New energy automobile power lithium battery performance detection method and system

Abstract

The invention provides a new energy automobile power lithium battery performance detection method and system, and relates to the technical field of lithium battery detection.The method comprises the steps that physical impact and voltage change data of a lithium battery during a new energy automobile collision test are obtained, and a potential damage area is predicted through simulation analysis; releasing a special marker in the area, and identifying the position of the marker by a detection part to determine the actual damage position in the battery; then, an electrochemical model is constructed by combining the potential damage area and the actual damage position, so that the electrical characteristics of the damage position are represented; and finally, through a low-current charging and discharging test, current activity and electric field abnormal change of a damage position are calculated based on the model, a battery overheating runaway risk assessment result is generated, and damage positioning after lithium battery collision and accurate assessment of the overheating runaway risk can be realized.

Description

Technical Field

[0001] This application relates to the field of lithium battery testing technology, and in particular to a method and system for testing the performance of power lithium batteries for new energy vehicles. Background Technology

[0002] Performance testing methods for lithium-ion batteries used in new energy vehicles are core technologies supporting the safe operation of these batteries, directly impacting the overall vehicle safety and market acceptance. With the rapid development of the new energy vehicle industry, the demand for power battery safety monitoring continues to rise, and the application scenarios for these testing methods are becoming increasingly widespread.

[0003] In existing technologies, the most commonly used method for testing the performance of lithium-ion batteries for new energy vehicles is the offline disassembly and testing method. This method requires removing the lithium battery from the vehicle and then using specialized instruments to disassemble and test the individual battery cells or modules. This method mainly obtains information about the battery's external damage and cell performance parameters through physical measurements and chemical analysis.

[0004] Offline disassembly and testing methods require damaging the original battery packaging structure, making it impossible to identify minute internal damage locations while the battery is in its original position. Furthermore, they struggle to replicate the dynamic electrical environment of a battery during actual operation. This results in test results that cannot accurately reflect the true extent of damage and potential risks after a collision, thus affecting the effectiveness of subsequent safety assessments. Therefore, existing technologies suffer from insufficient accuracy in assessing post-collision damage and risks to lithium batteries. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for testing the performance of lithium batteries for new energy vehicles, so as to solve the problem of insufficient accuracy in assessing the damage and risk of lithium batteries after collision in the prior art.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a method for testing the performance of power lithium batteries for new energy vehicles, comprising: To obtain physical impact data and voltage change data of lithium batteries during crash tests of new energy vehicles; Based on the physical impact data and the voltage change data, the potential damage area of ​​the lithium battery is predicted through simulation analysis. Within the potential damage area, a pre-set special marker is released, and the location of the special marker is identified by a detection component to determine the actual damage location inside the lithium battery. Based on the potential damage area and the actual damage location, an electrochemical model is constructed, which is used to characterize the electrical properties of the actual damage location in the internal environment of the battery. Using the electrochemical model, during low-current charge-discharge tests, the current activity process and abnormal electric field changes at the actual damage location are calculated, and an assessment result of the battery overheating runaway risk is generated based on the current activity process and the abnormal electric field changes.

[0007] Optionally, the step of using the electrochemical model to calculate the current activity process and abnormal electric field changes at the actual damage location during low-current charge-discharge testing, and generating an assessment result of the battery overheating runaway risk based on the current activity process and the abnormal electric field changes, includes: A small current charging and discharging signal is input into the electrochemical model to determine the current activity process and abnormal electric field changes at the actual damage site; Based on the current activity process, current activity characteristic data are obtained, and based on the abnormal changes in the electric field, abnormal electric field characteristic data are obtained. The current activity characteristic data and the electric field anomaly characteristic data are input into a preset risk assessment model to generate assessment results.

[0008] Optionally, the step of acquiring current activity characteristic data based on the current activity process and acquiring electric field anomaly characteristic data based on the electric field anomaly changes includes: During the current activity process, the signal of small current charging and discharging is captured, and the current flowing through the boundary of the actual damage location within the test cycle is measured; the rate of change of the current over time is calculated, and abnormal fluctuation points exceeding the preset fluctuation range in the rate of change are identified; the frequency and amplitude of the abnormal fluctuation points are statistically analyzed to generate current activity characteristic data. From the abnormal changes in the electric field, extract the electric field intensity distribution data within a preset radius centered on the actual damage location; calculate the rate of change of the electric field intensity distribution data in space, and identify abnormal regions where the rate of change exceeds a preset threshold; statistically analyze the volume and average intensity of the abnormal regions to generate electric field anomaly feature data.

[0009] Optionally, the step of predicting the potential damage area of ​​the lithium battery through simulation analysis based on the physical impact data and the voltage change data includes: Based on the physical impact data, the impact force distribution borne by the lithium battery at the moment of collision is reconstructed in the battery model corresponding to the lithium battery. Through simulation analysis, the internal stress response distribution of each component in the battery model under the action of the impact force distribution is calculated; Based on the physical tolerance limits of each component of the lithium battery, a first set of damage regions is determined from the internal stress response distribution; Based on the voltage change data and the connection relationship of the internal circuit of the battery model, the set of second damage regions is determined. The overlapping area between the first set of damaged areas and the second set of damaged areas is identified as a potential damage area.

[0010] Optionally, the step of identifying the location of the special marker by the detection component to determine the actual location of damage inside the lithium battery includes: By using detection components deployed inside the lithium battery, the physical field change signal triggered by the special marker is acquired; The physical field change signal is processed to generate a concentration distribution map of the special marker in the internal space of the lithium battery; Based on the distribution characteristics of signal intensity in the concentration distribution spectrum, the actual damage location inside the lithium battery is determined.

[0011] Optionally, processing the physical field change signal to generate a concentration distribution map of the special marker within the lithium battery includes: Based on the spatial position of each detection component and the intensity of the physical field change signal, the preliminary distribution position of the special marker inside the lithium battery is calculated using a signal tracing algorithm. Based on the battery model of the lithium battery, the distribution positions corresponding to the areas inside the lithium battery where the special marker cannot exist due to physical isolation are eliminated from the preliminary distribution positions to obtain the corrected distribution positions; Based on the corrected distribution location, a concentration distribution map of the special marker is generated in the internal space of the lithium battery.

