A method and system for stress assessment of battery rack weld swaying conditions

Abstract

This invention relates to the field of vibration engineering technology, and discloses a method and system for assessing the stress of battery frame welds under swaying conditions. The method includes: simultaneously acquiring vibration response signals and excitation reference signals from vibration sensors at key points of the weld when the containerized power unit is under swaying conditions; reconstructing the phase space of the vibration response signals to obtain the correlation dimension of the vibration signals; analyzing the dynamic equivalent stress amplitude at the key points of the weld based on the correlation dimension of the vibration signals; performing joint vibration deformation analysis on the vibration response signals and excitation reference signals to obtain the dynamic strain field distribution corresponding to the key points of the weld; incrementally iteratively analyzing the dynamic stress time-domain history of the key points of the weld based on the dynamic strain field distribution and the constitutive relation parameters of the battery frame material; and determining the stress of the battery frame weld under swaying conditions based on the stress data and dynamic equivalent stress amplitude in the dynamic stress time-domain history. This invention can improve the accuracy of stress assessment for battery frame welds under swaying conditions.

Description

Technical Field

[0001] This invention relates to the field of vibration engineering technology, and in particular to a method and system for assessing stress under swaying conditions at battery rack welds. Background Technology

[0002] In fields such as new energy storage and emergency power supply, containerized power units have become core equipment for outdoor and mobile power supply scenarios due to their advantages such as flexible deployment and scalable capacity. As the core load-bearing structure of the containerized power unit, the battery rack must withstand the weight of the battery modules, the bumps during transportation, and the swaying conditions during use for a long time. The welds of the battery rack are a key weak point in the structural strength, and their stress state directly determines the safety, stability, and service life of the entire power unit.

[0003] In existing technologies, the assessment of weld stress in battery racks often employs static stress calculation methods. This involves constructing a simplified geometric model of the battery rack and calculating the static stress value of the weld based on static loads (such as the battery's own weight). However, in practical applications, static analysis does not consider the dynamic load characteristics under swaying conditions and cannot reflect the impact effect of load changes over time on weld stress. This results in a significant deviation between the calculated results and the actual stress state, leading to insufficient accuracy. Summary of the Invention

[0004] This invention provides a method and system for assessing the stress of battery frame welds under swaying conditions, the main purpose of which is to solve the problem of low accuracy in assessing the stress of battery frame welds under swaying conditions.

[0005] To achieve the above objectives, the present invention provides a method for evaluating the stress of a battery rack weld under swaying conditions, comprising:

[0006] On the battery rack of the pre-designed containerized power unit, a weld evaluation area is determined, and key weld points in the weld evaluation area are identified.

[0007] When the pre-set containerized power supply device is in a swaying condition, the vibration response signal and excitation reference signal of the vibration sensor at the key point of the weld are collected simultaneously.

[0008] The vibration response signal is reconstructed in phase space to obtain the correlation dimension of the vibration signal, and the dynamic equivalent stress amplitude of the key points of the weld is analyzed based on the correlation dimension of the vibration signal.

[0009] By performing joint vibration deformation analysis on the vibration response signal and the excitation reference signal, the dynamic strain field distribution corresponding to the key points of the weld is obtained.

[0010] Based on the dynamic strain field distribution and constitutive parameters of the battery frame material, the dynamic stress time-domain history of the key points of the weld is analyzed incrementally and iteratively.

[0011] The stress of the battery frame weld under swaying condition is determined based on the stress data in the dynamic stress time-domain history and the dynamic equivalent stress amplitude.

[0012] To address the above problems, the present invention also provides a stress assessment system for battery rack weld swaying conditions, the system comprising:

[0013] A weld key point identification module is used to determine the weld evaluation area on the battery rack of a pre-set containerized power supply unit and identify the weld key points in the weld evaluation area.

[0014] The signal synchronization acquisition module is used to synchronously acquire the vibration response signal and excitation reference signal of the vibration sensor at the key point of the weld when the preset containerized power supply device is in a swaying condition.

[0015] The dynamic equivalent stress amplitude analysis module is used to reconstruct the phase space of the vibration response signal to obtain the correlation dimension of the vibration signal, and analyze the dynamic equivalent stress amplitude of the key points of the weld based on the correlation dimension of the vibration signal.

[0016] The joint vibration deformation analysis module is used to perform joint vibration deformation analysis on the vibration response signal and the excitation reference signal to obtain the dynamic strain field distribution corresponding to the key points of the weld.

[0017] The incremental iterative analysis module is used to incrementally iteratively analyze the dynamic stress time-domain history of the key points of the weld based on the dynamic strain field distribution and the constitutive parameters of the battery frame material.

[0018] The working condition stress determination module is used to determine the swing working condition stress of the battery frame weld seam based on the stress data in the dynamic stress time domain history and the dynamic equivalent stress amplitude.

[0019] This invention, through simultaneous acquisition of vibration response signals and excitation reference signals, combined with phase space reconstruction and joint vibration deformation analysis, can accurately quantify the dynamic stress distribution and equivalent stress amplitude of battery rack welds under swaying conditions. This effectively identifies fatigue weak points in the welds, improves the accuracy and reliability of structural safety assessments, provides data support for the optimized design and life prediction of containerized power units, and reduces the risk of failures caused by vibration. Therefore, the battery rack weld swaying stress assessment method and system proposed in this invention can solve the problem of low accuracy in assessing battery rack weld swaying stress. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a method for evaluating the stress of a battery rack weld under swaying conditions, provided in an embodiment of the present invention.

[0021] Figure 2This is a functional block diagram of a battery rack weld swaying stress assessment system provided in an embodiment of the present invention.

[0022] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0024] This application provides a method for assessing the stress of a battery rack weld under swaying conditions. The execution subject of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for assessing the stress of a battery rack weld under swaying conditions can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0025] Reference Figure 1 The diagram shown is a flowchart illustrating a method for assessing the stress of a battery rack weld under swaying conditions according to an embodiment of the present invention. In this embodiment, the method for assessing the stress of a battery rack weld under swaying conditions includes:

[0026] S1. Determine the weld evaluation area on the battery rack of the pre-set containerized power unit, and identify the key weld points in the weld evaluation area.

[0027] In this embodiment of the invention, the containerized power supply unit refers to a containerized power supply device that integrates battery modules, power distribution units, and control modules, enabling energy storage and output. It is widely used in new energy storage, emergency power supply, and other scenarios. The battery rack is a metal frame structure inside the containerized power supply unit used to fix and support the battery modules, typically welded from structural steel. The weld evaluation area refers to a specific area on the battery rack with high weld stress levels and a risk of structural failure, requiring focused monitoring and analysis.

[0028] In this embodiment of the invention, determining the weld evaluation area on the battery rack of a pre-set containerized power supply device includes:

[0029] A three-dimensional geometric model of the battery rack is constructed based on the structural features of the battery rack.

