A lithium ion battery impurity pollution detection method, system, terminal and medium

Abstract

A lithium ion battery impurity pollution detection method, system, terminal and medium, comprising: obtaining the electrochemical impedance spectrum of the battery to be tested and the same type normal battery as the first electrochemical impedance spectrum and the reference impedance spectrum, fitting as the first circular arc and the reference circular arc, judging whether the set condition is met based on the first circular arc and the reference circular arc, if yes, obtaining the electrochemical impedance spectrum of the battery to be tested under the adaptive second preset cycle number as the second electrochemical impedance spectrum, fitting as the second circular arc; based on whether the second circular arc is a new circular arc and other characteristics, judge whether the battery to be tested exists impurity pollution, if there is impurity pollution, the severity classification model output the severity of pollution through pre-training, and identify the impurity type based on the frequency corresponding to the vertex of the second circular arc. The present application provides reliable technical support for quality control in the manufacturing process of lithium ion battery and safety monitoring in the service stage.

Description

Technical Field

[0001] This invention belongs to the field of battery technology, and more specifically, relates to a method, system, terminal and medium for detecting impurities and contamination in lithium-ion batteries. Background Technology

[0002] Lithium-ion batteries, due to their high energy density and long cycle life, have become the core energy storage medium for electric vehicles and energy storage systems. However, during battery manufacturing, the incorporation of metallic foreign matter (such as iron, nickel, copper, zinc, chromium, and other metallic impurity particles) into the battery is a common source of defects. Foreign matter mixed between the positive and negative electrodes can puncture the separator, causing internal short circuits, or promote electrochemical side reactions, and in severe cases, lead to battery thermal runaway, posing significant safety hazards. Therefore, accurate and timely identification and assessment of impurity contamination during the initial stages of battery service and during cycling are crucial for ensuring battery system safety.

[0003] Currently, some progress has been made in the detection and assessment of metallic foreign matter defects in lithium-ion batteries. Some researchers simulate impurity contamination by artificially implanting copper powder of different particle sizes inside the battery and use electrochemical impedance spectroscopy (EIS), voltage curves, coulombic efficiency, and other testing methods to explore the evolution of defects. Existing research shows that metallic foreign matter defects often lead to changes in the capacitive arc in the mid-frequency region of the battery impedance spectrum, and even, as the defects worsen, new electrochemical reaction characteristics appear in the low-to-mid-frequency region.

[0004] However, current assessments largely rely on technicians' visual observation and manual comparison of capacitive arcs in EIS spectra. This method is not only inefficient and highly subjective, but also difficult to automate and standardize in battery management systems (BMS), failing to meet the needs of large-scale industrial battery testing. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method, system, terminal, and medium for detecting impurities in lithium-ion batteries.

[0006] The present invention adopts the following technical solution.

[0007] The first aspect of this invention provides a method for detecting impurities in lithium-ion batteries, comprising: The electrochemical impedance spectra of the battery under test and a normal battery of the same model are obtained at the first preset number of battery charge-discharge cycles as the first electrochemical impedance spectrum and the reference impedance spectrum; the first preset number of cycles is greater than or equal to 1 and less than the maximum number of cycles, and the maximum number of cycles is directly obtained through battery charge-discharge cycle experiments; The real and imaginary impedance data of the first electrochemical impedance spectrum and the reference impedance spectrum in the set mid-frequency characteristic range are fitted into circular arcs as the first and reference arcs. The equivalent diameter increase ratio, center coordinate shift, and arc vertex angular frequency shift of the first arc relative to the reference arc are calculated. If the equivalent diameter increase ratio and coordinate shift meet the set mid-frequency impedance increase condition or the arc vertex angular frequency shift meets the set early electrochemical anomaly condition, the electrochemical impedance spectrum of the battery under test at the second preset cycle number is obtained as the second electrochemical impedance spectrum. The second preset cycle number is dynamically adjusted according to the equivalent diameter increase ratio. The mid-frequency characteristic range is 100Hz to 1kHz. The real and imaginary parts of the impedance data of the second electrochemical impedance spectroscopy in the set mid-to-low frequency characteristic range are fitted into a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, it is determined whether the second arc is a newly added independent arc. The sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency are obtained based on the battery charge-discharge cycle. Combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency, it is determined whether the battery under test is contaminated with impurities. The mid-to-low frequency characteristic range is 0.1Hz~100Hz. If contamination is present, the impurity diameter is estimated based on the coulombic efficiency of the battery under test at different cycles. The geometric parameters of the second arc and the estimated impurity diameter are input into a pre-trained severity classification model, which outputs the severity of contamination. The impurity type is then identified based on the frequency corresponding to the vertex of the second arc. Preferably, the real and imaginary parts of the impedance data are fitted into an arc, specifically: On the complex impedance plane, the real and imaginary parts of the impedance data are fitted into a circular arc using the nonlinear least squares method; For the real and imaginary parts of each impedance, calculate the Euclidean distance between the real and imaginary parts and the center of the circle. Subtract the equivalent radius from the Euclidean distance to obtain the residual of the corresponding impedance. Add the squares of the residuals of all impedances in the corresponding frequency band to obtain the sum of squares of the residuals. The objective function of the nonlinear least squares method is to minimize the sum of squared residuals.

[0008] Preferably, if the increase in equivalent diameter and coordinate offset satisfy the set condition for increased impedance in the mid-frequency region, or if the angular frequency offset at the apex of the arc satisfies the set condition for early electrochemical anomalies, specifically: The increase in equivalent diameter is the difference between the equivalent diameter of the first arc and the equivalent diameter of the reference arc divided by the equivalent diameter of the reference arc. The coordinate offset is the difference between the x-coordinate of the center of the first arc and the x-coordinate of the center of the reference arc; If the increase in equivalent diameter exceeds the set suspected pollution threshold and the coordinate offset is positive, then the set condition for increasing impedance in the mid-frequency region is met. The angular frequency offset of the arc vertex is the absolute value of the difference between the angular frequency of the first arc vertex and the angular frequency of the reference arc vertex divided by the angular frequency of the reference arc vertex. If the angular frequency offset at the apex of the arc exceeds the set frequency offset threshold, then the set early electrochemical anomaly condition is met.

