Energy storage battery microstructure three-dimensional reconstruction simulation method and system based on electrochemical model, and computer equipment

Abstract

The invention discloses a battery microstructure three-dimensional reconstruction simulation method and system based on an electrochemical model and computer equipment. The method comprises the following steps: acquiring a lithium ion battery electrode scanning sequence image, and acquiring a three-dimensional voxel model and an electrode grid chart; s2, acquiring key microstructure parameters of the electrode; s3, performing mesh generation on three-phase voxels in the electrode mesh chart to obtain a mesh file for finite element calculation; and S4, comparing the experimental data of the battery with the simulation data of the three-dimensional microstructure electrochemical model, adjusting the precision of the three-dimensional microstructure electrochemical model, and optimizing related parameters. According to the invention, three-phase accurate segmentation is carried out by using filtering and denoising, image enhancement and connectivity and a watershed segmentation algorithm, and a three-dimensional voxel model containing real particle morphology, cracks and pores and an electrode grid chart are reconstructed; the problems that in the prior art, a lithium ion battery microstructure simulation model is large in idealization deviation and inconsistent with a real structure, and consequently performance prediction is inaccurate are solved.

Description

Technical Field

[0001] This invention relates to the field of lithium battery technology, and specifically to a method, system, and computer device for three-dimensional reconstruction simulation of the microstructure of energy storage batteries based on an electrochemical model. Background Technology

[0002] With the rapid development of the new energy industry, the demand for improved lithium-ion battery performance is becoming increasingly urgent, and its core performance is controlled by electrode materials and structure. Although simulation has become an important means of analyzing and optimizing battery performance, the current mainstream pseudo-two-dimensional homogenized models oversimplify the actual internal structure of the electrode, making it impossible to accurately analyze local reaction processes and reveal the influence mechanism of microstructural heterogeneity on macroscopic battery performance. Currently, the three-dimensional reconstruction simulation technology of lithium-ion battery microstructure mainly uses the Monte Carlo method / Gaussian random field to generate virtual electrode structures, and then uses LBM to simulate mesoscale ion transport. However, this method usually assumes that the particles are spherical or regular geometric shapes, while actual electrode particles have irregular edges, fragmentation, and surface defects (such as cracks), causing the generated microstructure to deviate from the real morphology, and the simulation results deviate significantly from the actual battery performance. In addition, to approximate the real structure, it is necessary to generate more than a million particles, and DEM / Monte Carlo simulation can take several weeks (e.g., a 50μm×50μm area). In the full-scale simulation of electrode-battery, there is also a scale discontinuity problem in the parameter transfer between the random model and the macroscopic P2D / 3D model.

[0003] Existing traditional processing methods based on CT / FIB-SEM images, such as thresholding and region growing, struggle to accurately distinguish the boundaries of different material phases (e.g., active particles, binders / additives, pores, cracks, SEI layers) when processing lithium-ion battery electrode materials (especially those with low contrast, artifacts, and multiphase mixtures). This results in reconstructed 3D geometric models containing numerous erroneous voxels or blurred phase interfaces. This not only distorts key structural parameters (e.g., tortuosity, specific surface area) but also leads to inaccurate input data for subsequent physics-based simulations, severely impacting the reliability of simulation results. Furthermore, when the original CT image has a low signal-to-noise ratio and insufficient resolution, existing general-purpose image processing algorithms exhibit poor robustness, leading to a sharp increase in segmentation errors and making it difficult to obtain a reliable 3D model suitable for effective physical simulations. Summary of the Invention

[0004] This invention addresses the problems in existing technologies by disclosing a method, system, and computer device for three-dimensional reconstruction simulation of battery microstructures based on an electrochemical model. This application employs filtering and denoising, image enhancement, and connectivity and watershed segmentation algorithms to achieve accurate three-phase segmentation, reconstructing a three-dimensional voxel model and electrode mesh containing realistic particle morphology, cracks, and pores. This fundamentally avoids model bias caused by idealized assumptions in traditional methods. Key parameters for subsequent parametric modeling are obtained, and redundant free data is removed during mesh processing, simplifying complex data, reducing the difficulty of subsequent mesh repair, and improving model processing speed, as well as the accuracy and reliability of battery performance prediction. This solves the problem of large idealization deviations and inaccurate performance predictions caused by discrepancies between existing lithium-ion battery microstructure simulation models and their actual structures.

[0005] This invention is achieved through the following technical solution: This invention first provides a three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model, comprising the following steps: S1. Obtain the scanning sequence image of lithium-ion battery electrode, filter and reduce noise, enhance contrast, divide the image into the optimal threshold and obtain smooth and stable independent voxels, and finally generate a three-dimensional voxel model and electrode mesh based on the threshold division result. S2. Obtain the key microstructure parameters of the electrode, including porosity, bulk porosity, tortuosity, average particle radius and phase volume fraction; S3. Mesh the three-phase voxels in the electrode mesh diagram, and perform lightweighting and closure processing on the mesh to obtain the mesh file for finite element calculation. S4. By importing the data from S2 and the mesh file from S3, a three-dimensional microstructure electrochemical model is constructed. By comparing the experimental data of the battery with the simulation data of the three-dimensional microstructure electrochemical model, the accuracy of the three-dimensional microstructure electrochemical model is adjusted and the relevant parameters are optimized by parametrically scanning the sensitivity parameters.