[0012] Optionally, constructing an electrochemical model based on the potential damage area and the actual damage location includes: Within the spatial range of the potential damage area, a key area containing the actual damage location is delineated; Based on the structural characteristics of the actual damage location, the boundary of the actual damage location is defined within the critical area; Within the critical region, evolution rules are established to describe the flow process of current and the evolution of the boundary of the actual damage location as it changes with the charging and discharging state. An electrochemical model is constructed based on the aforementioned evolutionary rules.

[0013] Secondly, this application provides a performance testing system for power lithium batteries in new energy vehicles, comprising: The acquisition module is used to acquire physical impact data and voltage change data of lithium batteries during crash tests of new energy vehicles; The prediction module is used to predict the potential damage area of ​​the lithium battery through simulation analysis based on the physical impact data and the voltage change data. The determination module is used to release a preset special marker within the potential damage area and identify the position of the special marker by a detection component to determine the actual damage location inside the lithium battery. A construction module is used to construct an electrochemical model based on the potential damage area and the actual damage location, wherein the electrochemical model is used to characterize the electrical properties of the actual damage location in the internal environment of the battery; The generation module is used to use the electrochemical model to calculate the current activity process and abnormal electric field changes at the actual damage location during low-current charge-discharge testing, and to generate an assessment result of the battery overheating runaway risk based on the current activity process and the abnormal electric field changes.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the new energy vehicle power lithium battery performance testing method as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the new energy vehicle power lithium battery performance testing method described in the first aspect above.

[0016] The performance testing method for new energy vehicle power lithium batteries provided in this application can directly collect the core state parameters of the battery under collision scenarios by acquiring physical impact data and voltage change data of the lithium battery during new energy vehicle collision tests, providing a precise and comprehensive basic data source for subsequent damage analysis. Based on the above data, the method can predict potential damage areas through simulation analysis, which can quickly narrow down the scope of damage investigation and reduce the blindness of subsequent detection. By releasing special markers in potential damage areas and identifying their positions through detection components, the method can accurately locate the actual damage location inside the battery, avoiding ineffective disassembly of the battery. By constructing an electrochemical model by combining potential damage areas and actual damage locations, the method can establish the correlation between damage locations and the internal electrical characteristics of the battery, providing theoretical model support for risk assessment. By using the electrochemical model in conjunction with small current charge and discharge tests to calculate the current activity process and abnormal electric field changes at the damage location and generate overheating runaway risk assessment results, the method can achieve quantitative judgment of battery safety risks and improve the scientificity and effectiveness of risk assessment.

[0017] Furthermore, when using an electrochemical model to conduct low-current charge-discharge tests to generate battery overheating runaway risk assessment results, a low-current charge-discharge signal is first input into the electrochemical model to determine the current activity process and abnormal electric field changes at the actual damage location. Then, the corresponding current activity characteristic data and electric field abnormality characteristic data are acquired separately. Finally, these two types of characteristic data are input into a preset risk assessment model to generate the final assessment result. This step improves the targeting of the risk assessment by extracting current and electric field characteristic data at the damage location step by step. By analyzing the characteristic data using a preset risk assessment model, the accuracy and reliability of the battery overheating runaway risk assessment results are further improved. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a method for testing the performance of a new energy vehicle power lithium battery, provided in an embodiment of this application; Figure 2 A flowchart illustrating another method for testing the performance of a new energy vehicle power lithium battery, provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a performance testing system for a new energy vehicle power lithium battery provided in an embodiment of this application. Detailed Implementation

[0020] Current performance testing of new energy vehicle power lithium batteries after collisions mostly adopts offline disassembly testing methods. This method requires removing the battery from the vehicle and destroying the original packaging structure. Not only can it not identify the location of minute internal damage in the battery's original state, but it also deviates from the dynamic electrical environment when the battery is actually working. As a result, the test results cannot accurately reflect the true degree of damage and potential safety risks of the battery after a collision, ultimately causing the problem of insufficient accuracy in assessing the damage and risks of lithium batteries after collisions.

[0021] To address the aforementioned issues, this invention proposes a performance testing method for lithium-ion batteries used in new energy vehicles. Its core lies in combining collision test data simulation with marker localization to achieve accurate damage identification, and relying on an electrochemical model to complete a quantitative risk assessment. Specifically, the method first acquires data on the physical impact and voltage changes during a battery collision, predicts potential damage areas through simulation analysis, and then uses special markers to determine the actual internal damage location. Subsequently, an electrochemical model is constructed by combining potential and actual damage information, and current activity and electric field anomalies at the damage location are calculated using low-current charge-discharge tests to generate an overheating runaway risk assessment result. This method can achieve in-situ damage localization without disassembling the battery, and simultaneously conducts risk assessments matching the actual electrical environment of the battery's operation. It fundamentally solves the problem of insufficient assessment accuracy caused by disassembly and destruction and detachment from actual operating conditions in existing technologies, significantly improving the scientific rigor and reliability of post-collision safety assessments of lithium-ion batteries.

[0022] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] The core of this application is to provide a method for testing the performance of power lithium batteries for new energy vehicles, and a flowchart of one specific implementation method is shown below. Figure 1 As shown, the method includes: S101. Obtain physical impact data and voltage change data of lithium batteries during crash testing of new energy vehicles.

[0024] The physical impact data includes information on the acceleration and force changes of the lithium battery at different locations and in different spatial directions during the collision, reflecting the intensity and distribution of the impact on the battery. The voltage change data includes the time-varying voltage value sequence of the positive and negative electrode output terminals of the lithium battery and information on the extreme points therein, used to characterize the instantaneous impact of the collision on the battery's electrical performance.

