Battery pack structural strength simulation test method

Abstract

The invention discloses a battery pack structural strength simulation test method, and relates to the technical field of data analysis, and the method comprises the steps: obtaining multilayer construction parameters of a test battery pack, building a multi-scale geometric simulation model of the battery pack, and generating multi-scale structural simulation parameters of the test battery pack; based on the multi-scale structure simulation parameters of the test battery pack, cross-scale constitutive correlation analysis of the test battery pack is carried out, and equivalent material parameter vectors of the test battery pack are evaluated; performing time-frequency domain load verification based on the equivalent material parameter vector of the test battery pack, establishing a time domain-frequency domain load coupling equation, and generating composite load distribution of the test battery pack; and according to the composite load distribution of the test battery pack, establishing a fatigue progressive failure fitting model of the test battery pack, and generating an intensity damage trend state of the test battery pack. The method has the advantages that the problems of scale splitting and load simplification in an existing method are solved, and the simulation test precision of the structural strength of the battery pack is improved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a battery pack structural strength simulation test method. Background Art

[0002] Existing battery pack structural strength simulations mostly use single-scale modeling methods, which cannot accurately characterize the cross-scale mechanical behavior from the macroscopic shell to the microscopic electrode material; traditional constitutive models ignore the influence of crystal orientation and grain boundary effects on material properties, resulting in distortion in the calculation of equivalent parameters; in load verification, time domain impact and frequency domain vibration are separated for analysis, making it difficult to simulate actual composite working conditions; fatigue assessment relies on static load assumptions and linear damage models, which can neither capture the nonlinear degradation characteristics of the material nor lack electrochemical-mechanical coupling analysis, resulting in poor accuracy in battery pack structural strength simulations. Summary of the Invention

[0003] In order to solve the above technical problems, a battery pack structural strength simulation test method is provided. This technical solution solves the above-mentioned problems that the existing battery pack structural strength simulation mostly adopts a single-scale modeling method, which cannot accurately characterize the cross-scale mechanical behavior from the macro shell to the micro electrode material; the traditional constitutive model ignores the influence of crystal orientation and grain boundary effects on material properties, resulting in distortion of equivalent parameter calculation; the time domain impact and frequency domain vibration are separated for analysis in load verification, which makes it difficult to simulate actual composite working conditions; fatigue assessment relies on static load assumptions and linear damage models, which can neither capture the nonlinear degradation characteristics of the material nor lack electrochemical-mechanical coupling analysis; resulting in poor accuracy of battery pack structural strength simulation.

[0004] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0005] A battery pack structural strength simulation test method, comprising:

[0006] S1. Obtain multi-layer structural parameters of the test battery pack, establish a multi-scale geometric simulation model of the battery pack, and generate multi-scale structural simulation parameters of the test battery pack;

[0007] S2. Based on the multi-scale structural simulation parameters of the test battery pack, perform cross-scale constitutive correlation analysis of the test battery pack and evaluate the equivalent material parameter vector of the test battery pack;

[0008] S3. Perform time-frequency domain load verification based on the equivalent material parameter vector of the test battery pack, establish a time-frequency domain load coupling equation, and generate a composite load distribution for the test battery pack;

[0009] S4. Based on the composite load distribution of the test battery pack, a fatigue progressive failure fitting model of the test battery pack is established to generate the strength damage trend state of the test battery pack.

[0010] Preferably, the multi-layer structural parameters of the test battery pack are obtained based on the CAD geometric drawing of the test battery pack;

[0011] Using NURBS surfaces, a macroscopic analytical method for the test battery pack is established. Geometric discretization and local network encryption of key areas are performed on the multi-layer structural parameters of the test battery pack to generate the macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack as follows:

[0012]

[0013] Among them, S(u,v) is the appearance thickness distribution of the test battery pack under the macroscopic finite element network node coordinates, P i,j N is the finite element network control coordinate point for testing the battery pack. i,A (u) is the A-order spline basis function of the finite element network node coordinate u of the test battery pack, N i,B (v) is the B-order spline basis function of the finite element network node coordinate u of the test battery pack, W i,j is the weight factor of the finite element network control coordinate points of the test battery pack, n is the total number of indexes of the control points in the u direction, and m is the total number of indexes of the control points in the v direction;

[0014] Obtaining battery module structural parameters of the test battery pack based on CT scan data of the test battery pack;

[0015] Based on Voronoi mesh generation, the initial seed points of the stress distribution are marked according to the battery module structural parameters of the test battery pack. The macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack is divided to obtain the microscopic finite element mesh of the test battery pack;

[0016] Based on the battery module construction parameters of the test battery pack, the material property values ​​of the corresponding units of the mesoscopic finite element grid of the test battery pack are determined, and the local stiffness distribution of the mesoscopic finite grid units of the test battery pack is established.

