Graphite negative electrode lithium extraction evaluation method and system

Abstract

The application discloses a graphite negative electrode lithium precipitation evaluation method and system, and relates to the technical field of lithium ion battery evaluation, and comprises the following steps: collecting voltage-time data of first stage relaxation of a graphite negative electrode half cell to be tested to obtain first relaxation voltage-time data, and collecting voltage-time data of second stage relaxation of the graphite negative electrode half cell to be tested to obtain second relaxation voltage-time data; performing differential processing on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain a first relaxation differential curve and a second relaxation differential curve; extracting lithium precipitation characteristic parameters from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data and the second relaxation voltage-time data; and determining whether the graphite negative electrode precipitates lithium and obtaining a lithium precipitation amount based on the extracted lithium precipitation characteristic parameters. The method overcomes the deficiency of the prior art in lithium precipitation quantitative evaluation.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion battery evaluation technology, and in particular to a method for evaluating lithium deposition on graphite anodes. Background Technology

[0002] Lithium-ion batteries, due to their high energy density and long cycle life, have been widely used in new energy vehicles, energy storage, and portable electronic devices. Graphite, as the mainstream anode material, possesses characteristics such as stable layered structure, high specific capacity, and low lithium intercalation potential. However, in practical applications, influenced by factors such as charging rate, ambient temperature, electrode fabrication process, and electrolyte performance, lithium ion deposition (lithium plating) easily occurs on the surface of graphite anodes, forming metallic lithium. Lithium plating not only consumes the active lithium source in the battery system, leading to rapid capacity decay and shortened cycle life, but also induces lithium dendrite growth, potentially piercing the separator and causing a short circuit between the positive and negative electrodes, resulting in serious safety accidents. Therefore, accurate and efficient evaluation of lithium plating behavior in graphite anodes is of great significance for ensuring battery safety and improving product quality.

[0003] Currently, methods for evaluating lithium plating in graphite anodes are mainly divided into two categories: offline detection and online detection. Offline detection methods include scanning electron microscopy, inductively coupled plasma atomic emission spectrometry (ICP-AES), and pixel image analysis. These methods require battery disassembly, cannot achieve in-situ detection, and are complex, time-consuming, and labor-intensive. Among online detection methods, voltage curve analysis is the most widely used technology in industry and laboratories due to its ease of operation and low cost. Existing voltage curve-based lithium plating evaluation methods typically monitor the voltage change curve during the post-discharge resting phase of charge-discharge cycles, perform differential processing, and determine whether lithium plating has occurred based on the presence or absence of characteristic peaks. However, this method can only achieve a qualitative judgment of lithium plating and cannot quantitatively assess the degree of lithium plating. In practical applications such as battery R&D, process optimization, and safety classification, simply knowing whether lithium plating has occurred is far from sufficient. Engineers need to accurately know the specific amount of lithium plating to assess the severity of the lithium plating hazard, determine whether the current safety status of the battery is controllable, and provide quantitative basis and direction for subsequent process improvements. The lack of quantitative capabilities is the most prominent drawback of existing lithium plating evaluation methods based on voltage curves.

[0004] In summary, how to provide a quantitative evaluation method for lithium plating in graphite anodes has become a pressing technical problem to be solved in this field. Summary of the Invention

[0005] This invention provides a method and system for evaluating lithium plating on graphite anodes. The technical problem it aims to solve is: how to provide an evaluation method that can overcome the shortcomings of existing methods in terms of lithium plating quantification assessment, early identification sensitivity, anti-interference ability, and adaptability to multiple temperature scenarios.

[0006] In a first aspect, embodiments of the present invention provide a method for evaluating lithium plating on graphite anodes, comprising: The voltage-time data of the first stage relaxation of the graphite anode half-cell under test is collected to obtain the first relaxation voltage-time data, and the voltage-time data of the second stage relaxation of the graphite anode half-cell under test is collected to obtain the second relaxation voltage-time data. The graphite anode half-cell under test is placed in a preset target test environment and the test is performed in a cycle of constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging. Differential processing is performed on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve; Lithium plating characteristic parameters are extracted from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data. Based on the extracted lithium plating characteristic parameters, it is determined whether the graphite anode has undergone lithium plating and the amount of lithium plating is obtained.

[0007] Optionally, the lithium plating characteristic parameters include a first characteristic parameter, a second characteristic parameter, a third characteristic parameter, and a fourth characteristic parameter; The first characteristic parameter includes the occurrence time and amplitude of the characteristic peak of the first relaxation differential curve; The second characteristic parameter includes the occurrence time and amplitude of the characteristic peak of the second relaxation differential curve; The third characteristic parameter includes the first relaxation voltage decay rate calculated based on the first relaxation voltage-time data. The fourth characteristic parameter includes the second relaxation voltage stability difference calculated based on the second relaxation voltage-time data.

[0008] Optionally, the characteristic peaks of the first relaxation differential curve and the second relaxation differential curve are determined based on a preset judgment criterion. The judgment criterion is: a bulge appears on the relaxation differential curve that rises first and then falls, and the amplitude of the bulge is greater than 3 times the amplitude of the background noise. The relaxation differential curve includes the first relaxation differential curve and the second relaxation differential curve.

[0009] Optionally, before acquiring the voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data, the method further includes: A coin cell was assembled using the graphite material to be tested as the working electrode and a lithium metal sheet as the counter electrode and reference electrode. The assembled coin cell was subjected to charge-discharge activation treatment to obtain the graphite negative electrode half-cell under test.

[0010] Optionally, the constant current charging rate is 0.5C to 5C, charging to a voltage of 2.0V; the first stage relaxation time after charging is 0.5 hours to 2 hours, and the relaxation process is a current-free resting state; the constant current discharging rate is a gradient increasing rate starting from 0.2C and increasing by 0.2C to 0.5C in each cycle, discharging to a voltage of 0.005V; the second stage relaxation time after discharging is 2 hours to 3 hours, and the relaxation process is a current-free resting state.

[0011] Optionally, before performing differentiation processing on the first relaxation voltage-time data and the second relaxation voltage-time data, the method further includes: The first relaxation voltage-time data and the second relaxation voltage-time data are subjected to noise reduction processing.

[0012] Optionally, determining whether the graphite anode has undergone lithium plating and obtaining the amount of lithium plating based on the extracted lithium plating characteristic parameters includes: The extracted lithium plating characteristic parameters are input into the lithium plating evaluation model, which then outputs the lithium plating grade and the amount of lithium plating.

[0013] Optionally, the lithium plating evaluation model is constructed using a machine learning algorithm, and the lithium plating level includes four levels: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating. The quantitative range of the amount of lithium plating is 0 to 5 milligrams per square centimeter.

[0014] Optionally, the temperature range of the target test environment is from -20 degrees Celsius to 40 degrees Celsius, and the temperature control accuracy is ±0.5 degrees Celsius.

[0015] Optionally, after obtaining the amount of lithium plating, the method further includes: The obtained amount of lithium plating is compared with the actual amount of lithium plating in the graphite anode half-cell under test determined by inductively coupled plasma atomic emission spectrometry to obtain the evaluation error. If the evaluation error is greater than 5%, then after adjusting at least one of the first stage relaxation time after charging, the second stage relaxation time after discharging, and the voltage acquisition interval, proceed to the step of acquiring the voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data, so as to re-evaluate the graphite anode half-cell under test.

[0016] Secondly, embodiments of the present invention provide a graphite anode lithium plating evaluation system, comprising: The environmental control module is used to control the graphite anode half-cell under test to be in a preset target test environment. The electrochemical testing module is used to perform cyclic testing with constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging as one cycle. It collects voltage-time data of the first stage relaxation to obtain first relaxation voltage-time data, and collects voltage-time data of the second stage relaxation to obtain second relaxation voltage-time data. The data processing module is used to perform differential processing on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve; extract lithium plating characteristic parameters from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data; and determine whether the graphite anode has lithium plating and obtain the amount of lithium plating based on the extracted lithium plating characteristic parameters.

[0017] This invention provides a method and system for evaluating lithium plating on graphite anodes. The method includes: acquiring voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain first relaxation voltage-time data; and acquiring voltage-time data of the second stage relaxation of the graphite anode half-cell under test to obtain second relaxation voltage-time data. The graphite anode half-cell under test is placed in a preset target test environment, and a cyclic test is performed, consisting of constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging. The first relaxation voltage-time data and the second relaxation voltage-time data are differentiated to obtain a first relaxation differential curve and a second relaxation differential curve. Lithium plating characteristic parameters are extracted from the first relaxation differential curve, the second relaxation differential curve, and the first and second relaxation voltage-time data. Based on the extracted lithium plating characteristic parameters, it is determined whether lithium plating has occurred on the graphite anode and the amount of lithium plating is obtained. This invention employs a two-stage relaxation voltage acquisition method, acquiring first and second relaxation voltage-time data respectively. This simultaneously captures early lithium plating signals caused by lithium deposition during charging and residual lithium plating characteristics reflected by the stripping of residual metallic lithium after discharge, significantly improving the early detection sensitivity for slight lithium plating. By differentiating the two sets of relaxation voltage-time data, first and second relaxation differential curves are obtained. Combined with the original voltage-time data, lithium plating characteristic parameters are extracted to form a multi-dimensional quantitative criterion covering voltage change rate anomalies, decay kinetics, and polarization recovery. Based on these characteristic parameters, it is possible to directly determine whether lithium plating has occurred and obtain the specific amount of lithium plating, achieving a leap from qualitative judgment to quantitative assessment and solving the problem that existing technologies cannot quantify the degree of lithium plating. Furthermore, the combination of two-stage relaxation data acquisition and differential processing effectively highlights lithium plating-related characteristic signals, suppresses background interference, and enhances the method's anti-interference capability. Furthermore, this method places the battery under test in a target test environment for cyclic testing. This environment can be set with different temperatures according to actual needs, thus adapting to various application scenarios such as room temperature and low temperature. In summary, this method effectively overcomes the shortcomings of existing technologies in terms of lithium plating quantification assessment, early identification sensitivity, anti-interference ability, and adaptability to multiple temperature scenarios. Attached Figure Description

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

[0019] Figure 1A schematic flowchart of the lithium plating evaluation method for graphite anodes provided in an embodiment of the present invention; Figure 2 This is a diagram showing the morphological characteristics of the graphite electrode sheets after disassembly of the button cell in Embodiment 1 of the present invention. Figure 3 This is a comparison of the characteristic peaks of the first and second relaxation differential curves of samples with different degrees of lithium plating in Example 1 of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0022] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0023] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if [described condition or event] is detected" may be interpreted, depending on the context, as "once determined," "in response to determination," "once [described condition or event] is detected," or "in response to detection of [described condition or event]."