[0025] In one specific implementation, multiple physical impact data acquisition units are pre-deployed at key stress points on the lithium battery casing, such as corners and the center of the sides, areas susceptible to impact. Simultaneously, voltage change data acquisition units are installed at the positive and negative output terminals of the lithium battery. When the new energy vehicle enters the crash test process, the test system issues a synchronous trigger command to activate all deployed acquisition units, ensuring that the recording times of the two types of data are perfectly aligned.

[0026] Subsequently, the physical impact data acquisition unit captures in real time the acceleration change sequence along the X, Y, and Z axes of space at each monitoring point during the entire collision process, as well as the magnitude and duration of the force experienced at the corresponding position; the voltage change data acquisition unit simultaneously records the real-time voltage fluctuation sequence between the lithium battery output terminals within the same time interval, and automatically extracts key extreme point information such as voltage peaks and valleys from the sequence, ultimately completing the synchronous acquisition and storage of the two types of data.

[0027] S102. Based on the physical impact data and the voltage change data, the potential damage area of ​​the lithium battery is predicted through simulation analysis.

[0028] S102 specifically includes: S1021. Based on the physical impact data, the impact force distribution borne by the lithium battery at the moment of collision is reconstructed in the battery model corresponding to the lithium battery.

[0029] Among them, the battery model is a three-dimensional simulation model built based on the actual structure, material properties and internal circuit connection relationship of lithium battery, which is used to simulate the mechanical and electrical response of battery during collision.

[0030] S1022. Through simulation analysis, calculate the internal stress response distribution of each component in the battery model under the action of the impact force distribution.

[0031] S1023. Based on the physical tolerance limits of each component of the lithium battery, determine the first set of damage regions from the internal stress response distribution.

[0032] The first set of damaged areas is a set of areas where the battery module may be damaged due to impact stress exceeding its bearing limit, based on mechanical characteristics.

[0033] S1024. Based on the voltage change data and the connection relationship of the internal circuit of the battery model, determine the set of second damage regions.

[0034] The second set of damaged areas is a set of areas inside the battery that may be damaged due to circuit abnormalities, based on electrical characteristics.

[0035] S1025. The overlapping area between the first set of damaged areas and the second set of damaged areas is determined as a potential damage area.

[0036] Among them, the potential damage area is the area of ​​the battery most likely to be damaged, which is determined by combining the results of both mechanical and electrical assessments, thereby improving the accuracy of damage prediction.

[0037] In this embodiment, mechanical damage is traced through mechanical data simulation and electrical abnormal damage is traced through electrical data. Then, potential damage areas are predicted through dual-region intersection verification. The entire process relies on a well-trained battery model to achieve deep linkage between data and simulation, which avoids the one-sidedness of a single mechanical or electrical judgment and improves the accuracy of prediction through step-by-step verification.

[0038] As an example, step S1021 is first executed. The triaxial acceleration and force time-series data collected from each monitoring point during the collision test are preprocessed to remove extreme outliers. Then, the data is input into an impact force reconstruction algorithm based on the finite element method. The algorithm first maps the physical impact data to the spatial mesh of the battery model, and then uses an interpolation algorithm to complete the data in areas where no sampling points are set up. Finally, it reconstructs the three-dimensional impact force distribution of various parts of the lithium battery, such as the casing, cell, and tabs, at the moment of impact. Taking a square lithium battery as an example, its casing has 20 sampling points and the cell surface has 10 sampling points. The algorithm generates a three-dimensional impact force cloud map covering the entire battery with a time resolution of 1ms by fitting the impact force values ​​at different times for each point, clearly showing the impact force in the corner areas. The impact force in the central area is The above example is merely one example of this application. In practical applications, the number of sampling points and the interpolation accuracy of the algorithm can be adjusted according to the battery size and testing requirements. This application does not limit this.

[0039] Secondly, in step S1022, the reconstructed three-dimensional impact force distribution is used as a load condition and precisely applied to the corresponding mesh elements according to the spatial position of each component in the battery model. The internal stress response distribution is calculated using ANSYS simulation software combined with linear elasticity algorithms, based on Hooke's Law. (1) In the formula, This refers to the internal stress of the component, expressed in Pa. This refers to the elastic modulus of the component material, expressed in Pa. The strain of the component is dimensionless.

[0040] During the simulation, the battery model is first finely meshed, with the outer shell mesh size set to 2mm and the cell mesh size set to 1mm. Then, values ​​are assigned to each component according to its material properties, such as the outer shell being made of aluminum alloy. Positive electrode material of battery cell Copper material for pole ears Subsequently, the strain values ​​of each grid element are obtained through iterative calculations, and the stress values ​​are solved by substituting them into the formula. Finally, the stress response distribution map of each component of the full battery is generated, such as the strain of the corner shell region of the aforementioned square lithium battery. The stress is calculated according to equation (1). Strain at the edge of the battery cell The stress is calculated according to equation (1). .

[0041] Subsequently, in step S1023, physical tolerance limits are pre-entered into the simulation system according to component type, establishing a correspondence between component type and tolerance limit, such as the tolerance limit of the shell. Cell withstand limit Electrode bearing limit The simulation system automatically traverses the stress values ​​of each grid cell in the stress response distribution map, compares them point by point with the bearing limit of its component, classifies grid cells exceeding the limit as candidate damage regions, and then merges adjacent candidate regions using a region connectivity algorithm to form the first set of damage regions. In the aforementioned example, the outer shell corner... The area did not exceed the limit, but the stress in a certain edge area of ​​the battery cell reached... If the damage exceeds the cell's tolerance limit, the affected area and the surrounding connected grid cells that exceed the limit will be included in the first damage area set.

[0042] Next, in step S1024, a circuit topology analysis algorithm is used to traverse the circuit connection relationships inside the battery model, clarify the node connection paths of the cells, tabs, busbars, and connecting harnesses, and simultaneously retrieve the voltage change data collected synchronously during the collision test. Using the voltage baseline under normal operating conditions as a reference, the difference algorithm is used to identify the time period and amplitude of voltage drops, rises, and extreme point anomalies. Then, the circuit topology path is traced back to determine the circuit components and related areas corresponding to the abnormal voltage. These areas are classified into a second set of damage areas. For example, in the aforementioned square lithium battery test, a voltage drop of 0.3V and two abnormal extreme points were detected in the positive busbar during the corresponding time period. Based on the circuit topology relationship, the busbar, the three cells connected to it, and the tab area are included in the second set of damage areas.