[0009] Preferably, the second preset number of cycles is dynamically adjusted according to the increase ratio of the equivalent diameter, specifically as follows: Different suspected contamination thresholds and rapid inspection thresholds are set for different battery types; If the increase in equivalent diameter is less than the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is equal to the first preset number of cycles or is the set first dynamic number of cycles. If the increase in equivalent diameter is greater than or equal to the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is the set second dynamic number of cycles. The first dynamic count is greater than the first preset cycle count, and the first preset cycle count is greater than the second dynamic count.

[0010] Preferably, the step of determining whether the second arc is a newly added independent arc based on the sum of squared residuals during the fitting of the second arc and the center of the arc specifically involves: If the minimum sum of squared residuals during fitting of the second arc is less than the set fitting threshold, and the center of the second arc is located outside the circle containing the reference arc, then the second arc is considered a newly added independent arc.

[0011] Preferably, the sawtooth fluctuation characteristic of the charge-discharge voltage curve obtained based on battery charge-discharge cycles is specifically as follows: Obtain the charge-discharge curve at the second preset number of cycles, apply Savitzky-Golay filtering to the charge-discharge curve, and use the filtered data as the normal charge-discharge trend curve. Subtract the voltage of the corresponding sampling point of the normal charge-discharge trend curve from the voltage of each sampling point of the charge-discharge curve to obtain the fluctuation value of the corresponding sampling point. Calculate the mean and standard deviation of the fluctuation values ​​of all sampling points, and use the sum of the mean and 3 times the standard deviation as the fluctuation threshold. If the fluctuation value of a sampling point is greater than the fluctuation threshold, the corresponding sampling point is taken as a transient peak point. The voltage amplitude of the transient peak point is recorded. The squares of the voltage amplitudes of all transient peak points are superimposed. The superposition result is input into the lg function. The output result of the lg function is multiplied by the total number of transient peak points and then divided by the total number of sampling points. The result of the division is taken as the sawtooth fluctuation characteristic of the charging and discharging voltage curve.

[0012] Preferably, the step of determining whether the battery under test has impurities by combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge and discharge voltage curve, and the coulombic efficiency is as follows: If the second arc is a newly added independent arc, the battery under test is considered to have impurities; or, if the sawtooth fluctuation characteristics of the charge and discharge voltage curve exceed the set intensity threshold and the coulombic efficiency of the battery under test is lower than the set normal coulombic efficiency threshold when the battery charge and discharge cycle reaches the set third preset number of cycles, the battery under test is considered to have impurities; the third preset number of cycles is greater than or equal to 1 and less than or equal to the maximum number of cycles.

[0013] Preferably, the estimation of impurity diameter based on the coulombic efficiency of the battery under test during different cycles specifically involves: The coulombic efficiency of the battery under test is obtained when the battery charge-discharge cycle reaches the third preset number of cycles. If the coulombic efficiency is in the first set range, the estimated impurity diameter is the first impurity diameter; if the coulombic efficiency is in the second set range, the estimated impurity diameter is the second impurity diameter; if the coulombic efficiency is in the third set range, the estimated impurity diameter is the third impurity diameter. The upper limit of the first interval is equal to the lower limit of the second interval, the upper limit of the second interval is equal to the lower limit of the third interval, and the upper limit of the third interval is equal to the set normal coulomb efficiency threshold. The diameter of the first impurity is smaller than the diameter of the second impurity, and the diameter of the second impurity is smaller than the diameter of the third impurity.

[0014] Preferably, the step of inputting the geometric parameters of the second arc and the estimated impurity diameter into the pre-trained severity classification model specifically involves: The geometric parameters of the second arc include the vertex angular frequency, equivalent diameter, and impedance real part span, wherein the impedance real part span is the maximum value of the impedance real part minus the minimum value of the impedance real part in the second arc. The severity classification model is a neural network model or a deep regression model.

[0015] Preferably, the identification of impurity types based on the frequency corresponding to the vertex of the second arc specifically involves: The frequency corresponding to the vertex of the second arc is taken as the characteristic frequency. If the characteristic frequency is greater than the set lower limit frequency of inert metal and less than or equal to the upper limit frequency of inert metal, it is determined to be contaminated by inert metal impurities. If the characteristic frequency is greater than the set upper limit frequency of inert metal and less than or equal to the upper limit frequency of active metal, it is determined to be contaminated by active metal impurities.

[0016] A second aspect of the present invention provides a lithium-ion battery impurity contamination detection system based on the method described in the first aspect of the present invention, comprising a first electrochemical impedance spectroscopy acquisition module, a second electrochemical impedance spectroscopy acquisition module, an impurity contamination determination module, and an impurity contamination degree and type determination module, specifically as follows: First electrochemical impedance spectroscopy acquisition module: used to acquire the electrochemical impedance spectra of the battery under test and a normal battery of the same type at the first preset number of battery charge-discharge cycles as the first electrochemical impedance spectrum and the reference impedance spectrum. The second electrochemical impedance spectroscopy (EIS) acquisition module is used to fit the real and imaginary impedance data of the first EIS and the reference EIS in a set mid-frequency characteristic range into a circular arc as the first and reference arcs. It calculates the equivalent diameter increase ratio, center coordinate shift, and arc vertex angular frequency shift of the first arc relative to the reference arc. If the equivalent diameter increase ratio and coordinate shift meet the set mid-frequency impedance increase conditions or the arc vertex angular frequency shift meets the set early electrochemical anomaly conditions, then the EIS of the battery under test at the second preset number of cycles is acquired as the second EIS. The second preset number of cycles is dynamically adjusted according to the equivalent diameter increase ratio. Impurity contamination determination module: This module fits the real and imaginary parts of the impedance data of the second electrochemical impedance spectrum in the set low-to-mid frequency characteristic range into a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, it determines whether the second arc is a newly added independent arc. It also obtains the sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency based on the battery charge-discharge cycle. Combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency, it determines whether the battery under test is contaminated with impurities. Impurity contamination degree and type judgment module: If impurity contamination exists, it is used to estimate the impurity diameter based on the coulombic efficiency of the battery under test in different cycles, input the geometric shape parameters of the second arc and the estimated impurity diameter into the pre-trained severity classification model, output the severity of contamination, and identify the impurity type based on the frequency corresponding to the vertex of the second arc.

[0017] A third aspect of the invention provides a terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor performing steps using the lithium-ion battery impurity contamination detection method of the first aspect of the invention.

[0018] A fourth aspect of the present invention provides a medium, which is a computer-readable storage medium storing a computer program that, when executed by a processor, uses the steps of the lithium-ion battery impurity contamination detection method described in the first aspect of the present invention.