[0006] As a further embodiment, S1 includes: S11. Perform tomographic scanning on the electrode slices to obtain the electrode microstructure scan image; S12. After filtering, noise reduction, image enhancement, and axis connectivity, the electrode microstructure scan image is used to obtain a three-phase slice scan image at the microscale of the battery cell. S13. The watershed segmentation algorithm is used to perform threshold segmentation on the three-phase slice scan image to distinguish the active material, binder / conductive agent and pore phases, and a three-dimensional voxel model is generated based on the segmentation results. S14. Optimize the surface mesh of different voxels generated from the three-dimensional voxel model to obtain a complete electrode mesh diagram.

[0007] As a further embodiment, S12 includes: S121. Filter and reduce noise in the electrode microstructure scan image: Use the Sigma Filter module to calculate the local average value of neighboring voxels within a set region around the central voxel of the 3D voxel model. For intensity values ​​far from the central voxel, the following must be met: I |I c -2σ,I c +2σ| Where I is the total filtering intensity value of the entire voxel, I c σ is the fixed filter intensity value set for a certain voxel, and σ is the set voxel threshold. It is important to note that when performing filtering operations, it is necessary to specify whether to filter the entire 3D voxel or to filter the image stack composed of multiple 2D images. In addition, it is necessary to set the kernel size pixels and standard deviation thresholds for different voxels on the X, Y, and Z axes. S122. Convert the grayscale image into a binary image by binarization; S123. Using connectivity processing for porosity connectivity: Given two parallel surfaces or a pair of binarized input images, synthesize a binarized image containing all paths of the two connected surfaces using the axis connectivity algorithm.

[0008] As a further step, a smoothing filter is used to filter and denoise the scanning image of the electrode microstructure.

[0009] As a further step, the Interactive Thresholding module displays the selected threshold in a self-defined color on the orthogonal slice-based representation, while the background is grayscale data. During this process, a linear transformation is used to adjust the Colormap to enhance the overall contrast of the slice. Then, by setting the Intensity Range in conjunction with the Colormap, the contrast of the local slice is adjusted. By performing this operation on all three phases of the electrode slice, a relatively clear three-phase slice binarized image can be obtained.

[0010] As a further embodiment, S2 includes: S21. The method for obtaining porosity and volume porosity is as follows: the pore portion of the electrode mesh diagram is separately segmented by Interactive Thresholding, and the Volume Fraction algorithm is called to obtain the volume porosity by taking the porosity phase as the numerator and the global domain as the denominator. S22. The method for obtaining tortuosity and average particle radius is as follows: The distance of the segmented image in S21 is obtained through the Auto Skeleton algorithm. Then, the binary image is thinned to obtain a series of connected voxels. Next, the voxel skeleton is converted into a "spatial graph". Each point in the "spatial graph" is stored in the spatial graph object as the nearest boundary. This value can be used as an estimate of the local thickness as a thickness attribute. The Trace Line under Output Options can be used to obtain the trajectory line of the image. The statistical information of the attached "spatial graph" data object is obtained by calling the Spatial Graph Statistics module, and a spreadsheet containing tortuosity and particle radius data is output. S23. Obtaining the volume fraction of the phase includes: S231. The Separate Object module is called to sequentially perform the watershed algorithm, distance transformation algorithm, and numerical reconstruction algorithm to obtain the watershed line of the electrode mesh map; the best fit between different phases in the electrode mesh map is obtained by using the gradient magnitude value, thereby realizing the labeling between different phases. For all the above results, the Generate Pore Network Model algorithm is called to generate a pore network model; the Generate Pore Network Model algorithm only accepts one input, namely a labeled image, which represents the separated pore space, and its output is the corresponding pore network model. S232. The pore network model includes: number of nodes, number of throats, pore volume fraction, and pore radius. The pore volume fraction of the phase can be output through the pore network model.

[0011] As a further embodiment, S232 includes: introducing a Pore Network Model View module to visualize the structure of the pore network model; and adjusting the Pore Scale Factor module and the Throat Scale Factor module to adjust the size of network nodes and the thickness of throats.

[0012] As a further embodiment, S3 includes: S31. Mesh the three-phase voxels in S1: The Generate surface algorithm is called to perform triangular approximation calculations on the interfaces between different regions or materials in the three-phase voxel. The Compactify Minimum edge length function in the algorithm is called to reduce the number of triangles in the interface. The Adjust Coords function of the Border module is used to precisely move the points adjacent to the boundary voxel to the nearest border boundary to perform closure processing on the surface of the interface. S32. Delete free voxels and voxels with a particle size of less than 10 μm in S31.

[0013] As a further embodiment, S4 includes: S41. Construct a three-dimensional microstructure electrochemical model by importing the data from S2 and the mesh file from S3: Convert the mesh file in S3 into a .mphbin file and import it into the COMSOL software. Locate the low-quality boundaries and domains, delete entities for the low-quality parts, select to delete adjacent low-dimensional entities, and construct a three-dimensional microstructure electrochemical model.