[0025] Please see Figure 1 This invention provides a method for evaluating lithium plating on graphite anodes, which includes the following steps: S1, collect the voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data, and collect the voltage-time data of the second stage relaxation of the graphite anode half-cell under test to obtain the second relaxation voltage-time data. The graphite anode half-cell under test is in a preset target test environment, and the test is performed in a cycle of constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging.

[0026] In practice, the graphite anode half-cell under test is kept in a preset target test environment. This target test environment is controlled by a temperature control device to maintain a stable ambient temperature during the test and avoid temperature fluctuations affecting the internal electrochemical reaction and lithium-ion migration rate of the battery.

[0027] Furthermore, after placing the graphite anode half-cell under test in the target test environment, a cyclic test was performed, consisting of constant current charging, first-stage relaxation after charging, constant current discharging, and second-stage relaxation after discharging. Constant current charging refers to charging the battery with a constant current until it reaches a preset voltage.

[0028] Furthermore, after charging is complete, the battery enters the first stage of relaxation. Relaxation refers to the process by which the battery's internal electrochemical system gradually approaches equilibrium under conditions of no current flow. This first stage of relaxation is carried out under conditions of no current flow, and the relaxation time is a preset duration. During the first stage of relaxation, voltage-time data are collected at preset time intervals to obtain the first relaxation voltage-time data. This data reflects the natural decay pattern of the battery's open-circuit voltage over time after charging.

[0029] Furthermore, after the first stage of relaxation, constant current discharge is performed. Constant current discharge refers to discharging the battery with a constant current until a preset voltage is reached. After discharge, the second stage of relaxation begins. This relaxation process is also carried out under no-current resting conditions, with a relaxation time of a preset duration. During the second stage of relaxation, voltage-time data are collected at the same time intervals to obtain the second relaxation voltage-time data, which reflects the change in the battery's open-circuit voltage over time after discharge. The above cycle test is performed a preset number of times to obtain multiple sets of relaxation voltage data under different discharge conditions, used to observe the dynamic process of lithium plating gradually developing with changes in discharge conditions.

[0030] In some preferred embodiments, the temperature range of the target test environment is from -20 degrees Celsius to 40 degrees Celsius, and the temperature control accuracy is ±0.5 degrees Celsius.

[0031] In practice, the target test environment has a temperature range of -20 degrees Celsius to 40 degrees Celsius, with a temperature control accuracy of ±0.5 degrees Celsius.

[0032] Furthermore, the graphite anode half-cell under test is placed in a temperature-controlled environment that provides a stable and controllable temperature environment. This temperature-controlled environment can be a battery temperature-controlled test chamber or a high-low temperature test cabinet, equipped with temperature sensors, heating elements, and cooling elements. A feedback control algorithm adjusts the heating or cooling power in real time to stabilize the internal temperature of the chamber near the set value. The temperature regulation range of the temperature-controlled environment covers -20 degrees Celsius to 40 degrees Celsius, simulating various practical application scenarios from low temperatures to room temperature.

[0033] For example, when simulating the charging scenario of new energy vehicles in low-temperature winter conditions, the target test temperature can be set to -10 degrees Celsius; when simulating lithium plating behavior under normal temperature conditions, the target test temperature can be set to 25 degrees Celsius.

[0034] Furthermore, during the testing process, the temperature control equipment monitors the ambient temperature in real time and adjusts the temperature through heating or cooling elements, keeping temperature fluctuations within ±0.5 degrees Celsius of the set value. For example, when the target test temperature is 25 degrees Celsius, the temperature control equipment maintains the temperature between 24.5 and 25.5 degrees Celsius; when the target test temperature is -10 degrees Celsius, the temperature control equipment maintains the temperature within 0.5 degrees Celsius above the set value, i.e., between -10.5 and -9.5 degrees Celsius. This precise temperature control ensures the stability of the ambient temperature during relaxation voltage data acquisition, preventing temperature fluctuations from interfering with the lithium-ion transport rate, electrode reaction kinetics, and relaxation voltage variation within the battery.

[0035] In some preferred embodiments, the constant current charging rate is 0.5C to 5C, charging to a voltage of 2.0V; the first stage relaxation time after charging is 0.5 hours to 2 hours, and the relaxation process is a current-free resting state; the constant current discharging rate is a gradient increasing rate starting from 0.2C and increasing by 0.2C to 0.5C in each cycle, discharging to a voltage of 0.005V; the second stage relaxation time after discharging is 2 hours to 3 hours, and the relaxation process is a current-free resting state.

[0036] In practice, the constant current charging rate ranges from 0.5C to 5C, charging to a voltage of 2.0V. This charging rate range covers the conventional charging rate range and can simulate lithium plating under different charging scenarios. The higher the charging rate, the faster the lithium ions are inserted into the graphite anode, and the easier it is for lithium ions to deposit on the anode surface to form metallic lithium. By testing within this rate range, the risk of lithium plating on the graphite anode under different charging conditions can be assessed.

[0037] Furthermore, the first relaxation stage after charging lasts for 0.5 to 2 hours, during which the battery remains stationary without current. During this relaxation stage, the battery is in an open-circuit state, with no external current flowing through it, to avoid interference from current on the natural changes in the relaxation voltage. The choice of relaxation time directly affects the amount of information in the relaxation voltage data; too short a time may not be able to fully record the voltage change pattern during the relaxation process, while too long a time increases the testing time. A relaxation time of 1 hour is preferred, ensuring both data sufficiency and testing efficiency.

[0038] Furthermore, the constant current discharge rate is a gradient increase from 0.2C to 0.5C in each cycle, until the voltage reaches 0.005V. This gradient discharge pattern means that the discharge rate is 0.2C in the first cycle, increases to 0.4C or 0.5C in the second cycle, continues to increase in the third cycle, and so on. For example, the rate can be set to increase by 0.2C per cycle, i.e., 0.2C, 0.4C, 0.6C, 0.8C, 1.0C, 1.2C. This gradient discharge pattern can gradually increase the discharge current intensity, simulating the discharge process under different load conditions. When the discharge rate increases to a certain level, the lithium-ion transport rate inside the graphite anode cannot meet the discharge requirements, potentially triggering lithium plating. By observing the relaxation voltage characteristics at different discharge rates, the critical discharge rate condition for lithium plating can be accurately captured, enabling early warning of lithium plating.

[0039] Furthermore, the second-stage relaxation time after discharge is 2 to 3 hours, during which the battery is left to rest without current. The second-stage relaxation time is longer than the first-stage relaxation time because the polarization state inside the battery is more complex after discharge, requiring more time to fully recover equilibrium. A relaxation time of 3 hours is preferred to ensure that the relaxation process is nearly complete after discharge, the voltage stabilizes, and thus the voltage stability difference is accurately obtained. This parameter reflects the degree of recovery of the battery's internal polarization during the post-discharge relaxation process and is positively correlated with the amount of lithium plating.

[0040] In some preferred embodiments, the number of cycles for the cyclic test is 3 to 10.

[0041] In practice, after completing a full cycle of constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging, the cycle is repeated 3 to 10 times.

[0042] For example, the number of cycles can be set to 6. The charging rate remains the same in each cycle, while the discharging rate increases gradually in a gradient manner, and the relaxation time and voltage sampling interval remain unchanged.

[0043] By conducting multiple cycle tests, relaxation voltage data under different discharge rates can be obtained, allowing observation of the dynamic process of lithium plating gradually developing with increasing discharge current. Too few cycles, such as less than 3, may fail to cover the complete process from no lithium plating to its occurrence, making it difficult to capture the critical conditions for lithium plating. Too many cycles, such as more than 10, may lead to excessively long test times or non-targeted performance degradation of the battery during multiple charge-discharge cycles, affecting the accuracy of the evaluation results. By limiting the number of cycles to between 3 and 10, sufficient data can be obtained while controlling the interference from test time and battery state changes.

[0044] In some preferred embodiments, before acquiring the voltage-time data of the first stage relaxation of the graphite anode half-cell under test and obtaining the first relaxation voltage-time data, the method further includes: assembling a coin cell using the graphite material under test as the working electrode and a lithium metal sheet as the counter electrode and reference electrode; and performing charge-discharge activation treatment on the assembled coin cell to obtain the graphite anode half-cell under test.

[0045] In specific implementation, when preparing the graphite negative electrode half-cell to be tested, the graphite material to be tested is used as the working electrode. The graphite material to be tested is mixed evenly with a binder and a conductive agent in a certain proportion, a solvent is added to form a slurry, which is then coated onto a copper foil, dried, and pressed to form the working electrode. For example, in one embodiment, the working electrode has a diameter of 12 mm and a thickness of 80 to 120 micrometers, preferably 85 micrometers. A lithium metal sheet is used as both the counter electrode and the reference electrode, with a diameter of 15.6 mm and a thickness of 50 micrometers.

[0046] Furthermore, the electrolyte is prepared; for example, in one embodiment, the electrolyte uses 1 mol / L lithium hexafluorophosphate, and the solvent is a mixed solvent of ethylene carbonate, diethyl carbonate, and dimethyl carbonate in a volume ratio of 1:1:1. This electrolyte system has good lithium-ion transport efficiency and can reduce the interference of the electrolyte on the relaxation voltage test.

