Battery lithium plating quality determination method, module and model training method, module

Abstract

The present disclosure provides a method and module for determining the quality of lithium deposition in a battery, as well as a model training method and module, which can be applied to the field of power battery safety detection technology. The determination method includes: obtaining an electrochemical curve of a target battery during the charge / discharge process; extracting a target electrochemical feature from the electrochemical curve, wherein the target electrochemical feature represents an electrochemical feature related to the quality of lithium deposition, and the target electrochemical feature includes time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve; inputting the target electrochemical feature into a trained target lithium deposition quality determination model to obtain the lithium deposition quality of the target battery, wherein the lithium deposition quality is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

Description

Technical Field

[0001] The present disclosure relates to the technical field of power battery safety detection, and more specifically, to a method and module for determining the quality of lithium deposition in a non-destructive battery, as well as a model training method and module. Background Art

[0002] With the rapid development of new energy vehicles and distributed energy storage systems, lithium-ion batteries have become a key energy device. However, battery safety has long been a factor hindering their widespread application. Consequently, the safety risks associated with lithium deposition within batteries have gradually attracted widespread attention.

[0003] In the process of realizing the concept of the present disclosure, the inventors discovered that the detection of lithium deposition status in the related art has at least technical problems of low sensitivity and low accuracy. Summary of the Invention

[0004] In view of this, the present disclosure provides a method and module for determining the quality of battery lithium deposition, as well as a method and module for model training.

[0005] One aspect of the present disclosure provides a method for determining the lithium deposition quality of a battery, comprising: obtaining an electrochemical curve of a target battery during a charge / discharge process; extracting a target electrochemical feature from the electrochemical curve, wherein the target electrochemical feature characterizes an electrochemical feature related to the lithium deposition quality, and the target electrochemical feature includes time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve; inputting the target electrochemical feature into a trained target lithium deposition quality determination model to obtain the lithium deposition quality of the target battery, wherein the lithium deposition quality is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

[0006] According to an embodiment of the present disclosure, the electrochemical curve includes a constant voltage current-time curve and a constant current voltage-time curve, the predetermined current interval includes a first current and a second current, and extracting the target electrochemical characteristics from the electrochemical curve includes: extracting a first moment corresponding to the first current and a second moment corresponding to the second current from the constant voltage current-time curve based on the first current and the second current; calculating the difference between the first moment and the second moment to obtain time difference data corresponding to the predetermined current interval; and calculating integral data corresponding to the predetermined current interval based on the first moment and the second moment.

[0007] Another aspect of the present disclosure provides a model training method, comprising:

[0008] Obtain a sample database, wherein the sample database includes a sample electrochemical feature set and a sample lithium deposition state set of multiple sample batteries corresponding to different predetermined aging conditions, the sample electrochemical feature set includes multiple sample electrochemical features, and the sample lithium deposition state set includes multiple sample lithium deposition state information; use the multiple sample lithium deposition state information and the multiple sample electrochemical features to train an initial model to obtain correlation information and an intermediate model between each sample electrochemical feature and the multiple lithium deposition state information; determine a target electrochemical feature from the sample electrochemical feature set based on the correlation information between each sample electrochemical feature and the multiple lithium deposition state information; train the intermediate model based on the target electrochemical feature to obtain a target lithium deposition quality determination model, wherein the target lithium deposition quality determination model is the trained target lithium deposition quality determination model.

[0009] According to an embodiment of the present disclosure, obtaining a sample database includes: performing cyclic charge / discharge processing on multiple sample batteries according to a first predetermined rate to obtain electrochemical curves of the multiple sample batteries; extracting multiple sample electrochemical characteristics from the electrochemical curves of the multiple sample batteries to obtain a sample electrochemical characteristic set; analyzing the multiple sample batteries to obtain a sample lithium deposition state set; and obtaining a sample database based on the sample electrochemical characteristic set and the sample lithium deposition state set.

[0010] According to an embodiment of the present disclosure, multiple sample batteries are analyzed to obtain a sample lithium deposition state set, including: performing scanning electron microscope observation and nuclear magnetic resonance analysis on the negative electrode pieces of the multiple sample batteries to obtain lithium deposition structure information; using titration gas chromatography technology to perform quantitative analysis on the negative electrode pieces of the multiple sample batteries to obtain lithium deposition weight information; based on the lithium deposition structure information, the lithium deposition weight information is verified to obtain the sample lithium deposition state set.

[0011] According to an embodiment of the present disclosure, the initial battery includes a first initial battery, a second initial battery, a third initial battery, and a fourth initial battery, and the multiple sample batteries are obtained by the following operations: maintaining the initial state of the first initial battery to obtain a first sample battery; performing a cyclic overcharge / discharge simulation on the second initial battery at a second predetermined rate to obtain a second sample battery; performing a cyclic charge / discharge simulation below the operating temperature on the third initial battery at the second predetermined rate to obtain a third sample battery; and performing a cyclic overcharge / discharge simulation below the operating temperature on the fourth initial battery at the second predetermined rate to obtain a fourth sample battery.

[0012] According to an embodiment of the present disclosure, multiple sample electrochemical characteristics include: voltage-related characteristics of the constant current charging process, current-related characteristics of the constant voltage charging process, diffusion resistance coefficient of the intermittent current interruption process, DC internal resistance and charging cut-off voltage. The current-related characteristics of the constant voltage charging process include time difference data and integral data corresponding to a predetermined current interval. According to the correlation information between each sample electrochemical characteristic and multiple lithium deposition state information, the target electrochemical characteristic is determined from the sample electrochemical characteristic set, including: sorting the correlation information between each sample electrochemical characteristic and multiple lithium deposition state information to obtain a sorting result; according to the sorting result, the sample electrochemical characteristics whose correlation information meets the predetermined threshold are screened from the sample electrochemical characteristic set to determine them as the target electrochemical characteristics.

[0013] According to an embodiment of the present disclosure, an initial model is trained using multiple sample lithium deposition state information and multiple sample electrochemical characteristics to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information, including: splitting the multiple sample lithium deposition state information and the multiple sample electrochemical characteristics into a verification sample set and a training sample set according to a predetermined ratio; using the training sample set to perform initial training on the initial model; using the verification sample set to optimize the parameters of the model after the initial training until the parameters meet predetermined conditions to obtain an intermediate model and correlation information.