[0014] S42. By comparing experimental data of lithium-ion batteries with simulation data of three-dimensional microstructure electrochemical models, the accuracy of the model is optimized by parametrically scanning the sensitive parameters.

[0015] As a further advancement, the three-dimensional microstructure electrochemical model adopts an NMC / Li half-cell. In terms of electrode particles, the geometry of the three-dimensional microstructure electrochemical model is heterogeneous, and the geometry includes a lithium metal foil electrode, a membrane domain, a porous electrode binder domain, and a positive electrode current collector. The model after CT scan reconstruction mainly includes three phases, namely the active particle domain, the binder domain, and the pore domain.

[0016] This invention also provides a three-dimensional reconstruction simulation system for battery microstructure based on an electrochemical model, including... The image acquisition and processing module is used to acquire scanned sequence images of the battery electrodes and to filter and enhance the contrast of the images. The 3D reconstruction module is used to generate 3D voxel models and electrode mesh maps using a watershed segmentation algorithm. The data processing module is used to acquire key microstructural parameters of the electrode, such as porosity, tortuosity, specific surface area, and particle size distribution. The mesh generation module is used to convert 3D voxel models into mesh files; The electrochemical simulation module is used to load the mesh model, set boundary conditions, solve the coupling equations, and complete parameter calibration.

[0017] As a further improvement, the mesh generation module includes an information display unit and a segmentation result display unit. The information display unit is used to display numerical information, and the segmentation result display unit is used to represent the result after segmenting the volume of the three-dimensional image.

[0018] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described.

[0019] (1) This invention performs three-phase accurate segmentation through filtering and denoising, image enhancement and connectivity and watershed segmentation algorithms, and reconstructs a three-dimensional voxel model and electrode mesh diagram containing real particle morphology, cracks and pores. This fundamentally avoids the model deviation caused by idealized assumptions in traditional methods; obtains key parameters for subsequent parametric modeling; and deletes redundant free data during mesh processing, simplifies redundant data, reduces the difficulty of subsequent mesh repair, improves the processing speed of the model and the accuracy and reliability of battery performance prediction, and solves the problem of inaccurate performance prediction caused by large idealization deviation and inconsistency with real structure in the existing lithium-ion battery microstructure simulation model.

[0020] (2) Based on real CT / FIB-SEM image data, the present invention performs three-phase accurate segmentation through an advanced watershed segmentation algorithm, which can reconstruct the microstructure containing real particle morphology, cracks and pores, fundamentally avoiding the model deviation caused by idealized assumptions in traditional methods.

[0021] (3) This invention realizes the automated and high-throughput extraction of key microstructure parameters (such as tortuosity and specific surface area), providing reliable data support for the quantitative analysis and performance optimization of electrode structures, and significantly improving design efficiency.

[0022] (4) This invention generates a high-quality simulation mesh while maintaining geometric features through efficient mesh repair and lightweight technology, effectively reducing the scale of finite element calculation, shortening the simulation time, and making high-precision three-dimensional microscopic simulation more practical in engineering.

[0023] (5) This invention closely integrates the real reconstructed microstructure with the electrochemical model and accurately calibrates the model using experimental data, so that the simulation results are highly consistent with the experimental data, which greatly improves the accuracy and reliability of battery performance prediction. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart of the three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model, as described in an embodiment of the present invention. Figure 2 The image shown is a binarized image of the electrode microstructure scan before filtering and noise reduction processing according to the embodiment of the present invention, wherein (a) represents pores; (b) represents active particles; and (c) represents binder. Figure 3 The image shown is a binarized image of the electrode microstructure scan image after filtering and noise reduction according to the embodiment of the present invention, wherein (a) represents pores; (b) represents active particles; and (c) represents binder. Figure 4 The image shows a three-phase voxel diagram of the battery cell according to an embodiment of the present invention, wherein (a) is the pore domain; (b) is the active particle domain; and (c) is the binder domain. Figure 5 The model geometry of the electrochemical-thermal coupling model described in the embodiments of the present invention; Figure 6 This is a comparison chart of the discharge curve measured experimentally under the 0.1C discharge condition described in this embodiment of the invention and the simulation results. Figure 7 for Figure 6 Enlarged view of part A in the middle. Detailed Implementation

[0026] To facilitate understanding of the present invention, a more comprehensive description of the present invention will be given below, and embodiments of the present invention will be provided, but this does not limit the scope of the present invention.

[0027] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0028] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0029] In existing technologies, mesh models generated by general-purpose 3D image processing software are directly used for multi-physics coupled simulations such as electrochemistry, stress, and heat. These models typically contain too many unnecessary details and excessively small elements, resulting in abnormally large model sizes (containing millions or even tens of millions of elements). This not only significantly increases the difficulty and time required for mesh repair, but also leads to low efficiency or even failure of general-purpose tools. Furthermore, it results in enormous computational resource consumption, convergence difficulties, and unacceptably long computation times for subsequent COMSOL simulations, greatly reducing work efficiency. This application provides a 3D reconstruction simulation method for battery microstructures based on electrochemical models, mainly used for the design and simulation of battery microstructures. Through filtering and denoising, image enhancement, and connectivity and watershed segmentation algorithms, accurate three-phase segmentation is performed to reconstruct a 3D voxel model and electrode mesh diagram containing realistic particle morphology, cracks, and pores. This fundamentally avoids model bias caused by idealized assumptions in traditional methods. It obtains key parameters for subsequent parametric modeling, greatly improving work efficiency. At the same time, during mesh processing, redundant free data is deleted, simplifying complex data, reducing the difficulty of subsequent mesh repair, and improving the model processing speed and the accuracy and reliability of battery performance prediction. This solves the problem of inaccurate performance prediction caused by large idealization deviations and discrepancies with the actual structure in existing lithium-ion battery microstructure simulation models.

