility, mechanical grinding may cause the binder to fail to break brittlely and instead be stretched into fine filaments, forming false height spikes. Using the lithium-ion battery electrode uniformity detection method provided in this application, these filament-drawing feature points can be identified and corrected by calculating local curvature, eliminating morphological noise and ensuring that the detection probe is focused on the true interface.

[0094] like Figure 2 The diagram shown is another flowchart illustrating the lithium-ion battery electrode uniformity detection method provided in this application embodiment, including the following steps:

[0095] S201. Control the mechanical grinding device to remove a layer of surface material from the lithium-ion battery electrode according to the preset depth parameters, exposing the interface to be tested.

[0096] S202. Control the ranging sensor to perform a grid-like scan of the interface to be measured, and obtain the height coordinate data of multiple discrete measuring points on the interface to be measured.

[0097] Steps S201-S202 and Figure 1 Steps S101-S102 in the illustrated embodiment are similar and can be found in the descriptions of steps S101-S102, which will not be repeated here.

[0098] S203. Construct the original surface topography model using the elevation coordinate data, and calculate the local spatial curvature of each measuring point in the original surface topography model.

[0099] Among them, the original surface morphology model refers to a three-dimensional digital surface that can reflect the micro-geometric characteristics of the interface under test, reconstructed by mathematical interpolation or meshing methods based on a discrete set of height coordinate data points; local spatial curvature refers to a geometric quantity used to describe the degree and direction of curvature of a surface at a certain point, usually including Gaussian curvature or mean curvature, reflecting the degree of protrusion or depression of that point relative to its surrounding neighborhood.

[0100] Specifically, firstly, the discrete point cloud data (x) obtained in S202... i y i , z i The topography function z = f(x, y) is constructed by mapping the topography function onto regular grid nodes. Next, to identify anomalous microstructures, the second derivative information of each measurement point is calculated. Using the finite difference approximation method, the first partial derivatives (f(x, y)) in the X and Y directions of each measurement point are calculated. x f y ) and second-order partial derivatives (f xx f yy f xy Based on these partial derivatives, point-by-point calculations are performed by substituting them into curvature calculation formulas (such as mean curvature). The larger the calculated local spatial curvature value, the more severe the surface curvature at that point, which may be a sharp protrusion or a deep pit.

[0101] In some embodiments, curvature calculation can be implemented in a variety of ways:

[0102] Optionally, a local polynomial fitting method can be used: First, select the target measurement point and its surrounding 3x3 or 5x5 neighborhood points; second, use the least squares method to fit a local quadratic surface z=ax. 2 +by 2 +cxy+dx+ey+f; Finally, the curvature parameter of the point is directly calculated based on the fitting coefficients a, b, c.

[0103] It is understandable that other methods can also be used, such as tensor calculation based on the rate of change of the normal vector, which is not limited here.

[0104] S204. Identify measurement points with local spatial curvature greater than the preset peak threshold as adhesive stringing feature points.

[0105] Among them, the preset peak threshold refers to the empirical critical value of curvature, which is usually derived from the statistical analysis of a large amount of grinding experimental data and is used to distinguish between normal surface roughness undulations and abnormal sharp burrs; the binder stringing characteristic point refers to the slender fibrous residue formed by the binder being stretched instead of being brittlely cut during the grinding process due to its high ductility, which is morphologically manifested as extremely high and sharp protrusions.

[0106] Specifically, while normal active material particles also exhibit undulations, their geometry is relatively rounded, and the curvature changes are gentler. In contrast, binder filaments typically appear as needle-like or fine threads, with extremely high spatial curvature at their tips. By iterating through the curvature matrix calculated by S203, the curvature value K at each measuring point is... i With the preset peak threshold K th Compare them. If K i >K th If the measurement point is found to have an abnormal morphology, it may be due to residual adhesive from grinding forming fibers. The coordinate index of this point will then be marked as abnormal.

[0107] In some embodiments, identification can be achieved in a variety of ways:

[0108] Optionally, a single threshold discrimination method can be used: First, a fixed upper limit value for curvature is set; second, all data points are traversed; finally, all points exceeding the value are directly marked as wire-drawing feature points.