[0047] Furthermore, assembly was carried out in an argon-filled glove box, where the water content and oxygen content were no more than 0.01 ppm to avoid affecting battery performance. The working electrode, lithium metal sheet, electrolyte, and separator were assembled into a coin cell, model CR2032. Multiple samples (e.g., three) of each graphite material were prepared in parallel to ensure test repeatability.

[0048] Further, the assembled coin cell half-cell is subjected to charge-discharge activation treatment to obtain the graphite anode half-cell to be tested. The activation temperature is 5°C to 45°C, preferably 25°C. The charge-discharge activation step adopts a stepped discharge rate, for example, in one embodiment, specifically 0.1C, 0.05C, 0.02C stepped discharge rates, and a 0.1C charge rate, and is carried out in a voltage range of 0.005V to 2.0V, for 3 cycles. The stepped discharge rate refers to discharging the battery with a gradually decreasing current. For example, in one embodiment, specifically, it is first discharged at 0.1C, then at 0.05C, and finally at 0.02C. This stepped discharge helps to form a stable and uniform solid electrolyte interface film on the graphite anode.

[0049] Furthermore, after activation, the electrode is left to stand for 30 minutes to remove the passivation layer on the electrode surface, allowing the graphite anode to form a stable solid electrolyte interface film. The battery after the above activation and standing treatment is the graphite anode half-cell under test, which can be used for subsequent charge-discharge relaxation cycle testing.

[0050] This embodiment, through step-rate charge-discharge activation treatment and a settling step, enables the formation of a stable and uniform solid electrolyte interface film on the graphite negative electrode surface, eliminating the interference of the initial passivation layer on the relaxation voltage signal. This provides a reliable battery state basis for the accurate acquisition of subsequent relaxation voltage data, thereby improving the repeatability and reliability of the evaluation results.

[0051] S2, perform differentiation processing on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve.

[0052] In practice, after obtaining the first relaxation voltage-time data and the second relaxation voltage-time data, differential processing is performed on these two sets of data. The differential processing employs a numerical differentiation algorithm to calculate the rate of change of voltage with time, i.e., the derivative of voltage with respect to time, thereby obtaining the first relaxation differential curve and the second relaxation differential curve. The vertical axis of the differential curve represents the rate of voltage change. This curve can amplify the subtle characteristics of voltage changes during relaxation, facilitating the identification of characteristic signals related to lithium plating.

[0053] In some preferred embodiments, before performing differentiation processing on the first relaxation voltage-time data and the second relaxation voltage-time data, the method further includes: performing noise reduction processing on the first relaxation voltage-time data and the second relaxation voltage-time data.

[0054] In practice, before performing differentiation processing on the first relaxation voltage-time data and the second relaxation voltage-time data, noise reduction processing is also performed on these two sets of data.

[0055] Specifically, after obtaining the first and second relaxation voltage-time data, wavelet denoising algorithm (or other denoising algorithms, which are not specifically limited in this invention) is applied to each set of data for denoising processing. The basic principle of wavelet denoising algorithm is to decompose the original signal into wavelet coefficients of different frequency components, perform threshold processing on the high-frequency components to remove high-frequency interference signals, and then reconstruct the signal to obtain the denoised data.

[0056] The specific implementation steps are as follows: First, select appropriate wavelet basis functions and decomposition levels. The choice of wavelet basis functions depends on the characteristics of the signal; for example, the db4 wavelet can be selected as the basis function, and the decomposition level can be set to 4 levels. The voltage-time data is decomposed into approximation coefficients and detail coefficients through wavelet transform. The approximation coefficients represent the low-frequency components of the signal, and the detail coefficients represent the high-frequency components. For the detail coefficients, a soft thresholding method is used, setting coefficients with absolute values ​​below a set threshold to zero and retaining coefficients above the threshold, thereby removing high-frequency noise components. The threshold setting can be determined based on the noise level of the signal; for example, it can be calculated using a general threshold formula. After thresholding, wavelet reconstruction is performed using the processed detail coefficients and the original approximation coefficients to obtain the denoised signal. Through this algorithm, interference signals with frequencies greater than or equal to 10 Hz are effectively filtered out, while the effective signal reflecting the relaxation voltage variation is retained. After denoising, smoothed first and second relaxation voltage-time data are obtained, which are used for subsequent differentiation processing and feature parameter extraction.

[0057] S3, extract lithium plating characteristic parameters from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data.

[0058] In specific implementation, lithium plating characteristic parameters are extracted from the first relaxation differential curve, the second relaxation differential curve, and the first relaxation voltage-time data and the second relaxation voltage-time data. These lithium plating characteristic parameters include characteristic parameters extracted from the first relaxation differential curve, characteristic parameters extracted from the second relaxation differential curve, characteristic parameters extracted from the first relaxation voltage-time data, and characteristic parameters extracted from the second relaxation voltage-time data.

[0059] The characteristic parameters of the first relaxation differential curve reflect the abnormal voltage change rate caused by lithium-ion deposition or stripping during the relaxation process after charging, including the time and intensity of this abnormality. The characteristic parameters of the second relaxation differential curve reflect the abnormal voltage change rate caused by residual lithium participating in the reaction during the relaxation process after discharging, also including the time and intensity of this abnormality. The characteristic parameters extracted from the first relaxation voltage-time data reflect the migration and redistribution rate of lithium ions within the graphite during the initial relaxation phase after charging. The characteristic parameters extracted from the second relaxation voltage-time data reflect the degree of recovery of internal polarization during the relaxation process after discharging.

[0060] For example, in some preferred embodiments, the lithium plating characteristic parameters include a first characteristic parameter, a second characteristic parameter, a third characteristic parameter, and a fourth characteristic parameter; the first characteristic parameter includes the characteristic peak occurrence time and characteristic peak amplitude of the first relaxation differential curve; the second characteristic parameter includes the characteristic peak occurrence time and characteristic peak amplitude of the second relaxation differential curve; the third characteristic parameter includes a first relaxation voltage decay rate calculated based on the first relaxation voltage-time data; and the fourth characteristic parameter includes a second relaxation voltage stability difference calculated based on the second relaxation voltage-time data.

[0061] In specific implementation, the first characteristic parameters include the occurrence time and amplitude of the characteristic peak of the first relaxation differential curve. On the first relaxation differential curve, the trend of the curve's change is scanned to find a convex structure that first rises and then falls. Specifically, starting from the beginning of relaxation, the change in the differential curve value is calculated point by point. When a value is detected to rise continuously to a local maximum and then fall continuously, the point corresponding to this local maximum is identified as a candidate characteristic peak. For each identified candidate characteristic peak, the time point corresponding to the peak value is recorded as the characteristic peak occurrence time, and the magnitude of the peak value is recorded as the characteristic peak amplitude. If no convex structure meeting the judgment criteria is identified on the first relaxation differential curve, the characteristic peak occurrence time is recorded as 0, and the characteristic peak amplitude is recorded as 0.

[0062] Furthermore, the second characteristic parameter includes the occurrence time and amplitude of the characteristic peak of the second relaxation differential curve. On the second relaxation differential curve, using the same identification method as the first characteristic parameter, a convex structure that rises first and then falls is identified, and the occurrence time and amplitude of the characteristic peak are recorded. If no convex structure meeting the judgment criteria is identified, it is also recorded as 0.

[0063] Furthermore, the third characteristic parameter includes the first relaxation voltage decay rate calculated based on the first relaxation voltage-time data. Voltage data within the first 30 minutes after the start of relaxation is extracted from the first relaxation voltage-time data. The voltage at the start of relaxation is recorded as the initial voltage, and the voltage 30 minutes after the start of relaxation is recorded as the 30-minute voltage. Subtracting the 30-minute voltage from the initial voltage yields the voltage decay amount. Dividing the voltage decay amount by 30 minutes yields the average decay rate, which is the first relaxation voltage decay rate, expressed in volts per minute. This decay rate reflects the migration and distribution of lithium ions within the graphite during the relaxation process after charging. A faster decay rate indicates a faster diffusion rate of lithium ions in the graphite anode, and better lithium intercalation kinetics in the graphite; conversely, a slower decay rate may indicate lithium plating or hindered diffusion.

[0064] Furthermore, the fourth characteristic parameter includes the second relaxation voltage stability difference calculated based on the second relaxation voltage-time data. From the second relaxation voltage-time data, the starting voltage at the beginning of relaxation and the ending voltage at the end of relaxation are extracted. Subtracting the starting voltage from the ending voltage yields the voltage difference, which is the second relaxation voltage stability difference, expressed in volts. This difference reflects the degree of polarization recovery within the battery during the relaxation process after discharge. If lithium plating occurs during discharge, the residual metallic lithium will continue to react with graphite after discharge, delaying the polarization recovery process and reducing the difference between the ending and starting voltages. The greater the amount of lithium plating, the smaller this difference; therefore, this parameter is negatively correlated with the amount of lithium plating.

[0065] It should be noted that the first, second, third, and fourth characteristic parameters selected in this invention are not artificially set empirical statistical parameters, but rather are established based on the intrinsic electrochemical nature of the graphite anode lithium intercalation / deintercalation kinetics, lithium metal deposition and self-dissolution behavior, SEI polarization decay law at the electrode interface, and open-circuit voltage relaxation kinetics, establishing a one-to-one correspondence. Each characteristic parameter has a clear electrochemical and physical meaning, and can characterize whether lithium plating occurs in the graphite anode, the severity of lithium plating, and the amount of lithium plating per unit area from a mechanistic perspective. Specific details are as follows.