[0030] The three-dimensional geometric model of the battery rack is meshed to obtain a finite element calculation mesh, and preset boundary conditions are applied to the finite element calculation mesh.

[0031] Based on the finite element calculation mesh after applying boundary conditions, a stress distribution cloud map of the battery frame under swaying conditions is generated.

[0032] Based on the stress distribution cloud map, stress concentration regions in the finite element calculation mesh are identified, and these stress concentration regions are marked as weld evaluation regions.

[0033] In detail, structural features refer to the key parameters that determine the geometric shape and mechanical properties of the battery rack, such as the type and size of the steel profiles, connection methods, and weld locations. The three-dimensional geometric model of the battery rack refers to a digital model constructed using 3D modeling software that matches the actual structure of the battery rack at a 1:1 scale. This involves using 3D modeling software to obtain the design drawings of the battery rack, then creating component models such as columns, beams, and support beams based on the structural features in the drawings, and finally assembling the components according to their actual welding positions to form a complete three-dimensional geometric model of the battery rack. The constructed 3D geometric model of the battery rack is subjected to finite element analysis. Hexahedral elements are selected to mesh the model. The weld area uses a fine mesh (element size set to 2mm) to improve the calculation accuracy, while other areas use a conventional mesh (element size set to 5mm). After meshing, the finite element calculation mesh is obtained. Then, according to the actual installation method of the battery rack in the container (such as being fixed to the bottom of the container with bolts), fixed constraints are applied at the bolt hole positions at the bottom of the finite element calculation mesh (restricting the translational degrees of freedom in the X, Y, and Z directions). At the same time, according to the swing condition requirements (such as ±15° swing), angular acceleration load along the X-axis is applied to complete the application of the preset boundary conditions.

[0034] Specifically, in the finite element analysis software, the dynamic structure analysis module is selected, and the calculation time step and total calculation time are set (e.g., 10s, covering one complete swing cycle). The software calculates the stress value of each element at different time steps based on the element mechanical parameters (e.g., elastic modulus, Poisson's ratio) of the finite element calculation mesh and the applied boundary conditions. Finally, the stress values ​​of all elements are output in the form of a cloud map, resulting in a stress distribution cloud map of the battery frame under swing conditions (e.g., red areas indicate stress values ​​greater than 300MPa, and blue areas indicate stress values ​​less than 100MPa). The stress distribution cloud map data file is read, and a stress threshold is set (e.g., based on the yield strength of Q235 steel, 235MPa, areas with stress values ​​greater than 180MPa are set as high-stress areas). The software's region selection function is used to filter out finite element elements with stress values ​​exceeding the threshold; the area composed of these elements is the stress concentration area. Then, the software's marking function is used to mark this stress concentration area as the weld evaluation area, and its coordinate range and element number are saved.

[0035] Furthermore, the weld assessment area is a large region encompassing multiple weld locations. Simply performing an overall stress assessment of this area is insufficient to accurately pinpoint the specific locations most prone to fatigue damage, such as stress abrupt changes at the weld root. Critical weld points, on the other hand, represent the most severe stress conditions within the assessment area and are the core nodes determining weld failure. Therefore, after defining the weld assessment area, it is necessary to further identify the critical weld points within it to improve the accuracy of stress assessment.

[0036] In this embodiment of the invention, the key point of the weld refers to the specific location in the weld evaluation area where stress changes suddenly and fatigue damage is likely to occur. It is the core monitoring point for stress evaluation.

[0037] In this embodiment of the invention, identifying key weld points in the weld evaluation area includes:

[0038] Identify the real-time stress distribution data of the weld evaluation area, and construct the stress distribution matrix of the weld evaluation area based on the real-time stress distribution data;

[0039] Principal component analysis is performed on the stress distribution matrix to obtain the principal eigenvectors and eigenvalues ​​of the stress distribution, and a stress gradient tensor field is constructed based on the principal eigenvectors.

[0040] The stress gradient change rate at each point in the weld evaluation area is calculated based on the stress gradient tensor field, and the stress gradient tensor field is decomposed into singular values ​​to obtain the main direction of stress distribution.

[0041] The stress abrupt boundary of the weld evaluation region is identified based on the principal direction and the stress gradient change rate.

[0042] The stress abrupt change location in the weld evaluation area is determined based on the stress abrupt change boundary and the stress gradient change rate, and the stress abrupt change location is identified as the weld critical point.

[0043] In detail, real-time stress distribution data refers to the stress values ​​at specific moments obtained from sensors at each monitoring point within the weld evaluation area. The stress calculation results (i.e., real-time stress distribution data) of all finite element elements within the weld evaluation area are extracted, such as the normal stress in the X direction of monitoring point 1 being 220 MPa, the normal stress in the Y direction being 150 MPa, and the shear stress in the XY direction being 80 MPa. The three stress components of each monitoring point are treated as a single row of data, and the stress data of all monitoring points are arranged in numerical order to form an m-row, 3-column stress distribution matrix (where m is the number of monitoring points, e.g., m=500). Principal component analysis is performed on the constructed stress distribution matrix. After inputting the stress distribution matrix, three eigenvalues ​​(e.g., ...) are output. ) and the corresponding 3 main feature vectors (such as , ), where the largest eigenvalue Corresponding main feature vector This is the principal eigenvector of the stress distribution; then, based on the principal eigenvector... By combining the coordinate information of the weld evaluation area, a stress gradient tensor field (such as tensor components) is constructed using tensor calculation methods (e.g., based on the gradient operator ∇). (corresponding to the X, Y, and Z directions respectively), this tensor field refers to the tensor field that describes the magnitude and direction distribution of stress gradient within the weld evaluation area. It can reflect the stress change trend and represent the rate of change of stress at each location along different directions.

[0044] Specifically, based on the constructed stress gradient tensor field, the rate of change of stress gradient at each monitoring point is calculated using numerical differentiation methods (such as the central difference method). For example, the rate of change of stress gradient at monitoring point A is 50 MPa / mm, and at monitoring point B it is 10 MPa / mm. A larger rate of change indicates a more drastic stress change at that point. Then, singular value decomposition (SVD) is performed on the tensor matrix of the stress gradient tensor field (SVD is a mathematical method that decomposes a matrix into singular values, left singular vectors, and right singular vectors, which can be used to extract the main directional information of the stress gradient tensor field), resulting in three singular values ​​(e.g., ...). , , ) and the corresponding left singular vector (e.g. , , ), where the maximum singular value The corresponding left singular vector This refers to the principal direction of stress distribution. The calculated principal direction of stress distribution (e.g., ...) Using the main direction as the analysis reference, all monitoring points within the weld evaluation area are traversed along this direction, and the stress gradient change rate at each point is recorded. A threshold for abrupt changes in the rate of change is set (e.g., the difference in the rate of change between two adjacent points exceeds 30 MPa / mm). When traversing along the main direction, if the difference in the rate of change of the stress gradient between this point and the previous point exceeds the threshold, this point is marked as a change point. Connecting all the change points in sequence, the resulting closed or semi-closed curve is the stress change boundary of the weld evaluation area. The stress change boundary refers to the boundary line formed by the location of a significant change in the rate of change of the stress gradient within the weld evaluation area, which is the transition region from a gentle to a severe stress state. The stress gradient change rate data of all monitoring points on the stress change boundary are extracted, and the monitoring point with the largest change rate is selected by comparison (e.g., the change rate of monitoring point C is 80 MPa / mm, the maximum value on the boundary). This point is the stress change location of the weld evaluation area. The coordinates (e.g., X, Y, Z) of this stress change location are recorded and marked as key weld points. The stress change location is the specific point on the stress change boundary with the largest rate of change of the stress gradient, which is the location with the most severe stress state.