[0019] The beneficial effects of this invention are as follows: Compared with the prior art, the technical solution of this application transforms the geometric characteristics of the mid-frequency capacitive arc into an objective equivalent diameter increase ratio through complex impedance plane fitting, overcoming the subjective limitations of the prior art that relies on technicians' visual observation to determine the arc enlargement. Simultaneously, by dynamically adjusting the timing of subsequent impedance spectrum acquisition based on the magnitude of this increase ratio, the rigid mode of fixed testing is broken, allowing monitoring during the critical period of defect deterioration and effectively avoiding the problem of missing key data in the early warning window of internal short circuits. Furthermore, by combining the appearance of new arcs with other geometric parameters for multi-dimensional judgment, the universality and accuracy of contamination judgment are significantly improved. In summary, this method constructs a closed-loop process from suspected screening to confirmed judgment, combining industrial usability with high sensitivity, providing reliable technical support for quality control in the lithium-ion battery manufacturing process and safety monitoring during service. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a framework diagram of the system of the present invention; Figure 3 This is a schematic diagram of the terminal of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0022] like Figure 1 As shown, Embodiment 1 of the present invention proposes a method for detecting impurities in lithium-ion batteries, comprising: S1. Obtain the electrochemical impedance spectra of the battery under test and normal batteries of the same model and batch at the first preset number of battery charge-discharge cycles as the first electrochemical impedance spectrum and the reference impedance spectrum. It should be noted that electrochemical impedance spectroscopy (EIS), as a non-destructive frequency domain measurement technique, can separate the physicochemical processes with different speeds inside the battery by applying small-amplitude AC perturbation signals of different frequencies. Specifically, in the Nyquist plot of a lithium-ion battery, the high-frequency region (1kHz~100kHz) mainly reflects the ohmic impedance of lithium ions in the electrolyte and separator; the mid-frequency region (100Hz~1kHz) mainly reflects the charge transfer process at the electrochemical reaction interface; and the mid-low frequency region (0.1Hz~100Hz) mainly reflects the diffusion process of lithium ions inside the electrode material (the oblique part of the low-frequency region) and the parasitic side reactions on the electrode surface (the arc part of the mid-low frequency region).

[0023] When metallic impurities (such as iron, nickel, copper, zinc, chromium, etc.) are introduced into the battery, these impurities dissolve and migrate in the electrolyte, and then interact electrochemically with the electrodes. This parasitic reaction alters the charge transfer characteristics of the original electrochemical interface and produces an independent impedance response at a specific characteristic frequency.

[0024] The first preset number of cycles is in the initial stage of the battery charge and discharge cycle, such as 1, 5 or 20. The first preset number of cycles can be determined manually or obtained through experiments. When set manually, the first preset number of cycles only needs to be an integer greater than or equal to 1 and less than the maximum number of cycles. The maximum number of cycles is obtained through charge-discharge cycle tests. If the first preset number of cycles is equal to the maximum number of cycles minus 1, when calculating the second preset number of cycles, if the second preset number of cycles needs to be greater than the first preset number of cycles, the second preset number of cycles can still be taken as the maximum number of cycles. Therefore, all the steps in this embodiment can still be carried out. However, the accuracy of determining whether there is pollution and the degree of pollution by taking the maximum number of cycles minus 1 is very low. The accuracy will be higher in the initial stage. When the accuracy is low, the technician can continuously adjust the first preset number of cycles until the accuracy cannot be improved by adjusting the first preset number of cycles or the accuracy is high enough (e.g., higher than 95% or higher than the accuracy of battery pollution judgment in the prior art). The maximum number of cycles for the ternary lithium-ion battery in the experiment of this invention is 200. Therefore, the optimal first preset number of cycles can be found between 1 and 199. More preferably, the optimal first preset number of cycles is set between 1 and 50. Alternatively, the experiment can be directly set by obtaining battery parameter curves such as capacity decay curves or internal resistance change curves of normal batteries of the same model as the test battery after different charge-discharge cycles. The inflection point where the slope of the curve changes significantly is the end time of the initial stage. The number of charge-discharge cycles before the end time of the initial stage can be the first preset cycle number. The first preset cycle number set in the experiment must also be greater than or equal to 1 and less than the maximum cycle number. More preferably, in this embodiment, an electrochemical workstation or an impedance measurement circuit integrated into the battery management system is used to measure the electrochemical impedance spectrum of the test battery under a set state of charge (preferably 100% SOC) in a constant temperature environment of 25°C.

[0025] S2. Based on the first electrochemical impedance spectrum and the reference impedance spectrum, make a preliminary judgment on whether the battery is contaminated. If so, proceed to S3; otherwise, determine that no impurity contamination has occurred. S2.1 Fit the real and imaginary impedance data of the first electrochemical impedance spectrum and the reference impedance spectrum in the set mid-frequency characteristic range into a circular arc as the first circular arc and the reference circular arc. In this preferred embodiment, the real and imaginary parts of the impedance data are fitted to an arc, specifically: On the complex impedance plane, the real and imaginary parts of the impedance data are fitted into a circular arc using the nonlinear least squares method; For the real and imaginary parts of each impedance, calculate the Euclidean distance between the real and imaginary parts and the center of the circle. Subtract the equivalent radius from the Euclidean distance to obtain the residual of the corresponding impedance. Add the squares of the residuals of all impedances in the corresponding frequency band to obtain the sum of squares of the residuals. The objective function of the nonlinear least squares method is to minimize the sum of squared residuals.

[0026] The objective function formula for the nonlinear least squares method is:

[0027] in, The total number of sampling points in the defined mid-frequency characteristic range; For the first i The impedance corresponding to each sampling point , They are respectively The real and imaginary parts; , These are the x and y coordinates of the center of the fitted circular arc, respectively. The equivalent radius of the fitted circular arc; Specifically, the mid-frequency characteristic range is 100Hz to 1kHz.

[0028] It should be noted that the core function of this step is to transform the original complex Nyquist graph into a computable quantitative indicator.