[0109] Understandably, other methods can also be used, such as a dual criterion method combining height threshold and curvature threshold, which is not limited here.

[0110] S205. Using the average height of non-feature points in the neighborhood around the adhesive stringing feature point, the height coordinate data of the adhesive stringing feature point is replaced and corrected to obtain the corrected height coordinate data.

[0111] Among them, the surrounding neighborhood refers to the area within a certain range centered on the feature point; non-feature points refer to normal measurement points within the neighborhood that are not marked as adhesive stringing feature points; replacement correction refers to the process of covering the original outlier with the calculated statistical value.

[0112] Specifically, if the height of the wire-drawing feature point is used directly for focusing, the detection probe will focus on the top of the wire, causing the underlying real interface to be out of focus. Furthermore, the wire itself has a single composition (pure binder), which will lead to compositional analysis bias. For each marked feature point, all normal points in its neighborhood are searched. The height values ​​of these normal points are read, and their arithmetic mean or median is calculated. Then, the original height value of the feature point is modified to this calculated average. Physically speaking, this is equivalent to aligning its height with the surrounding substrate surface at the data level, thereby reconstructing the true interface layer.

[0113] In some embodiments, the correction can be implemented in a variety of ways:

[0114] Optionally, a mean filtering replacement method can be used: First, determine the 8 neighboring pixels of the feature point; second, remove pixels in the neighborhood that are also marked as feature points; finally, calculate the average height of the remaining pixels and assign it to the center point.

[0115] Optionally, a median filtering replacement method can be used: First, obtain the height value sequence of all points in the neighborhood; second, sort the sequence and take the median; finally, replace the original height of the center point with the median. This method is more effective in removing isolated noise points.

[0116] It is understandable that other methods can also be used, such as local reconstruction based on Kriging interpolation, which is not limited here.

[0117] In some embodiments, when a large area of ​​continuous wire-drawing on the electrode surface results in a shortage of effective non-feature points in the local neighborhood, an anisotropic repair strategy based on texture directionality can be employed. Specifically, the grinding process is typically directional, causing the wire-drawing to occur in a consistent direction. First, the principal texture direction of the local region is calculated (using the structure tensor or gradient direction histogram). When it is found that most points in the neighborhood of a certain feature point are anomalous, instead of simply averaging, the nearest normal point is searched along the principal texture direction (i.e., perpendicular to the grinding mark direction). Using the height data of these laterally located normal points, the reference height that the feature point should have is calculated through one-dimensional spline interpolation or linear regression. This method avoids introducing incorrect wire-drawing heights along the wire-drawing direction, and can more accurately recover the true substrate morphology covered by large areas of wire-drawing, thereby improving the accuracy of subsequent spectral focusing.

[0118] S206. Generate a height distribution matrix based on the corrected height coordinate data.

[0119] Among them, the height distribution matrix refers to a two-dimensional numerical array that, after cleaning and correction, truly reflects the morphology of the substrate of the interface to be measured.

[0120] Specifically, after the S205 correction operation, the spike noise in the original data has been smoothed. All the corrected height data from the measurement points are reassembled, and the final height distribution matrix is ​​generated according to the grid index order during scanning. This matrix represents the ideal focusing plane. When the subsequent detection probe adjusts the Z-axis based on this matrix, it will ignore those spurious string heights and always keep the focus on the average plane containing the active material particles and the binder matrix.

[0121] In some embodiments, generation can be achieved in multiple ways:

[0122] Optionally, a direct recombination method can be used: First, create a new empty matrix; second, fill in the corrected data according to the row and column indices; finally, save it as a standard data format for the controller to call.

[0123] It is understandable that other methods can be used to achieve this, and no specific method is specified here.

[0124] S207. Control the detection probe to move to the target test position, and synchronously adjust the vertical height of the detection probe relative to the interface to be tested according to the height value corresponding to the target test position in the height distribution matrix.

[0125] S208. Control the spectrometer to collect spectral signal data at the target test location.

[0126] Steps S207-S208 and Figure 1 Steps S104-S105 in the illustrated embodiment are similar and can be found in the descriptions of steps S104-S105, which will not be repeated here.