[0066] The first characteristic parameter, namely the time of appearance and amplitude of the characteristic peak of the first relaxation differential curve, represents the first stage of relaxation after white charging, corresponding to the electrochemical behavior of lithium ion deposition and subsequent self-dissolution of deposited metallic lithium on the negative electrode surface during charging. Under high-rate charging and low-temperature charging conditions, the rate at which lithium ions migrate from the electrolyte to the graphite negative electrode surface is greater than the rate of lithium ion diffusion inside the graphite solid phase. Excess lithium ions cannot be intercalated into the graphite lattice in time and will undergo a reduction reaction on the negative electrode surface to generate elemental metallic lithium, forming surface lithium deposition and lithium dendrites. After charging is cut off and the system enters open-circuit relaxation without current input, the elemental metallic lithium deposited on the graphite surface will spontaneously undergo a self-dissolution reaction and undergo interfacial side reactions with the electrolyte and SEI film. At the same time, the high concentration of lithium intercalated on the graphite surface slowly diffuses into the solid phase inside the particles. After differentiation, the relaxation voltage-time curve can effectively amplify the subtle kinetic differences in open-circuit potential evolution. When lithium plating is present, the coupling effect of the lithium metal self-dissolution side reaction and the re-diffusion of excess lithium intercalation will form a significant characteristic peak that rises first and then falls on the first relaxation differential curve. The earlier the lithium plating begins and the greater the amount of lithium deposited, the earlier the corresponding characteristic peak appears and the higher the characteristic peak amplitude. In the absence of lithium plating, only normal graphite lithium intercalation solid-phase diffusion and conventional interface polarization relaxation exist, and the differential curve has no effective characteristic peak that meets the threshold. Therefore, the appearance time and amplitude of the characteristic peak of the first relaxation differential curve can directly characterize the timing and severity of lithium deposition on the graphite anode during the charging stage from an electrochemical mechanism perspective, and are the core sensitive features for identifying early weak lithium plating.

[0067] The second characteristic parameter, namely the time of appearance and amplitude of the characteristic peak in the second relaxation differential curve, is derived from the second stage of relaxation after discharge, corresponding to the electrochemical behavior of the continuous oxidation and dissolution of residual elemental lithium and interfacial side reactions after discharge. During constant current discharge, normal lithium intercalation within the graphite lattice can be successfully extracted, but elemental lithium deposited on the electrode surface and SEI interface during charging and cycling cannot be completely extracted during the normal delithiation process, remaining on the negative electrode surface as residual active lithium. After the discharge ends and a long period of open-circuit relaxation begins, this residual lithium will continue to undergo slow oxidation and dissolution, continuously inducing interfacial side reactions such as electrolyte decomposition and SEI reconstruction, continuously altering the evolution of the electrode open-circuit potential. Samples without lithium plating only show smooth voltage changes in double-layer polarization decay and solid-phase concentration polarization relaxation, with no abnormal bulges in the differential curve; samples with residual lithium plating, due to the potential perturbation contributed by the continuous electrochemical reaction of residual lithium, will form an effective characteristic peak on the second relaxation differential curve. The greater the residual lithium plating, the higher the active lithium reaction rate, the earlier the characteristic peak appears, and the larger the peak amplitude. This parameter characterizes the degree of lithium plating accumulation from the perspectives of residual lithium after discharge and reaction activity, making up for the inability to capture residual lithium plating signals by relying solely on single-stage relaxation after charging.

[0068] The third characteristic parameter, the first relaxation voltage decay rate, characterizes the degree of obstruction in the solid-phase diffusion kinetics of lithium ions within the graphite during the initial relaxation phase after charging, and is strongly correlated with lithium plating behavior. Under non-lithiation conditions, the graphite surface is enriched with a high concentration of intercalated lithium, which rapidly diffuses into the solid phase within the particles driven by the concentration gradient. The lithium ion concentration quickly reaches equilibrium, and the open-circuit voltage exhibits a rapid and regular natural decay, with the voltage decay rate remaining within a stable and normal range. When surface lithium plating occurs on the graphite anode, the deposited metallic lithium covers the active lithium intercalation sites in the graphite, blocking the lithium ion solid-phase diffusion channels. Simultaneously, elemental metallic lithium itself has a stable potential plateau effect, significantly suppressing the natural decay kinetics of the open-circuit voltage after charging. Macroscopically, this manifests as a significantly slower first relaxation voltage decay rate. Furthermore, the higher the degree of lithium plating and the denser the surface lithium coverage, the more severe the obstruction of lithium ion diffusion, and the lower the voltage decay rate. This parameter, from the perspective of lithium intercalation and diffusion kinetics, indirectly quantifies the inhibitory effect of lithium plating on the intrinsic lithium intercalation and deintercalation kinetics of the graphite anode, achieving a quantitative kinetic characterization of the degree of lithium plating.

[0069] The fourth characteristic parameter, the second relaxation voltage stability difference, characterizes the battery's overall electrochemical polarization recovery and dissipation capability during long-term relaxation after discharge, and is related to the long-term interfacial polarization effect induced by residual lithium plating. Under lithium-free conditions, the double-layer polarization, SEI interfacial polarization, and solid-phase concentration polarization generated during discharge can be fully dissipated after a sufficiently long second-stage relaxation, and the system quickly returns to electrochemical steady state. The difference between the relaxation initiation voltage and the termination steady-state voltage is significant, resulting in a large voltage stability difference. When lithium plating is present, the residual elemental lithium on the electrode surface continues to undergo oxidation and dissolution, as well as electrolyte side reactions, throughout the entire relaxation cycle. This continuously introduces additional interfacial and concentration polarization, hindering the battery system from reaching thermodynamic steady state. The greater the amount of lithium plating, the stronger the persistent side reaction effect, and the more pronounced the polarization retention. Even after relaxation ends, the steady-state potential cannot be fully restored, ultimately resulting in a significant reduction in the second relaxation voltage stability difference. Based on polarization recovery kinetics, this parameter shows a clear negative correlation with the amount of lithium plating and can be used for the correction and fitting of lithium plating grade classification and lithium plating quantification results.

[0070] In summary, this invention constructs a multi-dimensional lithium plating characteristic parameter system through four independent yet complementary electrochemical dimensions: lithium deposition characteristics after charging, residual lithium reaction characteristics after discharging, lithium intercalation solid-phase diffusion kinetics characteristics, and long-term polarization recovery characteristics. These four characteristic parameters are not empirical statistical parameters, but rather mechanistic parameters with clear electrochemical and physical significance. They can fundamentally reflect the occurrence, development, and quantification of lithium plating, providing a solid electrochemical theoretical foundation for the lithium plating evaluation method of this invention.

[0071] This embodiment clarifies the specific extraction methods for four lithium plating characteristic parameters. The first and second characteristic parameters extract characteristic peak information from the voltage change rate curves of post-charge relaxation and post-discharge relaxation, respectively. The characteristic peak appearing in the post-charge relaxation stage typically corresponds to the signal of the reaction between deposited metallic lithium and graphite during charging, while the characteristic peak appearing in the post-discharge relaxation stage typically corresponds to the signal of residual lithium that failed to escape during discharge continuing to participate in the reaction. The occurrence time and amplitude of these two characteristic peaks are directly related to the onset time and severity of lithium plating. Furthermore, the third characteristic parameter quantifies the rapid migration behavior of lithium ions in the early stages of relaxation by calculating the voltage decay rate in the first 30 minutes of post-charge relaxation. In the absence of lithium plating, lithium ions diffuse from the graphite surface to the interior, causing a rapid voltage drop; in the case of lithium plating, the metallic lithium deposited on the surface slows down the voltage drop rate. This parameter indirectly reflects the degree of lithium plating. Furthermore, the fourth characteristic parameter quantifies the degree of battery polarization recovery by calculating the voltage stability difference at the end of post-discharge relaxation. Without lithium plating, the battery polarization can be fully recovered after discharge, and the termination voltage is close to the starting voltage. With lithium plating, residual metallic lithium continues to participate in the reaction, resulting in a lower termination voltage, and this difference becomes smaller. These four characteristic parameters quantify the relaxation voltage data from different dimensions, forming a multi-dimensional and multi-faceted lithium plating characteristic description system, providing sufficient data support for the accurate determination of subsequent lithium plating states.

[0072] In some preferred embodiments, the characteristic peaks of the first relaxation differential curve and the second relaxation differential curve are determined based on a preset judgment criterion. The judgment criterion is that a bulge appears on the relaxation differential curve that rises first and then falls, and the amplitude of the bulge is greater than 3 times the amplitude of the background noise. The relaxation differential curve includes the first relaxation differential curve and the second relaxation differential curve.

[0073] In practice, for the first relaxation differential curve, its changing trend is observed. When a convex region appears on the curve that gradually rises, reaches a peak, and then gradually declines, this region is identified as a candidate characteristic peak. For the candidate characteristic peak, its amplitude is calculated, i.e., the vertical coordinate value corresponding to the peak point. Simultaneously, the background noise amplitude is calculated. The background noise amplitude is defined as the maximum fluctuation value of the relaxation differential curve within the first 10 minutes after relaxation begins; that is, within this 10-minute time window, the absolute value of the difference between the maximum and minimum values ​​on the differential curve is taken. This 10-minute time window is usually in the early stage of relaxation, at which time no characteristic signals related to lithium plating have appeared inside the battery. Therefore, the fluctuations during this period mainly originate from the inherent noise of the test system, including interference factors such as instrument errors and temperature fluctuations. The amplitude of the candidate characteristic peak is compared with three times the background noise amplitude. If the amplitude of the candidate characteristic peak is greater than three times the background noise amplitude, the candidate characteristic peak is determined to be a valid characteristic peak, and the time and amplitude of its appearance are recorded. If the amplitude of a candidate characteristic peak is not greater than three times the amplitude of the background noise, then the candidate characteristic peak is determined to be noise interference and is not considered a characteristic peak.

[0074] Furthermore, for the second relaxation differential curve, the same criteria are used to identify and determine the characteristic peaks.

[0075] S4. Based on the extracted lithium plating characteristic parameters, determine whether the graphite anode has undergone lithium plating and obtain the amount of lithium plating.

[0076] In practice, based on the extracted lithium plating feature parameters, it is determined whether the graphite anode has undergone lithium plating and the amount of lithium plating is obtained. Specifically, the extracted lithium plating feature parameters are combined to form a feature parameter vector, which is then input into the lithium plating evaluation model. The lithium plating evaluation model is constructed using a pre-defined machine learning algorithm. During the model training phase, a large number of graphite anode samples with known lithium plating states are used as training samples. Each sample contains the aforementioned feature parameters, as well as the corresponding actual lithium plating level and actual lithium plating amount. The machine learning algorithm learns the mapping relationship between the feature parameters and the lithium plating state. Furthermore, after the model training is completed, for the input feature parameter vector of the sample to be tested, the model calculates based on the learned mapping relationship and outputs the lithium plating level and amount of lithium plating of the graphite anode to be tested. The lithium plating level includes four levels: no lithium plating, slight lithium plating, moderate lithium plating, and heavy lithium plating. The amount of lithium plating is expressed as the mass of metallic lithium deposited per unit area.