[0043] In the final step S1052, based on the unified spatial coordinate system of the battery model, the coordinate ranges of the first and second damage region sets are precisely compared. The percentage of overlap area between each region in the two sets and the total area is calculated. Regions with an overlap ratio exceeding 50% are identified as overlapping regions, while regions with areas smaller than 50% are removed. By identifying small overlapping areas to avoid misjudgment, the potential damage area is finally obtained. In the example above, a certain edge area of ​​the battery cell simultaneously meets the correlation conditions of excessive stress and abnormal voltage, and the overlap ratio reaches 80%. This area is then identified as a potential damage area.

[0044] The training process for the battery model is as follows: First, basic data such as the actual size, material parameters, and circuit topology of the lithium battery are collected. The material parameters include the elastic modulus of the positive electrode material, the Poisson's ratio of the negative electrode material, and the yield strength of the shell material. The circuit topology includes the node positions, wire resistance, and connection methods. An initial three-dimensional solid model is constructed using SolidWorks and then imported into simulation software for mesh generation and parameter initialization. Subsequently, five sets of impact forces with different collision intensities were designed for physical tests. Each set of tests was repeated three times to obtain the mechanical response and voltage change data of various parts of the battery. The average value was taken as the standard test data. The experimental data and model simulation results were compared point by point. The least squares method was used to correct the model parameters and adjust parameters such as material stress and strain curve coefficients and circuit contact resistance until the error between the model simulation results and the experimental data was controlled within ±5%, thus completing the model training.

[0045] In another specific implementation, mechanical simulation can be performed using ABAQUS software, and electrical analysis can be performed using PSpice circuit simulation software. During data fusion, the first and second damage regions are assigned weights of 0.6 and 0.4 respectively. The potential damage region is determined by the weighted intersection algorithm, which further improves the adaptability of the prediction.

[0046] This application effectively eliminates misjudgment areas caused by single data judgment through mechanical and electrical data fusion simulation and dual-region intersection determination, improves the accuracy of potential damage area prediction, provides a precise range for subsequent actual damage location, and reduces the blindness of detection.

[0047] S103. Within the potential damage area, a preset special marker is released, and the location of the special marker is identified by a detection component to determine the actual damage location inside the lithium battery.

[0048] S103 specifically includes: S1031. The physical field change signal caused by the special marker is obtained by the detection component deployed inside the lithium battery.

[0049] S1032. Process the physical field change signal to generate a concentration distribution map of the special marker in the internal space of the lithium battery.

[0050] S1032 specifically includes: Based on the spatial position of each detection component and the intensity of the physical field change signal, the preliminary distribution position of the special marker inside the lithium battery is calculated using a signal tracing algorithm. According to the battery model of the lithium battery, the distribution positions corresponding to the areas inside the lithium battery where the special marker cannot exist due to physical isolation are removed from the preliminary distribution positions to obtain the corrected distribution positions. Based on the corrected distribution positions, a concentration distribution map of the special marker inside the lithium battery is generated.

[0051] S1033. Based on the distribution characteristics of signal intensity in the concentration distribution spectrum, determine the actual damage location inside the lithium battery.

[0052] In this embodiment, special markers are first released into the predicted potential damage areas. These markers are then used to mark the areas where they tend to accumulate at the gaps in the damage. Pre-deployed detection components simultaneously capture the physical field changes caused by the markers. The collected signals are then processed through source tracing calculations and model corrections to generate a concentration distribution map that visually represents the spatial aggregation of the markers. Finally, the signal intensity distribution pattern is analyzed based on the map to pinpoint the actual damage location. The entire process focuses on detecting the potential damage areas without involving the entire battery area, effectively avoiding redundancy caused by an excessively large detection range, and balancing positioning accuracy and detection efficiency.

[0053] As an example, in step S1031, a preset magnetic special marker is first released into the potentially damaged area of ​​the aforementioned square lithium battery, i.e., the edge area of ​​the cell. In this example, the special marker is a ferrite particle with a particle size of 5μm. The marker will accumulate at the tiny gaps in the damaged area. Eight magnetic sensors deployed around this area are simultaneously activated to collect the magnetic field strength change signal caused by the marker. The sampling frequency is set to 100Hz and the data is collected continuously for 10s to obtain the time-series data of the magnetic field strength of each sensor. The above example is only one example of this application. In practical applications, fluorescent, acoustic, and other types of markers can also be selected according to the type of damage. This application does not limit this.

[0054] Next, step S1032 is executed, which requires tracing, correcting, and generating a spectrum of the collected physical field change signal to obtain the concentration distribution spectrum of the special marker inside the battery. This process includes the following steps: The first step involves calculating the preliminary distribution location of the special markers inside the lithium battery based on the spatial positions of each detection component and the intensity of the physical field change signal using a signal tracing algorithm. The core of the signal tracing algorithm is based on a magnetic field strength attenuation model, and the model formula is as follows: (2) In the formula, B represents the magnetic field strength collected by a magnetic sensor, in mT. The initial magnetic field strength at the marker is expressed in mT, and k is the magnetic field attenuation coefficient of the internal medium of the battery; in this example, the value is taken as... d is the straight-line distance between the sensor and the marker, in mm.

[0055] Suppose a sensor collects B = 2.5 mT, and it is known that... =5mT, substituting the value into equation (2) and transforming it, we get: By combining the spatial coordinates of eight sensors preset in the battery model coordinate system, the preliminary distribution positions of the markers are calculated using triangulation, forming multiple candidate coordinate points. The above example is only one example of this application; in practical applications, the positioning accuracy can be adjusted according to the number of sensors, and this application does not limit this.