[0127] In some embodiments, when hard particles are detached from the electrode surface due to mechanical peeling, forming micro-voids, a micro-displacement obstacle avoidance sampling strategy based on spectral intensity feedback can be adopted to avoid invalid signal areas and ensure that the spectra of the physical material with statistical representativeness are collected, thereby improving the validity and confidence of the detection data.

[0128] Specifically, firstly, after the detection probe is positioned, the spectrometer performs a short-exposure pre-scan on the current target test location (e.g., the exposure time is only 1 / 10 of the formal acquisition time). The purpose of this scan is not to analyze composition, but to detect the presence of any material. The spectrometer quickly acquires the calibration spectral signal at this location, and the processor then performs a full-spectrum integration operation on this signal, summing the light intensity values ​​of all bands to obtain the total integrated intensity. This intensity value directly reflects the Raman scattering capability of the material within the detection area. Next, the calculated total integrated intensity is compared with a preset void anomaly threshold. This threshold is usually set based on the statistical value of the lowest scattering intensity of a normal electrode substrate (i.e., the area with material), for example, set to 20% of the normal average intensity. If the total integrated intensity is less than this threshold, it indicates that the current laser focus is very likely falling into a pit (void) left by particle shedding, or onto an exposed point of a metal current collector without a signal. In this case, if forced acquisition is performed, invalid data with extremely poor signal-to-noise ratio will be obtained. Therefore, an obstacle avoidance mechanism is triggered, controlling the XY-axis displacement platform to move the detection probe within a preset neighborhood of the target position (e.g., a small circular area with a radius of 2 μm) by performing micro-step displacements (e.g., moving 0.5 μm each time). After each step, the above pre-scanning and judgment process is repeated. Once the intensity of the acquired calibration spectral signal exceeds the threshold, it indicates that the probe has moved out of the cavity and is aligned with the effective active material or binder area. At this point, the current probe position is immediately locked, and the spectrometer is controlled to acquire spectral signal data for subsequent feature peak extraction according to the formal acquisition parameters (e.g., long exposure, high power).

[0129] By employing the above technical solution, the common problem of powder shedding and void interference after machining of heterogeneous materials is solved through intelligent optimization of pre-scanning and fine-tuning. This avoids misjudging background noise at void locations as extremely low component content, ensuring that every recorded data point originates from the actual electrode material. This improves the completeness and accuracy of the final component distribution map and eliminates detection blind spots caused by sample preparation defects.

[0130] S209. Extract the first characteristic peak intensity value of the active substance and the second characteristic peak intensity value of the binder from the spectral signal data, and calculate the ratio of the second characteristic peak intensity value to the first characteristic peak intensity value to obtain the component distribution value at the target test location.

[0131] Step S209 and Figure 1 Step S106 in the illustrated embodiment is similar and can be found in the description of step S106, which will not be repeated here.

[0132] In some embodiments, when the electrode exhibits longitudinal migration of the binder and mechanical grinding is prone to interlayer coating interference, a longitudinal gradient correction and evaluation strategy based on multilayer difference can be adopted to eliminate false uniformity or false abrupt changes, thereby quantifying the true composition distribution law of the electrode along the thickness direction.

[0133] Specifically, the process first determines whether the current grinding depth has reached the preset total electrode thickness threshold (i.e., whether it has ground through to the current collector). If not, the depth parameters are automatically updated (e.g., the removal depth is increased by 5 μm from the current layer to the next gradient), and the mechanical grinding device continues to remove the surface material of the next gradient. Subsequently, the scanning, focusing, acquisition, and calculation steps described above are repeated until the same target test location is obtained at multiple different depths (d1, d2, ..., d...). n The numerical sequence of component distribution at the target depth is used. Based on this sequence, a longitudinal component distribution model is constructed, and the rate of change of component ratio between adjacent depths (i.e., longitudinal gradient) is calculated using a difference algorithm. A preset smearing effect threshold is introduced, which is a gradient limit value calibrated through rheological experiments based on the maximum theoretical diffusion rate of the binder during drying and the dragging characteristics of the grinding head on the soft material. If the calculated rate of change exceeds this threshold, it indicates that the abrupt change in component distribution within that depth range does not conform to the natural drying process, and it is possible that the grinding head has smeared a high-content binder from the previous layer onto the next. This range is marked as an abnormal gradient region, and the values ​​within this region are weighted and corrected using a smoothing algorithm or neighborhood interpolation to reduce the weight of abnormal points. Finally, by combining the corrected full-thickness data, a longitudinal uniformity evaluation index reflecting the overall consistency of the electrode is generated.