[0030] like Figures 1 to 7 As shown, a three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model includes... S1. Acquire CT or FIB-SEM scan sequence images of lithium-ion battery electrodes, and achieve optimal threshold division through image noise reduction, image contrast enhancement, connectivity and watershed algorithm extraction to obtain smooth and stable independent voxels. Finally, generate a three-dimensional voxel model and electrode mesh map based on the threshold division results.

[0031] S11. Slice the electrode and obtain the electrode microstructure scan image by tomographic scanning of the slice; In some embodiments, the voxel size of the slice is 39μm×39μm×119μm.

[0032] S12. After filtering, noise reduction, image enhancement, and axis connectivity, a three-phase slice scan image of the cell at the microscale is obtained; S121. Since noise is inevitably introduced during the tomographic scanning process in S1, affecting the image quality, image filtering and noise reduction are required. The specific method is as follows: The sigma filter calculation module is used to calculate the local average value of neighboring voxels within a defined region around the central voxel. For intensity values ​​far from the central voxel, the following must be satisfied: I |I c -2σ,I c +2σ| Where I is the total filtering intensity value of the entire voxel, I c σ is the fixed filter intensity value set for a certain voxel, and σ is the set voxel threshold.

[0033] Statistically, approximately 95.45% of the voxels fall within range I. c -2σ,I c Within the range of +2σ, while the intensity value is less than I c -2σ or greater than I c Neighboring voxels with a +2σ value are either noise or edges of different tissues / structures. Filtering out "abnormal voxels" (noise or edge interference) with excessively large intensity differences from the central voxel achieves edge-preserving denoising, avoiding the "blurred edges" problem of traditional mean filtering. It's important to note that when performing filtering, it's necessary to specify whether filtering is applied to the entire 3D voxel or to a stack of multiple 2D images. Additionally, the kernel size (in pixels) and standard deviation thresholds for different voxels on the X, Y, and Z axes need to be set.

[0034] After some experimentation, it was found that using a smoothing filter can significantly reduce noise, but at the cost of image blurring, the edges of the target may also be difficult to discern due to excessive smoothing.

[0035] S122. To overcome this limitation and enhance edge features, additional contrast enhancement is generally required to obtain a clearer outline. To address this, the Interactive Thresholding module is introduced. Binarization converts the grayscale image into a binary image. In this case, the selected threshold can be displayed in a user-defined color on the orthogonal slice-based representation, while the background is grayscale data. During this process, linear transformation is used to adjust the Colormap to improve the overall contrast of the slice. Furthermore, by setting the Intensity Range in conjunction with the Colormap, the contrast of local slices can be adjusted. Performing this operation on all three phases of the electrode slice yields a relatively clear three-phase slice binarized image. A comparison of the slices before and after processing is shown below. Figure 2 and Figure 3 As shown.

[0036] S123. To enhance pore connectivity in the electrode microstructure scanning image and make the acquired data more accurate, pore connectivity processing methods include: The Axis Connectivity algorithm is used to synthesize a binary image containing all paths of the two connected surfaces, given two parallel surfaces or a pair of binarized input images.

[0037] If the target image is a pair of binary images, the algorithm will retain each labeled region in this area. If there are identical labeled regions on both faces but they are not connected, these regions will also be retained by the algorithm, even if they do not span the entire volume. When processing voxel-to-voxel connections, the connection type should be fully considered. A six-flux setting indicates that voxels with a common face are considered connected; an eighteen-flux setting indicates that voxels with at least one common boundary are considered connected; and a twenty-six-flux setting indicates that voxels with at least one common vertex are considered connected. It can be seen that the larger the number of port setting categories, the lower the required channel conditions and the higher the generated connectivity.

[0038] S13. The watershed segmentation algorithm is used to perform threshold segmentation on the preprocessed image to accurately distinguish the three phases of active material, binder / conductive agent, and pores, and an accurate three-dimensional voxel model is generated based on the segmentation results; specifically including: First, the optimal threshold range for each phase is determined through gray-level histogram analysis; then, watershed transformation is applied to solve the problem of particle boundary adhesion; finally, connected component analysis is used to ensure the independence and accuracy of the three-phase structure.

[0039] S14. Optimize the surface mesh of the generated three-dimensional voxel models of different voxels to obtain a complete electrode mesh diagram. The electrode mesh diagram is used to import into the finite element software for three-dimensional reconstruction simulation.

[0040] The Volume Rendering Settings algorithm is used to display different voxels in the three-phase slice scan image. In some embodiments, different colors are used to distinguish the granular phase, porous phase, and binder phase to more clearly show them. Specific three-phase voxels are shown below. Figure 4 As shown.