[0056] The second step involves calling up a lithium battery model and, based on the precise coordinate range of the physical barriers such as the cell separator and casing within the model, verifying each candidate point in the initial distribution location and eliminating coordinate points falling within the barrier areas. Since markers cannot penetrate the separator and other physical structures, these candidate points are deemed unreasonable and are eliminated, resulting in a more realistic corrected distribution location.

[0057] The third step involves processing the corrected coordinate points and assigning marker concentration values ​​based on the magnetic field strength at each location; higher magnetic field strength results in higher concentration values. Subsequently, an interpolation algorithm is used to complete the uncovered coordinate data within the region, generating a two-dimensional concentration distribution map. The map uses different colors to distinguish between high and low concentrations, with red areas corresponding to the regions with the highest marker concentrations.

[0058] Finally, step S1033 is executed to analyze the concentration distribution spectrum. The red area with the highest signal intensity and concentrated aggregation range is identified as the actual damage location. This area corresponds to the tiny crack at the edge of the battery cell, which is consistent with the aggregation characteristics of the marker.

[0059] The training process of the signal source tracing algorithm is as follows: Select a sample battery of the same type as the target lithium battery, artificially create damage of different sizes at different locations of the sample battery, release special markers, and collect a large amount of signal intensity and actual marker location data through a detection component to construct a training dataset; Using sensor coordinates and signal strength as inputs and the actual position of the marker as output, a gradient descent-based prediction model is trained. Parameters such as the attenuation coefficient k are iteratively optimized until the model prediction error is controlled within ±3mm, thus completing the algorithm training.

[0060] In another specific implementation, fluorescent special markers can be selected, and optical detection components can be used to collect light field change signals. The signal tracing algorithm is replaced with a light intensity attenuation model, and the concentration distribution spectrum is generated by visualizing the fluorescence intensity. At the same time, the light transmission characteristics of the battery model are combined to eliminate the obstructed areas, which can adapt to the detection needs of different types of markers.

[0061] This application achieves precise location of actual damage within lithium batteries by analyzing marker aggregation characteristics and signal source tracing, without disassembling the battery, effectively avoiding blind testing and improving the accuracy and targeting of damage location.

[0062] S104. Based on the potential damage area and the actual damage location, an electrochemical model is constructed, which is used to characterize the electrical properties of the actual damage location in the internal environment of the battery.

[0063] S104 specifically includes: S1041. Within the spatial range of the potential damage area, delineate a key area that includes the actual damage location; The critical area is a local spatial range defined around the actual damage location within the potential damage area. It is used to focus on the model building object, reduce interference from irrelevant areas, and improve the efficiency and accuracy of model building.

[0064] S1042. Based on the structural characteristics of the actual damage location, define the boundary of the actual damage location within the critical area; S1043. Within the critical region, establish evolution rules to describe the flow process of current and the changes in the boundary of the actual damage location with the charging and discharging state. Among them, the evolution rules are a set of rules that describe the current flow patterns in key areas and the morphological or characteristic changes of the actual damage location boundary with the charging and discharging state, providing the core computational logic for the electrochemical model.

[0065] S1044. Based on the evolution rules, construct an electrochemical model.

[0066] Among them, the electrochemical model is a simulation model built based on the physical structure and evolution rules of key regions. It can accurately simulate the electrical behavior of actual damage locations during battery charging and discharging, and reflect the impact of damage on battery electrical characteristics.

[0067] In this embodiment, the model is based on the potential damage area and the actual damage location. By delineating key areas to lock the modeling range, defining the boundary in combination with the damage structure characteristics, and then establishing evolution rules that fit the actual working conditions, an electrochemical model that can accurately characterize the electrical properties of the damage location is finally constructed. The model focuses on the core area throughout the process to avoid redundant calculations.

[0068] As an example, in step S1041, based on the potential damage area and actual damage location of the aforementioned square lithium battery (i.e., the cell edge area and the micro-cracks at the cell edge), a region expansion algorithm is used to delineate the critical region. First, the coordinates of the center point of the actual damage location are determined. Using this center point as a reference, the region is uniformly expanded by 5 mm in each of the three spatial directions, ultimately forming a cubic critical region with a side length of 10 mm. This not only completely covers the damage location and the surrounding potentially affected area but also effectively excludes irrelevant parts within the potential damage area, controlling the model's computational scale. The above example is merely one example of this application. In practical applications, the expansion range can be adjusted according to the damage size and modeling accuracy requirements; this application does not limit this.

[0069] In step S1042, a complete structural image of the actual damage location is first obtained through high-precision scanning. Then, an edge detection algorithm is used to process the image and extract structural features such as the morphology, size, and spatial orientation of the damage. The damage is detected as a linear crack, 0.1 mm wide and 3 mm long, extending obliquely along the edge of the battery cell. Combining the spatial coordinate system of the key area, the damage boundary is transformed into a continuous set of broken line coordinates. The relative positions of the crack and components such as the positive electrode, negative electrode, and separator of the battery cell are defined point by point, clarifying the boundary between the damaged area and the normal area, laying the foundation for subsequent simulation calculations of the current flow path.

[0070] Next, through step S1043, two types of core evolution rules are established based on the electrical characteristics and damage evolution patterns of the key regions: First, the current flow rules are derived based on Ohm's law, using the formula for the current density distribution in the key region. The formula is as follows: (3) In the formula, J is the current density, and the unit is amperes per square meter; The conductivity of the battery material in the critical area is expressed in Siemens per meter. The electric field strength is expressed in volts per meter. Considering the difference in material properties between the damaged and normal areas, the conductivity at the crack is set to 10 siemens per meter, while the conductivity of the normal cell material is set to 1000 siemens per meter. At the same time, the attenuation law and detour path of the current when it comes into contact with the damaged boundary are defined, that is, after the current encounters the crack, it splits along the edge to both sides, and the current density decreases proportionally to the material conductivity during the splitting process.