[0014] Another aspect of the present disclosure provides a device for determining the quality of lithium deposition in a battery, comprising:

[0015] A first acquisition module is used to obtain the electrochemical curve of the target battery during the charge / discharge process;

[0016] An extraction module is used to extract target electrochemical characteristics from the electrochemical curve, wherein the target electrochemical characteristics represent electrochemical characteristics related to the quality of lithium deposition, and the target electrochemical characteristics include time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve;

[0017] The first obtaining module is used to input the target electrochemical characteristics into the trained target lithium deposition mass determination model to obtain the lithium deposition mass of the target battery, wherein the lithium deposition mass is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

[0018] Another aspect of the present disclosure provides a model training device, comprising:

[0019] a second acquisition module, configured to acquire a sample database, wherein the sample database includes a sample electrochemical feature set and a sample lithium deposition state set of a plurality of sample batteries corresponding to different predetermined aging conditions, the sample electrochemical feature set including a plurality of sample electrochemical features, and the sample lithium deposition state set including a plurality of sample lithium deposition state information;

[0020] A training module is used to train the initial model using the lithium deposition state information and electrochemical characteristics of multiple samples to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information;

[0021] a determination module, configured to determine a target electrochemical feature from a set of sample electrochemical features based on correlation information between the electrochemical features of each sample and a plurality of lithium deposition state information;

[0022] The second obtaining module is used to train the intermediate model according to the target electrochemical characteristics to obtain the target lithium deposition mass determination model, wherein the target lithium deposition mass determination model is the trained target lithium deposition mass determination model.

[0023] According to the embodiments of the present disclosure, by extracting target electrochemical characteristics from the electrochemical curve of the target battery during the charge / discharge process and inputting the target electrochemical characteristics into a trained target lithium deposition quality determination model, the mass of metallic lithium deposited on the graphite surface of the target battery can be obtained. Since the lithium deposition quality can be output by inputting the target electrochemical characteristics into the model, the specific lithium deposition quality can be obtained without being restricted by the battery operating conditions and without damaging the battery. This solves the problem that the related technology can only identify a larger amount of lithium deposition after the middle period, thereby improving the sensitivity and accuracy of detecting the lithium deposition quality of the target battery, and then using the obtained lithium deposition quality to more accurately and efficiently predict or monitor the safety of the target battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0025] Figure 1 A flow chart of a method for determining the quality of lithium deposition in a battery according to an embodiment of the present disclosure is schematically shown.

[0026] Figure 2 The flowchart of the model training method according to an embodiment of the present disclosure is schematically shown.

[0027] Figure 3 Schematic diagrams of scanning electron microscope images of negative electrode sheets of sample batteries in different lithium deposition states according to an embodiment of the present disclosure are shown.

[0028] Figure 4A schematic diagram of an electrochemical curve according to an embodiment of the present disclosure is schematically shown.

[0029] Figure 5 A schematic diagram schematically illustrates the diffusion resistance coefficients of sample batteries in different lithium deposition states according to an embodiment of the present disclosure.

[0030] Figure 6 A schematic diagram schematically illustrates the lithium deposition quality of sample batteries in different lithium deposition states before and after cycle testing according to an embodiment of the present disclosure.

[0031] Figure 7 Schematic diagram of sample batteries in different lithium deposition states according to an embodiment of the present disclosure 7 Li spectrum.

[0032] Figure 8 The figure schematically shows the hydrogen content of a sample battery according to an embodiment of the present disclosure.

[0033] Figure 9 The figure schematically shows the experimental results of the titration gas chromatography method and the weighing results on a balance according to an embodiment of the present disclosure.

[0034] Figure 10 The following schematically shows a result diagram of sorting association information according to an embodiment of the present disclosure.

[0035] Figure 11 A block diagram of a device for determining battery lithium deposition quality according to an embodiment of the present disclosure is schematically shown.

[0036] Figure 12 A block diagram of a device for determining battery lithium deposition quality according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0037] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the detailed description below, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0038] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise," "include," etc. used herein indicate the presence of the features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0039] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0040] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0041] In the embodiments of this disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of all data involved (including, but not limited to, user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and maintain the security of user personal information and network security.

[0042] The lithium deposition reaction in lithium-ion batteries differs from the conventional insertion process. Even a small amount of lithium deposition can cause capacity loss and increase internal resistance. As dendrites grow further, they can pierce the separator, causing short circuits, fires, or even explosions. These safety incidents seriously affect the reliability of lithium batteries in practical applications and user safety.

[0043] In order to ensure the safety of lithium-ion batteries, relevant technologies have made a lot of explorations in battery safety monitoring and early warning, mainly focusing on non-invasive in-situ physical characterization of lithium and safety characterization based on external electrical signals. Among them, in-situ physical characterization can be used to identify the process and mechanism of metallic lithium and lithium precipitation reaction in the complex components of the battery, mainly using synchrotron X-ray diffraction, surface-enhanced Raman, electron paramagnetic resonance and nuclear magnetic resonance technology. However, these in-situ physical characterization instruments and equipment are relatively complex, which will cause the battery structure to deviate from the battery structure of actual application. At the same time, these characterizations rely on high-cost experimental equipment, and it is difficult to achieve online monitoring, which makes it difficult to meet the safety testing requirements of power batteries under complex working conditions.

[0044] In addition, characterization based on external electrical signals of the battery, such as ohmic drop analysis, differential voltage analysis (DVA), relaxation voltage analysis (VRP) and incremental capacity analysis (ICA), etc., are simpler, less expensive and more portable than the above-mentioned in-situ physical characterization, making them more suitable for industrial applications. However, they still have certain limitations. For example, there is a lack of diagnosis for early lithium deposition, and only mid-to-late lithium deposition can be detected. It still relies on external voltage signals, making it difficult to effectively detect the lithium deposition state inside the battery; secondly, the repeatability of data testing is poor, the accuracy is low, and the test results are greatly affected by the battery working conditions, battery structure and battery system type; at the same time, it is difficult to achieve quantitative lithium deposition based on electrochemical test characterization, and the detection sensitivity is low.