[0017] Preferably, obtaining material microstructure image data of the test battery pack according to SEM scanning;

[0018] Based on the Phase-Field method, the local stiffness distribution of the microscopic finite grid cells of the test battery pack is correlated with the material microstructure image data of the test battery pack to verify the material phase change mobility rate of the test battery pack, calculate the local stress influence value of the crystal evolution of the test battery pack, and determine the microscopic grain boundary energy density in the high stress area of ​​the test battery pack;

[0019] The FFT fast Fourier transform is used to analyze the image texture direction in the material microstructure image data of the test battery pack, and the microscopic grain orientation distribution in the high stress area of ​​the test battery pack is obtained.

[0020] Preferably, based on the local stiffness distribution of the mesoscopic finite grid unit of the test battery pack, the mesoscopic geometric dimensions of the test battery pack are determined, and the local stiffness matrix of the mesoscopic finite grid unit of the test battery pack is established;

[0021] Using the local stiffness matrix of the mesoscopic finite grid cells of the test battery pack, displacement constraints are imposed on the corresponding boundaries of the mesoscopic finite grid cells of the test battery pack. Substituting the finite element discretization into a system of linear algebraic equations and solving them, the characteristic displacement field of the mesoscopic finite grid cells of the test battery pack is obtained.

[0022] The Gaussian integral method is used to calculate the integrand of each unit sampling point in the characteristic displacement field of the microscopic finite grid unit of the test battery pack, and the equivalent elastic tensor of the microscopic finite grid unit of the test battery pack is obtained. The equivalent elastic tensor is converted into an engineering constant to obtain the macroscopic-microscopic correlated homogenized equivalent elastic tensor of the test battery pack.

[0023] Preferably, based on the microscopic grain orientation distribution in the high stress area of ​​the test battery pack, the microscopic grain orientation distribution coordinate system in the high stress area is determined to perform a rotation operation, and a crystal orientation rotation matrix of the test battery pack is established;

[0024] Based on the crystal orientation rotation matrix of the test battery pack, the initial slip direction and initial slip plane normal of the test battery pack are determined, and the crystal slip coefficient of the test battery pack is calculated as follows:

[0025]

[0026] Where α is the crystal slip coefficient of the test battery pack, To test the initial slip direction of the battery pack crystal, is the normal direction of the initial slip plane of the crystal of the test battery pack, R(θ) is the crystal orientation rotation matrix of the test battery pack;

[0027] Based on the microscopic grain boundary energy density in the high stress area of ​​the test battery pack, the initial crystal migration resistance of the test battery pack is determined, and the crystal slip strength coefficient of the test battery pack is calculated as follows:

[0028]

[0029] Among them, G α is the crystal slip strength coefficient of the test battery pack, G0 is the initial crystal migration resistance of the test battery pack, k is the interface energy coupling coefficient, and γ is the microscopic grain boundary energy density coefficient in the high stress area of ​​the test battery pack;

[0030] Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack, the local stress tensor of the test battery pack is determined, and the crystal slip coefficient and crystal slip strength coefficient of the test battery pack are used for macro stress projection to obtain the decomposed shear stress of the crystal slip coefficient of the test battery pack;

[0031] Based on the Power-Law model, the decomposed shear stress of the crystal slip coefficient of the test battery pack and the crystal slip strength coefficient of the test battery pack are used as input, and the crystal slip system shear rate of the test battery pack is used as output. The crystal slip system shear rates of each test battery pack are superimposed to generate the meso-micro crystal plastic strain rate of the test battery pack;

[0032] Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the meso-micro crystal plastic strain rate of the test battery pack, the equivalent material parameter vector of the test battery pack is constructed.