[0029] S2.2 Calculate the increase ratio of the equivalent diameter of the first arc relative to the reference arc, the offset of the center coordinates, and the offset of the arc vertex angular frequency; if the increase ratio of the equivalent diameter and the coordinate offset meet the set condition for increased impedance in the mid-frequency region or the offset of the arc vertex angular frequency meets the set condition for early electrochemical anomaly, then proceed to S3; otherwise, it is determined that no battery contamination has occurred. In this preferred embodiment, if the increase in equivalent diameter and coordinate offset satisfy the set condition for increased impedance in the mid-frequency region, or if the angular frequency offset at the apex of the arc satisfies the set condition for early electrochemical anomalies, specifically: (1) If the increase in equivalent diameter exceeds the set suspected pollution threshold and the coordinate offset is positive, then the set condition for increasing impedance in the mid-frequency region is met. The increase in the equivalent diameter is calculated by dividing the difference between the equivalent diameter of the first arc and the equivalent diameter of the reference arc by the equivalent diameter of the reference arc. It should be noted that the equivalent diameter is twice the equivalent radius. Equivalent diameter increase ratio The formula is:

[0030] in, Let be the equivalent diameter of the first arc; The equivalent diameter of the reference arc; The coordinate offset is the difference between the x-coordinate of the center of the first arc and the x-coordinate of the center of the reference arc, calculated using the formula:

[0031] in, For coordinate offset; Let x be the x-coordinate of the center of the first arc; The x-coordinate of the center of the reference arc; It should be noted that the most significant early sign of the introduction of metallic foreign matter is an increase in the charge transfer impedance in the mid-frequency region of the battery, which is reflected in the impedance spectrum as "the mid-frequency capacitive reactance arc increases and shifts to the right". According to the experimental verification of this application, compared with normal batteries, the mid-frequency arc of the impedance spectrum of batteries with metallic foreign matter defects is significantly larger (for example, more than 10%, the specific values ​​in this embodiment are shown below) and shifted to the right, indicating that the self-discharge phenomenon increases the impedance in the mid-frequency region of the battery. Therefore, it is determined that the increase in the equivalent diameter and the coordinate shift meet the set condition for the increase in mid-frequency impedance.

[0032] Specifically, different pollution suspected thresholds are set for different batteries, with the pollution suspected threshold set in the range of 9% to 15%. In this embodiment, if the battery is a lithium iron phosphate square cell, the pollution suspected threshold is 12%; if it is a ternary lithium-ion battery, the threshold is lowered to 9% because the baseline charge transfer impedance of the ternary system battery LFP is lower.

[0033] (2) If the angular frequency shift at the apex of the arc exceeds the set frequency shift threshold, then the set early electrochemical anomaly condition is met. In this preferred embodiment, the angular frequency offset at the arc vertex is the absolute value of the difference between the angular frequency of the first arc vertex and the angular frequency of the reference arc vertex, divided by the angular frequency of the reference arc vertex, as shown in the formula:

[0034] in, For the angular frequency offset of the arc vertex; Let be the angular frequency at the vertex of the first arc; The angular frequency at the vertex of the reference arc; Specifically, in this embodiment, the frequency offset threshold is set to 5%, and the range of the frequency offset threshold is 1% to 20%. The frequency offset threshold can be determined in the following ways: statistically analyze the arc vertex angular frequency of multiple normal batteries of the same model under the same number of cycles, calculate its coefficient of variation (standard deviation / mean), and use 3 to 6 times the coefficient of variation as the frequency offset threshold; or, those skilled in the art can manually set it in the range of 1% to 20% through a limited number of experiments according to the balance requirements of alarm sensitivity and false alarm rate, and 5% is taken as an example.

[0035] It should be noted that, in order to further improve the sensitivity and specificity of early warning, this embodiment also adds the determination of early anomalies based on angular frequency shift; using only (1) to determine the condition of increased impedance in the mid-frequency region; or using only (2) to determine the condition of early electrochemical anomalies are all within the scope of protection of this invention.

[0036] S3. Obtain the electrochemical impedance spectrum of the battery under test at the second preset number of cycles as the second electrochemical impedance spectrum; the second preset number of cycles is dynamically adjusted according to the increase ratio of the equivalent diameter. It should be noted that the second preset number of loops is greater than the first preset number of loops; In this embodiment, different suspected contamination thresholds and rapid inspection thresholds are set for different battery types. If the increase in equivalent diameter is less than the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is equal to the first preset number of cycles or is the set first dynamic number of cycles. If the increase in equivalent diameter is greater than or equal to the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is the set second dynamic number of cycles. The first dynamic count is greater than the first preset cycle count, and the first preset cycle count is greater than the second dynamic count.

[0037] It should be noted that when the increase in the equivalent diameter of the test battery is less than the corresponding rapid inspection threshold, it indicates that the contamination is in the latent period at the first preset number of cycles, and the original preset number of cycles is maintained or the number of cycles is slightly increased; when it is greater than or equal to the corresponding rapid inspection threshold, it indicates that the intermediate frequency impedance has drifted significantly, and EIS data at a smaller number of cycles is obtained to ensure data during the critical window period of defect deterioration.

[0038] Specifically, different rapid inspection thresholds are set for different batteries. The rapid inspection threshold is greater than the suspected contamination threshold. In this embodiment, if the battery is a lithium iron phosphate square cell, the rapid inspection threshold is 30%; if it is a ternary lithium-ion battery, the threshold is lowered to 25% because the baseline charge transfer impedance of the ternary system battery LFP is lower. In this embodiment, the first preset cycle number is 20, the first dynamic number is 25, and the set second dynamic number is 5 or 10.

[0039] It should be noted that the suspected contamination threshold and rapid inspection threshold in this embodiment are thresholds set based on experimental data. This embodiment uses ternary lithium-ion batteries and lithium iron phosphate square cells as examples. To simulate the situation of metal welding slag mixed in during battery manufacturing, spherical copper powder of different diameters is implanted between the positive electrode sheet and the upper separator of the battery. Specifically, 50μm, 100μm, and 200μm copper powder metal impurities are implanted. Copper is one of the most dangerous and frequently occurring metal contaminants in battery production lines, characterized by its wide source, low dissolution potential, and fast dissolution rate. After the defective batteries are prepared, formation is performed at a rate of 0.02C. Each experimental group has 4 parallel samples, including 3 metal foreign object defective batteries in the experimental group and 1 normal battery in the control group.