[0045] Furthermore, by constructing a three-dimensional geometric model of the battery frame, performing finite element analysis and mathematical calculations, the weld evaluation area and key points of the weld were accurately determined. This solved the problem that static analysis could not locate high-risk weld areas and key points, providing a clear target for subsequent measured data collection and stress analysis, and improving the pertinence and accuracy of stress assessment.

[0046] S2. When the preset containerized power supply device is in a swaying condition, the vibration response signal and excitation reference signal of the vibration sensor at the key point of the weld are collected simultaneously.

[0047] In this embodiment of the invention, the swaying condition refers to the working state in which the pre-set containerized power supply device sways periodically laterally or longitudinally with the carrier during transportation or use. The vibration response signal refers to the electrical signal (such as an acceleration signal) collected by the vibration sensor that reflects the vibration state of key weld points under the swaying condition. The excitation reference signal refers to the signal used to reflect the characteristics of the excitation source under the swaying condition, and can be used as a reference for analyzing the correlation between the vibration response signal and the excitation source (such as the operating status signal of the power converter).

[0048] In this embodiment of the invention, the synchronous acquisition of vibration response signals and excitation reference signals from vibration sensors at key points of the weld includes:

[0049] The layout topology diagram of the vibration sensor is constructed based on the preset multi-channel data acquisition unit, and the data acquisition sequence is configured based on the layout topology diagram;

[0050] Each vibration sensor is activated sequentially according to the data acquisition sequence, and vibration response signals are acquired through the activated vibration sensors.

[0051] The operating status signal of the power converter is synchronously acquired based on the acquisition timestamp of the vibration response signal, and the operating status signal is used as the excitation reference signal.

[0052] In detail, determine the number of channels in the multi-channel data acquisition unit (e.g., select an 8-channel data acquisition unit), and then assign one vibration sensor to each key weld point according to the number of key weld points. Number the sensors (A, B, C, D, E), corresponding to the five key weld points respectively. Generate a layout topology diagram based on the sensor number, installation coordinates, and connection relationship with each channel of the data acquisition unit. Based on this layout topology diagram, configure the data acquisition sequence: set the acquisition order as A→B→C→D→E, the single acquisition duration of each sensor is 10s (covering one complete swing cycle), the activation interval between adjacent sensors is 0.1s (ensuring continuous acquisition), and the sampling frequency is set to 1000Hz to avoid signal aliasing. The multi-channel data acquisition unit with the configured data acquisition sequence is physically connected to each vibration sensor (already installed at the corresponding key points of the weld). Then, the containerized power supply is placed on the swing test bench (simulating the swing condition of the battery rack weld, with a swing angle of ±15° and a swing period of 10s). The data acquisition unit is started, and it sends activation signals to each vibration sensor in sequence according to the preset acquisition sequence (A→B→C→D→E). The activated sensors convert the vibration acceleration signal at the key point of the weld into an electrical signal, which is transmitted to the data acquisition unit through the connection line. The data acquisition unit filters the received electrical signal (using a low-pass filter with a cutoff frequency of 200Hz to filter out high-frequency noise) and performs analog-to-digital conversion (converting the analog signal into a digital signal with a quantization bit of 16 bits). A timestamp is added to each digital signal (e.g., the signal acquired by sensor A on 2024-05-20 14:30:00.000 is timestamped at that time).

[0053] Specifically, a current sensor is installed at the output of the power converter of the containerized power supply unit to collect the output current signal (i.e., the operating status signal) of the power converter. The current sensor is connected to an idle channel of a multi-channel data acquisition unit, and a timestamp synchronization acquisition function is set in the software of the data acquisition unit: when the data acquisition unit adds a timestamp to the vibration response signal, the idle channel is synchronously triggered to collect the operating status signal output by the current sensor, and the same timestamp as the corresponding vibration response signal is added to the operating status signal. After the acquisition is completed, the vibration response signal with the same timestamp is associated and stored with the operating status signal. At this time, the operating status signal is the excitation reference signal, realizing the time synchronization between the vibration response signal and the excitation reference signal.

[0054] Furthermore, by simultaneously acquiring vibration response signals and excitation reference signals at key weld points under oscillating conditions, the measured data and actual working conditions were combined, solving the problem of the disconnect between theoretical analysis and actual working conditions. This provided reliable raw data for subsequent stress calculations based on measured data, ensuring the authenticity of the stress assessment results.

[0055] S3. Perform phase space reconstruction on the vibration response signal to obtain the correlation dimension of the vibration signal, and analyze the dynamic equivalent stress amplitude of the key points of the weld based on the correlation dimension of the vibration signal.

[0056] In this embodiment of the invention, the vibration signal correlation dimension refers to a quantitative index (without units) that reflects the complexity of the vibration signal, obtained by analyzing the geometric characteristics of the attractor in the high-dimensional reconstructed phase space. The larger the correlation dimension, the more complex the vibration signal and the more intense the dynamic response of the key points of the weld.

[0057] In this embodiment of the invention, the step of reconstructing the phase space of the vibration response signal to obtain the correlation dimension of the vibration signal includes:

[0058] Non-stationary time series data are extracted from the vibration response signal, and empirical mode decomposition is performed on the non-stationary time series data to obtain multiple intrinsic mode function components.

[0059] The target intrinsic mode function component with the highest complexity is selected from multiple intrinsic mode function components, and the target intrinsic mode function component is used as the target vibration feature sequence.

[0060] The phase space of the target vibration feature sequence is reconstructed using the preset delay coordinate method to obtain a high-dimensional reconstructed phase space, and the geometric properties of the attractors in the high-dimensional reconstructed phase space are calculated.

[0061] The chaotic properties of the attractor are verified based on the geometric characteristics, and the correlation integral function in the high-dimensional reconstructed phase space is calculated based on the verified attractor.

[0062] In a preset double logarithmic coordinate system, the linear relationship interval between the correlation integral function and the preset scale parameter is analyzed, and the slope of the linear relationship interval is used as the correlation dimension of the vibration signal.