[0033] Preferably, the shell curvature radius of the test battery pack is initialized based on the macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack;

[0034] Based on the type of test battery pack, the standardized test impact waveform parameters in the corresponding test standard are screened out, and a matching test transient force-time history curve for the test battery pack is established;

[0035] According to the matching test transient force time history curve of the test battery pack, the shell curvature radius of the test battery pack is updated and initialized in real time according to the inverse ratio between the basic impact frequency of the matching test transient force time history curve and the shell curvature radius of the test battery pack;

[0036] Based on Morlet wavelet synthesis, the shell curvature radius of the test battery pack is initialized by real-time updating to establish the amplitude wavelet function of the test battery pack. The basic impact frequency that matches the test transient force time history curve is verified, and the impact force wavelet amplitude is associated with the peak acceleration to generate the time domain impact force time series of the test battery pack.

[0037] Preferably, based on the type of the tested battery pack, a standardized broadband random vibration signal in the corresponding test standard is screened out;

[0038] Determine the fundamental vibration frequency of the test battery pack based on the proportionality between the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the standardized broadband random vibration signal;

[0039] Verify the test battery pack according to the basic vibration frequency of the test battery pack and collect the frequency domain parameters of the basic vibration frequency of the test battery pack;

[0040] Discretize the frequency domain parameters of the basic vibration frequency of the test battery pack and synthesize the frequency domain and time domain signals of the basic vibration frequency of the test battery pack;

[0041] Based on the time domain impact force time series of the test battery pack and the frequency domain and time domain signals of the basic vibration frequency of the test battery pack, the unit impulse response of the test battery pack per unit time is modally analyzed through the IRF impact response function;

[0042] According to the unit pulse response of the test battery pack per unit time, the time domain impact force of the test battery pack acts on the basic vibration frequency of the test battery pack, a coupled load time history simulation function is established, the dynamic load spectrum of the test battery pack is generated, and the composite load distribution of the test battery pack is determined.

[0043] Preferably, the basic load capacity of the test battery pack is initialized according to the multi-layer structural parameters of the test battery pack using linear regression;

[0044] Based on linear algebra, vector transformation is performed on the basic load capacity of the test battery pack to obtain the load vector distribution of the test battery pack;

[0045] The moment when the composite load distribution of the screening test battery pack exceeds the load vector distribution of the test battery pack is recorded as the abnormal moment of the composite load distribution of the test battery pack;

[0046] Taking the unit time as the observation window and the overload value at the moment of abnormal composite load distribution of the test battery pack as the observation object, the overload characteristic time series data of the composite load distribution of the test battery pack at the moment of abnormal composite load distribution is collected;

[0047] Based on the SVR linear support regression vector machine, the overload feature time series data of the test battery pack at the abnormal moment of the composite load distribution is used as input to train the strength damage quantification hyperplane boundary of the test battery pack, correct the basic load capacity of the initialized test battery pack, and determine the strength damage trend state of the test battery pack.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] This paper proposes a battery pack structural strength simulation test solution. This solution achieves high-precision characterization of the macro-microstructure through the multi-scale fusion of NURBS surfaces and Voronoi meshes. It also combines crystal plasticity theory with a multi-level homogenization algorithm to establish equivalent material parameters that account for grain boundary evolution. Morlet wavelets and IRF functions are used to simulate the time-frequency domain coupling of impact-vibration combined loads. Finally, an SVR dynamic damage model is used to accurately predict progressive failure trends. This solution addresses the scale fragmentation and load simplification issues inherent in traditional methods, improving the accuracy of battery pack structural strength simulation tests. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The figure is a flow chart of a battery pack structural strength simulation test method. DETAILED DESCRIPTION

[0051] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0052] Reference Figure 1 As shown, a battery pack structural strength simulation test method includes:

[0053] Step 1: Obtain the multi-layer structural parameters of the test battery pack, establish a multi-scale geometric simulation model of the battery pack, and generate multi-scale structural simulation parameters of the test battery pack;

[0054] The step 1 includes the following:

[0055] Based on the CAD geometric drawing of the test battery pack, obtain the multi-layer structural parameters of the test battery pack; the multi-layer structural parameters of the test battery pack include: battery pack housing, module frame, and cooling plate;

[0056] Using NURBS surfaces, a macroscopic analytical method for the test battery pack is established. Geometric discretization and local network encryption of key areas are performed on the multi-layer structural parameters of the test battery pack to generate the macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack as follows:

[0057]

[0058] Among them, S(u,v) is the appearance thickness distribution of the test battery pack under the macroscopic finite element network node coordinates, P i,j N is the finite element network control coordinate point for testing the battery pack. i,A (u) is the A-order spline basis function of the finite element network node coordinate u of the test battery pack, N i,B (v) is the B-order spline basis function of the finite element network node coordinate u of the test battery pack, W i,j is the weight factor of the finite element network control coordinate points of the test battery pack, n is the total number of indexes of the control points in the u direction, and m is the total number of indexes of the control points in the v direction;