[0040] S4. Fit the real and imaginary parts of the second electrochemical impedance spectrum in the set low-to-mid frequency range to form a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, determine whether the second arc is a newly added independent arc. Based on the battery charge-discharge cycle, obtain the sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency. Combine whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency to determine whether the battery under test is contaminated with impurities. In this preferred embodiment, the step of determining whether the second arc is a newly added independent arc based on the sum of squared residuals during the fitting of the second arc and the center of the arc specifically involves: If the minimum sum of squared residuals during fitting of the second arc is less than the set fitting threshold, and the center of the second arc is located outside the circle containing the reference arc, then the second arc is considered a newly added independent arc.

[0041] It should be noted that when fitting the second circular arc, if the minimum sum of squared residuals is less than the set fitting threshold, it means that the data segment can be fitted into a circular arc. In this embodiment, the fitting threshold is set to 10. -6 It should be noted that the value of this fitting threshold is not fixed; it is a convergence metric for the least squares method and is usually a small value, such as 10. -6 ~10 -8 In practical engineering applications, the value can be set as the mean of the sum of squared residuals when fitting the second arc of several contaminated batteries. The fact that the center of the second arc is located outside the circle containing the reference arc indicates that this arc is an independent new arc. This step is the core step in achieving highly specific impurity contamination judgment in this application. The typical evolution process of metallic impurities inside a battery is as follows: small-particle impurities migrate to the edge of the electrode under the influence of an electric field and concentration gradient, undergo electrochemical dissolution, and trigger micro-short circuits or parasitic electrochemical reactions between the positive and negative electrodes. As self-discharge continues, a new arc independent of the main capacitive arc will appear in a specific frequency band of the impedance spectrum. In this embodiment, experiments were conducted on a 50μm diameter copper powder defect battery. With the increase of charge-discharge cycles, a clearly observable new arc appeared in the low-to-mid frequency region of the battery impedance spectrum. The location and shape characteristics of this new arc are highly correlated with the internal short circuit and self-discharge process caused by impurity contamination. However, when experiments are conducted with medium-to-large-particle impurities (100μm, 200μm copper powder), obvious separation of the new arc may not necessarily occur. This invention combines the judgment of the newly added arc with the sawtooth fluctuation characteristics of the subsequent discharge voltage curve and the judgment of the coulombic efficiency to comprehensively determine whether impurity contamination has occurred.

[0042] In this preferred embodiment, the step of obtaining the sawtooth fluctuation characteristic of the charge-discharge voltage curve based on battery charge-discharge cycles specifically refers to: Obtain the charge-discharge curves at the second preset number of cycles, apply Savitzky-Golay filtering to the charge-discharge curves, and use the filtered data as the normal charge-discharge trend curve. Subtract the voltage of the corresponding sampling point of the normal charge-discharge trend curve from the voltage of each sampling point of the charge-discharge curve to obtain the fluctuation value of the corresponding sampling point, so as to eliminate the voltage change caused by the normal decrease in state of charge (SOC); calculate the mean and standard deviation of the fluctuation values ​​of all sampling points, and use the sum of the mean and 3 times the standard deviation as the fluctuation threshold. If the fluctuation value of a sampling point is greater than the fluctuation threshold, the corresponding sampling point is taken as a transient peak point; and the voltage amplitude of the transient peak point is recorded. The squares of the voltage amplitudes of all transient peak points are superimposed, the superposition result is input into the lg function, the output result of the lg function is multiplied by the total number of transient peak points, and then divided by the total number of sampling points. The result of the division is taken as the sawtooth fluctuation characteristic of the charging and discharging voltage curve. The formula for the sawtooth fluctuation characteristic of the charge / discharge voltage curve is:

[0043] in, The charge-discharge curve exhibits sawtooth fluctuation characteristics, and the charge-discharge curve is a voltage charge-discharge curve; This represents the total number of transient spikes. This represents the total number of sampling points; For the first j The voltage amplitude at each transient peak point.

[0044] It should be noted that, Indicates the first j The energy of a transient peak voltage, therefore This indicator comprehensively reflects the severity of the sawtooth pattern in the voltage curve by multiplying the logarithm of the transient peak energy by the peak density. The charge / discharge curve exhibits spikes and sawtooth-like fluctuations. The physical mechanism is as follows: when an internal short circuit first occurs, the battery's terminal voltage drops sharply due to the short circuit; the aluminum current collector at the short circuit location melts rapidly due to the enormous heat generated by the short circuit current, breaking the internal short circuit path; simultaneously, the internal porous protective material continues to support the battery structure, preventing further contact between the negative and positive electrodes, leading to the recovery of the battery's terminal voltage. Therefore, the larger the sawtooth fluctuation characteristic value of the charge / discharge voltage curve, the higher the frequency and intensity of repeated internal short circuit-melting events, meaning the more severe the micro-short circuits caused by impurities.

[0045] In this preferred embodiment, the presence of impurity contamination in the battery under test is determined by combining whether the second arc fitted by the second electrochemical impedance spectroscopy is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency. Specifically: If the second arc is a newly added independent arc, the battery under test is considered to have impurities; or, if the sawtooth fluctuation characteristics of the charge and discharge voltage curve exceed the set intensity threshold and the coulombic efficiency of the battery under test is lower than the set normal coulombic efficiency threshold when the battery charge and discharge cycle reaches the set third preset number of cycles, the battery under test is considered to have impurities.

[0046] It should be noted that the intensity threshold can be obtained through experiments or set manually based on experience. For example, it can be set to be greater than 0 manually. Similarly, the accuracy can be continuously adjusted manually to obtain the optimal value. Specifically, the experiment involves: since different batteries have different fluctuation characteristics, several contaminated batteries of the same model are selected for testing, and their charge-discharge voltage curve sawtooth fluctuation characteristics are extracted. The value corresponding to any percentile between the 95th and 99th percentiles of the mean of all charge-discharge voltage curve sawtooth fluctuation characteristics is used as the threshold. Alternatively, the mean of all charge-discharge voltage curve sawtooth fluctuation characteristics can be subtracted from the standard deviation of a set multiple. The set multiple is generally 3. In practice, as long as the set multiple is greater than 0 and the standard deviation of the mean minus the set multiple is not less than 0, it is acceptable.

[0047] Specifically, the third preset number of cycles is greater than or equal to 1 and less than or equal to the maximum number of cycles. Even in the initial stage, there is still a difference in coulombic efficiency between normal and contaminated batteries, but the difference is small, resulting in low accuracy and a high risk of misjudgment. However, it is still possible to make a judgment. More preferably, for better results, this embodiment selects any number of cycles in which the coulombic efficiency of different impurity diameters can be clearly distinguished as the third preset number of cycles. For example, in this embodiment, the distinction becomes clear after 60 tests (if the difference in coulombic efficiency between two impurity diameters is less than 1%, the distinction is not clear). Therefore, 60 cycles or more are sufficient. Specifically, in this embodiment, the third preset number of cycles is set to 60; the normal coulombic efficiency threshold is set to 99%.