[0041] S2. Obtain key microstructural parameters of the electrode, such as porosity, tortuosity, specific surface area, and particle size distribution, for subsequent parametric modeling or optimization.

[0042] S21. Obtain porosity and volumetric porosity; The porous portion of the electrode mesh is segmented separately using Interactive Thresholding; the VolumeFraction algorithm is called to calculate the volume porosity by using the porosity phase as the numerator and the global domain as the denominator.

[0043] S22. In order to match the obtained data with the final domain and improve the accuracy of the results, it is necessary to obtain the tortuosity and average particle radius. S221. To extract the centerline of the filamentary structure from the image data of the electrode grid diagram, the Auto Skeleton algorithm is called in the image after thresholding and connectivity calculation binarization, specifically including the following methods: The distances of the segmented image are obtained, and then the binary image is thinned to obtain a series of connected voxels. The voxel skeleton is then converted into a "space map." For each point in the "space map," the nearest boundary is stored in the space map object as a thickness attribute; this value can be used as an estimate of the local thickness. The image's trajectory line can be obtained by enabling Trace Line under Output Options.

[0044] S222: Obtain statistical information from the accompanying "spatial graph" data object by calling the Spatial Graph Statistics algorithm. The output of this module depends on the "Output" port. By default, it generates a spreadsheet containing information such as tortuosity and particle radius. Note that under the Output interface, for better visualization and intuitive understanding of the image, both Spreadsheet and Spatial Graph must be selected.

[0045] S23. Obtain the volume fraction of the phase; S231. The Seprate Object module is invoked to sequentially combine the watershed algorithm, distance transformation algorithm, and numerical reconstruction algorithm to accurately calculate the watershed line of the electrode mesh map. The best fit between different phases in the electrode mesh map is obtained through the gradient magnitude, thereby realizing the labeling between different phases. For all the above results, the Generate PoreNetwork Model algorithm (PNM generation algorithm) is invoked to generate a pore network model.

[0046] Specifically, the Generate Pore Network Model algorithm accepts only one input, namely a labeled image representing the separated pore space, and its output is the corresponding pore network model.

[0047] S232. The pore network model includes various statistical data, such as the number of nodes, the number of throats, the pore volume fraction, the pore radius, etc. The pore volume fraction of the phase can be output through the pore network model.

[0048] In some embodiments, the Pore NetworkModel View module is required to visualize the structure of the pore network model; the size of network nodes and the thickness of throats can be adjusted by adjusting the Pore Scale Factor module and the Throat Scale Factor module.

[0049] S3. Mesh the three-phase voxels in S1, and perform lightweighting and closure processing on the mesh to finally generate a volume mesh file that can be used for finite element calculation.

[0050] S31. Mesh the three-phase voxels in S1. First, convert the result into a surface, then perform lightweighting and closure processing on the mesh. Specific methods include: The Generate surface algorithm is called to perform triangular approximation calculations on the interfaces between different regions or materials. The coordinates can be uniformly distributed or stacked. The algorithm's Compactify Minimum edge length feature (a specialized post-processing boundary shrinking technique to reduce the number of triangles in the generated surface) allows a minimum allowed edge length range of 0 to 0.8 relative to the input voxel size without selecting the "Compactify" option. Typically, a value of 0.4 yields good results. To ensure the generated surface is closed, enable the Border module under the algorithm, select Adjust Coords, and precisely move points adjacent to the boundary voxels to the nearest border boundary. This results in a clearly truncated surface boundary.

[0051] S32. Delete isolated fragmented meshes in S1 to ensure mesh integrity and computational efficiency.

[0052] Because the surface mesh composed of the binder phase is too fragmented, it will bring great difficulties to subsequent mesh processing. Therefore, it is necessary to remove free and excessively small voxels, call the Label Ananlysis module, and interpret the target volume elements in three dimensions.

[0053] The measurement method is determined based on the factors that need to be adjusted in the model. For example, if the parameter to be adjusted is defined as "basic," then the measurement objects under "basic" can be Volume3d, Area3d, Mean, etc. Of course, if there are further parameter requirements, a new measure group needs to be created, and the parameters can be obtained in the Native Measures window. Alternatively, a custom algorithm can be defined, for example, using Custom Measures to edit the algorithm. For instance, the pore size can be defined as length3d / width3d, and the defined algorithm can be re-imported into the newly defined variables.

[0054] Furthermore, to filter for smaller voxels, the Sieve Analysis algorithm is invoked. This algorithm primarily generates new label images by grouping the components of the input label image. These groups are defined based on a custom range of physical measures (e.g., volume) of any class. The algorithm requires the label image as input data and an analysis spreadsheet generated by "Label Analysis" containing physical measurements; both support the entire algorithm. For this group of slices, the upper and lower limits of Values ​​in the Sieve window are adjusted to filter voxels with sizes between 10μm and 52μm, while smaller fragmented voxels smaller than 10μm are deleted, generating a new mesh file for finite element calculations. This completes the data volume processing stage.

[0055] S4. By importing the data from S2 and the mesh file from S3, an electrochemical model is used to simulate the electrical performance of the battery cell based on the sliced ​​real structure. By comparing the experimental data of the ternary battery with the simulation data of the constructed three-dimensional microstructure electrochemical model, the relevant parameters are further optimized to improve the accuracy of the model.