[0071] Secondly, the damage boundary change rule is established. Combining the dynamic changes in internal temperature and pressure of the battery during charging and discharging, a threshold for the expansion of the boundary with the number of charge-discharge cycles is set. When the cumulative number of charge-discharge cycles reaches 50, the crack width expands to a maximum of 0.12 mm. After exceeding this threshold, the boundary no longer expands, thus simulating the natural evolution trend of the damage location under actual working conditions.

[0072] Finally, in step S1044, the defined critical region, the defined damage boundary coordinates, and the two types of evolution rules are imported into the COMSOL Multiphysics simulation software. Based on the material properties of each component within the critical region, including conductivity and dielectric constant, the evolution rules are embedded into the model's computational logic to construct a three-dimensional electrochemical model. The model can accurately simulate current flow, electric field distribution, and the impact of damage boundary changes on electrical properties during charging and discharging within the critical region, fully reproducing the electrical behavior of the damaged location within the battery's internal environment.

[0073] The training process for the electrochemical model is as follows: Select cells of the same type as the target lithium battery, artificially create cracks of the same size, and conduct 10 sets of low-current charge-discharge tests with different charge-discharge cycles.

[0074] In each group of experiments, the current density, electric field strength and boundary change data at the damage location were collected in real time, and the average value of each group of experiments was taken as the standard reference data.

[0075] The standard reference data and model simulation results were compared point by point. The model parameters were optimized using a genetic algorithm, with a focus on adjusting key parameters such as the material conductivity coefficient and boundary expansion threshold, until the deviation between the model simulation results and the experimental reference data was within a reasonable range, ensuring that the model had sufficient accuracy and completing the training.

[0076] In another specific implementation, an electrochemical model can be constructed using ANSYS Fluent software. The current flow rules can be optimized and improved by combining them with the Nernst equation. At the same time, the temperature field coupling effect is introduced, so that the model can synchronously reflect the influence of temperature changes on the electrical properties and boundary morphology of the damage site, and adapt to more complex actual working scenarios.

[0077] This application constructs an electrochemical model by focusing on key areas, accurately correlates the damage location structure with electrical properties, and provides reliable simulation support for subsequent risk assessment, balancing the accuracy of the model with computational efficiency.

[0078] S105. Using the electrochemical model, during low-current charge-discharge testing, calculate the current activity process and abnormal electric field changes at the actual damage location, and generate an assessment result of the battery overheating runaway risk based on the current activity process and the abnormal electric field changes.

[0079] S105 specifically includes: S1051. Input a small current charging and discharging signal into the electrochemical model to determine the current activity process and abnormal electric field changes at the actual damage location.

[0080] S1052. Based on the current activity process, obtain current activity characteristic data, and based on the abnormal changes in the electric field, obtain electric field abnormality characteristic data.

[0081] Among them, current activity characteristic data is key data extracted from the current activity process that reflects abnormal current changes at the damage location, including the frequency and amplitude of abnormal fluctuation points, and is used to quantify the degree of abnormality in current activity. Electric field anomaly characteristic data is core data extracted from abnormal electric field changes that characterizes the abnormal electric field distribution in the damage area, including the volume and average intensity of the abnormal region, and is used to reflect the range and intensity of the electric field anomaly.

[0082] S1052 specifically includes: During the current activity, the signal of small-current charging and discharging is captured, and the current flowing through the boundary of the actual damage location within the test cycle is recorded. The rate of change of the current over time is calculated, and abnormal fluctuation points exceeding a preset fluctuation range are identified. The frequency and amplitude of the abnormal fluctuation points are statistically analyzed to generate current activity characteristic data. From the abnormal changes in the electric field, the electric field intensity distribution data within a preset radius centered on the actual damage location is extracted. The rate of change of the electric field intensity distribution data in space is calculated, and abnormal regions with a rate of change exceeding a preset threshold are identified. The volume and average intensity of the abnormal regions are statistically analyzed to generate electric field anomaly characteristic data.

[0083] S1053. Input the current activity characteristic data and the electric field anomaly characteristic data into the preset risk assessment model to generate assessment results.

[0084] Among them, the risk assessment model is a prediction model trained based on historical damage data and corresponding overheating risk results, which is used to integrate the two types of characteristic data to output accurate risk assessment conclusions.

[0085] In the embodiments of this application, such as Figure 2 As shown, the system relies on the established electrochemical model to simulate low-current charging and discharging conditions. First, it acquires dynamic data on current activity and electric field anomalies at the damage location. Then, it extracts two types of characteristic data to quantify the degree of anomalies. Finally, it integrates and analyzes the data through a risk assessment model to output the battery overheating runaway risk level, thus achieving a closed loop from operating condition simulation to risk assessment.

[0086] As an example, in step S1051, for the aforementioned electrochemical model of a square lithium battery, a small current charge / discharge signal is input. The charging current is set to 0.2C, the charging cutoff voltage to 3.65 volts, the discharging current to 0.2C, and the discharging cutoff voltage to 2.5 volts. The test period is 1 hour, and the sampling frequency is 10 Hz. Based on the built-in current flow rules and damage boundary evolution rules, the battery model fully simulates the dynamic behavior of ion migration and charge transport throughout the entire charge / discharge process, accurately outputting the current activity process and abnormal electric field change data at the actual damage location. The current activity process is reflected as a continuous current time-series fluctuation curve, clearly showing the changing trend of current at different charge / discharge stages; the abnormal electric field change is reflected as a dynamic map of the spatial distribution of electric field intensity in the damaged area and its surroundings, intuitively presenting the spatial variation law of electric field intensity over time. The above example is only one example of this application. In practical applications, the small current rate, charge / discharge cutoff voltage, and test period can be adjusted according to test requirements, and this application does not limit these adjustments.

[0087] Secondly, through step S1052, the feature data extraction and generation are completed in two steps: The first step is to use the model's boundary monitoring module to accurately capture the time-series data of the current flowing through the boundary of the damage location within the test cycle, covering the constant current and constant voltage processes during the charging phase and the constant current process during the discharging phase. Then, the rate of change of current over time is calculated using the following formula: (4) In the formula, The rate of change of current is expressed in amperes per second. The difference in current between two adjacent sampling times is expressed in amperes. The sampling time interval is in seconds.