[0134] By employing the above technical solution and introducing vertical dimension data iteration and differential analysis, the limitations of single-layer detection are overcome, and interlayer smear artifacts caused by the physical carrying of the cutting head are identified and corrected. The true component migration trend inside the electrode is separated from processing noise, achieving high-fidelity reproduction of the vertical heterogeneity characteristics of the electrode, such as top-rich and bottom-poor composition, providing key depth structural parameters for battery process optimization.

[0135] In the above embodiments, a morphology model is constructed using height data and local spatial curvature is calculated to identify wire-drawing feature points caused by the ductility of the binder. By performing neighborhood-based statistical replacement correction on these abnormal peaks, the interference of physical burrs on the surface morphology data is eliminated. Digital smoothing avoids the detection probe from misfocusing on the wire-drawing tip, causing defocusing of the actual interface, ensuring that subsequent spectral acquisition is always aligned with the effective matrix material, thereby improving the signal accuracy and data confidence of microscopic composition analysis on rough grinding interfaces.

[0136] The following describes an exemplary lithium-ion battery electrode uniformity detection system 300 provided in an embodiment of this application. Figure 3 This is an exemplary hardware structure diagram of the lithium-ion battery electrode uniformity detection system 300 provided in the embodiments of this application.

[0137] In some embodiments, the lithium-ion battery electrode uniformity detection system 300 is a computer device or includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data. The network interface communicates with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, it can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods described in the embodiments of this application.

[0138] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0139] 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.

[0140] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as "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 "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0141] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0142] 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 detecting the uniformity of lithium-ion battery electrode sheets, characterized in that, A system for detecting the uniformity of lithium-ion battery electrode sheets, the system comprising a mechanical grinding device, a distance sensor, a detection probe, and a spectrometer connected to the detection probe, the method comprising: The mechanical grinding device is controlled to remove a layer of surface material from the lithium-ion battery electrode according to preset depth parameters, exposing the interface to be tested. The ranging sensor is controlled to perform a gridded scan of the interface to be measured, and the height coordinate data of multiple discrete measuring points on the interface to be measured are obtained. The original surface topography model is constructed using the height coordinate data, and the local spatial curvature of each measuring point in the original surface topography model is calculated. The measuring points with local spatial curvature greater than a preset peak threshold are identified as adhesive stringing feature points; The height coordinate data of the adhesive drawing feature point is replaced and corrected by using the average height of non-feature points in the neighborhood around the adhesive drawing feature point to obtain the corrected height coordinate data. Based on the corrected height coordinate data, a height distribution matrix is ​​generated; The detection probe is controlled to move to the target test position, and the vertical height of the detection probe relative to the interface to be tested is synchronously adjusted according to the height value corresponding to the target test position in the height distribution matrix. Control the spectrometer to acquire spectral signal data at the target test location; The first characteristic peak intensity value of the active substance and the second characteristic peak intensity value of the binder are extracted from the spectral signal data, and the ratio of the second characteristic peak intensity value to the first characteristic peak intensity value is calculated to obtain the component distribution value of the target test location.

2. The method according to claim 1, characterized in that, The control of the spectrometer to acquire spectral signal data at the target test location specifically includes: Based on the height distribution matrix, the local surface curvature of the target test location is calculated; Determine whether the curvature of the local surface is concave or convex. When the local surface curvature is concave, the target test location is determined to be a thermally sensitive area, and a low-power excitation command is generated. When the curvature of the local surface is a convex shape, the target test position is determined to be a light scattering enhancement region, and a high-power excitation command is generated; The spectrometer is controlled to emit laser light of corresponding power toward the target test position and acquire the original spectral signal according to the low-power excitation command or the high-power excitation command; The laser power value corresponding to the original spectral signal is obtained, and the intensity of the original spectral signal is normalized using the laser power value to obtain spectral signal data.