[0077] This invention employs a dual-stage relaxation voltage acquisition method, acquiring first and second relaxation voltage-time data respectively. This simultaneously captures early lithium plating signals caused by lithium deposition during charging and residual lithium plating characteristics reflected by the stripping of residual metallic lithium after discharge, significantly improving the early detection sensitivity for slight lithium plating. By differentiating the two sets of relaxation voltage-time data, first and second relaxation differential curves are obtained. Combined with the original voltage-time data, lithium plating characteristic parameters are extracted to form a multi-dimensional quantitative criterion covering voltage change rate anomalies, decay kinetics, and polarization recovery. Based on these characteristic parameters, it is possible to directly determine whether lithium plating has occurred and obtain the specific amount of lithium plating, achieving a leap from qualitative judgment to quantitative assessment and solving the problem that existing technologies cannot quantify the degree of lithium plating. The combination of dual-stage relaxation data acquisition and differential processing effectively highlights lithium plating-related characteristic signals, suppresses background interference, and enhances the method's anti-interference capability. Furthermore, this method places the battery under test in a target test environment for cyclic testing. This environment can be set with different temperatures according to actual needs, thus adapting to various application scenarios such as room temperature and low temperature. In summary, this method effectively overcomes the shortcomings of existing technologies in terms of lithium plating quantification assessment, early identification sensitivity, anti-interference ability, and adaptability to multiple temperature scenarios.

[0078] In some preferred embodiments, determining whether the graphite anode has undergone lithium deposition and obtaining the amount of lithium deposition based on the extracted lithium deposition characteristic parameters includes: inputting the extracted lithium deposition characteristic parameters into a lithium deposition evaluation model, and having the lithium deposition evaluation model output the lithium deposition grade and the amount of lithium deposition.

[0079] In practice, the step of determining whether the graphite anode has undergone lithium deposition and obtaining the amount of lithium deposition based on the extracted lithium deposition characteristic parameters is achieved through a lithium deposition evaluation model.

[0080] The extracted lithium plating characteristic parameters, including the peak occurrence time and amplitude of the first relaxation differential curve, the peak occurrence time and amplitude of the second relaxation differential curve, the first relaxation voltage decay rate, and the second relaxation voltage stability difference, are collectively constituted into a characteristic parameter vector. This characteristic parameter vector contains all the key information extracted from the relaxation voltage data and serves as the input to the lithium plating evaluation model.

[0081] Furthermore, the feature parameter vector is input into a pre-trained lithium plating evaluation model. Specifically, the lithium plating evaluation model can be constructed using the random forest algorithm. The random forest algorithm is an ensemble learning method that improves prediction accuracy and generalization ability by constructing multiple decision trees and combining the prediction results of each decision tree. During the model training phase, a large number of graphite anode samples with known lithium plating states are used as training samples. Each sample contains the above six feature parameters and the corresponding actual lithium plating grade and actual lithium plating amount. The actual lithium plating grade and actual lithium plating amount are obtained by observing the battery after disassembly using scanning electron microscopy and measuring it using inductively coupled plasma atomic emission spectrometry. The random forest algorithm learns the mapping relationship between feature parameters and lithium plating states. The algorithm constructs multiple decision trees, each of which is trained based on a different feature subset and sample subset. The final prediction result is the average of the prediction results of all decision trees or the majority vote result.

[0082] Furthermore, after the lithium plating evaluation model is trained, it calculates based on the learned mapping relationship for the input feature parameter vector of the sample to be tested. Internally, the feature parameter vector is input into each decision tree, and each tree judges according to its own node splitting rules, outputting a prediction result. For the lithium plating level, a majority voting method is used, that is, the lithium plating level output by all decision trees is counted, and the level with the most votes is used as the final output; for the lithium plating amount, an average method is used, that is, the average of the lithium plating amount output by all decision trees is calculated as the final output. Finally, the lithium plating evaluation model outputs the lithium plating level and lithium plating amount of the graphite anode to be tested. The lithium plating level is divided into four levels: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating, and the quantification range of the lithium plating amount is 0 to 5 milligrams per square centimeter.

[0083] Furthermore, in this embodiment of the invention, the lithium plating evaluation model is constructed using the following systematic training, verification, and generalization verification methods to ensure its prediction accuracy and applicability to multiple scenarios.

[0084] In terms of constructing the model training sample set, a large number of coin half-cell samples were prepared using different graphite anode materials, different test temperatures, different charge-discharge rates, and different cycle numbers to construct a sample set covering multiple operating conditions. Three types of data were simultaneously acquired for each sample: First, six lithium plating characteristic parameters from the aforementioned four categories (i.e., the appearance time and amplitude of the characteristic peak of the first relaxation differential curve, the appearance time and amplitude of the characteristic peak of the second relaxation differential curve, the decay rate of the first relaxation voltage, and the stability difference of the second relaxation voltage) were used as the model input feature vector; Second, lithium plating level labels, visually graded by scanning electron microscopy (SEM), including four levels: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating, were used as the ground truth labels for the classification task; Third, the amount of lithium plating per unit area (mg / cm²) measured by inductively coupled plasma atomic emission spectrometry (ICP). 2 The data is used as the true value for the regression task. The sample set is divided into training and testing sets in a 7:3 ratio, and the samples are guaranteed to cover all working conditions, including room temperature / low temperature, different graphite materials, different rate gradients, and different degrees of lithium plating, to ensure the sufficiency and representativeness of the model learning.

[0085] The specific training process of the random forest model includes the following steps: (1) The above six feature parameters form a multidimensional input feature vector and input it into the random forest model; (2) The model undertakes both classification and regression tasks: For the lithium plating level classification task, the four-level lithium plating labels calibrated by SEM are used as the output, and the majority voting mechanism is adopted to output the final lithium plating level by voting by all decision trees; For the lithium plating amount regression task, the measured lithium plating amount of ICP is used as the fitting true value, and the average value of the output values ​​of all decision trees is used as the quantified lithium plating amount; (3) During the training process, the bootstrap sampling strategy is used to randomly sample the training set with replacement, generate multiple differentiated training subsets, and randomly select the feature subset for optimal splitting at each node split, so as to avoid overfitting of a single decision tree and enhance the generalization ability of the model; The number of decision trees, the maximum tree depth, the minimum number of samples in the leaf node and other hyperparameters are determined by grid optimization until the model converges; (4) The training termination condition is: the prediction error of the test set converges and no longer decreases, the model fit reaches the set threshold, and the optimal training model is saved for subsequent inference evaluation of the test samples.

[0086] Regarding model validation evaluation metrics and actual performance, quantitative validation was conducted after training in three dimensions: classification accuracy, regression error, and sample repeatability. First, the classification accuracy for lithium plating levels was no less than 95% on the test set, classifying no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating, with no significant misclassification. Second, the quantification accuracy of lithium plating was no more than 5% relative to the model's predicted lithium plating amount compared to the actual lithium plating amount measured by ICP, meeting the accuracy requirements for industrial evaluation. Third, sample repeatability was demonstrated by inputting parallel samples from the same batch into the model, resulting in consistent output lithium plating levels and a relative standard deviation of no more than 6% for the lithium plating amount, indicating excellent model stability. Furthermore, the model's goodness of fit R0 was [not specified in the original text]. 2 A goodness of fit of 0.95 or higher indicates a strong correlation between the characteristic parameters and the lithium plating grade and amount, demonstrating reliable model fitting. For example, in Example 1 of this invention, the model fit reached 0.96, and the error between the predicted lithium plating amount and the measured ICP value was only 3.6%; in Example 2, the prediction error was only 2.2%, and the relative standard deviation of parallel samples was no greater than 1.2%, both of which fully verified the actual evaluation effect of the model.

[0087] Regarding the model's applicability and generalization ability across multiple scenarios, the trained random forest evaluation model possesses excellent generalization and adaptation capabilities because the training sample set was intentionally constructed to cover various boundary conditions. Specifically, this is reflected in the following aspects: First, temperature scenario adaptation: the training samples cover high and low temperature environments from -20℃ to 40℃, allowing the model to adapt to actual vehicle and energy storage battery application scenarios such as low-temperature charging in winter and normal-temperature cycling. Second, material adaptation: the training samples cover mainstream graphite anode systems such as artificial graphite and natural graphite, ensuring that the model does not fail to evaluate due to differences in graphite microstructure, particle size, or preparation process. Third, rate condition adaptation: the training samples adapt to charging rates from 0.5C to 5C and gradient-increasing discharge rates, covering different application rate scenarios such as regular charging and fast charging. Fourth, cycle life adaptation: the training samples include graphite anodes at different cycling stages, enabling the model to identify and quantify lithium plating behavior at different cycling stages (early, middle, and late stages), unaffected by relaxation characteristic shifts caused by battery cycle aging.

[0088] In practical inference applications, simply input the multidimensional lithium plating feature parameters extracted from the measured half-cell under test into the trained and converged random forest model. The model automatically traverses each decision tree branch internally, and through voting and mean regression calculations, directly outputs the lithium plating grade and a value of 0 to 5 mg / cm³. 2 The quantitative lithium plating amount within the range can be determined without manually setting the criterion threshold, realizing automated, standardized, and reproducible intelligent evaluation of lithium plating.

[0089] In some preferred embodiments, the lithium plating evaluation model is constructed using a machine learning algorithm, and the lithium plating level includes four levels: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating. The quantitative range of the amount of lithium plating is 0 to 5 milligrams per square centimeter.

[0090] In specific implementation, the lithium plating evaluation model is constructed using a preset machine learning algorithm (e.g., random forest algorithm, which is not specifically limited in this invention), and the output of lithium plating level and amount is as follows.