[0088] Based on extensive experimental data, a preset fluctuation range was set at ±0.05 amperes per second. Values ​​exceeding this range were considered abnormal fluctuation points. A specific sampling period within the test cycle was selected for calculation; the current at two adjacent sampling moments within this period were 0.32 amperes and 0.332 amperes, respectively. The sampling time interval is 0.012 amperes. The value is 0.2 seconds. Substituting this into equation (4), we get... =0.06 amperes per second. This value exceeds the upper limit of the preset fluctuation range and is judged as an abnormal fluctuation point.

[0089] Using this method, the rate of change of current at all sampling moments within the test cycle is calculated point by point. After identifying all abnormal fluctuation points, the identification results are statistically analyzed. It is found that the frequency of abnormal fluctuation points within the test cycle is 12 times per hour, and the average rate of change of all abnormal fluctuation points, i.e., the average amplitude, is 0.07 amperes per second. These statistical data are integrated to generate current activity characteristic data.

[0090] The second step involves defining a spherical region with a radius of 3 mm, using the center point of the actual damage location as a reference. The spatial data extraction module of the model is then used to extract the electric field intensity distribution data for all spatial nodes within this region, covering the core location of the damage area and the surrounding affected areas. Subsequently, the rate of change of the electric field intensity in space is calculated using the following formula: (5) In the formula, The rate of change of electric field intensity in space is expressed in volts per cubic meter. This represents the difference in electric field strength between two adjacent spatial nodes, expressed in volts per meter. Spacing between nodes, in meters.

[0091] Based on extensive experimental data, a preset threshold of 50 volts per cubic meter was set. Areas with spatial change rates exceeding this threshold were identified as areas of abnormal electric fields. A specific spatial region within the defined area was selected for calculation; the electric field strengths at two adjacent nodes in this region were 2.1 volts per meter and 2.2 volts per meter, respectively. The voltage is 0.1 volts per meter, and the spatial node spacing is... The value is 0.001 meters. Substituting this into the formula, we get... =100 volts per cubic meter. This value exceeds the preset threshold, and the area is designated as an abnormal electric field area.

[0092] Following this method, all spatial nodes within the defined range are traversed. After identifying all abnormal electric field regions, the identification results are statistically analyzed. The total volume of abnormal electric field regions within the test period is 0.02 cubic centimeters, and the average electric field strength of all abnormal regions is 200 volts per meter. These statistical data are then integrated to generate abnormal electric field characteristic data.

[0093] Finally, in step S1053, the generated current activity characteristic data and electric field anomaly characteristic data are input into a preset risk assessment model. The model first normalizes the two types of characteristic data to eliminate the influence of different data dimensions. Then, based on the built-in algorithm logic, it analyzes the correlation between abnormal fluctuation frequency, amplitude, and abnormal region volume and average intensity. Combining the correspondence between characteristic data and actual overheating risk in historical sample data, it outputs the corresponding overheating runaway risk level. The risk level is divided into three categories: low risk, medium risk, and high risk. In the aforementioned example, the model comprehensively considers the anomaly degree of the two types of characteristic data and ultimately outputs a medium-risk assessment result, indicating that the damaged location has a certain overheating risk during subsequent charging and discharging processes, requiring further monitoring and handling.

[0094] The training process for the risk assessment model is as follows: First, a large amount of test data of the same type of lithium battery was collected, including damage data of different sizes and locations, current activity characteristics data and electric field abnormality characteristics data in the corresponding low current charge and discharge test, as well as overheating runaway after actual charge and discharge cycles. Sample datasets containing low, medium and high risk levels were constructed to ensure that the datasets cover different degrees of anomalies and risk types.

[0095] Then, using current activity characteristic data and electric field anomaly characteristic data as the input layer and the actual overheating runaway risk level as the output layer, a risk assessment model based on the random forest algorithm was trained. During training, the classification accuracy of the model was continuously optimized by adjusting parameters such as the number of decision trees, the depth of decision trees, and the feature sampling ratio. At the same time, cross-validation was used to verify the generalization ability of the model until the model's accuracy in judging the risk level of the validation set samples reached the preset requirements, thus completing the model training.

[0096] In another specific implementation, the risk assessment model can be constructed using a neural network algorithm. Temperature change data during battery charging and discharging can be added to the input layer as an auxiliary input feature, allowing the model to comprehensively assess risk based on both electrical and thermal characteristics. This further optimizes the accuracy of risk assessment and makes it suitable for application scenarios with higher risk assessment requirements.

[0097] This application achieves precise quantification of the risk of overheating and runaway at the site of battery damage through electrochemical model simulation and feature data-driven risk assessment, providing a reliable basis for battery safety management.

[0098] Figure 3 This is a schematic diagram of a specific implementation of a performance testing system for new energy vehicle power lithium batteries provided in this application, with reference to... Figure 3 The system may include: The acquisition module 31 is used to acquire physical impact data and voltage change data of lithium batteries during the collision test of new energy vehicles; Prediction module 32 is used to predict the potential damage area of ​​the lithium battery through simulation analysis based on the physical impact data and the voltage change data. The determination module 33 is used to release a preset special marker within the potential damage area and identify the position of the special marker by the detection component in order to determine the actual damage location inside the lithium battery. The construction module 34 is used to construct an electrochemical model based on the potential damage area and the actual damage location, wherein the electrochemical model is used to characterize the electrical properties of the actual damage location in the internal environment of the battery. The generation module 35 is used to use the electrochemical model to calculate the current activity process and abnormal electric field changes at the actual damage location during low-current charge-discharge testing, and to generate an assessment result of the battery overheating runaway risk based on the current activity process and the abnormal electric field changes.

[0099] The new energy vehicle power lithium battery performance testing system of this application embodiment is used to implement the aforementioned new energy vehicle power lithium battery performance testing method. Therefore, the specific implementation of the new energy vehicle power lithium battery performance testing system can be found in the embodiment section of the new energy vehicle power lithium battery performance testing method above. The specific implementation can be referred to the description of the corresponding embodiment, and will not be repeated here.