3. The method according to claim 2, characterized in that, The control of the spectrometer to emit laser light of corresponding power towards the target test location and acquire the original spectral signal according to the low-power excitation command or the high-power excitation command specifically includes: The surface gradient modulus of the target test location is calculated using the neighborhood height data centered on the target test location in the height distribution matrix. If the surface gradient modulus is greater than a preset scattering threshold, an instruction to extend the integration time is generated; If the surface gradient modulus is less than or equal to the scattering threshold, a standard integration time command is generated. The spectrometer is controlled to acquire the original spectral signal according to the determined low-power excitation command or high-power excitation command, and based on the exposure duration corresponding to the extended integration time command or the standard integration time command; The exposure time value corresponding to the original spectral signal is obtained, and the intensity of the original spectral signal is time-normalized using the exposure time value.

4. The method according to claim 1, characterized in that, The process of controlling the mechanical grinding device to remove a layer of surface material from the lithium-ion battery electrode according to preset depth parameters, thereby exposing the interface to be tested, specifically includes: The thickness sensor in the system is controlled to measure the initial thickness of the lithium-ion battery electrode at the current grinding position; Calculate the target thickness threshold for grinding termination based on the initial thickness value and the preset depth parameter; The mechanical grinding device is controlled to start rotary grinding, and the negative pressure adsorption module in the system is simultaneously turned on to collect dust with the first adsorption power. During the grinding process, the current thickness value fed back by the thickness sensor is acquired in real time. When the current thickness value reaches the target thickness threshold, the mechanical grinding device is controlled to stop rotating and lift. The negative pressure adsorption module is controlled to switch its adsorption power to a second adsorption power that is higher than the first adsorption power, and then shuts down after maintaining the preset cleaning time.

5. The method according to claim 1, characterized in that, After calculating the ratio of the second characteristic peak intensity value to the first characteristic peak intensity value to obtain the component distribution value at the target test location, the method further includes: If the current grinding depth does not reach the preset total electrode thickness threshold, the preset depth parameter is reduced to obtain a new depth parameter, and the mechanical grinding device is controlled to remove the surface material of the next gradient in combination with the new depth parameter, until the new composition distribution numerical sequence of the target test position at multiple different depths is obtained again. Based on the new component distribution numerical sequence, a longitudinal component distribution model of the target test location is constructed, and the rate of change of component ratio between adjacent depths is calculated. The depth range where the rate of change of the component ratio is greater than the preset smearing effect threshold is marked as an abnormal gradient region, and the component distribution values ​​in the abnormal gradient region are weighted and corrected to generate an evaluation index for the longitudinal uniformity of the electrode.

6. The method according to claim 1, characterized in that, The control of the spectrometer to acquire spectral signal data at the target test location specifically includes: The spectrometer is controlled to perform a short-exposure pre-scan of the target test location to acquire a verification spectral signal; Calculate the total integrated intensity of the full spectrum of the verification spectral signal; When the total integrated intensity of the full spectrum is less than the preset void anomaly threshold, the detection probe is controlled to perform micro-step displacement within the preset neighborhood of the target test position until the obtained verification spectrum signal meets the void anomaly threshold. Lock the current probe position and control the spectrometer to acquire spectral signal data for extracting characteristic peak intensity values.

7. A lithium-ion battery electrode uniformity detection system, characterized in that, The lithium-ion battery electrode uniformity detection 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 lithium-ion battery electrode uniformity detection system to perform the method as described in any one of claims 1-6.

8. A computer program product containing instructions, characterized in that, When the computer program product is run on the lithium-ion battery electrode uniformity detection system, the lithium-ion battery electrode uniformity detection system performs the method as described in any one of claims 1-6.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the lithium-ion battery electrode uniformity detection system, the lithium-ion battery electrode uniformity detection system performs the method as described in any one of claims 1-6.

Images