[0059] Obtaining battery module structural parameters of the test battery pack based on CT scan data of the test battery pack; the battery module structural parameters include: battery pack electrodes and battery pack separators;

[0060] Based on Voronoi mesh generation, the initial seed points of the stress distribution are marked according to the battery module structural parameters of the test battery pack. The macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack is divided to obtain the microscopic finite element mesh of the test battery pack;

[0061] As a further aspect, the initial seed points of the stress distribution are marked according to the structural parameters of the battery module of the test battery pack. The electrode points of the battery pack in the battery pack can be used as the decision points, or the distribution of the battery pack separator can be used as the decision points. This is well known to those skilled in the art and will not be described in detail here.

[0062] Based on the battery module structural parameters of the test battery pack, determine the material property values ​​of the corresponding units of the mesoscopic finite element mesh of the test battery pack, and establish the local stiffness distribution of the mesoscopic finite element mesh of the test battery pack;

[0063] Obtaining material microstructure image data of the test battery pack according to SEM scanning; the material microstructure image of the test battery pack includes: grain size, grain orientation, and grain phase distribution;

[0064] Based on the Phase-Field method, the local stiffness distribution of the microscopic finite grid cells of the test battery pack is correlated with the material microstructure image data of the test battery pack to verify the material phase change mobility rate of the test battery pack, calculate the local stress influence value of the crystal evolution of the test battery pack, and determine the microscopic grain boundary energy density in the high stress area of ​​the test battery pack;

[0065] Using FFT (Fast Fourier Transform), we analyze the image texture direction in the material microstructure image data of the test battery pack and obtain the micro grain orientation distribution in the high stress area of ​​the test battery pack.

[0066] When using, combine the contents in the above steps.

[0067] As a further content, the traditional battery pack strength simulation method has the following problems: macroscopic single-scale modeling ignores the differences in multi-layer structures, resulting in distorted stress prediction; insufficient geometric discretization and simplified material properties (such as failure to consider micro grain boundaries and phase changes) affect the calculation accuracy; cross-scale data fragmentation, insufficient microscopic characterization (SEM images lack quantitative analysis) and low efficiency of phase field method in multi-scale modeling; local stress prediction relies on empirical assumptions, misses the influence of micro grain boundary energy density, and has insufficient ability to simulate material phase change response under dynamic load.

[0068] This solution achieves high-precision geometric representation of the battery pack from macro to micro through multi-scale fusion modeling of NURBS surfaces and Voronoi meshes. Combined with the Phase-Field method and FFT texture analysis, it quantitatively correlates the evolution of microscopic grain boundaries with macroscopic mechanical properties and accurately locates high-stress areas. Through automatic multi-scale parameter transfer and dynamic stiffness updating, it significantly improves computational efficiency and dynamic load adaptability. At the same time, it uses SEM image data to drive modeling, effectively solving the problems of cross-scale segmentation, insufficient microscopic representation, and local stress prediction distortion in traditional methods, providing a more accurate analysis solution for battery pack structural strength assessment.

[0069] Step 2: Based on the multi-scale structural simulation parameters of the test battery pack, perform cross-scale constitutive correlation analysis of the test battery pack to evaluate the equivalent material parameter vector of the test battery pack;

[0070] The second step includes the following:

[0071] Based on the local stiffness distribution of the mesoscopic finite grid unit of the test battery pack, the mesoscopic geometric dimensions of the test battery pack are determined, and the local stiffness matrix of the mesoscopic finite grid unit of the test battery pack is established;

[0072] Using the local stiffness matrix of the mesoscopic finite grid cells of the test battery pack, displacement constraints are imposed on the corresponding boundaries of the mesoscopic finite grid cells of the test battery pack. Substituting the finite element discretization into a system of linear algebraic equations and solving them, the characteristic displacement field of the mesoscopic finite grid cells of the test battery pack is obtained.

[0073] Using the Gaussian integral method, the integrand of each sampling point in the characteristic displacement field of the microscopic finite grid unit of the test battery pack is calculated to obtain the equivalent elastic tensor of the microscopic finite grid unit of the test battery pack. This is converted into an engineering constant to obtain the macroscopic-microscopic correlation homogenized equivalent elastic tensor of the test battery pack.