[0045] In view of this, an embodiment of the present disclosure provides a method for determining the quality of lithium deposition in a battery, comprising:

[0046] Obtain an electrochemical curve of a target battery during a charge / discharge process; extract a target electrochemical feature from the electrochemical curve, wherein the target electrochemical feature characterizes an electrochemical feature related to lithium deposition quality, and the target electrochemical feature includes time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve; input the target electrochemical feature into a trained target lithium deposition quality determination model to obtain the lithium deposition quality of the target battery, wherein the lithium deposition quality is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

[0047] Figure 1 The flowchart of the method for determining the quality of lithium deposition in a battery according to an embodiment of the present disclosure is schematically shown.

[0048] like Figure 1 As shown, the method includes operations S110 to S130.

[0049] In operation S110 , an electrochemical curve of a target battery during a charge / discharge process is acquired.

[0050] In operation S120 , a target electrochemical feature is extracted from the electrochemical curve.

[0051] In operation S130 , the target electrochemical characteristics are input into the trained target lithium deposition quality determination model to obtain the lithium deposition quality of the target battery.

[0052] According to an embodiment of the present disclosure, the electrochemical curve of the target battery during the charge / discharge process is acquired in real time by corresponding testing equipment, and may include a constant voltage current-time curve and a constant current voltage-time curve.

[0053] According to an embodiment of the present disclosure, the target electrochemical characteristics characterize the electrochemical characteristics related to the quality of lithium deposition, and the target electrochemical characteristics include time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve, for example: the predetermined current interval is 2.5A~3.5A.

[0054] According to the embodiments of the present disclosure, the lithium plating reaction of the target battery is different from the conventional embedding process. When the electrochemical reduction rate of lithium ions on the graphite surface of the target battery does not match the diffusion transmission rate, the lithium ions will be deposited on the graphite surface in the form of metallic lithium, and the lithium plating mass is the mass of metallic lithium produced by the lithium plating reaction.

[0055] According to an embodiment of the present disclosure, the target lithium deposition quality determination model can be obtained by training based on a machine learning method, wherein the independent variables in the training process are various electrochemical characteristics including the target electrochemical characteristics, and the dependent variable is the lithium deposition quality.

[0056] According to embodiments of the present disclosure, machine learning methods include, but are not limited to, supervised learning algorithms, unsupervised learning algorithms, and deep learning algorithms, such as support vector machines, deep learning neural networks, ensemble learning (Adaptive Boosting), and decision tree models.

[0057] According to the embodiments of the present disclosure, for lithium-ion batteries, when the lithium deposition mass increases, there is a risk that the generated lithium dendrites will pierce the battery separator, causing a short circuit and triggering a more serious thermal runaway problem. At the same time, the threshold temperature for thermal runaway will also be significantly lowered. Therefore, the lithium deposition mass output by the target lithium deposition mass determination model can be used to assess the safety of the target battery.

[0058] According to the embodiments of the present disclosure, by extracting target electrochemical characteristics from the electrochemical curve of the target battery during the charge / discharge process and inputting the target electrochemical characteristics into a trained target lithium deposition quality determination model, the mass of metallic lithium deposited on the graphite surface of the target battery can be obtained. Since the lithium deposition quality can be output by inputting the target electrochemical characteristics into the model, the specific lithium deposition quality can be obtained without being restricted by the battery operating conditions and without damaging the battery. This solves the problem that the related technology can only identify a larger amount of lithium deposition after the middle period, thereby improving the sensitivity and accuracy of detecting the lithium deposition quality of the target battery, and then using the obtained lithium deposition quality to more accurately and efficiently predict or monitor the safety of the target battery.

[0059] According to an embodiment of the present disclosure, the electrochemical curve includes a constant voltage current-time curve and a constant current voltage-time curve, the predetermined current interval includes a first current and a second current, and extracting the target electrochemical characteristics from the electrochemical curve includes: extracting a first moment corresponding to the first current and a second moment corresponding to the second current from the constant voltage current-time curve based on the first current and the second current; calculating the difference between the first moment and the second moment to obtain time difference data corresponding to the predetermined current interval; and calculating integral data corresponding to the predetermined current interval based on the first moment and the second moment.

[0060] For example, the first current in the predetermined current interval is 2.5A, and the second current is 3.5A. According to the constant voltage current-time curve, the current is 2.5A at 3000s, that is, the first moment is 3000s, and the current is 3.5A at 3500s, that is, the first moment is 3500s. The difference between the first moment and the second moment is 500s, which is the time difference data. The curve in this interval is integrated to obtain the integral data corresponding to the predetermined current interval.

[0061] According to an embodiment of the present disclosure, taking the random forest model as an example, statistical methods can be used to convert the target electrochemical features into numerical vectors, such as calculating the maximum value, minimum value, mean, standard deviation, skewness, etc., or using dimensionality reduction techniques (such as PCA) to extract the main features.

[0062] Figure 2 The flowchart of the model training method according to an embodiment of the present disclosure is schematically shown.

[0063] like Figure 2 As shown, the method includes operations S210 to S240.

[0064] In operation S210 , a sample database is acquired.

[0065] In operation S220 , the initial model is trained using the lithium deposition state information of the plurality of samples and the electrochemical characteristics of the plurality of samples to obtain correlation information between the electrochemical characteristics of each sample and the plurality of lithium deposition state information and an intermediate model.

[0066] In operation S230 , a target electrochemical feature is determined from the sample electrochemical feature set according to association information between the electrochemical features of each sample and a plurality of lithium deposition state information.

[0067] In operation S240 , the intermediate model is trained according to the target electrochemical characteristics to obtain a target lithium deposition quality determination model.

[0068] According to an embodiment of the present disclosure, the sample database includes a sample electrochemical feature set and a sample lithium deposition state set of multiple sample batteries corresponding to different predetermined aging conditions. The sample electrochemical feature set includes multiple sample electrochemical features, and the sample lithium deposition state set includes multiple sample lithium deposition state information.