[0100] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of any of the above-described new energy vehicle power lithium battery performance testing methods.

[0101] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described methods for testing the performance of new energy vehicle power lithium batteries.

[0102] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0103] The embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the embodiments of the new energy vehicle power lithium battery performance testing method.

[0104] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0105] The above provides a detailed description of the performance testing method and system for new energy vehicle power lithium batteries provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for testing the performance of power lithium batteries for new energy vehicles, characterized in that, include: To obtain physical impact data and voltage change data of lithium batteries during crash tests of new energy vehicles; Based on the physical impact data and the voltage change data, the potential damage area of ​​the lithium battery is predicted through simulation analysis. Within the potential damage area, a pre-set special marker is released, and the location of the special marker is identified by a detection component to determine the actual damage location inside the lithium battery. Based on the potential damage area and the actual damage location, an electrochemical model is constructed, which is used to characterize the electrical properties of the actual damage location in the internal environment of the battery. Using the electrochemical model, during low-current charge-discharge tests, the current activity process and abnormal electric field changes at the actual damage location are calculated, and an assessment result of the battery overheating runaway risk is generated based on the current activity process and the abnormal electric field changes.

2. The method according to claim 1, characterized in that, The electrochemical model is used to calculate the current activity process and abnormal electric field changes at the actual damage location during low-current charge-discharge testing. Based on the current activity process and abnormal electric field changes, an assessment result of the battery overheating runaway risk is generated, including: A small current charging and discharging signal is input into the electrochemical model to determine the current activity process and abnormal electric field changes at the actual damage site; Based on the current activity process, current activity characteristic data are obtained, and based on the abnormal changes in the electric field, abnormal electric field characteristic data are obtained. The current activity characteristic data and the electric field anomaly characteristic data are input into a preset risk assessment model to generate assessment results.

3. The method according to claim 2, characterized in that, The step of acquiring current activity characteristic data based on the current activity process and acquiring electric field anomaly characteristic data based on the electric field anomaly changes includes: During the current activity process, the signal of small current charging and discharging is captured, and the current flowing through the boundary of the actual damage location within the test cycle is measured; the rate of change of the current over time is calculated, and abnormal fluctuation points exceeding the preset fluctuation range in the rate of change are identified; the frequency and amplitude of the abnormal fluctuation points are statistically analyzed to generate current activity characteristic data. From the abnormal changes in the electric field, extract the electric field intensity distribution data within a preset radius centered on the actual damage location; calculate the rate of change of the electric field intensity distribution data in space, and identify abnormal regions where the rate of change exceeds a preset threshold; statistically analyze the volume and average intensity of the abnormal regions to generate electric field anomaly feature data.

4. The method according to claim 1, characterized in that, Based on the physical impact data and the voltage change data, the potential damage area of ​​the lithium battery is predicted through simulation analysis, including: Based on the physical impact data, the impact force distribution borne by the lithium battery at the moment of collision is reconstructed in the battery model corresponding to the lithium battery. Through simulation analysis, the internal stress response distribution of each component in the battery model under the action of the impact force distribution is calculated; Based on the physical tolerance limits of each component of the lithium battery, a first set of damage regions is determined from the internal stress response distribution; Based on the voltage change data and the connection relationship of the internal circuit of the battery model, the set of second damage regions is determined. The overlapping area between the first set of damaged areas and the second set of damaged areas is identified as a potential damage area.

5. The method according to claim 1, characterized in that, The step of identifying the location of the special marker by a detection component to determine the actual location of damage inside the lithium battery includes: By using detection components deployed inside the lithium battery, the physical field change signal triggered by the special marker is acquired; The physical field change signal is processed to generate a concentration distribution map of the special marker in the internal space of the lithium battery; Based on the distribution characteristics of signal intensity in the concentration distribution spectrum, the actual damage location inside the lithium battery is determined.

6. The method according to claim 5, characterized in that, The process of processing the physical field change signal to generate a concentration distribution map of the special marker within the lithium battery includes: Based on the spatial position of each detection component and the intensity of the physical field change signal, the preliminary distribution position of the special marker inside the lithium battery is calculated using a signal tracing algorithm. Based on the battery model of the lithium battery, the distribution positions corresponding to the areas inside the lithium battery where the special marker cannot exist due to physical isolation are eliminated from the preliminary distribution positions to obtain the corrected distribution positions; Based on the corrected distribution location, a concentration distribution map of the special marker is generated in the internal space of the lithium battery.

7. The method according to claim 1, characterized in that, The construction of an electrochemical model based on the potential damage area and the actual damage location includes: Within the spatial range of the potential damage area, a key area containing the actual damage location is delineated; Based on the structural characteristics of the actual damage location, the boundary of the actual damage location is defined within the critical area; Within the critical region, evolution rules are established to describe the flow process of current and the evolution of the boundary of the actual damage location as it changes with the charging and discharging state. An electrochemical model is constructed based on the aforementioned evolutionary rules.

8. A performance testing system for power lithium batteries in new energy vehicles, characterized in that, include: The acquisition module is used to acquire physical impact data and voltage change data of lithium batteries during crash tests of new energy vehicles; The prediction module is used to predict the potential damage area of ​​the lithium battery through simulation analysis based on the physical impact data and the voltage change data. The determination module is used to release a preset special marker within the potential damage area and identify the position of the special marker by a detection component to determine the actual damage location inside the lithium battery. A construction module is used to construct an electrochemical model based on the potential damage area and the actual damage location, wherein the electrochemical model is used to characterize the electrical properties of the actual damage location in the internal environment of the battery; The generation module is used to use the electrochemical model to calculate the current activity process and abnormal electric field changes at the actual damage location during low-current charge-discharge testing, and to generate an assessment result of the battery overheating runaway risk based on the current activity process and the abnormal electric field changes.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the new energy vehicle power lithium battery performance testing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the implementation of the performance testing method for new energy vehicle power lithium batteries as described in any one of claims 1 to 7.

Images