[0091] The lithium plating evaluation model is constructed using the random forest algorithm. The random forest algorithm is an ensemble learning method. Its basic principle is to construct multiple decision trees. Each decision tree is trained using a subset of samples randomly drawn with replacement from the original dataset. At each node, a subset of features is randomly selected for optimal splitting, thus ensuring diversity among the trees. The final prediction result is obtained by combining the predictions of all decision trees. For classification tasks, majority voting is used, and for regression tasks, the average value method is used.

[0092] Furthermore, during model training, the model is trained using supervised learning, with feature parameters as input and actual lithium plating level and actual lithium plating amount as output. The actual lithium plating level is determined by observing the surface morphology of the graphite electrode using a scanning electron microscope. The specific criteria are as follows: no lithium plating indicates a smooth graphite anode surface without metallic lithium deposits; slight lithium plating indicates scattered metallic lithium particles on the graphite anode surface, not forming a continuous coverage; moderate lithium plating indicates significant metallic lithium deposition on the graphite anode surface, forming a localized continuous coverage area; and severe lithium plating indicates the graphite anode surface is completely covered by a dense metallic lithium layer. The actual lithium plating amount is determined by inductively coupled plasma atomic emission spectrometry (ICP-AES). Specifically, the graphite electrode is disassembled, the deposited metallic lithium is dissolved in an acid solution, and the lithium ion concentration is measured using ICP-AES to calculate the lithium plating mass per unit area.

[0093] Furthermore, the lithium plating level is divided into four grades: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating. The quantification range of lithium plating amount is 0 to 5 mg / cm². When the model outputs, the lithium plating level is output as a classification label, such as "moderate lithium plating"; the lithium plating amount is output as a continuous numerical value, such as "2.3 mg / cm²". The quantification range of lithium plating amount covers the complete range from no lithium plating to severe lithium plating, which can meet the lithium plating assessment needs in different application scenarios.

[0094] In some preferred embodiments, after obtaining the amount of lithium plating, the method further includes: comparing the obtained amount of lithium plating with the actual amount of lithium plating of the graphite anode half-cell under test determined by inductively coupled plasma atomic emission spectrometry to obtain the evaluation error; if the evaluation error is greater than 5%, then after adjusting at least one of the first stage relaxation time after charging, the second stage relaxation time after discharging, and the voltage acquisition interval, the method proceeds to the step of acquiring the voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data, so as to re-evaluate the graphite anode half-cell under test.

[0095] In practice, after obtaining the amount of lithium deposited in the graphite anode half-cell using the lithium deposition evaluation model, this amount is compared with the actual amount of lithium deposited in the same graphite anode half-cell determined by inductively coupled plasma atomic emission spectrometry (ICP-AES). ICP-AES is a quantitative analysis method. Its principle involves exciting the sample in a high-temperature plasma torch, causing elemental atoms to emit characteristic spectra, and then quantitatively analyzing the elemental content by detecting the intensity of these characteristic spectra. To determine the actual amount of lithium deposited, the graphite anode half-cell needs to be disassembled, the graphite electrode removed, and the metallic lithium deposited on the graphite surface dissolved using an acid solution such as dilute hydrochloric acid or dilute nitric acid. The dissolved solution is then diluted to a certain volume, and the lithium ion concentration is measured using ICP-AES. Based on the concentration, solution volume, and the area of ​​the graphite electrode, the actual amount of lithium deposited per unit area is calculated, expressed in milligrams per square centimeter. This actual amount of lithium deposited serves as a benchmark value to verify the accuracy of the model output.

[0096] Furthermore, the lithium plating amount output by the model is compared with the actual lithium plating amount measured by inductively coupled plasma atomic emission spectrometry (ICP-AES), and the evaluation error is calculated. The evaluation error is calculated as follows: the absolute value of the difference between the model output lithium plating amount and the actual lithium plating amount is divided by the actual lithium plating amount, and then multiplied by 100% to obtain the percentage error. After obtaining the evaluation error, it is determined whether the evaluation error is greater than 5%.

[0097] Furthermore, if the evaluation error is no greater than 5%, the evaluation result is valid, and the amount of lithium plating is the final evaluation result, which can be used for performance evaluation of graphite anode materials or determination of battery safety level.

[0098] Furthermore, if the evaluation error exceeds 5%, it indicates that the accuracy of the evaluation results under the current test parameter settings is insufficient, and the test parameters need to be adjusted. The adjusted parameters include at least one of the following: the first-stage relaxation time after charging, the second-stage relaxation time after discharging, and the voltage sampling interval. The voltage sampling interval refers to the time interval between two adjacent voltage sampling actions when the electrochemical test module (or evaluation system) continuously samples the open-circuit voltage of the half-cell during the relaxation stage (including the first and second stages of relaxation). For example, the relaxation time can be appropriately extended from 1 hour to 1.5 hours to obtain more complete relaxation data; or the voltage sampling interval can be shortened from 200 milliseconds to 100 milliseconds to improve data resolution. After the parameter adjustment is completed, return to the step of collecting the voltage-time data of the first-stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data. Then, re-perform the charge-discharge relaxation cycle test, data processing, feature extraction, and model determination of the graphite anode half-cell under test until the evaluation error is no greater than 5%, thus obtaining a valid evaluation result.

[0099] This embodiment introduces inductively coupled plasma atomic emission spectrometry (ICP-AES) as a benchmark verification method, establishing a verification and feedback adjustment mechanism for the evaluation results. ICP-AES is characterized by high sensitivity and accuracy, enabling precise determination of the actual lithium deposition on the graphite anode surface, serving as a verification benchmark for the model output. When the error between the model-output lithium deposition and the actual lithium deposition exceeds 5%, adjusting key parameters such as relaxation time and voltage acquisition interval and retesting effectively reduces the evaluation error, ensuring the accuracy of the final evaluation results. For example, if the initial relaxation time is too short, resulting in insufficient relaxation, the measured voltage stability difference may be inaccurate. Extending the relaxation time can yield more accurate relaxation data. This verification and adjustment mechanism forms a closed-loop control, enabling the evaluation method to self-calibrate and optimize, adapting to different sample characteristics and testing conditions, and ensuring the reliability and consistency of the evaluation results.

[0100] Accordingly, embodiments of the present invention provide a graphite anode lithium plating evaluation system, comprising: The environmental control module is used to control the graphite anode half-cell under test to be in a preset target test environment. The electrochemical testing module is used to perform cyclic testing with constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging as one cycle. It collects voltage-time data of the first stage relaxation to obtain first relaxation voltage-time data, and collects voltage-time data of the second stage relaxation to obtain second relaxation voltage-time data. The data processing module is used to perform differential processing on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve; extract lithium plating characteristic parameters from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data; and determine whether the graphite anode has lithium plating and obtain the amount of lithium plating based on the extracted lithium plating characteristic parameters.

[0101] In practice, the graphite anode lithium plating evaluation system includes an environmental control module, an electrochemical testing module, and a data processing module.

[0102] The environmental control module is used to control the graphite anode half-cell under test within a preset target test environment. This module can be a temperature-controlled device, such as a battery temperature control chamber or a high / low temperature test cabinet, which contains temperature sensors, heating elements, and cooling elements. The environmental control module uses a feedback control algorithm to adjust the heating or cooling power in real time, stabilizing the ambient temperature of the graphite anode half-cell under test near the preset target test temperature. The target test environment temperature range is -20°C to 40°C, with a temperature control accuracy of ±0.5°C. Through the environmental control module, various practical application scenarios, such as room temperature and low temperature, can be simulated, avoiding interference from temperature fluctuations on the internal electrochemical reactions and lithium-ion migration rates of the battery.

[0103] Furthermore, the electrochemical testing module is electrically connected to the graphite anode half-cell under test, and is used for cyclic testing consisting of constant current charging, first-stage relaxation after charging, constant current discharging, and second-stage relaxation after discharging. The electrochemical testing module can be an electrochemical workstation capable of precisely controlling the charge / discharge rate, voltage range, and relaxation time. During constant current charging, the electrochemical testing module charges the graphite anode half-cell under test at a preset charging rate until it reaches a preset voltage. After charging, the electrochemical testing module stops current output, allowing the battery to enter the first-stage relaxation, which is performed under no-current resting conditions. During the first-stage relaxation, the electrochemical testing module collects voltage-time data at preset time intervals to obtain the first relaxation voltage-time data. After the first-stage relaxation, the electrochemical testing module discharges the graphite anode half-cell under test using constant current discharging until it reaches a preset voltage. The discharge rate increases in a gradient from 0.2C to 0.5C with each cycle. After discharge, the electrochemical testing module stops current output again, allowing the battery to enter the second stage of relaxation, which is also carried out under no-current resting conditions. During the second stage of relaxation, the electrochemical testing module collects voltage-time data at the same time intervals to obtain the second relaxation voltage-time data. The electrochemical testing module outputs the collected first and second relaxation voltage-time data to the data processing module.

[0104] Further, the data processing module receives the first relaxation voltage-time data and the second relaxation voltage-time data sent by the electrochemical testing module. The data processing module can be, for example, a computer device. The data processing module first preprocesses these two sets of data. Preprocessing includes using a wavelet denoising algorithm to remove interference signals with frequencies greater than or equal to 10 Hz, resulting in smoothed first and second relaxation voltage-time data. Then, the data processing module performs differential processing on the smoothed first and second relaxation voltage-time data, respectively, using a numerical differential algorithm to calculate the rate of voltage change over time, obtaining the first and second relaxation differential curves. Next, the data processing module extracts lithium plating characteristic parameters from the first and second relaxation differential curves, as well as the first and second relaxation voltage-time data. These lithium plating characteristic parameters include the occurrence time and amplitude of the characteristic peak of the first relaxation differential curve, the occurrence time and amplitude of the characteristic peak of the second relaxation differential curve, the decay rate of the first relaxation voltage calculated from the first relaxation voltage-time data, and the stability difference of the second relaxation voltage calculated from the second relaxation voltage-time data. The criterion for determining the characteristic peak is: a bulge appears on the relaxation differential curve that first rises and then falls, and the amplitude of the bulge is greater than three times the amplitude of the background noise, where the amplitude of the background noise is the maximum fluctuation value of the relaxation differential curve within the first 10 minutes after the start of relaxation.