[0069] According to an embodiment of the present disclosure, the sample batteries may include batteries in different lithium deposition states, for example, no lithium deposition state, slight lithium deposition state, moderate lithium deposition state, and severe lithium deposition state.

[0070] In order to better reflect the structure of the negative electrode of the sample battery with different lithium deposition states, the following Figure 3 Make a comparison.

[0071] Figure 3 Schematic diagrams of scanning electron microscope images of negative electrode sheets of sample batteries in different lithium deposition states according to an embodiment of the present disclosure are shown.

[0072] like Figure 3 As shown in the figure, a sample battery was disassembled, the graphite negative electrode sheet was removed, and the structure of the negative electrode was characterized using a scanning electron microscope (SEM). The SEM observations revealed the structural characteristics of the deposited metallic lithium, including the morphology and distribution of the deposited lithium. A indicates no lithium deposition, B indicates a small amount of lithium deposition, C indicates moderate lithium deposition, and D indicates severe lithium deposition. This shows that the lithium deposition quality and structure of the negative electrode sheets differ depending on the lithium deposition state.

[0073] According to an embodiment of the present disclosure, the correlation information between the electrochemical characteristics of each sample and multiple lithium deposition state information represents the importance of the sample electrochemical characteristics to the lithium deposition state information, and different sample electrochemical characteristics have different predictive capabilities for the lithium deposition quality output by the target lithium deposition quality determination model.

[0074] According to the embodiments of the present disclosure, the obtained initial sample electrochemical characteristics can be preprocessed to remove missing values ​​and outliers to obtain the sample electrochemical characteristics. At the same time, the scales of the electrochemical characteristics of different samples vary greatly. To prevent the model from being affected by scale differences, the electrochemical characteristics of multiple samples can be standardized, such as with zero mean and unit variance, or normalized, such as scaling all features to the interval [0, 1], to ensure more stable and efficient model training.

[0075] According to an embodiment of the present disclosure, one-hot encoding or embedding should also be used to convert the electrochemical features of the sample into numerical features so that the machine learning algorithm can process them.

[0076] According to an embodiment of the present disclosure, the sample lithium deposition state information may include sample lithium deposition mass information and sample lithium deposition structure information. The target electrochemical characteristic is the electrochemical characteristic that has the greatest influence on the lithium deposition state information among the electrochemical characteristics of each sample.

[0077] According to an embodiment of the present disclosure, the hyperparameters of the intermediate model can be optimized through network search or random search, and the K-fold cross-validation method can be used to verify the model and evaluate the generalization ability of the model to avoid overfitting or underfitting of the model.

[0078] According to an embodiment of the present disclosure, the initial model is trained using lithium deposition state information of multiple samples and electrochemical characteristics of multiple samples in a sample database to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information, and the target electrochemical characteristics are determined. The intermediate model is trained based on the target electrochemical characteristics to obtain a target lithium deposition quality determination model, thereby establishing a correlation between the electrochemical characteristics and the lithium deposition quality, so that the lithium deposition quality can be output by inputting the target electrochemical characteristics into the model, and the specific lithium deposition quality can be obtained without being restricted by the battery operating conditions and without damaging the battery, thereby solving the problem that the related technology can only identify a larger amount of lithium deposition after the middle stage, thereby improving the sensitivity and accuracy of detecting the lithium deposition quality of the target battery, and then being able to use the obtained lithium deposition quality to more accurately and efficiently predict or monitor the safety of the target battery.

[0079] According to an embodiment of the present disclosure, obtaining a sample database includes: performing cyclic charge / discharge processing on multiple sample batteries according to a first predetermined rate to obtain electrochemical curves of the multiple sample batteries; extracting multiple sample electrochemical characteristics from the electrochemical curves of the multiple sample batteries to obtain a sample electrochemical characteristic set; analyzing the multiple sample batteries to obtain a sample lithium deposition state set; and obtaining a sample database based on the sample electrochemical characteristic set and the sample lithium deposition state set.

[0080] According to an embodiment of the present disclosure, multiple sample electrochemical characteristics include: voltage-related characteristics of the constant current charging process, current-related characteristics of the constant voltage charging process, diffusion resistance coefficient of the intermittent current interruption process, DC internal resistance and charging cut-off voltage, and the current-related characteristics of the constant voltage charging process include time difference data and integral data corresponding to a predetermined current interval.

[0081] For example: The electrochemical characteristics of the constant current section include different voltage intervals (V i To V i+1 ) of the charge (ΔQ), the slope of the current curve (dV / dt(t i ))、Integral area of ​​current curve(S i ) and the average voltage V a , the electrochemical characteristics of the constant voltage section include different current intervals (I i to Ii+1 ) time difference (Δt), voltage curve slope (dI / dt(t i ))、Integral area of ​​voltage curve(S v ) and average current I a The diffusion resistance coefficient during the intermittent current interruption process is the diffusion resistance coefficient (k) obtained by the intermittent current interruption method (ICI) test, the DC internal resistance (R) is obtained by the DC internal resistance (DCIR) test near the fully charged state, and the charge cut-off voltage is the charge cut-off voltage after the relaxation is completed (EOCV).

[0082] For example, the constant current charge rate can be 0.7 C. When the battery is charged to 4.2 V, an intermittent current interruption test is performed. This involves applying a 500 mA charge current for 5 minutes, followed by a 10-second relaxation period, and cycling until the voltage reaches the upper limit. The constant current charge cutoff voltage is 4.45 V, followed by constant voltage charging until the current decreases to 100 mA. A DC internal resistance test is then performed at full charge, applying a 5A current for 1 second. The relaxation time can be 5 minutes.

[0083] Figure 4 A schematic diagram of an electrochemical curve according to an embodiment of the present disclosure is schematically shown.

[0084] like Figure 4 As shown, the horizontal axis is time, the vertical axis is current or voltage, and the electrochemical curves of the constant current charging stage, the intermittent current interruption stage, the constant voltage charging stage and the DC internal resistance test stage are shown in sequence. a =S v / ∆t v , V a =S a / ∆t i .

[0085] Figure 5 A schematic diagram schematically illustrates the diffusion resistance coefficients of sample batteries in different lithium deposition states according to an embodiment of the present disclosure.