[0063] In detail, the statistical characteristics that change over time are extracted from the vibration response signal, i.e., the non-stationary time series data. Using the empirical mode decomposition function, this non-stationary time series data is decomposed, with the decomposition termination condition set at a residual error of less than 0.01 between two adjacent decompositions. After decomposition, multiple intrinsic mode function components (IMF1-IMF5) and one residual component (reflecting the signal trend) are obtained. Each IMF component corresponds to a vibration at a specific characteristic scale (e.g., IMF1 corresponds to high-frequency vibration, and IMF5 corresponds to low-frequency vibration), thus separating the vibration components at different characteristic scales in the non-stationary signal. The approximate entropy value of each IMF component (IMF1-IMF5) is calculated using an approximate entropy calculation function. The component with the largest entropy value is selected; this component is the target IMF component with the highest complexity. The signal values ​​of this target IMF component are then arranged in chronological order to form the target vibration feature sequence. The delay time is determined using the autocorrelation method. The autocorrelation function of the target vibration characteristic sequence is calculated. The time when the autocorrelation function first drops to 1 / e of its maximum value (approximately 0.368) is the delay time. The embedding dimension is determined using the spurious nearest neighbor method, and the embedding dimension is gradually increased. Calculate the proportion of spurious neighbors in the high-dimensional phase space. When the proportion drops below 5%, the corresponding... That is, the embedding dimension; based on a determined delay time and embedding dimension, the target vibration feature sequence is... Transform into dimensional vector , The space formed by these vectors is the high-dimensional reconstructed phase space; and the geometric properties of the attractor in this phase space are calculated, such as obtaining the volume of the attractor by calculating the convex hull volume of the vector set, and quantifying the dynamic law of the attractor through geometric properties.

[0064] Specifically, based on the calculated geometric properties of the attractor (such as fractal dimension, where non-integer fractal dimension is a typical characteristic of chaotic attractors), combined with the Lyapunov exponent, such as the maximum Lyapunov exponent being 0.3 > 0, it is shown that the attractor is sensitive to initial conditions, thus verifying that the attractor possesses chaotic properties; based on the verified chaotic attractor, the correlation integral function is calculated. That is, setting preset scale parameters The value range is 0.1-10 (m / s²), for each Calculate all vector pairs in the high-dimensional reconstructed phase space. Euclidean distance ,statistics The number of vector pairs, associated integral function Ultimately, different results were obtained. corresponding Value. Based on a double logarithmic coordinate system (x-axis is...). The vertical axis is , calculate Data points are labeled in a coordinate system; the data points are fitted, and the goodness of fit is found. The linear interval; the slope of the fitted line in this linear interval is the correlation dimension of the vibration signal.

[0065] In this embodiment of the invention, the dynamic equivalent stress amplitude refers to the dynamic stress (which changes with time) at the key point of the weld under swaying conditions being equivalent to a constant stress amplitude, which is used to measure the maximum level of dynamic stress.

[0066] In this embodiment of the invention, the step of analyzing the dynamic equivalent stress amplitude of the key points of the weld based on the correlation dimension of the vibration signal includes:

[0067] Construct a database mapping relationship between the preset material damage level and the vibration signal dimension;

[0068] Query the damage state corresponding to the associated dimension of the vibration signal in the mapping relationship database;

[0069] The reduction coefficient of the load-bearing area of ​​the material at the critical points of the weld is determined based on the damage state.

[0070] The theoretical stress amplitude at the critical point of the weld is corrected based on the bearing area reduction coefficient, and the corrected theoretical stress amplitude is used as the dynamic equivalent stress amplitude at the critical point of the weld.

[0071] In detail, the preset material refers to the metal material used in the battery frame, and the damage degree refers to the damage level of the preset material under dynamic load, which is usually divided into four levels: no damage, slight damage, moderate damage, and severe damage. Steel specimens of the same material as the battery frame are selected to generate multiple sets of standard tensile specimens (3 specimens per set). The specimens are placed on a fatigue testing machine, and dynamic loads of different amplitudes (e.g., 100MPa, 200MPa, 300MPa, 400MPa) are applied. Vibration sensors are installed on the specimen surface to collect vibration response signals under different damage degrees. The collected vibration response signals are reconstructed in phase space and their correlation dimensions are calculated to obtain the correlation dimensions corresponding to different damage degrees (e.g., 1.0 for no damage, 1.8 for slight damage, 2.5 for moderate damage, and 3.2 for severe damage). The damage degrees (no damage, slight damage, moderate damage, severe damage) and their corresponding correlation dimensions (1.0, 1.8, 2.5, 3.2) are organized into a mapping relationship table to form a mapping relationship database. The correlation dimension of the vibration signal at the key point of the weld is used as the query condition to send a query request to the mapping database. The database matches the corresponding damage degree according to the query condition. When the correlation dimension is 2.5, the query result is moderate damage, which is the damage state of the material at the key point of the weld.

[0072] Specifically, a correspondence between damage state and bearing area reduction coefficient is established based on material mechanics test data (e.g., coefficient is 1.0 for no damage, 0.9 for slight damage, 0.7 for moderate damage, and 0.4 for severe damage). Based on the obtained damage state, the bearing area reduction coefficient of the material at key weld points is determined from the correspondence, thus providing key parameters for correcting the theoretical stress amplitude. The formula for calculating the theoretical stress amplitude is as follows: ,in The dynamic load at the critical point of the weld (calculated from the collected excitation reference signal). The intact load-bearing area at the critical weld point (calculated based on the weld size). Based on the excitation reference signal (the power converter's operating status signal, such as an output current of 50A), and combined with the structural mechanics model of the battery rack, the dynamic load at the critical weld point is calculated. Then, calculate the intact bearing area based on the actual dimensions of the weld. Thus, the theoretical stress amplitude is obtained. Substitute the bearing area reduction factor into the correction formula ,in The reduced area coefficient is used to obtain the corrected theoretical stress amplitude. This value is the dynamic equivalent stress amplitude at the critical point of the weld.

[0073] Furthermore, by reconstructing the phase space, the one-dimensional vibration response signal is transformed into a high-dimensional phase space attractor, and the correlation dimension of the vibration signal is calculated, thus quantifying the complexity of the vibration signal. Then, by combining the pre-set mapping relationship database between the degree of material damage and the correlation dimension, the damage state and bearing area reduction coefficient of the key points of the weld are determined, and finally the dynamic equivalent stress amplitude is corrected to ensure that the obtained dynamic equivalent stress amplitude can truly reflect the actual stress level of the key points of the weld under the swaying condition, providing an accurate amplitude index for subsequent comprehensive stress assessment.

[0074] S4. Perform joint vibration deformation analysis on the vibration response signal and the excitation reference signal to obtain the dynamic strain field distribution corresponding to the key points of the weld.

[0075] In this embodiment of the invention, dynamic strain field distribution refers to the distribution state of strain as time and spatial location changes in the key points of the weld and its surrounding area under swaying conditions.