[0074] Based on the microscopic grain orientation distribution in the high stress area of ​​the test battery pack, the coordinate system of the microscopic grain orientation distribution in the high stress area is determined for rotation operation to establish the crystal orientation rotation matrix of the test battery pack;

[0075] Based on the crystal orientation rotation matrix of the test battery pack, the initial slip direction and initial slip plane normal of the test battery pack are determined, and the crystal slip coefficient of the test battery pack is calculated as follows:

[0076]

[0077] Where α is the crystal slip coefficient of the test battery pack, To test the initial slip direction of the battery pack crystal, is the normal direction of the initial slip plane of the crystal of the test battery pack, R(θ) is the crystal orientation rotation matrix of the test battery pack;

[0078] Based on the microscopic grain boundary energy density in the high stress area of ​​the test battery pack, the initial crystal migration resistance of the test battery pack is determined, and the crystal slip strength coefficient of the test battery pack is calculated as follows:

[0079]

[0080] Among them, G α is the crystal slip strength coefficient of the test battery pack, G0 is the initial crystal migration resistance of the test battery pack, k is the interface energy coupling coefficient, and γ is the microscopic grain boundary energy density coefficient in the high stress area of ​​the test battery pack;

[0081] Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack, the local stress tensor of the test battery pack is determined, and the crystal slip coefficient and crystal slip strength coefficient of the test battery pack are used for macro stress projection to obtain the decomposed shear stress of the crystal slip coefficient of the test battery pack;

[0082] Based on the Power-Law model, the decomposed shear stress of the crystal slip coefficient of the test battery pack and the crystal slip strength coefficient of the test battery pack are used as input, and the crystal slip system shear rate of the test battery pack is used as output. The crystal slip system shear rates of each test battery pack are superimposed to generate the meso-micro crystal plastic strain rate of the test battery pack;

[0083] Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the meso-micro crystal plastic strain rate of the test battery pack, the equivalent material parameter vector of the test battery pack is constructed;

[0084] When using, combine the contents in the above steps.

[0085] As a further content, the traditional battery pack strength simulation method has the following key defects: the homogenization method oversimplifies the macro equivalent parameters and ignores the influence of the microstructure, resulting in calculation deviations; the cross-scale constitutive correlation only realizes one-way data transmission and lacks macro-micro bidirectional coupling; the crystal plasticity model has an idealized assumption on the strength of the slip system and does not consider the influence of grain boundary energy density; at the same time, the existing method is not adaptable enough to dynamic loads, neither considers the strain rate effect, nor lacks the real-time correlation between the dynamic evolution of micro defects and slip strength, resulting in inaccurate fatigue prediction.

[0086] This solution achieves high-precision cross-scale constitutive correlation through a multi-level homogenization algorithm and dynamic mapping of crystal orientation, and adopts grain boundary energy density-driven slip strength and Power-Law rate-sensitive coupling to significantly improve the accuracy of the crystal plasticity model. At the same time, it specifically solves key defects of traditional methods such as one-way transmission of cross-scale data and idealization of slip strength, realizes macro-fine-micro bidirectional coupling and accurate prediction under dynamic load, and provides a more accurate solution and evaluation method for battery pack structural strength analysis.

[0087] Step 3: Perform time-frequency domain load verification based on the equivalent material parameter vector of the test battery pack, establish a time-frequency domain load coupling equation, and generate a composite load distribution of the test battery pack;

[0088] The step three includes the following:

[0089] Initialize the shell curvature radius of the test battery pack based on the appearance thickness distribution of the macroscopic finite element network node coordinates of the test battery pack;

[0090] Based on the type of test battery pack, the standardized test impact waveform parameters in the corresponding test standard are screened out, and a matching test transient force-time history curve for the test battery pack is established;

[0091] According to the matching test transient force time history curve of the test battery pack, the shell curvature radius of the test battery pack is updated and initialized in real time according to the inverse ratio between the basic impact frequency of the matching test transient force time history curve and the shell curvature radius of the test battery pack;

[0092] Based on Morlet wavelet synthesis, the shell curvature radius of the test battery pack is initialized by real-time updating to establish the amplitude wavelet function of the test battery pack. The fundamental impact frequency that matches the test transient force time history curve is verified, and the impact force wavelet amplitude is correlated with the peak acceleration to generate the time domain impact force time series of the test battery pack.