[0086] like Figure 5 As shown, the horizontal axis represents different lithium deposition states, and the vertical axis represents the diffusion resistance coefficient.

[0087] According to an embodiment of the present disclosure, the electrochemical characteristics of the sample may further include the battery's electrical energy density obtained from the integral or gradient of the voltage-capacity curve, or the decay rate and decay capacity extracted based on battery cycle charging data.

[0088] According to an embodiment of the present disclosure, the initial battery includes a first initial battery, a second initial battery, a third initial battery, and a fourth initial battery, and the plurality of sample batteries are obtained by the following operations: maintaining the initial state of the first initial battery to obtain a first sample battery; performing an overcharge / discharge cycle simulation on the second initial battery at a second predetermined rate to obtain a second sample battery; performing a charge / discharge cycle simulation on the third initial battery at a temperature below the operating temperature at the second predetermined rate to obtain a third sample battery; and performing an overcharge / discharge cycle simulation on the fourth initial battery at a temperature below the operating temperature at the second predetermined rate to obtain a fourth sample battery.

[0089] According to an embodiment of the present disclosure, the initial battery can be a plurality of commercial power batteries with a battery capacity of 5 Ah, the positive electrode material of the battery is lithium cobalt oxide and the negative electrode material is graphite, and the initial battery includes a plurality of different lithium deposition states.

[0090] According to an embodiment of the present disclosure, the first sample battery does not need to undergo aging simulation and can be kept in its initial factory state.

[0091] According to an embodiment of the present disclosure, the overcharge voltage of the second sample battery may be 4.6 V, 4.65 V, 4.7 V, etc.

[0092] According to an embodiment of the present disclosure, the predetermined rate can be 0.7 C charging and 0.2 C discharging. Accordingly, the electrochemical characteristics of the sample can be extracted from the electrochemical curve by performing an intermittent current interruption test when the constant current charging stage is close to the full charge state. The battery is first charged at 500 mA for 5 minutes, followed by a 10-second relaxation, and the cycle is repeated until the voltage reaches the upper limit. The voltage response to time is recorded to extract the lithium ion diffusion resistance coefficient. After reaching 100% state of charge (SOC), a current of 5 A is applied for 1 second to measure the DC internal resistance of the battery. The battery is left to stand for 5 minutes, and the open circuit voltage is recorded.

[0093] According to an embodiment of the present disclosure, the number of charge / discharge cycles of each sample battery can include 0 cycles, 5 cycles, 10 cycles, and 30 cycles. Cyclic testing can be performed on batteries with different lithium deposition states to verify the degree of impact of cyclic charge / discharge on the lithium deposition quality of the battery.

[0094] Figure 6 A schematic diagram schematically illustrates the lithium deposition quality of sample batteries in different lithium deposition states before and after cycle testing according to an embodiment of the present disclosure.

[0095] like Figure 6 As shown in the figure, the horizontal axis is the lithium deposition state, and the vertical axis is the lithium deposition quality. By comparison, it can be seen that the lithium deposition quality of sample batteries with different lithium deposition states remains basically unchanged after the cycle test, that is, the impact of cyclic charge / discharge on lithium deposition quality is relatively small.

[0096] According to an embodiment of the present disclosure, by simulating predetermined aging conditions on initial batteries in different lithium deposition states at a predetermined rate, a plurality of sample batteries can be obtained, and the changes in lithium deposition quality of the plurality of sample batteries can be compared to obtain the influence of the predetermined aging conditions, i.e., the battery health, on the lithium deposition state.

[0097] According to an embodiment of the present disclosure, multiple sample batteries are analyzed to obtain a sample lithium deposition state set, including: performing scanning electron microscope observation and nuclear magnetic resonance analysis on the negative electrode pieces of the multiple sample batteries to obtain lithium deposition structure information; using titration gas chromatography technology to perform quantitative analysis on the negative electrode pieces of the multiple sample batteries to obtain lithium deposition weight information; based on the lithium deposition structure information, the lithium deposition weight information is verified to obtain the sample lithium deposition state set.

[0098] According to the embodiments of the present disclosure, the microscopic morphology of the negative electrode of a sample battery can be observed using a scanning electron microscope (SEM). 0.5 cm × 0.5 cm pieces can be randomly cut from electrodes with different lithium deposition states to preliminarily analyze the growth of lithium deposition in the negative electrode. The change in the proportion of lithium metal in the lithium-containing species can be measured using solid-state nuclear magnetic resonance.

[0099] Figure 7 Schematic diagram of sample batteries in different lithium deposition states according to an embodiment of the present disclosure 7 Li spectrum.

[0100] like Figure 7 As shown in the figure, 3mg~4mg of powder was scraped from the surface of the negative electrode in different lithium deposition states and tested. 7 The Li spectrum is used to preliminarily analyze the proportion of metallic lithium in lithium-containing species based on the difference in chemical shift, that is, to determine the lithium precipitation state information. 7 The Li spectrum shows peaks corresponding to metallic lithium, as well as other peaks corresponding to lithiated graphite precursors and lithium salts. By comparison, the peak value of the sample battery corresponding to the severe lithium deposition state (D) is the highest.

[0101] According to the embodiments of the present disclosure, to further quantitatively analyze the "dead lithium" precipitated from the sample battery, the sample battery can be disassembled in a glove box, and the negative electrode and separator can be cut into strips. These strips are then divided into 10 equal portions, each placed in a 1L headspace vial and sealed with a rubber stopper and aluminum ring, maintaining the pressure in the glove box at approximately 1.2 atm. 10 mL of water is added to each headspace vial using a syringe. After thorough shaking to allow it to react with the electrode and separator, 2 mL of gas is extracted using a sampling needle and injected into the inlet of a gas chromatograph to measure the hydrogen content.

[0102] According to an embodiment of the present disclosure, titration gas chromatography (TGC) is used to quantitatively analyze the negative electrode samples in each lithium deposition state to obtain the specific content range of metallic lithium deposited in different lithium deposition states, and multiple samples are measured for each lithium deposition state to verify consistency.

[0103] Figure 8 The figure schematically shows the hydrogen content of a sample battery according to an embodiment of the present disclosure.