[0076] In this embodiment of the invention, the step of performing joint vibration deformation analysis on the vibration response signal and the excitation reference signal to obtain the dynamic strain field distribution corresponding to the key points of the weld includes:

[0077] Modal decomposition is performed on the vibration response signal to obtain the vibration modes;

[0078] The dominant excitation frequency is obtained by performing spectral analysis on the excitation reference signal;

[0079] Generate a coupling relationship model between the vibration mode and the dominant excitation frequency, and calculate the vibration dynamic participation factor of the vibration mode based on the coupling relationship model;

[0080] The dynamic strain field distribution corresponding to the key points of the weld is generated based on the vibration dynamic participation factor and the vibration mode.

[0081] In detail, modal decomposition is performed on the vibration response signal to identify the first five vibration modes of the battery rack structure (e.g., the first mode has a natural frequency of 10Hz and a mode shape of overall lateral bending of the battery rack; the second mode has a natural frequency of 25Hz and a mode shape of local vibration in the weld area). The natural frequency, mode shape vector (describing the relative magnitude of vibration displacement at each location), and damping ratio (e.g., the damping ratio of the first mode is 0.02) of each vibration mode are output, thereby extracting the inherent vibration characteristics of the battery rack structure and clarifying the modal composition of the vibration at key weld points. A complete steady-state signal is extracted from the excitation reference signal, and spectral analysis is performed on this steady-state signal. The frequency resolution is set to 0.2Hz (determined by the sampling frequency and signal length; frequency resolution = sampling frequency / number of signal points). The frequency domain amplitude spectrum of the signal is obtained (the horizontal axis is frequency, and the vertical axis is amplitude). The frequency component with the largest amplitude is selected from the amplitude spectrum; this frequency is the dominant excitation frequency.

[0082] Specifically, the coupling relationship model is as follows: ,in For the first Vibration dynamic participation factors of first-order vibration modes, For the first First-order vibration mode at the dominant excitation frequency The amplitude of the frequency response function at that point. For all vibration modes in The sum of the amplitudes of the frequency response functions at each point. Based on the vibration modes and the dominant excitation frequency, the frequency response function is tested (using the hammer impact method, applying pulse excitation to the battery rack, collecting response signals, and calculating the frequency response functions of each mode) to obtain the first... First mode in frequency response function amplitude at point And calculate the sum of the amplitudes of the frequency response functions for all modes. This allows us to obtain the vibration dynamic participation factors for each mode, quantify the contribution of each vibration mode to the vibration of the key weld point, focus on the core influencing mode, and provide a weighting basis for the subsequent generation of the dynamic strain field distribution. Based on the mode shape vectors, we calculate the strain distribution of the key weld point and its surrounding area (a radius of 50 mm centered on the key weld point) under each mode (e.g., under the first mode, the strain at the key weld point is...). The area within 5mm around is In the second-order mode, the strain at the critical point of the weld is: The area within 5mm around is The strain distribution of each mode is weighted according to its vibration dynamic participation factor to obtain the total strain distribution of the key points of the weld and its surrounding area; the change of the total strain distribution over time (combined with the 10s period of the oscillating condition) is dynamically simulated to obtain the dynamic strain field distribution (presented in the form of a cloud map, with different colors representing different strain values, such as red indicating strain > 10s). Blue indicates strain. This allows for a direct visualization of the dynamic strain spatial distribution patterns at key weld points and their surrounding areas.

[0083] Furthermore, by performing modal decomposition on the vibration response signal to extract the vibration modes, performing spectral analysis on the excitation reference signal to determine the dominant excitation frequency, and then constructing a coupling relationship model between the two to calculate the vibration dynamic participation factor, a dynamic strain field distribution is finally generated. This realizes the correlation analysis between the vibration response and the excitation source, solving the problems of not being able to clearly define the contribution of each vibration mode to the weld deformation and lacking information on the spatial distribution of strain. The obtained dynamic strain field distribution can intuitively present the strain change law of the key points of the weld and its surrounding area, providing comprehensive spatial strain data support for the subsequent calculation of the dynamic stress time-domain history.

[0084] S5. Based on the dynamic strain field distribution and constitutive parameters of the battery frame material, perform incremental iterative analysis of the dynamic stress time-domain history of the key points of the weld.

[0085] In this embodiment of the invention, constitutive parameters refer to parameters describing the mechanical properties of the battery frame material (such as Q235 steel), including elastic modulus, Poisson's ratio, yield strength, strain rate hardening coefficient, cyclic hardening coefficient, etc., used to establish the relationship between stress and strain. The dynamic stress time-domain history refers to a complete process curve (horizontal axis: time, vertical axis: stress) describing the stress change over time at key weld points throughout the entire oscillation cycle.

[0086] In this embodiment of the invention, the incremental iterative analysis of the dynamic stress time-domain history of the key points of the weld based on the dynamic strain field distribution and the constitutive relation parameters of the battery frame material includes:

[0087] A dynamic constitutive model of the battery frame material is constructed based on the preset strain rate strengthening characteristics and cyclic hardening characteristics, and an adaptive iteration step size for enhanced iteration is generated.

[0088] The dynamic strain field distribution is discretized into strain increments in a time series, and the strain increments at each time step are updated using the dynamic constitutive relation model and the adaptive iteration step size.

[0089] The updated stress increment is superimposed with the stress state of the previous time step to obtain the target stress state.

[0090] By fusing the target stress states at all time steps, the dynamic stress time-domain history of the key points of the weld is obtained.

[0091] In detail, parameters of strain rate strengthening characteristics (such as the strain rate strengthening coefficient of steel) are obtained through tensile testing, and parameters of cyclic hardening characteristics (such as the cyclic hardening coefficient) are obtained through fatigue testing. Standard specimens of Q235 steel, the material for battery frames, are selected and subjected to different strain rates (e.g., strain rate coefficient of steel) on an MTS fatigue testing machine. , , Tensile tests were conducted to obtain strain rate strengthening characteristic parameters (strain rate strengthening coefficient, reference strain rate); symmetrical cyclic fatigue tests (stress ratio R = -1) were conducted to obtain cyclic hardening characteristic parameters (cyclic hardening coefficient, number of saturated cycles); based on the dynamic constitutive relation model, the above strain rate strengthening characteristic parameters and cyclic hardening characteristic parameters were substituted into the model formula: ,in For dynamic stress, Static yield strength (Q235 steel) , The strain hardening index ( ), To represent the equivalent plastic strain, a dynamic constitutive model of the battery frame material was constructed. Based on the dynamic strain field distribution, the rate of change of strain over time (strain rate) was calculated, and an adaptive iteration step size rule was set: when the strain rate > At that time, step length When the strain rate is ≤1 At that time, step length Generate an adaptive iterative step size sequence (e.g., the strain rate is high in the first 2 seconds, with a step size of 0.005s; the strain rate is low in the next 8 seconds, with a step size of 0.01s).