[0048] S5. If contamination is present, the impurity diameter is estimated based on the coulombic efficiency of the battery under test in different cycles. The geometric parameters of the second arc and the estimated impurity diameter are input into the pre-trained severity classification model, which outputs the severity of contamination. The impurity type is then identified based on the frequency corresponding to the vertex of the second arc. It should be noted that the frequency corresponding to the vertex of the second arc is the vertex angular frequency of the second arc divided by 2π.

[0049] In this preferred embodiment, the estimation of impurity diameter based on the coulombic efficiency of the battery under test during different cycles specifically involves: The coulombic efficiency of the battery under test is obtained when the battery charge-discharge cycle reaches the third preset number of cycles. If the coulombic efficiency is in the first set range, the estimated impurity diameter is the first impurity diameter; if the coulombic efficiency is in the second set range, the estimated impurity diameter is the second impurity diameter; if the coulombic efficiency is in the third set range, the estimated impurity diameter is the third impurity diameter. The upper limit of the first interval is equal to the lower limit of the second interval, the upper limit of the second interval is equal to the lower limit of the third interval, and the upper limit of the third interval is equal to the set normal coulomb efficiency threshold. The diameter of the first impurity is smaller than that of the second impurity, and the diameter of the second impurity is smaller than that of the third impurity. It should be noted that although the radii of 100µm and 200µm copper powder are large, they do not pierce the separator, so their reduction in coulombic efficiency is not as severe as that of 50µm copper powder. In the verification of this embodiment, the coulombic efficiencies of normal batteries at 20, 40, and 60 cycles were 99.7%, 99.5%, and 99.1%, respectively, maintaining a high level; while the coulombic efficiencies of defective batteries all showed a significant decrease, and as shown in Table 1, the smaller the impurity diameter, the more significant the decrease in coulombic efficiency. Table 1

[0050] Specifically, the first interval is set as The second interval is set as The third interval is set as It should be noted that the range of the first, second, and third intervals is set by experiments and a third preset number of cycles, as shown in Table 1. The first, second, and third intervals are set based on the average coulombic efficiency of different impurity diameters obtained from experiments. The average coulombic efficiency of the corresponding impurity at the third preset number of cycles must be within the corresponding interval. The third interval is the normal coulombic efficiency threshold.

[0051] It should be noted that the intensity threshold, the first preset number of cycles, the third preset number of cycles, and the first, second, and third intervals in this embodiment can all be set directly through manual experience. As long as they are within the corresponding ranges (intensity threshold greater than 0, first preset number of cycles greater than or equal to 1 and less than the maximum number of cycles, third preset number of cycles greater than or equal to 1 and less than or equal to the maximum number of cycles, first, second, and third intervals between 0 and 1, and no overlapping intervals between the first, second, and third intervals), the steps of this embodiment can be performed to finally output the pollution degree judgment. However, the classification accuracy of the severity classification model output by different values ​​is different. Technicians can continuously adjust these parameters until the accuracy can no longer be improved or has reached a sufficiently high level. That is, these values ​​or thresholds are equivalent to hyperparameters, and hyperparameters can be optimized by using the accuracy obtained through experiments or simulations. The optimization process (i.e., how to adjust these thresholds or values ​​each time) can be carried out using various methods such as grid search, random search, or rules of thumb. The specific values ​​listed above in this embodiment are the results after optimization in this embodiment.

[0052] In this preferred embodiment, the step of inputting the geometric shape parameters of the second arc and the estimated impurity diameter into the pre-trained severity classification model specifically involves: The geometric parameters of the second arc include the vertex angular frequency, equivalent diameter, and impedance real part span, wherein the impedance real part span is the maximum value of the impedance real part minus the minimum value of the impedance real part in the second arc. The severity classification model is a neural network model or a deep regression model.

[0053] Specifically, in this embodiment, a three-layer fully connected MLP neural network is selected as the severity classification model. The geometric shape parameters of the second arc and the estimated impurity diameter are min-max normalized and then concatenated into a 4-dimensional feature vector, which is then input into the neural network.

[0054] In this preferred embodiment, the identification of impurity types based on the frequency corresponding to the vertex of the second arc specifically involves: The frequency corresponding to the vertex of the second arc is taken as the characteristic frequency. If the characteristic frequency is greater than the set lower limit frequency of inert metal and less than or equal to the upper limit frequency of inert metal, it is determined to be contaminated by inert metal impurities. If the characteristic frequency is greater than the set upper limit frequency of inert metal and less than or equal to the upper limit frequency of active metal, it is determined to be contaminated by active metal impurities.

[0055] Specifically, the lower limit frequency for inert metals is set at 0.1 Hz, the upper limit frequency for inert metals is set at 10 Hz, and the upper limit frequency for active metals is set at 100 Hz. It should be noted that if the frequency is not in the 0.1Hz-10Hz range, it indicates that there may be a problem with the previous judgment on whether impurity contamination has occurred. An error should be reported and the judgment on whether impurity contamination has occurred should be re-performed. If it is exactly in the 10Hz range, or in the 10Hz±0.5Hz range, it can be further confirmed by personnel in this technical field based on experience and in combination with other experimental parameters.