[0105] Furthermore, the data processing module determines whether the graphite anode has undergone lithium plating and obtains the amount of lithium plating based on the extracted lithium plating feature parameters. Specifically, the data processing module constructs a feature parameter vector from the extracted lithium plating feature parameters and inputs this feature parameter vector into the lithium plating evaluation model. The lithium plating evaluation model is constructed using a random forest algorithm. During the model training phase, a large number of graphite anode samples with known lithium plating states are used as training samples. Each sample contains the aforementioned multiple feature parameters, as well as the corresponding actual lithium plating level and actual lithium plating amount. The model learns the mapping relationship between feature parameters and lithium plating state through machine learning algorithms. After the model training is completed, for the input feature parameter vector of the sample to be tested, the model calculates based on the learned mapping relationship and outputs the lithium plating level and lithium plating amount of the graphite anode to be tested. The lithium plating level includes four levels: no lithium plating, slight lithium plating, moderate lithium plating, and heavy lithium plating. The lithium plating amount is expressed as the mass of metallic lithium deposited per unit area, with a quantification range of 0 to 5 milligrams per square centimeter. The data processing module outputs the determination results as system output for operators to view or for subsequent analysis.

[0106] Furthermore, in some preferred embodiments, the graphite anode lithium plating evaluation system also includes a sample carrier module. The sample carrier module is used to simultaneously hold multiple graphite anode half-cells to be tested, enabling parallel testing of batch samples. The sample carrier module works in conjunction with the electrochemical testing module, which can sequentially or simultaneously perform charge-discharge relaxation tests and data acquisition on multiple cells, thereby significantly improving testing efficiency. The sample carrier module can be, for example, a sample support frame with multiple testing stations.

[0107] Furthermore, in some preferred embodiments, the graphite anode lithium plating evaluation system also includes an alarm module. The alarm module is connected to the data processing module. When the data processing module outputs a lithium plating level of severe lithium plating, the alarm module automatically issues an audible and visual alarm signal to remind the operator to promptly handle battery samples with a serious risk of lithium plating. The alarm module may be, for example, a buzzer or indicator light; this invention is not specifically limited to these.

[0108] Through the coordinated work of the above modules, the graphite anode lithium plating evaluation system can automatically complete the entire process from environmental control, charge and discharge relaxation test, data acquisition, data processing, feature extraction to lithium plating state determination, realizing the automation, standardization and batch processing of graphite anode lithium plating evaluation.

[0109] Example 1 This embodiment provides a method for evaluating lithium plating on graphite anodes based on multi-stage relaxation voltage analysis. A coin cell is used, and the test environment is room temperature (25 degrees Celsius). The specific steps are as follows.

[0110] Step 1: Commercially available artificial graphite powder was selected as the graphite material to be tested. A binder and conductive agent were added, and the mixture was thoroughly mixed at a mass ratio of 94.5:4:1.5. NMP solvent was then added to form a slurry, which was coated onto copper foil and vacuum-dried at 80°C for 12 hours. The slurry was then pressed into a working electrode. The working electrode had a diameter of 12 mm and a thickness of 85 μm. A lithium metal sheet was used as the counter electrode, with a diameter of 15.6 mm and a thickness of 50 μm. The electrolyte was 1 mol / L lithium hexafluorophosphate, and the solvent was a mixture of ethylene carbonate, diethyl carbonate, and dimethyl carbonate in a volume ratio of 1:1:1. CR2032 coin cell half-cells were assembled in an argon glove box, where the water content and oxygen content were no more than 0.01 ppm. Three copies of each sample were prepared in parallel to ensure test repeatability.

[0111] Step 2: Activate the battery. Place the assembled half-cell in a constant temperature chamber at 25°C ± 0.5°C, and charge and discharge it at a rate of 0.1C within a voltage range of 0.005V to 2.0V for 3 cycles to complete the activation. After activation, let it stand for 30 minutes to remove the passivation layer on the electrode surface.

[0112] Step 3: Perform charge-discharge relaxation tests. Place the activated half-cell in a constant-temperature environment at 25 °C with a temperature control accuracy of ±0.5 °C. Take one cycle of constant-current charging, the first-stage relaxation after charging, constant-current discharging, and the second-stage relaxation after discharging, and conduct 6-cycle tests. The conditions for constant-current charging are: charging rate of 0.5C, charging to a voltage of 2.0V. The relaxation time in the first stage after charging is 1 hour, and relaxation voltage data is collected at a time interval of 200 milliseconds. The conditions for constant-current discharging are: the discharging rates for the 1st to 6th cycles are 0.2C, 0.4C, 0.6C, 0.8C, 1.0C, and 1.2C in sequence, and each time discharge to a voltage of 0.005V. The relaxation time in the second stage after discharging is 3 hours, and relaxation voltage data is collected at a time interval of 200 milliseconds. Use the evaluation system supporting this application, and the electrochemical test module completes charge-discharge and data acquisition, while the constant-temperature control module maintains the environmental temperature stable.

[0113] Step 4: Data processing. Through the data processing module of the evaluation system, use the wavelet denoising algorithm to perform denoising processing on the collected relaxation voltage data in the first stage and the second stage, remove interference signals with frequencies greater than or equal to 10 Hz, and obtain smooth first relaxation voltage-time curve and second relaxation voltage-time curve. Subsequently, perform numerical differentiation on the two curves to obtain the corresponding first relaxation differential curve and second relaxation differential curve.

[0114] Step 5: Extraction of lithium plating characteristic parameters. Through the data processing module, extract 4 characteristic parameters from the differential curve and the relaxation voltage curve. The test data of 3 parallel samples are as shown in the characteristic parameter data table of 3 parallel samples in Example 1 below, that is, Table 1. The determination basis for the characteristic peak is: the convex amplitude is greater than 3 times the background noise amplitude (0.0008 volts per second), and it is determined as an effective characteristic peak. The relative standard deviation is not greater than 5.6%, indicating that the repeatability of this method for testing is good.

[0115]

[0116] Table 1. Characteristic parameter data table of 3 parallel samples in Example 1 To further intuitively display the correspondence between the characteristic peaks of the relaxation differential curve and the lithium plating state in this example, now in combination with Figure 3 it is described as follows. Figure 3 This is the comparison spectrum of the characteristic peaks of the first relaxation differential curve and the second relaxation differential curve of samples with different lithium plating degrees in Example 1 of the present invention.

[0117] Figure 3In the baseline data for the lithium-free standard sample, both the first and second relaxation differential curves were smooth curves, with no protruding structures meeting the judgment criteria. Only background noise signals with an amplitude not exceeding 0.0008 V / s were observed. This indicates that, in the absence of lithium plating, the open-circuit voltage during the relaxation processes after charging and discharging follows normal kinetic decay or recovery patterns, without any abnormal voltage change rate caused by additional electrochemical reactions.

[0118] For the sample with slight lithium plating, its first relaxation differential curve shows a distinct characteristic peak at approximately 28 min, which rises first and then falls, with a peak amplitude of 0.0025 V / s, exceeding three times the background noise threshold (0.0024 V / s). Based on the criteria of this invention, this peak is confirmed as a valid characteristic peak. The appearance of this characteristic peak confirms that the lithium metal deposited during the charging stage undergoes a self-dissolution reaction in the early relaxation phase, causing a non-monotonic recovery in the voltage change rate. Simultaneously, the second relaxation differential curve of this sample with slight lithium plating shows a characteristic peak at approximately 45 min, with a peak amplitude of 0.0032 V / s, also exceeding three times the background noise threshold. This characteristic peak indicates that the lithium metal remaining on the negative electrode surface after discharge continues to undergo oxidation and dissolution reactions during the second relaxation stage, triggering potential disturbances during the voltage recovery process after discharge.

[0119] For samples with moderate lithium plating, the characteristic peaks on the first and second relaxation differential curves appeared significantly earlier than those of samples with slight lithium plating, and the peak amplitudes were significantly increased. This spectral characteristic directly confirms the positive correlation between the degree of lithium plating and the characteristic peaks of the differential curves: the more severe the lithium plating, the more deposited or residual metallic lithium, and the more intense the electrochemical side reactions, resulting in earlier appearance and higher amplitude of the characteristic peaks.

[0120] pass Figure 3 As shown in the comparison graphs, the dual-stage relaxation differential curves employed in this invention can intuitively and visually distinguish different lithium plating states, such as no lithium plating, slight lithium plating, moderate lithium plating, and heavy lithium plating, from two dimensions: lithium deposition characteristics after charging and residual lithium reaction characteristics after discharging. These graphs provide clear visual evidence and mechanistic support for the feature parameter extraction in this embodiment, the rationality verification of the feature peak determination criteria of this invention, and the subsequent training of the lithium plating evaluation model.

[0121] Step Six: Lithium Plating Determination and Quantification. The four characteristic parameters mentioned above are input into the pre-trained lithium plating evaluation model through the lithium plating discrimination module of the evaluation system. The lithium plating evaluation model is constructed using the random forest algorithm, and the model fit is 0.96. The model outputs that the lithium plating level of the graphite anode under test is slight lithium plating, with a lithium plating amount of 0.8 mg / cm².

[0122] Step 7: Verification of evaluation results. The actual lithium precipitation amount of the graphite negative electrode to be measured is determined by inductively coupled plasma emission spectrometry. The actual lithium precipitation amounts of 3 parallel samples are 0.83 mg / cm², 0.82 mg / cm², and 0.84 mg / cm² respectively, and the average value is 0.83 mg / cm². The evaluation error of this method is 3.6%, which is not greater than 5%, and it is determined that the evaluation results are valid.