[0092] Specifically, based on the generated adaptive iterative step size sequence (such as...) The dynamic strain field distribution is discretized into strain increments corresponding to the time step. (i=1, 2, ..., N, where N is the total number of time steps), strain increment , For the first Strain values ​​for each time step For the first Strain values ​​for each time step; for each time step , will increase the strain Current strain rate Substituting the stress into the constructed dynamic constitutive model, the stress increment at that time step is calculated. This process updates the stress for each strain increment at each time step, thus transforming the continuous strain field distribution into discrete strain increments. The initial stress state is then set. (Before the start of the swing condition, there is no stress at the critical points of the weld); for the first time step Target stress state For the second time step First, based on the dynamic strain field distribution, we obtain... , Then, substituting into the dynamic constitutive relation model, we can calculate... Then compare it with the stress state of the previous time step. Superposition yields the target stress state. ; Calculate the target stress state for each time step in this manner. (i=1, 2, ..., N), thus obtaining the stress value at the end of each time step, and gradually constructing the stress change process over time. The target stress state for all time steps is then determined. (i=1, 2, ..., N) and the corresponding time ( Organize into time-stress data pairs ( , ); by time The x-axis is represented by stress (unit: seconds, range: 0-10 seconds). Plot the time-stress curve with the vertical axis (unit: MPa, range: 0-50 MPa); smooth the curve (using the moving average method, window size: 5) to remove computational noise and obtain a smooth dynamic stress time-domain history curve of the weld key point (e.g., the stress reaches a peak value of 45 MPa at t=2.5s; the stress reaches a valley value of 5 MPa at t=7.5s), thus obtaining the complete stress time history of the weld key point throughout the entire oscillation cycle.

[0093] Furthermore, based on the time-series strain increments obtained by discretizing the dynamic strain field distribution, and combined with a dynamic constitutive relation model that considers strain rate strengthening and cyclic hardening characteristics, the stress increments at each time step are calculated step by step through an adaptive incremental iteration method and superimposed to obtain the target stress state. Finally, the dynamic stress time-domain history is formed, which can capture the stress changes over time. Moreover, the dynamic stress time-domain history fully reflects the stress fluctuation law of the key points of the weld during the entire oscillation cycle.

[0094] S6. Determine the stress of the battery frame weld under swaying condition based on the stress data in the dynamic stress time-domain history and the dynamic equivalent stress amplitude.

[0095] In this embodiment of the invention, the stress of the battery frame weld under swaying conditions refers to the comprehensive stress level that the weld actually bears under swaying conditions, which is obtained by mapping through the comprehensive stress evaluation index. It is the final basis for judging whether the weld meets the strength requirements.

[0096] In this embodiment of the invention, determining the stress of the battery frame weld under swaying conditions based on the stress data in the dynamic stress time-domain history and the dynamic equivalent stress amplitude includes:

[0097] Rainflow counting was performed on the dynamic stress time-domain history to obtain stress cycle characteristics;

[0098] The cumulative damage degree of the battery frame weld is calculated based on the stress cycle characteristics described above;

[0099] The safety margin factor of the battery frame weld is generated based on the dynamic equivalent stress amplitude and the material fatigue limit of the battery frame.

[0100] Based on the cumulative damage degree and the safety margin coefficient, the comprehensive stress evaluation index of the battery frame weld is analyzed, and the comprehensive stress evaluation index is mapped to the stress of the battery frame weld under swaying conditions.

[0101] In detail, the rainflow counting function is called, and the stress threshold for rainflow counting is set to 1 MPa (stress fluctuations below this threshold are not counted in the cycle). The counting process is initiated, starting from the peak or trough of the stress time-domain history, and sequentially counting stress cycles along the rainflow direction (e.g., from a stress peak of 45 MPa to a trough of 5 MPa, then rising back to a peak of 42 MPa, forming a complete stress cycle). After counting, the stress cycle characteristics are output, including the number of stress cycles (e.g., 10 complete cycles), the stress amplitude of each cycle (e.g., the first cycle amplitude is 20 MPa, the second is 18 MPa), and the average stress (e.g., the first cycle average stress is 25 MPa, the second is 23 MPa), thereby quantifying the cyclic load characteristics in the dynamic stress time-domain history. The cumulative damage degree is calculated using Miner's linear cumulative damage theory, with the theoretical formula being... ,in To accumulate damage, For the first The number of cycles corresponding to each stress amplitude. For the material in the first Fatigue life at a given stress amplitude (obtainable from the material's S / N curve). Obtain the S / N curve of the battery frame material Q235 steel (obtained through fatigue testing, such as stress amplitude). At that time, fatigue life ; hour, ); based on stress cycle characteristics (such as the first cycle) , Second cycle; , (Time), retrieve the corresponding fatigue life from the SN curve. Second-rate, Next, substitute the values ​​into the Miner's theoretical formula to calculate the cumulative damage. If the cumulative damage is much less than 1, it indicates that the current damage level is extremely low.

[0102] Specifically, the fatigue limit of the Q235 steel used for the battery rack was determined. (Determined through standard fatigue testing); Obtain the dynamic equivalent stress amplitude at key weld points. Based on the safety margin coefficient calculation formula The safety margin coefficient is obtained. If S>1, it indicates that the weld stress amplitude is much lower than the material fatigue limit, and the safety margin is high. The formula for the comprehensive stress assessment index is: ,in As a comprehensive stress assessment index, The maximum safety margin factor is preset (set to 100, representing the ideal safety state). The cumulative damage is calculated using the comprehensive stress assessment index formula. Safety margin coefficient Dynamic equivalent stress amplitude The corresponding comprehensive stress assessment index. Establish a mapping relationship between the comprehensive stress assessment index and the stress under swaying conditions at the battery frame weld (since the values ​​of the two are close in the current calculations, the mapping relationship is set to...). ),Will Mapped to the stress of the battery frame weld under swaying conditions .

[0103] Furthermore, stress cycle characteristics are obtained by processing the dynamic stress time-domain history through rainflow counting. The cumulative damage degree is calculated based on Miner's linear cumulative damage theory. A safety margin coefficient is generated by combining the dynamic equivalent stress amplitude and the material fatigue limit. The stress of the battery rack weld under swaying conditions is obtained by mapping a comprehensive stress assessment index. This achieves a comprehensive assessment of both weld stress damage accumulation and safety margin, solving the problems of incomplete single-index assessment and inability to quantify the overall stress risk of the weld in existing technologies. The obtained swaying condition stress can be directly used as a basis for judging whether the weld meets the strength requirements under swaying conditions, providing key technical support for the safe design and maintenance of pre-designed containerized power unit battery racks.

[0104] like Figure 2 The diagram shown is a functional block diagram of a battery rack weld swaying stress assessment system provided in an embodiment of the present invention.