[0104] like Figure 8 As shown in the figure, the horizontal axis is the lithium deposition mass, and the vertical axis is the detected hydrogen area. It can be seen that the higher the lithium deposition mass, the larger the hydrogen area.

[0105] In order to verify the rationality of the TGC method in the quantitative analysis of lithium deposition, the balance weighing results of the lithium deposition mass can be compared with the quantitative results of the TCG method through experiments.

[0106] Figure 9 The figure schematically shows the experimental results of the titration gas chromatography method and the weighing results on a balance according to an embodiment of the present disclosure.

[0107] like Figure 9 As shown in the figure, the horizontal axis is the sample number of the sample battery, and the vertical axis is the mass of the deposited metallic lithium. By comparison, it can be seen that the mass of metallic lithium obtained by titration gas chromatography is roughly equal to the mass of metallic lithium weighed on a balance. Therefore, the TGC method has a certain rationality in the quantitative analysis of lithium deposition.

[0108] According to the embodiments of the present disclosure, lithium deposition structural information is obtained by nuclear magnetic resonance analysis, and lithium deposition mass information is obtained by quantitative analysis through titration gas chromatography technology, thereby achieving qualitative and quantitative analysis of sample batteries respectively, and obtaining a more comprehensive sample database to ensure the accuracy of model training.

[0109] According to an embodiment of the present disclosure, a target electrochemical feature is determined from a set of sample electrochemical features based on correlation information between the electrochemical features of each sample and multiple lithium deposition state information, including: sorting the correlation information between the electrochemical features of each sample and multiple lithium deposition state information to obtain a sorting result; based on the sorting result, screening, from the set of sample electrochemical features, the electrochemical features of the samples whose correlation information meets a predetermined threshold, and determining them as the target electrochemical features.

[0110] According to an embodiment of the present disclosure, the sample electrochemical characteristics may include 43 sample electrochemical characteristics as shown in Table 1 below.

[0111] Table 1

[0112]

[0113] Figure 10The following schematically shows a result diagram of sorting association information according to an embodiment of the present disclosure.

[0114] like Figure 10 As shown, in the order of the magnitude of the correlation information of features 1 to 43, the correlation information of features 40 and 39 is the largest, and thus can be used as target electrochemical features.

[0115] According to the embodiments of the present disclosure, the target electrochemical characteristics are determined by statistically analyzing the correlation information of the electrochemical characteristics of different samples and sorting them, so that the electrochemical characteristics of the samples with the highest correlation with the lithium deposition quality can be obtained. Then, the trained target lithium deposition quality determination model and target electrochemical characteristics are used to output the lithium deposition quality of the target battery and then evaluate the safety of the target battery.

[0116] According to an embodiment of the present disclosure, an initial model is trained using multiple sample lithium deposition state information and multiple sample electrochemical characteristics to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information, including: splitting the multiple sample lithium deposition state information and the multiple sample electrochemical characteristics into a verification sample set and a training sample set according to a predetermined ratio; using the training sample set to perform initial training on the initial model; using the verification sample set to optimize the parameters of the initially trained model until the parameters meet predetermined conditions, thereby obtaining an intermediate model and correlation information.

[0117] According to an embodiment of the present disclosure, the predetermined ratio may be 2:8, that is, for 100% of the sample lithium deposition state information and 100% of the sample electrochemical characteristics, 20% may be divided into a verification sample set and 80% may be divided into a training sample set.

[0118] According to an embodiment of the present disclosure, the performance of the intermediate model can be evaluated by cross-validation or by using an independent validation sample set. Common evaluation indicators include mean square error (MSE), coefficient of determination (R²), etc., until the evaluation indicators can reach a high and stable level, that is, meet the predetermined conditions.

[0119] Figure 11 A block diagram of a device for determining battery lithium deposition quality according to an embodiment of the present disclosure is schematically shown.

[0120] like Figure 11 As shown, the apparatus 1100 includes a first acquisition module 1110 , an extraction module 1120 and a first obtaining module 1130 .

[0121] The first acquisition module 1110 is configured to acquire an electrochemical curve of a target battery during a charge / discharge process.

[0122] The extraction module 1120 is used to extract target electrochemical characteristics from the electrochemical curve, wherein the target electrochemical characteristics represent electrochemical characteristics related to the quality of lithium deposition, and the target electrochemical characteristics include time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve.

[0123] The first obtaining module 1130 is used to input the target electrochemical characteristics into the trained target lithium deposition mass determination model to obtain the lithium deposition mass of the target battery, wherein the lithium deposition mass is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

[0124] According to an embodiment of the present disclosure, the extraction module 1120 includes an extraction submodule, an acquisition submodule, and a calculation submodule.

[0125] The extraction submodule is configured to extract a first moment corresponding to the first current and a second moment corresponding to the second current from the constant-voltage current-time curve according to the first current and the second current.

[0126] The obtaining submodule is used to calculate the difference between the first moment and the second moment to obtain time difference data corresponding to the predetermined current interval.

[0127] The calculation submodule is configured to calculate integral data corresponding to a predetermined current interval according to the first moment and the second moment.

[0128] Figure 12 A block diagram of a device for determining battery lithium deposition quality according to an embodiment of the present disclosure is schematically shown.

[0129] like Figure 12 As shown, the apparatus 1200 includes a second acquisition module 1210 , a training module 1220 , a determination module 1230 and a second obtaining module 1240 .

[0130] The second acquisition module 1210 is configured to acquire a sample database, wherein the sample database includes a sample electrochemical characteristic set and a sample lithium deposition state set of multiple sample batteries corresponding to different predetermined aging conditions, wherein the sample electrochemical characteristic set includes multiple sample electrochemical characteristics, and the sample lithium deposition state set includes multiple sample lithium deposition state information.

[0131] Training module 1220, for using multiple sample lithium precipitation state information and multiple sample electrochemical characteristics, to train the initial model, and obtain the correlation information and intermediate model of each sample electrochemical characteristic and multiple lithium precipitation state information.

[0132] The determination module 1230 is configured to determine a target electrochemical feature from the sample electrochemical feature set according to correlation information between the electrochemical features of each sample and a plurality of lithium deposition state information.