[0093] Based on the type of battery pack being tested, the standardized broadband random vibration signal in the corresponding test standard is selected;

[0094] Determine the fundamental vibration frequency of the test battery pack based on the proportionality between the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the standardized broadband random vibration signal;

[0095] Verify the test battery pack according to the basic vibration frequency of the test battery pack and collect the frequency domain parameters of the basic vibration frequency of the test battery pack;

[0096] Discretize the frequency domain parameters of the basic vibration frequency of the test battery pack and synthesize the frequency domain and time domain signals of the basic vibration frequency of the test battery pack;

[0097] Based on the time domain impact force time series of the test battery pack and the frequency domain and time domain signals of the basic vibration frequency of the test battery pack, the unit impulse response of the test battery pack per unit time is modally analyzed through the IRF impact response function;

[0098] Based on the unit pulse response of the test battery pack per unit time, the time domain impact force of the test battery pack acts on the basic vibration frequency of the test battery pack, a coupled load time history simulation function is established, the dynamic load spectrum of the test battery pack is generated, and the composite load distribution of the test battery pack is determined;

[0099] When using, combine the contents in the above steps.

[0100] As a further aspect, the traditional battery pack load verification method has the following problems: the time domain impact and frequency domain vibration are analyzed separately, and the coupling effect between the two is ignored; the standardized waveform is mechanically applied without dynamic adjustment based on the geometric characteristics of the battery pack; a fixed base frequency is used without considering the change in material stiffness, resulting in distorted fatigue assessment; there is a lack of an impact-vibration joint simulation mechanism, and it is impossible to simulate actual composite working conditions; at the same time, there are problems such as rough frequency domain discretization that loses high-frequency details, and static assumptions about curvature radius that ignore geometric nonlinearity, which seriously affect the simulation accuracy.

[0101] Step 4: Based on the composite load distribution of the test battery pack, a fatigue progressive failure fitting model of the test battery pack is established to generate the strength damage trend state of the test battery pack;

[0102] The step 4 includes the following contents:

[0103] Using linear regression, the basic load capacity of the test battery pack is initialized according to the multi-layer structural parameters of the test battery pack;

[0104] Based on linear algebra, vector transformation is performed on the basic load capacity of the test battery pack to obtain the load vector distribution of the test battery pack;

[0105] The moment when the composite load distribution of the screening test battery pack exceeds the load vector distribution of the test battery pack is recorded as the abnormal moment of the composite load distribution of the test battery pack;

[0106] Taking the unit time as the observation window and the overload value at the moment of abnormal composite load distribution of the test battery pack as the observation object, the overload characteristic time series data of the composite load distribution of the test battery pack at the moment of abnormal composite load distribution is collected;

[0107] Based on the SVR linear support regression vector machine, the overload feature time series data of the abnormal composite load distribution of the test battery pack is used as input to train the strength damage quantification hyperplane boundary of the test battery pack, correct the initial basic load capacity of the test battery pack, and determine the strength damage trend status of the test battery pack;

[0108] When using, combine the contents in the above steps.

[0109] As a further content, the traditional battery pack fatigue assessment method has the following key defects: the static load assumption is adopted and the dynamic degradation of the material is ignored, resulting in the missed detection of early damage; the reliance on rough threshold methods to identify abnormal loads has a high misjudgment rate and cannot capture the timing correlation effect; the damage model using only linear regression is difficult to characterize nonlinear cumulative damage, and lacks electrochemical-mechanical multi-physics field coupling analysis, resulting in insufficient fatigue life prediction accuracy.

[0110] This solution uses support vector regression (SVR) to establish a dynamic damage assessment model. Through hyperplane boundary correction and intelligent extraction of time series features, it achieves high-precision prediction of the progressive failure process of the battery pack. At the same time, through load vector space mapping and multi-scale damage association, it effectively integrates the macro-stress and micro-structure evolution characteristics, solving the static load assumption and linear regression limitations of traditional methods, and providing a more accurate assessment solution for battery pack strength status assessment.