[0123] After completing the above evaluation, disassemble the coin-type half-cell in an argon glove box, take out the graphite electrode sheet, and observe the microscopic morphology of the surface of the graphite electrode sheet, such as Figure 2 shown. Through the intuitive observation of the electrode sheet morphology, the accuracy of the evaluation results of this method is further verified.

[0124] Example 2 The difference between this example and Example 1 is that the target test environment is low temperature (minus 10 degrees Celsius), and the lithium precipitation state of the graphite negative electrode is tested. The specific adjustments are as follows: In Step 3, place the activated half-cell in a constant temperature environment of minus 10 degrees Celsius, and control the temperature accuracy to be ±0.5 degrees Celsius through the constant temperature control module of the evaluation system. The number of cycles is 8 times, and the discharge rates of the 1st to 8th cycles are 0.2C, 0.4C, 0.6C, 0.8C, 1.0C, 1.2C, 1.4C, and 1.6C in sequence. The relaxation time in the first stage after charging is 1 hour, and the relaxation time in the second stage after discharging is 3 hours. The remaining steps are the same as those in Example 1, and 3 parallel samples are prepared for each sample. After testing, the characteristic parameter data extracted from the 3 parallel samples by the data processing module are as shown in the characteristic parameter data table of the 3 parallel samples in Example 2 below, that is, Table 2.

[0125]

[0126] Table 2. Characteristic parameter data table of 3 parallel samples in Example 2 After inputting the above characteristic parameters into the evaluation model through the lithium precipitation discrimination module, the model outputs a lithium precipitation level of moderate lithium precipitation, and the lithium precipitation amount is 2.3 mg / cm². The actual lithium precipitation amount is determined to be 2.25 mg / cm² by inductively coupled plasma emission spectrometry. The evaluation error is 2.2%, which is not greater than 5%, and the evaluation results are valid. Repeat the test 3 times, and the relative standard deviation is not greater than 1.2%, indicating that this method still has good accuracy and repeatability in low-temperature scenarios.

[0127] Example 3 In this example, the graphite negative electrode lithium precipitation evaluation system supporting this application is used to conduct batch tests on 10 graphite negative electrode samples with different lithium precipitation degrees to verify the automation degree and test efficiency of the system.

[0128] In this embodiment, a lithium plating evaluation system for graphite anodes that is配套 with this application is used to conduct batch tests on 10 graphite anode samples with different degrees of lithium plating (numbered 1 to 10, covering no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating) to verify the automation level, test efficiency, and accuracy of the system. The specific steps are as follows.

[0129] Step 1: Prepare 10 coin-type half-cell samples with different degrees of lithium plating according to the method of Example 1. Among them, Sample 1 has no lithium plating, Samples 2 to 4 have slight lithium plating, Samples 5 to 7 have moderate lithium plating, and Samples 8 to 10 have severe lithium plating.

[0130] Step 2: Place the 10 samples on the sample carrier module of the system respectively, and set the test parameters through the system: temperature 25 °C, number of cycles 6, discharge rate gradient starting from 0.2C and increasing by 0.2C (0.2C to 1.2C), first relaxation time 1 hour, second relaxation time 3 hours, voltage acquisition interval 200 milliseconds.

[0131] Step 3: After starting the system, the electrochemistry test module automatically completes charge and discharge, two-stage relaxation operations, and voltage data acquisition; the data processing module automatically completes noise reduction, differentiation, and extraction of characteristic parameters, outputs the lithium plating grade and lithium plating amount of each sample, and stores all test results. The test results are as shown in the test result table of the 10 samples in Example 3 below, that is, Table 3.

[0132] The test results show that the evaluation error of all samples is not greater than 5%, and the consistency with the manual test results reaches more than 98%, indicating that the evaluation system of this application can achieve automated, efficient, and accurate batch testing, meeting the requirements of industrial scale applications.

[0133]

[0134] Table 3. Test result table of 10 samples in Example 3 Comparative example Adopt the method of the existing technology (Chinese Patent CN115097341B) to conduct lithium plating evaluation on the graphite anode to be tested in Example 1. The specific steps are as follows.

[0135] Step 1: Prepare a coin-type half-cell and activate it using the same method as in Example 1.

[0136] Step 2: Take one cycle of discharging, standing after discharging, charging, and standing after charging, and conduct 6 cycle tests. The discharge rate is the same as that in Example 1, the standing time after discharging is 1 hour, and the standing time after charging is 30 minutes.

[0137] Step 3: Monitor the voltage-time curve during the static period after discharge, differentiate it to obtain the relaxation differential curve, and determine whether lithium plating has occurred based on the presence or absence of characteristic peaks.

[0138] Test results show that the existing method can only determine the presence of lithium plating in the graphite anode, but cannot quantify the amount of lithium plating. Furthermore, no characteristic peaks were detected in the first and second cycles (discharge rates of 0.2C and 0.4C), leading to a misjudgment of no lithium plating. In contrast, the method of this application detected a slight lithium plating signal in the aforementioned cycles, indicating that the method of this application has higher early detection sensitivity and better accuracy.

[0139] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0140] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Since these modifications and variations fall within the scope of the claims and their equivalents, this invention also intends to include these modifications and variations.

[0141] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating lithium plating on graphite anodes, characterized in that, include: The voltage-time data of the first stage relaxation of the graphite anode half-cell under test is collected to obtain the first relaxation voltage-time data, and the voltage-time data of the second stage relaxation of the graphite anode half-cell under test is collected to obtain the second relaxation voltage-time data. The graphite anode half-cell under test is placed in a preset target test environment and the test is performed in a cycle of constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging. Differential processing is performed on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve; Lithium plating characteristic parameters are extracted from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data. Based on the extracted lithium plating characteristic parameters, it is determined whether the graphite anode has undergone lithium plating and the amount of lithium plating is obtained.

2. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, The lithium plating characteristic parameters include a first characteristic parameter, a second characteristic parameter, a third characteristic parameter, and a fourth characteristic parameter; The first characteristic parameter includes the occurrence time and amplitude of the characteristic peak of the first relaxation differential curve; The second characteristic parameter includes the occurrence time and amplitude of the characteristic peak of the second relaxation differential curve; The third characteristic parameter includes the first relaxation voltage decay rate calculated based on the first relaxation voltage-time data. The fourth characteristic parameter includes the second relaxation voltage stability difference calculated based on the second relaxation voltage-time data.

3. The method for evaluating lithium plating on graphite anodes according to claim 2, characterized in that, The characteristic peaks of the first relaxation differential curve and the second relaxation differential curve are determined based on a preset judgment criterion. The judgment criterion is that a bulge appears on the relaxation differential curve that first rises and then falls, and the amplitude of the bulge is greater than 3 times the amplitude of the background noise. The relaxation differential curve includes the first relaxation differential curve and the second relaxation differential curve.

4. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, Before acquiring the first relaxation voltage-time data of the graphite anode half-cell under test, the method further includes: A coin cell was assembled using the graphite material to be tested as the working electrode and a lithium metal sheet as the counter electrode and reference electrode. The assembled coin cell was subjected to charge-discharge activation treatment to obtain the graphite negative electrode half-cell under test.

5. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, The constant current charging rate is 0.5C to 5C, charging to a voltage of 2.0V; the first stage relaxation time after charging is 0.5 hours to 2 hours, and the relaxation process is a current-free resting state; the constant current discharging rate is a gradient increasing rate starting from 0.2C and increasing by 0.2C to 0.5C in each cycle, discharging to a voltage of 0.005V; the second stage relaxation time after discharging is 2 hours to 3 hours, and the relaxation process is a current-free resting state.

6. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, Before performing differentiation processing on the first relaxation voltage-time data and the second relaxation voltage-time data, the method further includes: The first relaxation voltage-time data and the second relaxation voltage-time data are subjected to noise reduction processing.

7. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, The process of determining whether the graphite anode has undergone lithium plating and obtaining the amount of lithium plating based on the extracted lithium plating characteristic parameters includes: The extracted lithium plating characteristic parameters are input into the lithium plating evaluation model, which then outputs the lithium plating grade and the amount of lithium plating.

8. The method for evaluating lithium plating on graphite anodes according to claim 7, characterized in that, The lithium plating evaluation model is constructed using a machine learning algorithm. The lithium plating level includes four levels: no lithium plating, slight lithium plating, moderate lithium plating, and severe lithium plating. The quantitative range of the amount of lithium plating is 0 to 5 milligrams per square centimeter.

9. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, The target test environment has a temperature range of -20 degrees Celsius to 40 degrees Celsius, with a temperature control accuracy of ±0.5 degrees Celsius.

10. The method for evaluating lithium plating on graphite anodes according to claim 1, characterized in that, After obtaining the amount of lithium plating, the method further includes: The obtained amount of lithium plating is compared with the actual amount of lithium plating in the graphite anode half-cell under test determined by inductively coupled plasma atomic emission spectrometry to obtain the evaluation error. If the evaluation error is greater than 5%, then after adjusting at least one of the first stage relaxation time after charging, the second stage relaxation time after discharging, and the voltage acquisition interval, proceed to the step of acquiring the voltage-time data of the first stage relaxation of the graphite anode half-cell under test to obtain the first relaxation voltage-time data, so as to re-evaluate the graphite anode half-cell under test.

11. A graphite anode lithium plating evaluation system, characterized in that, include: The environmental control module is used to control the graphite anode half-cell under test to be in a preset target test environment. The electrochemical testing module is used to perform cyclic testing with constant current charging, first stage relaxation after charging, constant current discharging, and second stage relaxation after discharging as one cycle. It collects voltage-time data of the first stage relaxation to obtain first relaxation voltage-time data, and collects voltage-time data of the second stage relaxation to obtain second relaxation voltage-time data. The data processing module is used to perform differential processing on the first relaxation voltage-time data and the second relaxation voltage-time data respectively to obtain the first relaxation differential curve and the second relaxation differential curve; extract lithium plating characteristic parameters from the first relaxation differential curve, the second relaxation differential curve, the first relaxation voltage-time data, and the second relaxation voltage-time data; and determine whether the graphite anode has lithium plating and obtain the amount of lithium plating based on the extracted lithium plating characteristic parameters.

Images