[0133] The second obtaining module 1240 is used to train the intermediate model according to the target electrochemical characteristics to obtain a target lithium deposition mass determination model, wherein the target lithium deposition mass determination model is the trained target lithium deposition mass determination model of claim 1 or 2.

[0134] According to an embodiment of the present disclosure, the second acquisition module 1210 includes an acquisition submodule, an extraction submodule, a first obtaining submodule, and a second obtaining submodule.

[0135] The acquisition submodule is used to obtain electrochemical curves of multiple sample batteries.

[0136] The extraction submodule is used to extract multiple sample electrochemical features from the electrochemical curves of multiple sample batteries to obtain a sample electrochemical feature set.

[0137] The first obtaining submodule is used to analyze multiple sample batteries to obtain a sample lithium deposition state set.

[0138] The second submodule is used to obtain a sample database based on the sample electrochemical feature set and the sample lithium deposition state set.

[0139] According to an embodiment of the present disclosure, the first obtaining submodule includes an analyzing unit, a quantitative analyzing unit, and a verifying unit.

[0140] The analysis unit is used to perform scanning electron microscope observation and nuclear magnetic resonance analysis on the negative electrode sheets of multiple sample batteries to obtain lithium deposition structure information.

[0141] The quantitative analysis unit is used to perform quantitative analysis on the negative electrode plates of multiple sample batteries using titration gas chromatography technology to obtain lithium deposition weight information.

[0142] The verification unit is used to verify the lithium deposition weight information according to the lithium deposition structure information to obtain a sample lithium deposition state set.

[0143] According to an embodiment of the present disclosure, the determination module 1230 includes a sorting submodule and a determination submodule.

[0144] The sorting submodule is used to sort the correlation information between the electrochemical characteristics of each sample and multiple lithium deposition state information to obtain a sorting result.

[0145] The determination submodule is used to screen, from the sample electrochemical feature set, sample electrochemical features whose associated information meets a predetermined threshold value according to the sorting result, and determine them as target electrochemical features.

[0146] According to an embodiment of the present disclosure, the training module 1220 includes a partitioning submodule, a training submodule, and an optimization submodule.

[0147] The division submodule is used to split the lithium deposition state information of multiple samples and the electrochemical characteristics of multiple samples into a verification sample set and a training sample set according to a predetermined ratio.

[0148] The training submodule is used to perform initial training on the initial model using the training sample set.

[0149] The optimization submodule is used to optimize the parameters of the initially trained model using the validation sample set until the parameters meet the predetermined conditions, thereby obtaining an intermediate model and associated information.

[0150] According to the embodiments of the present invention, any number of modules, sub-modules, units, and sub-units, or at least part of the functions of any number of them, can be implemented in one module. According to the embodiments of the present invention, any one or more of the modules, sub-modules, units, and sub-units can be split into multiple modules for implementation. According to the embodiments of the present invention, any one or more of the modules, sub-modules, units, and sub-units can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, according to the embodiments of the present invention, one or more of the modules, sub-modules, units, and sub-units can be at least partially implemented as a computer program module, which can perform the corresponding functions when the computer program module is executed.

[0151] For example, any multiple of the first acquisition module 1110, the extraction module 1120, and the first obtaining module 1130 can be combined into a single module / unit / sub-unit, or any one of these modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functionality of one or more of these modules / units / sub-units can be combined with at least part of the functionality of other modules / units / sub-units and implemented in a single module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the first acquisition module 1110, the extraction module 1120, and the first obtaining module 1130 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented in hardware or firmware by any other reasonable means of circuit integration or packaging, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or any appropriate combination of any of these. Alternatively, at least one of the first acquisition module 1110 , the extraction module 1120 and the first obtaining module 1130 may be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function may be performed.

[0152] For example, any number of the second acquisition module 1210, the training module 1220, the determination module 1230, and the second acquisition module 1240 can be combined into a single module / unit / sub-unit, or any one of these modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functionality of one or more of these modules / units / sub-units can be combined with at least part of the functionality of other modules / units / sub-units and implemented in a single module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the second acquisition module 1210, the training module 1220, the determination module 1230, and the second acquisition module 1240 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented in hardware or firmware by any other reasonable means of integrating or packaging circuits, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or any appropriate combination of any of them. Alternatively, at least one of the second acquisition module 1210 , the training module 1220 , the determination module 1230 and the second obtaining module 1240 may be at least partially implemented as a computer program module, which may perform corresponding functions when executed.

[0153] It should be noted that the portion of the device for determining the quality of lithium deposition in the embodiments of the present disclosure corresponds to the portion of the method for determining the quality of lithium deposition in the embodiments of the present disclosure. For a description of the device for determining the quality of lithium deposition in the embodiments of the present disclosure, please refer to the portion of the method for determining the quality of lithium deposition in the embodiments of the present disclosure, and will not be repeated here. The portion of the model training device in the embodiments of the present disclosure corresponds to the portion of the model training method in the embodiments of the present disclosure. For a description of the model training device, please refer to the portion of the model training method, and will not be repeated here.

[0154] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, which contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, as well as the combination of boxes in the block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified functions or operations, or can be implemented using a combination of dedicated hardware and computer instructions. It will be understood by those skilled in the art that the features described in the various embodiments of the present disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features described in the various embodiments of the present disclosure may be combined and / or coupled in various ways, and all of these combinations and / or couplings fall within the scope of the present disclosure.

[0155] The above describes the embodiments of the present disclosure. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although each embodiment has been described separately above, this does not mean that the measures in each embodiment cannot be advantageously used in combination. Without departing from the scope of the present disclosure, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A method for determining the quality of lithium deposition in a battery, characterized in that: include: Obtain the electrochemical curve of the target battery during the charge / discharge process; Extracting a target electrochemical feature from the electrochemical curve, wherein the target electrochemical feature represents an electrochemical feature related to lithium deposition quality, and the target electrochemical feature includes time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve; The target electrochemical characteristics are input into the trained target lithium deposition mass determination model to obtain the lithium deposition mass of the target battery, wherein the lithium deposition mass is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

2. The method according to claim 1, characterized in that The electrochemical curve includes a constant voltage current-time curve and a constant current voltage-time curve, the predetermined current interval includes a first current and a second current, and extracting a target electrochemical feature from the electrochemical curve includes: extracting a first moment corresponding to the first current and a second moment corresponding to the second current from the constant-voltage current-time curve according to the first current and the second current; Calculating the difference between the first moment and the second moment to obtain time difference data corresponding to the predetermined current interval; According to the first moment and the second moment, integral data corresponding to the predetermined current interval is calculated.