[0056] The electrochemical model uses a series of partial differential equations and algebraic equations to describe the diffusion and migration of lithium ions inside the battery, electrochemical reactions on the surface of active particles, Ohm's law, and charge conservation, among other physical and chemical phenomena. Specifically, it includes the following equations: (1) Diffusion and migration equations used to describe the distribution of lithium-ion liquid phase substances:

[0057] in, C e This refers to the lithium ion concentration in the electrolyte. D e The liquid-phase diffusion coefficient in lithium ions; S a Specific surface area; j loc This represents the local current density at the electrolyte interface. F It is Faraday's constant; t + The transport number of lithium ions in the electrolyte; ε e This refers to the volume fraction of the electrolyte. γ e denoted by the Brugmann coefficient of the electrolyte; The effective liquid phase diffusion coefficient; The degree of curvature.

[0058] (2) Ohm's law for liquid phase used to describe potential distribution:

[0059] in, i e This represents the current density of lithium ions in the electrolyte; σ e eff The effective conductivity in the electrolyte; R This is the universal gas constant; f T is the particle activity coefficient in the electrolyte; T is the battery temperature. C e This represents the lithium ion concentration in the electrolyte.

[0060] (3) Using a rare substance transfer interface to describe the concentration gradient change caused by lithium diffusion:

[0061] in, C s This refers to the concentration of lithium ions in the solid phase. D s The solid-phase lithium-ion diffusion coefficient; r The radius of the lithium-ion particle is denoted as .

[0062] (4) Ohm's law for solid-state potential distribution:

[0063] in, i s This represents the current density of solid-phase lithium ions; σs eff The effective electrical conductivity in the solid phase; The potential in the solid phase; ε s This refers to the volume fraction of the electrodes. γ s denoted as Brugmann's coefficient for the solid phase.

[0064] (5) The Butler-Volmer kinetic equation, used to describe the rate of lithium ion insertion or extraction at the solid or liquid interface of the positive or negative electrode and to characterize the rate of the main or side reaction:

[0065] in, i 0 It is the reference exchange current density; k 0 It is the reaction rate constant; α a and α c These are the charge transfer coefficients of the anode and cathode, respectively; C s,max and C s,surf These represent the maximum lithium-ion concentration in the solid phase and the surface concentration, respectively. η This is a local overpotential; j loc This represents the local current density at the electrolyte interface. F It is Faraday's constant; C e This refers to the lithium ion concentration in the electrolyte. i This represents the local current density.

[0066] The present invention involves many physical quantities in its specific implementation. For ease of understanding, these physical quantities are presented in Table 1.

[0067] Table 1

[0068] Specific methods include: S41. Construct a three-dimensional microstructure electrochemical model; The three-dimensional microstructure electrochemical model uses an NMC / Li half-cell. Regarding the electrode particles, the geometry of the electrochemical-thermal coupling model is heterogeneous, comprising a lithium metal foil electrode, a membrane domain, a porous electrode binder domain, and a positive electrode current collector. The model, reconstructed after CT scanning, mainly consists of three phases: the active particle domain, the binder domain, and the porous domain. The model geometry is as follows: Figure 5 As shown.

[0069] The software used to build this model is not limited here, but shall be deemed to be capable of implementing the technical solution of this application.

[0070] In some embodiments, the three-dimensional microstructure electrochemical model is constructed using the electrochemical module of Comsol software.

[0071] In some embodiments, the generated mesh is converted into a .mphbin file and imported into the COMSOL software. Low-quality boundaries and domains are located in the three-dimensional microstructure electrochemical model. For low-quality parts, entity deletion operations are performed, and adjacent low-dimensional entities are deleted and objects are constructed.

[0072] In some embodiments, for ease of understanding, component 1 is used as an example: A membrane domain is drawn in component 1. Geometry 1 from the created 3D microstructure electrochemical model is combined with the imported mesh 1, and mesh nodes are added. Geometry 1 is then meshed to generate mesh 2, which is imported into mesh 1. Under Boolean operations, union is selected to merge mesh 1 and mesh 2, creating domain conditions for merging integrated meshes with the same interface. It is worth noting that the directly imported mesh 1 does not require geometry creation, while mesh 2 can be edited to suit the needs of the physical field settings.

[0073] Preferably, the slice surface is divided into free triangular meshes, and the remaining domain is divided into free tetrahedral meshes to generate a continuous mesh.

[0074] S42. By comparing experimental data of lithium-ion batteries with simulation data of three-dimensional microstructure electrochemical models, the parametric scanning function is further utilized to adjust the accuracy of the model by parametrically scanning sensitive parameters, thereby optimizing relevant parameters and improving the accuracy of the model.

[0075] Specifically, with the goal of minimizing the root mean square error of the battery experimental voltage and the simulated voltage of the three-dimensional microstructure electrochemical model, an iterative method is used to identify and optimize the parameters of the electrochemical model, including the positive electrode SOC reaction range, positive electrode diffusion coefficient, positive and negative electrode conductivity, and positive and negative electrode reaction rates, thereby improving the accuracy of the model.

