Method and system for online diagnosis of battery degradation mode

Abstract

The invention discloses a method and a system for online diagnosis of a battery degradation mode. The method comprises the following steps: S1, testing battery characteristics; s2, macroscopically diagnosing a battery degradation mode; s3, microscopically diagnosing the degradation mode of the battery; s4, performing macroscopic-microscopic correlation quantitative diagnosis on the battery degradation mode based on the diagnosis results of S2 and S3, and verifying the correctness of the quantitative diagnosis result made by the battery degradation mode diagnosis model; and S5, carrying out online diagnosis on the battery degradation mode. The accuracy and reliability of battery degradation mode diagnosis are improved, the limitations that an existing macroscopic external characteristic method is difficult to reveal a degradation mechanism and a microcosmic physicochemical disintegration method is large in destructiveness and cannot perform quantitative diagnosis are improved, lossless, quantitative and online real-time diagnosis is achieved, the problem that a complex nonlinear degradation mechanism in a battery is difficult to accurately and quantitatively diagnose is solved, and the reliability of battery degradation mode diagnosis is improved. And the requirements of accurate and efficient diagnosis of the degradation mode in an actual battery health management system are met.

Description

Technical Field

[0001] This invention relates to a method and system for online diagnosis of battery degradation modes, belonging to the field of battery degradation diagnosis technology. Background Technology

[0002] As a typical nonlinear, dynamic, and time-varying system, batteries possess extremely complex degradation mechanisms, which can be broadly categorized into three types: active material loss (LAM), active lithium-ion loss (LLI), and electrolyte loss (LE). These degradation modes originate from complex physical or chemical side reactions within the battery and are often interconnected or mutually reinforcing. Generally, active material loss and active lithium-ion loss are considered the primary degradation modes, while electrolyte loss is regarded as a secondary degradation mode.

[0003] Currently, diagnostic methods for battery degradation are mainly divided into two categories: macroscopic external characteristic methods and microscopic physical analysis methods. External characteristic methods primarily characterize degradation behavior by exploring the correlation between signals collected during battery operation, such as voltage, current, and temperature, and degradation modes. Common analytical techniques include open-circuit voltage (OCV) curve analysis, incremental capacity (IC) curve analysis, and differential voltage (DV) curve analysis. In recent years, machine learning and deep learning methods have also been increasingly applied to the identification and diagnosis of degradation modes. Common methods include Support Vector Machines (SVM), Convolutional Neural Networks (CNN), and Long Short-Term Memory Networks (LSTM). External characteristic methods can achieve non-destructive detection of degradation modes and have the advantages of simple application and real-time diagnosis. However, this method struggles to reveal the internal degradation mechanisms of the battery, and the diagnostic results lack direct verification.

[0004] The physical-to-discrete method focuses on directly exploring the internal physical or chemical changes of a battery through advanced experimental and analytical techniques. For example, it utilizes X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and Raman spectroscopy to qualitatively analyze the structural evolution, elemental distribution, and by-reaction products of electrode materials. The physical-to-discrete method can directly reveal the essential mechanisms of battery degradation and has high research value. However, this type of method can cause irreversible damage to the battery, and it is difficult to quantitatively diagnose the battery's degradation patterns. Summary of the Invention

[0005] Purpose of the invention: To address the shortcomings of existing technologies, this invention provides a method and system for online diagnosis of battery degradation modes. This invention diagnoses battery degradation modes based on the characteristic peaks of the IC curve and performs rate correction on the results to diagnose battery degradation modes online.

[0006] Technical solution: A method for online diagnosis of battery degradation modes, comprising the following steps:

[0007] S1. Battery characteristic testing: Macroscopic and microscopic characteristic tests are conducted on the battery. Macroscopic characteristic tests include cycle testing, capacity calibration testing, and charge-discharge testing at different rates. Microscopic characteristic tests include half-cell testing and X-ray diffraction testing, in order to obtain the macroscopic and microscopic characteristics of the battery.

[0008] S2. Macroscopic diagnosis of battery degradation mode: Based on the macroscopic characteristics of the battery obtained in S1, a battery degradation mode diagnosis model is established based on the open circuit voltage (OCV) method. Diagnostic parameters are determined and identified, and the diagnostic parameters are decoupled to make quantitative diagnoses of active material loss (LAM) and active lithium ion loss (LLI).

[0009] S3. Microscopic diagnosis of battery degradation mode: Based on the microscopic characteristics obtained in S1, the half-cell capacity change and the rate of change of the maximum lithium content of the battery are diagnosed.