[0056] like Figure 2 As shown, Embodiment 2 of the present invention proposes a lithium-ion battery impurity contamination detection system based on the method described in Embodiment 1 of the present invention, including a first electrochemical impedance spectroscopy acquisition module, a second electrochemical impedance spectroscopy acquisition module, an impurity contamination determination module, and an impurity contamination degree and type determination module, specifically as follows: First electrochemical impedance spectroscopy acquisition module: used to acquire the electrochemical impedance spectra of the battery under test and a normal battery of the same type at the first preset number of battery charge-discharge cycles as the first electrochemical impedance spectrum and the reference impedance spectrum. The second electrochemical impedance spectroscopy (EIS) acquisition module is used to fit the real and imaginary impedance data of the first EIS and the reference EIS in a set mid-frequency characteristic range into a circular arc as the first and reference arcs. It calculates the equivalent diameter increase ratio, center coordinate shift, and arc vertex angular frequency shift of the first arc relative to the reference arc. If the equivalent diameter increase ratio and coordinate shift meet the set mid-frequency impedance increase conditions or the arc vertex angular frequency shift meets the set early electrochemical anomaly conditions, then the EIS of the battery under test at the second preset number of cycles is acquired as the second EIS. The second preset number of cycles is dynamically adjusted according to the equivalent diameter increase ratio. Impurity contamination determination module: This module fits the real and imaginary parts of the impedance data of the second electrochemical impedance spectrum in the set low-to-mid frequency characteristic range into a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, it determines whether the second arc is a newly added independent arc. It also obtains the sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency based on the battery charge-discharge cycle. Combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency, it determines whether the battery under test is contaminated with impurities. Impurity contamination degree and type judgment module: If impurity contamination exists, it is used to estimate the impurity diameter based on the coulombic efficiency of the battery under test in different cycles, input the geometric shape parameters of the second arc and the estimated impurity diameter into the pre-trained severity classification model, output the severity of contamination, and identify the impurity type based on the frequency corresponding to the vertex of the second arc.

[0057] Embodiment 3 of the present invention provides a terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor performs the steps of the lithium-ion battery impurity contamination detection method described in Embodiment 1 of the present invention.

[0058] Specifically, such as Figure 3 As shown, the memory is non-volatile, and the terminal also includes a network interface for transmitting detection results.

[0059] Embodiment 4 of the present invention proposes a medium, which is a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it uses the steps of the lithium-ion battery impurity contamination detection method described in Embodiment 1 of the present invention.

[0060] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0061] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0062] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0063] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for detecting impurities in lithium-ion batteries, characterized in that, include: The electrochemical impedance spectra of the battery under test and a normal battery of the same model are obtained at the first preset number of charge-discharge cycles and used as the first electrochemical impedance spectrum and the reference impedance spectrum. The first preset number of cycles is greater than or equal to 1 and less than the maximum number of cycles. The maximum number of cycles is obtained directly through battery charge-discharge cycle experiments. The real and imaginary impedance data of the first electrochemical impedance spectrum and the reference impedance spectrum in the set mid-frequency characteristic range are fitted into circular arcs as the first and reference arcs. The equivalent diameter increase ratio, center coordinate shift, and arc vertex angular frequency shift of the first arc relative to the reference arc are calculated. If the equivalent diameter increase ratio and center coordinate shift meet the set mid-frequency impedance increase condition or the arc vertex angular frequency shift meets the set early electrochemical anomaly condition, the electrochemical impedance spectrum of the battery under test at the second preset cycle number is obtained as the second electrochemical impedance spectrum. The second preset cycle number is dynamically adjusted according to the equivalent diameter increase ratio. The mid-frequency characteristic range is 100Hz to 1kHz. The real and imaginary parts of the impedance data of the second electrochemical impedance spectroscopy in the set mid-to-low frequency characteristic range are fitted into a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, it is determined whether the second arc is a newly added independent arc. The sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency are obtained based on the battery charge-discharge cycle. Combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency, it is determined whether the battery under test is contaminated with impurities. The mid-to-low frequency characteristic range is 0.1Hz~100Hz. If contamination is present, the impurity diameter is estimated based on the coulombic efficiency of the battery under test in different cycles. The geometric parameters of the second arc and the estimated impurity diameter are input into a pre-trained severity classification model, which outputs the severity of contamination. The impurity type is then identified based on the frequency corresponding to the vertex of the second arc. The frequency corresponding to the vertex of the second arc is taken as the characteristic frequency. If the characteristic frequency is greater than the set lower limit frequency of inert metal and less than or equal to the upper limit frequency of inert metal, it is determined to be contaminated by inert metal impurities. If the characteristic frequency is greater than the set upper limit frequency of inert metal and less than or equal to the upper limit frequency of active metal, it is determined to be contaminated by active metal impurities.

2. The method for detecting impurities in lithium-ion batteries according to claim 1, characterized in that: The real and imaginary parts of the impedance are fitted to an arc, specifically as follows: On the complex impedance plane, the real and imaginary parts of the impedance data are fitted into a circular arc using the nonlinear least squares method; For the real and imaginary parts of each impedance, calculate the Euclidean distance between the real and imaginary parts and the center of the circle. Subtract the equivalent radius from the Euclidean distance to obtain the residual of the corresponding impedance. Add the squares of the residuals of all impedances in the corresponding frequency band to obtain the sum of squares of the residuals. The objective function of the nonlinear least squares method is to minimize the sum of squared residuals.

3. The method for detecting impurities in lithium-ion batteries according to claim 1, characterized in that: The condition that the equivalent diameter increase ratio and the center coordinate shift satisfy the set mid-frequency impedance increase condition or the arc vertex angular frequency shift satisfies the set early electrochemical anomaly condition is as follows: The increase in equivalent diameter is the difference between the equivalent diameter of the first arc and the equivalent diameter of the reference arc divided by the equivalent diameter of the reference arc. The center coordinate offset is the difference between the x-coordinate of the center of the first arc and the x-coordinate of the center of the reference arc; If the increase in equivalent diameter exceeds the set suspected pollution threshold and the center coordinate offset is positive, then the set condition for increasing impedance in the mid-frequency region is met. The angular frequency offset of the arc vertex is the absolute value of the difference between the angular frequency of the first arc vertex and the angular frequency of the reference arc vertex divided by the angular frequency of the reference arc vertex. If the angular frequency offset at the apex of the arc exceeds the set frequency offset threshold, then the set early electrochemical anomaly condition is met.

4. The method for detecting impurities in lithium-ion batteries according to claim 3, characterized in that: The second preset number of cycles is dynamically adjusted according to the increase ratio of the equivalent diameter, specifically as follows: Different suspected contamination thresholds and rapid inspection thresholds are set for different battery types; If the increase in equivalent diameter is less than the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is equal to the first preset number of cycles or is the set first dynamic number of cycles. If the increase in equivalent diameter is greater than or equal to the rapid inspection threshold corresponding to the test battery, then the second preset number of cycles is the set second dynamic number of cycles. The first dynamic count is greater than the first preset cycle count, and the first preset cycle count is greater than the second dynamic count.