[0076] The calibrated three-dimensional microstructure electrochemical model was used to simulate the 1C discharge of the battery, and the results were compared with experimental data under the same conditions to verify the accuracy of the model.

[0077] Figure 6 The comparison between the experimentally measured discharge curve and the simulation results under the 0.1C discharge condition shows that the simulation results and experimental results are in high agreement under the low current discharge condition, with very small error.

[0078] This application further calibrates the model using experimental data to improve its accuracy. When the model's accuracy meets the standards, the battery's capacity, voltage, and temperature can be predicted by adjusting the battery's design parameters, thus guiding the battery design.

[0079] In summary, the method presented in this application, based on real CT / FIB-SEM image data, employs an advanced watershed segmentation algorithm for precise three-phase segmentation. This enables the reconstruction of microstructures containing realistic particle morphology, cracks, and pores, fundamentally avoiding model biases caused by idealized assumptions (such as spherical particles) in traditional methods. It achieves automated, high-throughput extraction of key microstructural parameters (such as tortuosity and specific surface area), providing reliable data support for quantitative analysis and performance optimization of electrode structures, significantly improving design efficiency. Through efficient mesh repair and lightweighting techniques, a high-quality simulation mesh is generated while maintaining geometric features, effectively reducing the scale of finite element calculations and shortening simulation time, making high-precision three-dimensional microscopic simulations more practically applicable in engineering. By closely integrating the realistically reconstructed microstructure with the electrochemical model and accurately calibrating the model using experimental data, the simulation results closely match the experimental data (e.g., the voltage curve error can be less than 5% at a 0.1C discharge rate), greatly improving the accuracy and reliability of battery performance prediction.

[0080] This embodiment also provides a three-dimensional reconstruction simulation system for battery microstructure based on an electrochemical model, including... The image acquisition and processing module is used to acquire scanned sequence images of the battery electrodes and to filter and enhance the contrast of the images. The 3D reconstruction module is used to generate 3D voxel models and electrode mesh maps using a watershed segmentation algorithm. The data processing module is used to acquire key microstructural parameters of the electrode, such as porosity, tortuosity, specific surface area, and particle size distribution. The mesh generation module is used to convert 3D voxel models into mesh files; The electrochemical simulation module is used to load the mesh model, set boundary conditions, solve the coupling equations, and complete parameter calibration.

[0081] In one embodiment, the mesh generation module includes an information display unit and a segmentation result display unit. The information display unit is similar to a read-only spreadsheet, which can display the numerical calculation results of interest in the control panel, plot them, or display them via scripts. The segmentation result display unit is used to represent the result of segmenting the 3D image volume. It is a regular cubic mesh with the same size as the underlying image volume. For each voxel, it contains a label indicating the region to which the voxel belongs. Labeled images can be created from the image data stack using the "multi-threshold" module.

[0082] This embodiment also provides a computer device applicable to a three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model as proposed in the above embodiment.

[0083] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model as proposed in the above embodiments.

[0084] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0085] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0086] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0087] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0088] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0090] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A three-dimensional reconstruction simulation method for battery microstructure based on an electrochemical model, characterized in that: Includes the following steps: S1. Obtain the lithium-ion battery electrode scanning sequence image, filter and reduce noise, enhance contrast, divide the image into the optimal threshold and obtain smooth and stable independent voxels, and finally generate a three-dimensional voxel model and electrode mesh based on the threshold division result. S2. Obtain the key microstructure parameters of the electrode, including porosity, bulk porosity, tortuosity, average particle radius and phase volume fraction; S3. Mesh the three-phase voxels in the electrode mesh diagram, and perform lightweighting and closure processing on the mesh to obtain the mesh file for finite element calculation. S4. By importing the data from S2 and the mesh file from S3, a three-dimensional microstructure electrochemical model is constructed. After comparing the experimental data of the battery with the simulation data of the three-dimensional microstructure electrochemical model, the accuracy of the three-dimensional microstructure electrochemical model is adjusted and the relevant parameters are optimized by parametrically scanning the sensitivity parameters.

2. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 1, characterized in that: S1 includes: S11. Perform tomographic scanning on the electrode slices to obtain the electrode microstructure scan image; S12. After filtering, noise reduction, image enhancement, and axis connectivity, the electrode microstructure scan image is used to obtain a three-phase slice scan image at the microscale of the battery cell. S13. The watershed segmentation algorithm is used to perform threshold segmentation on the three-phase slice scan image to distinguish the active material, binder / conductive agent and pore phases, and a three-dimensional voxel model is generated based on the segmentation results. S14. Optimize the surface mesh of different voxels generated from the three-dimensional voxel model to obtain a complete electrode mesh diagram.

3. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 2, characterized in that: S12 includes: S121. Filter and reduce noise in the electrode microstructure scan image: Use the Sigma Filter module to calculate the local average value of neighboring voxels within a set region around the central voxel of the 3D voxel model. For intensity values ​​far from the central voxel, the following must be met: I |I c -2σ,I c +2σ| Where I is the total filtering intensity value of the entire voxel, I c σ is the fixed filter intensity value set for a certain voxel, and σ is the set voxel threshold. S122. Convert the grayscale image into a binary image by binarization; S123. Using connectivity to process porosity connectivity: Given two parallel surfaces or a pair of binarized input images, synthesize a binarized image containing all paths of the two connected surfaces using the axis connectivity algorithm; Preferably, a smoothing filter is used to filter and denoise the electrode microstructure scanning image; Preferably, the specific method for converting a grayscale image into a binary image is as follows: the selected threshold is displayed in a self-defined color on the representation map based on orthogonal slices through the Interactive Thresholding module, while the background is grayscale data. During this period, the overall contrast of the slice is improved by adjusting the Colormap through linear transformation. Then, the contrast of the local slice is adjusted by setting the Intensity Range in combination with the Colormap. By performing this operation on all three phases of the electrode slice, a relatively clear three-phase slice binary image can be obtained.

4. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 1, characterized in that: S2 includes: S21. The method for obtaining porosity and volume porosity is as follows: the pore portion of the electrode mesh diagram is separately segmented by Interactive Thresholding, and the Volume Fraction algorithm is called to obtain the volume porosity by taking the porosity phase as the numerator and the global domain as the denominator. S22. The method for obtaining tortuosity and average particle radius is as follows: The distance of the segmented image in S21 is obtained through the Auto Skeleton algorithm. Then, the binary image is thinned to obtain a series of connected voxels. Next, the voxel skeleton is converted into a "space map". Each point in the "space map" is stored in the space map object as the nearest boundary. This value can be used as an estimate of the local thickness as a thickness attribute. The Trace Line under OutputOptions can be used to obtain the trajectory line of the image. The statistical information of the accompanying "spatial graph" data object is obtained by calling the Spatial Graph Statistics module, and a spreadsheet containing data on tortuosity and particle radius is output. S23. Obtaining the volume fraction of the phase includes: S231. The Separate Object module is called to sequentially perform the watershed algorithm, distance transformation algorithm, and numerical reconstruction algorithm to obtain the watershed line of the electrode mesh map; the best fit between different phases in the electrode mesh map is obtained by using the gradient magnitude value, thereby realizing the labeling between different phases. For all the above results, the Generate Pore Network Model algorithm is called to generate a pore network model; the Generate Pore Network Model algorithm only accepts one input, namely a labeled image, which represents the separated pore space, and its output is the corresponding pore network model. S232. The pore network model includes: number of nodes, number of throats, pore volume fraction, and pore radius. The pore volume fraction of the phase can be output through the pore network model.

5. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 4, characterized in that: S232 includes: introducing a Pore Network Model View module to visualize the structure of the pore network model; and adjusting the Pore Scale Factor module and the Throat Scale Factor module to adjust the size of network nodes and the thickness of throats.

6. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 1, characterized in that: S3 includes: S31. Mesh the three-phase voxels in S1: The Generate surface algorithm is called to perform triangular approximation calculations on the interfaces between different regions or materials in the three-phase voxel. The Compactify Minimum edge length function in the algorithm is called to reduce the number of triangles in the interface. The Adjust Coords function of the Border module is used to precisely move the points adjacent to the boundary voxel to the nearest border boundary to perform closure processing on the surface of the interface. S32. Delete free voxels and voxels with a particle size of less than 10 μm in S31.

7. The method for three-dimensional reconstruction simulation of battery microstructure based on an electrochemical model according to claim 1, characterized in that: S4 includes: S41. Construct a three-dimensional microstructure electrochemical model by importing the data from S2 and the mesh file from S3: Convert the mesh file in S3 into a .mphbin file and import it into the COMSOL software. Locate the low-quality boundaries and domains, delete entities for the low-quality parts, select to delete adjacent low-dimensional entities, and construct a three-dimensional microstructure electrochemical model. S42. By comparing experimental data of lithium-ion batteries with simulation data of three-dimensional microstructure electrochemical models, the accuracy of the model is optimized by parametrically scanning the sensitive parameters. Preferably, the three-dimensional microstructure electrochemical model adopts an NMC / Li half-cell. In terms of electrode particles, the geometry of the three-dimensional microstructure electrochemical model is heterogeneous, and the geometry includes a lithium metal foil electrode, a membrane domain, a porous electrode binder domain, and a positive electrode current collector. The model after CT scan reconstruction mainly includes three phases, namely the active particle domain, the binder domain, and the pore domain.

8. A three-dimensional reconstruction simulation system for battery microstructure based on an electrochemical model, characterized in that: include: The image acquisition and processing module is used to acquire scanned sequence images of the battery electrodes and to filter and enhance the contrast of the images. The 3D reconstruction module is used to generate 3D voxel models and electrode mesh maps using a watershed segmentation algorithm. The data processing module is used to acquire key microstructural parameters of the electrode, such as porosity, tortuosity, specific surface area, and particle size distribution. The mesh generation module is used to convert 3D voxel models into mesh files; The electrochemical simulation module is used to load the mesh model, set boundary conditions, solve the coupling equations, and complete parameter calibration.

9. The three-dimensional reconstruction simulation system for battery microstructure based on an electrochemical model according to claim 8, characterized in that: The mesh generation module includes an information display unit and a segmentation result display unit. The information display unit is used to display numerical information. The segmentation result display unit is used to show the result after the volume of the three-dimensional image is segmented.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

Images