3. A model training method, characterized in that: include: Acquire a sample database, wherein the sample database includes a sample electrochemical feature set and a sample lithium deposition state set of a plurality of sample batteries corresponding to different predetermined aging conditions, the sample electrochemical feature set includes a plurality of sample electrochemical features, and the sample lithium deposition state set includes a plurality of sample lithium deposition state information; Using the lithium deposition state information of the multiple samples and the electrochemical characteristics of the multiple samples, an initial model is trained to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information; determining a target electrochemical feature from the set of sample electrochemical features according to correlation information between each of the sample electrochemical features and the plurality of lithium deposition state information; According to the target electrochemical characteristics, the intermediate model is trained to obtain a target lithium deposition mass determination model, wherein the target lithium deposition mass determination model is the trained target lithium deposition mass determination model according to claim 1 or 2.

4. The method according to claim 3, characterized in that The obtaining of the sample database includes: Performing a cyclic charge / discharge process on the plurality of sample batteries at a first predetermined rate to obtain electrochemical curves of the plurality of sample batteries; Extracting the plurality of sample electrochemical features from the electrochemical curves of the plurality of sample batteries to obtain the sample electrochemical feature set; Analyzing the plurality of sample batteries to obtain the sample lithium deposition state set; The sample database is obtained according to the sample electrochemical feature set and the sample lithium deposition state set.

5. The method according to claim 4, characterized in that The analyzing the plurality of sample batteries to obtain the sample lithium deposition state set includes: Performing scanning electron microscopy observation and nuclear magnetic resonance analysis on the negative electrode sheets of the plurality of sample batteries to obtain lithium deposition structure information; Using titration gas chromatography technology, quantitatively analyzing the negative electrode sheets of the plurality of sample batteries to obtain lithium deposition weight information; The lithium deposition weight information is verified according to the lithium deposition structure information to obtain the sample lithium deposition state set.

6. The method according to claim 4, characterized in that The initial batteries include a first initial battery, a second initial battery, a third initial battery, and a fourth initial battery, and the plurality of sample batteries are obtained by the following operations: maintaining the initial state of the first initial battery to obtain a first sample battery; performing an overcharge / discharge cycle simulation on the second initial battery at a second predetermined rate to obtain a second sample battery; performing a charge / discharge cycle simulation below the operating temperature on the third initial battery according to the second predetermined rate to obtain a third sample battery; According to the second predetermined rate, an overcharge / discharge cycle simulation below the operating temperature is performed on the fourth initial battery to obtain a fourth sample battery.

7. The method according to claim 3, characterized in that The multiple sample electrochemical characteristics include: voltage-related characteristics of a constant current charging process, current-related characteristics of a constant voltage charging process, a diffusion resistance coefficient, a DC internal resistance, and a charging cut-off voltage during an intermittent current interruption process; the current-related characteristics of the constant voltage charging process include time difference data and integral data corresponding to a predetermined current interval; and determining a target electrochemical characteristic from the sample electrochemical characteristic set based on association information between each of the sample electrochemical characteristics and the multiple lithium deposition state information includes: Sorting the correlation information between the electrochemical characteristics of each sample and the plurality of lithium deposition state information to obtain a sorting result; According to the ranking result, the sample electrochemical features whose associated information meets a predetermined threshold are screened from the sample electrochemical feature set and determined as the target electrochemical features.

8. The method according to claim 3, characterized in that The initial model is trained using the lithium deposition state information of the multiple samples and the electrochemical characteristics of the multiple samples to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the multiple lithium deposition state information, including: Splitting the lithium deposition state information of the plurality of samples and the electrochemical characteristics of the plurality of samples into a verification sample set and a training sample set according to a predetermined ratio; Performing initial training on the initial model using the training sample set; The validation sample set is used to optimize the parameters of the initially trained model until the parameters meet predetermined conditions to obtain the intermediate model and the associated information.

9. A device for determining the quality of lithium deposition in a battery, characterized in that: include: A first acquisition module is used to obtain the electrochemical curve of the target battery during the charge / discharge process; an extraction module, configured to extract a target electrochemical feature from the electrochemical curve, wherein the target electrochemical feature represents an electrochemical feature related to the quality of lithium deposition, and the target electrochemical feature includes time difference data and integral data corresponding to a predetermined current interval in the electrochemical curve; The first obtaining module is used to input the target electrochemical characteristics into the trained target lithium deposition mass determination model to obtain the lithium deposition mass of the target battery, wherein the lithium deposition mass is the mass of metallic lithium deposited on the graphite surface when the electrochemical reduction rate of lithium ions on the graphite surface does not match the diffusion transmission rate during the charge / discharge process of the target battery.

10. A model training device, characterized in that: include: a second acquisition module, configured to acquire a sample database, wherein the sample database includes a sample electrochemical feature set and a sample lithium deposition state set of a plurality of sample batteries corresponding to different predetermined aging conditions, the sample electrochemical feature set including a plurality of sample electrochemical features, and the sample lithium deposition state set including a plurality of sample lithium deposition state information; A training module, configured to train an initial model using the lithium deposition state information of the plurality of samples and the electrochemical characteristics of the plurality of samples, to obtain correlation information and an intermediate model between the electrochemical characteristics of each sample and the lithium deposition state information; a determination module, configured to determine a target electrochemical feature from the set of sample electrochemical features based on correlation information between each of the sample electrochemical features and the plurality of lithium deposition state information; The second obtaining module is used to train the intermediate model according to the target electrochemical characteristics to obtain a target lithium deposition mass determination model, wherein the target lithium deposition mass determination model is the trained target lithium deposition mass determination model according to claim 1 or 2.

Images