[0111] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A battery pack structural strength simulation test method, characterized in that: include: S1. Obtain multi-layer structural parameters of the test battery pack, establish a multi-scale geometric simulation model of the battery pack, and generate multi-scale structural simulation parameters of the test battery pack; S2. Based on the multi-scale structural simulation parameters of the test battery pack, perform cross-scale constitutive correlation analysis of the test battery pack and evaluate the equivalent material parameter vector of the test battery pack; S3. Perform time-frequency domain load verification based on the equivalent material parameter vector of the test battery pack, establish a time-frequency domain load coupling equation, and generate a composite load distribution for the test battery pack; S4. Based on the composite load distribution of the test battery pack, a fatigue progressive failure fitting model of the test battery pack is established to generate the strength damage trend state of the test battery pack.

2. A battery pack structural strength simulation test method according to claim 1, characterized in that: Said S1 comprises: Based on the CAD geometric drawings of the test battery pack, obtain the multi-layer structural parameters of the test battery pack; Using NURBS surfaces, a macroscopic analytical method for the test battery pack is established. Geometric discretization and local network encryption of key areas are performed on the multi-layer structural parameters of the test battery pack to generate the macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack as follows: Among them, S(u,v) is the appearance thickness distribution of the test battery pack under the macroscopic finite element network node coordinates, P i,j N is the finite element network control coordinate point for testing the battery pack. i,A (u) is the A-order spline basis function of the finite element network node coordinate u of the test battery pack, N i,B (v) is the B-order spline basis function of the finite element network node coordinate u of the test battery pack, W i,j is the weight factor of the finite element network control coordinate points of the test battery pack, n is the total number of indexes of the control points in the u direction, and m is the total number of indexes of the control points in the v direction; Obtaining battery module structural parameters of the test battery pack based on CT scan data of the test battery pack; Based on Voronoi mesh generation, the initial seed points of the stress distribution are marked according to the battery module structural parameters of the test battery pack. The macroscopic finite element network node coordinate appearance thickness distribution of the test battery pack is divided to obtain the microscopic finite element mesh of the test battery pack; Based on the battery module construction parameters of the test battery pack, the material property values ​​of the corresponding units of the mesoscopic finite element grid of the test battery pack are determined, and the local stiffness distribution of the mesoscopic finite grid units of the test battery pack is established.

3. A battery pack structural strength simulation test method according to claim 2, characterized in that: Said S1 further comprises: Obtain material microstructure image data of the test battery pack based on SEM scanning; Based on the Phase-Field method, the local stiffness distribution of the microscopic finite grid cells of the test battery pack is correlated with the material microstructure image data of the test battery pack to verify the material phase change mobility rate of the test battery pack, calculate the local stress influence value of the crystal evolution of the test battery pack, and determine the microscopic grain boundary energy density in the high stress area of ​​the test battery pack; The FFT fast Fourier transform is used to analyze the image texture direction in the material microstructure image data of the test battery pack, and the microscopic grain orientation distribution in the high stress area of ​​the test battery pack is obtained.

4. A battery pack structural strength simulation test method according to claim 3, characterized in that: The S2 includes: Based on the local stiffness distribution of the mesoscopic finite grid unit of the test battery pack, the mesoscopic geometric dimensions of the test battery pack are determined, and the local stiffness matrix of the mesoscopic finite grid unit of the test battery pack is established; Using the local stiffness matrix of the mesoscopic finite grid cells of the test battery pack, displacement constraints are imposed on the corresponding boundaries of the mesoscopic finite grid cells of the test battery pack. Substituting the finite element discretization into a system of linear algebraic equations and solving them, the characteristic displacement field of the mesoscopic finite grid cells of the test battery pack is obtained. The Gaussian integral method is used to calculate the integrand of each unit sampling point in the characteristic displacement field of the microscopic finite grid unit of the test battery pack, and the equivalent elastic tensor of the microscopic finite grid unit of the test battery pack is obtained. The equivalent elastic tensor is converted into an engineering constant to obtain the macroscopic-microscopic correlated homogenized equivalent elastic tensor of the test battery pack.