[0105] The battery rack weld swaying stress assessment system 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the battery rack weld swaying stress assessment system 100 may include a weld key point identification module 101, a signal synchronous acquisition module 102, a dynamic equivalent stress amplitude analysis module 103, a combined vibration deformation analysis module 104, an incremental iterative analysis module 105, and a working condition stress determination module 106. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0106] In this embodiment, the functions of each module / unit are as follows:

[0107] The weld key point identification module 101 is used to determine the weld evaluation area on the battery rack of the preset containerized power device and identify the weld key points in the weld evaluation area.

[0108] The signal synchronization acquisition module 102 is used to synchronously acquire the vibration response signal and excitation reference signal of the vibration sensor at the key point of the weld when the preset containerized power supply device is in a swaying condition.

[0109] The dynamic equivalent stress amplitude analysis module 103 is used to reconstruct the phase space of the vibration response signal to obtain the vibration signal correlation dimension, and analyze the dynamic equivalent stress amplitude of the key points of the weld based on the vibration signal correlation dimension.

[0110] The joint vibration deformation analysis module 104 is used to perform joint vibration deformation analysis on the vibration response signal and the excitation reference signal to obtain the dynamic strain field distribution corresponding to the key points of the weld.

[0111] The incremental iterative analysis module 105 is used to incrementally iteratively analyze the dynamic stress time-domain history of the key points of the weld based on the dynamic strain field distribution and the constitutive parameters of the battery frame material.

[0112] The working condition stress determination module 106 is used to determine the swing working condition stress of the battery frame weld seam based on the stress data in the dynamic stress time domain history and the dynamic equivalent stress amplitude.

[0113] In detail, the modules in the battery rack weld swaying stress assessment system 100 described in this embodiment of the invention employ the same methods as described above during use. Figure 1 The method used is the same as the stress assessment method for the swaying condition of the battery rack weld described in the previous section, and can produce the same technical effect, so it will not be repeated here.

[0114] In the several embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0115] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0116] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0117] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0118] Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects. The scope of the invention is not limited to the foregoing description, and all variations within the meaning and scope of equivalents falling within the protection scope are intended to be included in the invention.

[0119] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0120] Furthermore, it is clear that the word "including" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems described in a system item may also be implemented by a single unit or system through software or hardware. Terms such as "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0121] Finally, 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.

Claims

1. A method for evaluating battery rack weld swing work condition stress, characterized in that, The method comprises: constructing a battery rack three-dimensional geometric model based on the structural features of the battery rack; performing meshing on the battery rack three-dimensional geometric model to obtain a finite element calculation grid, applying a fixed constraint at the bottom bolt hole position of the finite element calculation grid according to the actual installation mode of the battery rack in the container, and at the same time applying an angular acceleration load along the X axis according to the swing working condition requirement, to complete the application of the preset boundary condition; setting a calculation time step and a total calculation time, calculating the stress value of each element at different time steps according to the element mechanical parameters of the finite element calculation grid and the applied boundary condition, outputting the stress values of all elements in the form of a cloud chart to obtain a stress distribution cloud chart of the battery rack under the swing working condition; identifying a stress concentration area in the finite element calculation grid according to the stress distribution cloud chart, marking the stress concentration area as a weld evaluation area, and identifying a weld key point in the weld evaluation area; When the preset container-type power supply device is in a swing working condition, the number of channels of the preset multi-channel data acquisition unit is determined; a layout topology relationship diagram is generated according to the number, installation position coordinates and connection relationship with each channel of the data acquisition unit of each sensor; based on the layout topology relationship diagram, a data acquisition sequence is configured; the multi-channel data acquisition unit with the configured data acquisition sequence is physically connected with each vibration sensor, and the battery rack weld swing working condition is simulated; the data acquisition unit is started, and an activation signal is sequentially sent to each vibration sensor according to the preset acquisition sequence; the activated sensor converts the vibration acceleration signal at the weld key point into an electric signal, which is transmitted to the data acquisition unit through a connection line; the data acquisition unit filters and digitizes the received electric signal, and adds a collection time stamp to each digital signal; a current sensor is installed at the output end of the power converter of the container-type power supply device to collect the output current signal of the power converter; the current sensor is connected with an idle channel of the multi-channel data acquisition unit; when the data acquisition unit adds a time stamp to the vibration response signal, the idle channel is triggered to collect the working state signal output by the current sensor, and the working state signal is added with the same time stamp as the corresponding vibration response signal; after the collection is completed, the vibration response signal and the working state signal with the same time stamp are associated and stored, and the working state signal at this time is the excitation reference signal; The vibration response signal is reconstructed in phase space to obtain a vibration signal correlation dimension, and the dynamic equivalent stress amplitude of the weld key point is analyzed according to the vibration signal correlation dimension; The vibration response signal is decomposed in mode to obtain a vibration mode, the excitation reference signal is analyzed in frequency spectrum to obtain a dominant excitation frequency, and a coupling relationship model of the vibration mode and the dominant excitation frequency is generated, wherein the coupling relationship model is: wherein is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the sum of the frequency response function amplitudes of all orders of vibration mode at the dominant excitation frequency According to the mode shape vector of the vibration mode, the strain distribution of the weld key point and the surrounding area under each order mode is calculated, the strain distribution of each order mode is weighted and calculated according to the vibration dynamic participation factor to obtain the total strain distribution of the weld key point and the surrounding area; the change of the total strain distribution with time is dynamically simulated in combination with the cycle of the swing working condition to obtain the dynamic strain field distribution; According to the preset strain rate strengthening characteristics and the cyclic hardening characteristics, a dynamic constitutive relationship model of the battery rack material is constructed, and an enhanced iterative adaptive iteration step is generated, wherein the dynamic constitutive relationship model comprises: wherein, is the dynamic stress, is the static yield strength, is the equivalent plastic strain, is the strain rate hardening characteristic, is the cyclic hardening characteristic, is the strain hardening exponent; The dynamic strain field distribution is discretized into a time sequence of strain increments, and the stress of each time step strain increment is updated using the dynamic constitutive relationship model and the adaptive iteration step; the updated stress increment is superimposed with the stress state of the previous time step to obtain the target stress state; the target stress states of all time steps are fused to obtain the dynamic stress time domain history of the weld key point; According to the stress data in the dynamic stress time domain history and the dynamic equivalent stress amplitude, the weld seam swing working condition stress of the battery rack is determined.

2. The method of battery rack weld seam walk-off stress evaluation of claim 1, wherein, The weld key point in the weld evaluation area is identified, including: Real-time stress distribution data of the weld evaluation area is identified, and a stress distribution matrix of the weld evaluation area is constructed according to the real-time stress distribution data; Principal component analysis is performed on the stress distribution matrix to obtain principal feature vectors and eigenvalues of the stress distribution, and a stress gradient tensor field is constructed based on the principal feature vectors; According to the stress gradient tensor field, the stress gradient change rate of each point in the weld evaluation area is calculated, and singular value decomposition is performed on the stress gradient tensor field to obtain the principal direction of the stress distribution; Based on the principal direction and the stress gradient change rate, the stress mutation boundary of the weld evaluation area is identified; According to the stress mutation boundary and the stress gradient change rate, the stress mutation position of the weld evaluation area is determined, and the stress mutation position is determined as the weld key point.