5. The method for detecting impurities in lithium-ion batteries according to claim 2, characterized in that: The determination of whether the second arc is a newly added independent arc based on the sum of squared residuals during the fitting of the second arc and the center of the arc is as follows: If the minimum sum of squared residuals during fitting of the second arc is less than the set fitting threshold, and the center of the second arc is located outside the circle containing the reference arc, then the second arc is considered a newly added independent arc.

6. The method for detecting impurities in lithium-ion batteries according to claim 1, characterized in that: The sawtooth fluctuation characteristic of the charge / discharge voltage curve obtained based on battery charge / discharge cycles is specifically as follows: Obtain the charge-discharge curve at the second preset number of cycles, apply Savitzky-Golay filtering to the charge-discharge curve, and use the filtered data as the normal charge-discharge trend curve. Subtract the voltage of the corresponding sampling point of the normal charge-discharge trend curve from the voltage of each sampling point of the charge-discharge curve to obtain the fluctuation value of the corresponding sampling point. Calculate the mean and standard deviation of the fluctuation values ​​of all sampling points, and use the sum of the mean and 3 times the standard deviation as the fluctuation threshold. If the fluctuation value of a sampling point is greater than the fluctuation threshold, the corresponding sampling point is taken as a transient peak point. The voltage amplitude of the transient peak point is recorded. The squares of the voltage amplitudes of all transient peak points are superimposed. The superposition result is input into the lg function. The output result of the lg function is multiplied by the total number of transient peak points and then divided by the total number of sampling points. The result of the division is taken as the sawtooth fluctuation characteristic of the charging and discharging voltage curve.

7. The method for detecting impurities in lithium-ion batteries according to claim 6, characterized in that: The method of determining whether the battery under test has impurities by combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge and discharge voltage curve, and the coulombic efficiency is as follows: If the second arc is a newly added independent arc, the battery under test is considered to have impurities; or, if the sawtooth fluctuation characteristics of the charge and discharge voltage curve exceed the set intensity threshold and the coulombic efficiency of the battery under test is lower than the set normal coulombic efficiency threshold when the battery charge and discharge cycle reaches the set third preset number of cycles, the battery under test is considered to have impurities; the third preset number of cycles is greater than or equal to 1 and less than or equal to the maximum number of cycles.

8. The method for detecting impurities in lithium-ion batteries according to claim 7, characterized in that: The estimation of impurity diameter based on the coulombic efficiency of the battery under test during different cycles is as follows: The coulombic efficiency of the battery under test is obtained when the battery charge-discharge cycle reaches the third preset number of cycles. If the coulombic efficiency is in the first set range, the estimated impurity diameter is the first impurity diameter; if the coulombic efficiency is in the second set range, the estimated impurity diameter is the second impurity diameter; if the coulombic efficiency is in the third set range, the estimated impurity diameter is the third impurity diameter. The upper limit of the first interval is equal to the lower limit of the second interval, the upper limit of the second interval is equal to the lower limit of the third interval, and the upper limit of the third interval is equal to the set normal coulomb efficiency threshold. The diameter of the first impurity is smaller than the diameter of the second impurity, and the diameter of the second impurity is smaller than the diameter of the third impurity.

9. The method for detecting impurities in lithium-ion batteries according to claim 1, characterized in that: The step of inputting the geometric parameters of the second arc and the estimated impurity diameter into the pre-trained severity classification model is as follows: The geometric parameters of the second arc include the vertex angular frequency, equivalent diameter, and impedance real part span, wherein the impedance real part span is the maximum value of the impedance real part minus the minimum value of the impedance real part in the second arc. The severity classification model is a neural network model or a deep regression model.

10. A lithium-ion battery impurity contamination detection system based on the lithium-ion battery impurity contamination detection method according to any one of claims 1-9, comprising a first electrochemical impedance spectroscopy acquisition module, a second electrochemical impedance spectroscopy acquisition module, an impurity contamination determination module, and an impurity contamination degree and type determination module, characterized in that: First electrochemical impedance spectroscopy acquisition module: used to acquire the electrochemical impedance spectra of the battery under test and a normal battery of the same type at the first preset number of battery charge-discharge cycles as the first electrochemical impedance spectrum and the reference impedance spectrum. The second electrochemical impedance spectroscopy (EIS) acquisition module is used to fit the real and imaginary impedance data of the first EIS and the reference EIS in a set mid-frequency characteristic range into a circular arc as the first and reference arcs. It calculates the equivalent diameter increase ratio, center coordinate shift, and arc vertex angular frequency shift of the first arc relative to the reference arc. If the equivalent diameter increase ratio and center coordinate shift meet the set mid-frequency impedance increase condition or the arc vertex angular frequency shift meets the set early electrochemical anomaly condition, then the EIS of the battery under test at the second preset number of cycles is acquired as the second EIS. The second preset number of cycles is dynamically adjusted according to the equivalent diameter increase ratio. Impurity contamination determination module: This module fits the real and imaginary parts of the impedance data of the second electrochemical impedance spectrum in the set low-to-mid frequency characteristic range into a circular arc as the second arc. Based on the sum of squared residuals and the center of the second arc during fitting, it determines whether the second arc is a newly added independent arc. It also obtains the sawtooth fluctuation characteristics of the charge-discharge voltage curve and the coulombic efficiency based on the battery charge-discharge cycle. Combining whether the second arc is a newly added independent arc, the sawtooth fluctuation characteristics of the charge-discharge voltage curve, and the coulombic efficiency, it determines whether the battery under test is contaminated with impurities. Impurity contamination degree and type judgment module: If impurity contamination exists, it is used to estimate the impurity diameter based on the coulombic efficiency of the battery under test in different cycles. The geometric shape parameters of the second arc and the estimated impurity diameter are input into the pre-trained severity classification model, and the severity of contamination is output. The impurity type is identified based on the frequency corresponding to the vertex of the second arc: the frequency corresponding to the vertex of the second arc is used as the characteristic frequency. If the characteristic frequency is greater than the set lower limit frequency of inert metal and less than or equal to the upper limit frequency of inert metal, it is determined to be inert metal impurity contamination. If the characteristic frequency is greater than the set upper limit frequency of inert metal and less than or equal to the upper limit frequency of active metal, it is determined to be active metal impurity contamination.

11. A terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor performs the steps of the lithium-ion battery impurity contamination detection method according to any one of claims 1 to 9.

12. A medium, characterized in that, The medium is a computer-readable storage medium storing a computer program, which, when executed by a processor, uses the steps of the lithium-ion battery impurity contamination detection method according to any one of claims 1 to 9.

Images