5. A battery pack structural strength simulation test method according to claim 4, characterized in that: Said S2 further comprises: Based on the microscopic grain orientation distribution in the high stress area of ​​the test battery pack, the coordinate system of the microscopic grain orientation distribution in the high stress area is determined for rotation operation to establish the crystal orientation rotation matrix of the test battery pack; Based on the crystal orientation rotation matrix of the test battery pack, the initial slip direction and initial slip plane normal of the test battery pack are determined, and the crystal slip coefficient of the test battery pack is calculated as follows: Where α is the crystal slip coefficient of the test battery pack, To test the initial slip direction of the battery pack crystal, is the normal direction of the initial slip plane of the crystal of the test battery pack, R(θ) is the crystal orientation rotation matrix of the test battery pack; Based on the microscopic grain boundary energy density in the high stress area of ​​the test battery pack, the initial crystal migration resistance of the test battery pack is determined, and the crystal slip strength coefficient of the test battery pack is calculated as follows: Among them, G α is the crystal slip strength coefficient of the test battery pack, G0 is the initial crystal migration resistance of the test battery pack, k is the interface energy coupling coefficient, and γ is the microscopic grain boundary energy density coefficient in the high stress area of ​​the test battery pack; Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack, the local stress tensor of the test battery pack is determined, and the crystal slip coefficient and crystal slip strength coefficient of the test battery pack are used for macro stress projection to obtain the decomposed shear stress of the crystal slip coefficient of the test battery pack; Based on the Power-Law model, the decomposed shear stress of the crystal slip coefficient of the test battery pack and the crystal slip strength coefficient of the test battery pack are used as input, and the crystal slip system shear rate of the test battery pack is used as output. The crystal slip system shear rates of each test battery pack are superimposed to generate the meso-micro crystal plastic strain rate of the test battery pack; Based on the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the meso-micro crystal plastic strain rate of the test battery pack, the equivalent material parameter vector of the test battery pack is constructed.

6. A battery pack structural strength simulation test method according to claim 5, characterized in that: The S3 includes: Initialize the shell curvature radius of the test battery pack based on the appearance thickness distribution of the macroscopic finite element network node coordinates of the test battery pack; Based on the type of test battery pack, the standardized test shock waveform parameters in the corresponding test standard are screened out, and a matching test transient force-time history curve for the test battery pack is established; According to the matching test transient force time history curve of the test battery pack, the shell curvature radius of the test battery pack is updated and initialized in real time according to the inverse ratio between the basic impact frequency of the matching test transient force time history curve and the shell curvature radius of the test battery pack; Based on Morlet wavelet synthesis, the shell curvature radius of the test battery pack is initialized by real-time updating to establish the amplitude wavelet function of the test battery pack. The basic impact frequency that matches the test transient force time history curve is verified, and the impact force wavelet amplitude is associated with the peak acceleration to generate the time domain impact force time series of the test battery pack.

7. A battery pack structural strength simulation test method according to claim 6, characterized in that: Said S3 further comprises: Based on the type of battery pack being tested, the standardized broadband random vibration signal in the corresponding test standard is selected; Determine the fundamental vibration frequency of the test battery pack based on the proportionality between the macro-micro correlation homogenized equivalent elastic tensor of the test battery pack and the standardized broadband random vibration signal; Verify the test battery pack according to the basic vibration frequency of the test battery pack and collect the frequency domain parameters of the basic vibration frequency of the test battery pack; Discretize the frequency domain parameters of the basic vibration frequency of the test battery pack and synthesize the frequency domain and time domain signals of the basic vibration frequency of the test battery pack; Based on the time domain impact force time series of the test battery pack and the frequency domain and time domain signals of the basic vibration frequency of the test battery pack, the unit impulse response of the test battery pack per unit time is modally analyzed through the IRF impact response function; According to the unit pulse response of the test battery pack per unit time, the time domain impact force of the test battery pack acts on the basic vibration frequency of the test battery pack, a coupled load time history simulation function is established, the dynamic load spectrum of the test battery pack is generated, and the composite load distribution of the test battery pack is determined.

8. A battery pack structural strength simulation test method according to claim 7, characterized in that: The S4 includes: Using linear regression, the basic load capacity of the test battery pack is initialized according to the multi-layer structural parameters of the test battery pack; Based on linear algebra, vector transformation is performed on the basic load capacity of the test battery pack to obtain the load vector distribution of the test battery pack; The moment when the composite load distribution of the screening test battery pack exceeds the load vector distribution of the test battery pack is recorded as the abnormal moment of the composite load distribution of the test battery pack; Taking the unit time as the observation window and the overload value at the moment of abnormal composite load distribution of the test battery pack as the observation object, the overload characteristic time series data of the composite load distribution of the test battery pack at the moment of abnormal composite load distribution is collected; Based on the SVR linear support regression vector machine, the overload feature time series data of the test battery pack at the abnormal moment of the composite load distribution is used as input to train the strength damage quantification hyperplane boundary of the test battery pack, correct the basic load capacity of the initialized test battery pack, and determine the strength damage trend state of the test battery pack.

Images