3. The method of battery rack weld seam walk-off stress evaluation of claim 1, wherein, The phase space of the vibration response signal is reconstructed to obtain the correlation dimension of the vibration signal, including: Non-stationary time series data in the vibration response signal is extracted, and empirical mode decomposition is performed on the non-stationary time series data to obtain a plurality of intrinsic mode function components; The most complex target intrinsic mode function component is selected from the plurality of intrinsic mode function components, and the target intrinsic mode function component is taken as a target vibration feature sequence; The target vibration feature sequence is reconstructed in the phase space by using a preset delay coordinate method to obtain a high-dimensional reconstructed phase space, and the geometric characteristics of an attractor in the high-dimensional reconstructed phase space are calculated; The chaotic characteristics of the attractor are verified according to the geometric characteristics, and the correlation integral function in the high-dimensional reconstructed phase space is calculated based on the verified attractor; The linear relationship interval of the correlation integral function and a preset scale parameter is analyzed in a preset double logarithmic coordinate system, and the slope of the linear relationship interval is taken as the correlation dimension of the vibration signal.

4. The method of battery rack weld seam walk-off stress evaluation of claim 1, wherein, The dynamic equivalent stress amplitude of the weld key point is analyzed according to the correlation dimension of the vibration signal, including: construct a mapping relationship database of a preset material damage degree and the correlation dimension of the vibration signal; query the damage state corresponding to the correlation dimension of the vibration signal in the mapping relationship database; determine a load area reduction coefficient of the material at the weld key point according to the damage state; correct the theoretical stress amplitude at the weld key point based on the load area reduction coefficient, and take the corrected theoretical stress amplitude as the dynamic equivalent stress amplitude of the weld key point.

5. The method of battery rack weld seam walk-off stress evaluation of claim 1, wherein, The determination of the battery rack weld swing working condition stress according to the stress data in the dynamic stress time domain history and the dynamic equivalent stress amplitude comprises: perform rainflow counting processing on the dynamic stress time domain history to obtain stress cycle characteristics; calculate the cumulative damage degree of the battery rack weld based on the stress cycle characteristics; generate a safety margin coefficient of the battery rack weld according to the dynamic equivalent stress amplitude and the material fatigue limit of the battery rack; analyze the comprehensive stress evaluation index of the battery rack weld based on the cumulative damage degree and the safety margin coefficient, and map the comprehensive stress evaluation index to the battery rack weld swing working condition stress.

6. A battery rack weld seam swing out stress evaluation system characterized by, The system for performing the battery rack weld swing working condition stress evaluation method as claimed in any one of claims 1-5 comprises: a weld key point identification module configured to construct a three-dimensional geometric model of the battery rack based on the structural characteristics of the battery rack, perform meshing on the three-dimensional geometric model of the battery rack to obtain a finite element calculation mesh, apply a fixed constraint at the bottom bolt hole position of the finite element calculation mesh according to the actual installation mode of the battery rack in the container, and simultaneously apply an angular acceleration load along the X axis according to the swing working condition requirement to complete the application of a preset boundary condition, set a calculation time step and a total calculation time, calculate the stress value of each element at different time steps according to the element mechanical parameters of the finite element calculation mesh and the applied boundary condition, output the stress values of all elements in the form of a cloud chart to obtain a stress distribution cloud chart of the battery rack under the swing working condition, identify a stress concentration area in the finite element calculation mesh according to the stress distribution cloud chart, mark the stress concentration area as a weld evaluation area, and identify a weld key point in the weld evaluation area. The signal synchronous acquisition module is configured to: determine a channel number of a preset multi-channel data acquisition unit when a preset container-type power supply device is in a swing working condition; generate a layout topology graph according to a number of each sensor, a mounting position coordinate, and a connection relationship with each channel of the data acquisition unit; configure a data acquisition sequence based on the layout topology graph; physically connect the multi-channel data acquisition unit with the configured data acquisition sequence and each vibration sensor, simulate a battery rack weld seam swing working condition; start the data acquisition unit, and sequentially send an activation signal to each vibration sensor according to a preset acquisition sequence; the activated sensor converts a vibration acceleration signal at a weld seam key point into an electric signal, which is transmitted to the data acquisition unit through a connection line; the data acquisition unit filters and digitizes the received electric signal, and adds a collection time stamp to each digital signal; a current sensor is installed at an output end of a power converter of the container-type power supply device, configured to collect an output current signal of the power converter; the current sensor is connected to an idle channel of the multi-channel data acquisition unit; when the data acquisition unit adds a time stamp to a vibration response signal, an idle channel is triggered to collect a working state signal output by the current sensor, and the working state signal is added with the same time stamp as the corresponding vibration response signal; after the collection is completed, the vibration response signal and the working state signal with the same time stamp are stored in association, and the working state signal at this time is an excitation reference signal; The dynamic equivalent stress amplitude analysis module is configured to perform phase space reconstruction on the vibration response signal to obtain a vibration signal correlation dimension, and analyze a dynamic equivalent stress amplitude of the weld seam key point according to the vibration signal correlation dimension; The joint vibration deformation analysis module is configured to perform modal decomposition on the vibration response signal to obtain a vibration mode, perform frequency spectrum analysis on the excitation reference signal to obtain a dominant excitation frequency, and generate a coupling relationship model of the vibration mode and the dominant excitation frequency, wherein the coupling relationship model is: wherein is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency is the frequency response function amplitude of the nth order vibration mode at the dominant excitation frequency According to a vibration mode vector of the vibration mode, a strain distribution of the weld seam key point and a surrounding area thereof under each order mode is calculated, and a total strain distribution of the weld seam key point and the surrounding area thereof is obtained by weighted calculation of strain distributions of each order mode according to a vibration dynamic participation factor; a dynamic strain field distribution is obtained by dynamically simulating a change of the total strain distribution over time and a cycle of the swing working condition; The incremental iteration analysis module is configured to construct a dynamic constitutive relationship model of the battery rack material according to a preset strain rate strengthening characteristic and a cyclic hardening characteristic, and generate an adaptive iteration step of enhanced iteration, wherein the dynamic constitutive relationship model is: wherein, is the dynamic stress, is the static yield strength, is the equivalent plastic strain, is the strain rate hardening characteristic, is the cyclic hardening characteristic, is the strain hardening exponent; The dynamic strain field distribution is discretized into a time sequence of strain increments, and the dynamic constitutive relationship model and the adaptive iteration step are used to update stress of each time step of the strain increment; a target stress state is obtained by superimposing the updated stress increment and a stress state of a previous time step; and the target stress states of all time steps are fused to obtain a dynamic stress time domain history of the weld seam key point. The working condition stress determination module is configured to determine the battery rack weld swing working condition stress according to the stress data in the dynamic stress time domain history and the dynamic equivalent stress amplitude.

Images