[0010] S4. Based on the diagnostic results of S2 and S3, perform a quantitative diagnosis of the macro-micro correlation of battery degradation modes to verify the correctness of the quantitative diagnostic results made by the battery degradation mode diagnostic model.

[0011] S5. Online diagnosis of battery degradation mode: Construct batteries with different degradation levels, establish OCV curves, and obtain IC curves based on OCV curves. Diagnose degradation modes by analyzing the area of ​​their characteristic peaks, and use capacity and rate correction coefficients to correct and optimize the diagnostic results.

[0012] In a preferred embodiment, S1 specifically comprises:

[0013] Cyclic test: The battery is charged with constant current and constant voltage in sequence, and then discharged with constant current. This cycle is repeated. After at least 50 cycles, a capacity calibration test is performed to obtain the current health status of the battery.

[0014] Capacity calibration test: The battery is simultaneously charged with constant current and charged with constant voltage, and then discharged with constant voltage. This cycle is repeated, and the average discharge amount during the cycle is taken as the capacity calibration value of the battery in its current state.

[0015] Different rate charge and discharge tests: The battery is charged and discharged at different rates to obtain the battery's charge and discharge characteristics. The specific test rates include C / 25, C / 10, C / 5, C / 3, C / 2, and 1C.

[0016] Half-cell test: Charge batteries in different health states to full charge and discharge them to empty state, then disassemble them to obtain the positive and negative electrode plates. Then, take samples of the complete parts of the electrode plates to assemble half-cells, and then perform C / 25 rate charge or discharge tests to obtain the capacity of the positive and negative electrode plates in different health states.

[0017] X-ray diffraction test: Samples are taken from intact parts of the positive and negative electrode plates, and X-ray diffractometer is used to test the electrodes in different health conditions. The scanning range is 10-80 degrees, and the scanning speed is 0.013 degrees per step.

[0018] In a preferred embodiment, S2 specifically comprises:

[0019] Battery degradation modes include loss of active lithium ions (LLI) and loss of positive electrode active material (LAM). PE and loss of LAM of negative electrode active material NE The positive electrode active material loses LAM PE Divided into lithium-containing cathode material loss LAM liPE and loss of LAM without lithium cathode material dePE The negative electrode active material loses LAM NE Lithium-containing anode material loss LAM liNE and loss of LAM without lithium anode material deNE ;

[0020] The establishment of the battery degradation mode diagnostic model specifically involves:

[0021] The electrode potential curves of the positive and negative electrodes of the battery are derived from different coordinate systems, namely, the positive electrode lithium intercalation rate-positive electrode potential coordinate system x. PE -y PE Negative electrode lithium insertion rate - negative electrode potential coordinate system x NE -y NE The polar coordinate system x can be used. APE -y APE and the available negative polar coordinate system x ANE -y ANE ;

[0022] Fitted electrode potential curve:

[0023] For a half-cell made of a positive electrode, the potential curve mainly consists of two single-phase regions and a two-phase coexistence region sandwiched in between. To improve the accuracy of the potential curve fitting, two sets of hyperbolic tangent functions are introduced to fit the electrode potential curve. The specific fitting expression is as follows:

[0024]

[0025] In the formula, U p This represents the positive electrode potential, and the letter af represents the fitting parameters.

[0026] For a half-cell made of a negative electrode, there are many phase transition regions. To improve the accuracy of the potential curve fitting, three sets of hyperbolic tangent functions are introduced to fit the characteristics of the phase transition regions. The specific fitting expression is as follows:

[0027]

[0028] In the formula, U n Indicates the negative electrode potential;

[0029] Since the OCV of a battery is determined by the potentials of its positive and negative electrodes and the matching relationship between them, it is necessary to unify the positive and negative coordinate systems to the same coordinate system through coordinate transformation.

[0030] The lithium insertion rate of its positive and negative electrodes, i.e., x p x n Transform into x ANE By subtracting the magnitudes of the potentials of the positive and negative electrodes, the full cell OCV can be obtained, which can then be used to construct a battery degradation mode diagnostic model. The battery degradation mode diagnostic model is shown below:

[0031]

[0032] In the formula, U p U represents the positive electrode potential; n Indicates negative electrode potential; C PE Indicates positive electrode capacity; C NE Indicates the negative electrode capacity;

[0033] The determination and identification of diagnostic parameters specifically refers to:

[0034] For the positive electrode, its usable capacity is affected by C. PE LAM liPE LAM dePE The combined effect; therefore, the available capacity of the positive electrode is defined as the first diagnostic parameter X1, as follows:

[0035] X1 = C APE =C PE -LAM liPE -LAM dePE (4)

[0036] For the positive electrode, its usable capacity is affected by C. NE LAM liNE LAM deNE The combined effect; therefore, the available capacity of the positive electrode is defined as the second diagnostic parameter X2, as follows:

[0037] X2 = C ANE =C NE -LAM liNE -LAM deNE (5)

[0038] The position where the available negative electrode lithium insertion rate is zero is affected by the matching relationship between the positive and negative electrodes; therefore, the offset of the coordinate system is defined as the third diagnostic parameter X3, as follows:

[0039] X3 = LAM liNE -LAM dePE +LLI (6)

[0040] The combined formulas (3)-(6) simplify the battery degradation mode diagnostic model as follows:

[0041]

[0042] Since the lithium intercalation rate of the electrode cannot be directly measured experimentally, it is necessary to establish the relationship between the lithium intercalation rate of the battery electrode and the battery capacity. When the battery reaches the discharge cutoff voltage, the lithium intercalation rate of the negative electrode can be defined as the fourth diagnostic parameter X4, resulting in the following equation:

[0043]

[0044] In the formula, LR neg,100 This indicates the lithium intercalation rate of the available negative electrode when the battery reaches the charging cutoff voltage; Q aging Indicates the capacity of the degraded battery;

[0045] The relationship between the available lithium intercalation rate of the negative electrode and the battery capacity is as follows:

[0046]

[0047] In the formula, Capacity(i) represents the battery's charging capacity; x ANE,i This represents the corresponding lithium intercalation rate, and i represents the charging time.

[0048] By integrating equations (8)-(9), we obtain:

[0049]

[0050] The optimal diagnostic parameters are determined based on the minimum root mean square error between the constructed OCV and the measured OCV; the optimization objective is as follows:

[0051]

[0052] In the formula, OCV(i) is obtained through low-rate discharge experiments; OCV(x) ANE,i The data is constructed using electrode potential curves; n is the total amount of data.

[0053] The decoupling diagnostic parameters, which provide quantitative diagnosis of active material loss (LAM) and active lithium ion loss (LLI), are as follows:

[0054] Assuming the probability of breakage of the active material is equal under different lithium insertion rates, then the lithium insertion rate of the isolated portion in one cycle should be equal to the average lithium insertion rate of the usable negative electrode; therefore, the loss of the lithium-containing active material is obtained as follows:

[0055]

[0056] In the formula, x crk,p,fresh Indicates the lithium content when the positive electrode of a brand-new battery is broken; x crk,p,aging Indicates the lithium content when the positive electrode of an aged battery breaks; x crk,n,fresh Indicates the lithium content when the negative electrode of a brand-new battery is broken; x crk,n,aging Indicates the lithium content when the negative electrode of an aged battery breaks; X 1,fresh Indicates the usable capacity of a brand new battery positive terminal; X 1,aging Indicates the usable capacity of the positive electrode of an aged battery; X 2,fresh Indicates the usable capacity of a brand new battery negative terminal; X 2,aging Indicates the usable capacity of the negative electrode of an aged battery;

[0057] The simultaneous equations (6)-(7), (10), and (12) can achieve quantitative diagnosis of different degradation modes; based on the diagnosis results, the active lithium ion loss (LLI) and the negative electrode active material loss (LAM) are obtained. NE This is the main degradation mode of batteries.

[0058] In a preferred embodiment, S3 specifically includes:

[0059] Based on the positive and negative electrode capacities obtained from the half-cell test in S1 under different health states, it was determined that the loss of LAM due to the negative electrode active material was due to... NE This causes the capacity of the negative electrode half-cell to gradually decrease as the battery degrades;

[0060] Based on the diffraction peak distribution of the positive and negative electrodes obtained from the X-ray diffraction test in S1, it is found that the diffraction peak distribution of the negative electrode is more concentrated and it is easier to distinguish the different component types. Therefore, the diffraction peak of the negative electrode is used to characterize the battery degradation mode.

[0061] By utilizing the diffraction peak intensity of the lithium-carbon compound in the negative electrode and normalizing the carbon content, the maximum lithium content corresponding to the battery's state from full charge to full discharge is calculated. The calculation formula is as follows:

[0062]

[0063] In the formula, x1 is the diffraction intensity of the LiC6 peak; x2 is the diffraction intensity of the LiC6 peak. 12 The diffraction intensity of the peak; Li x C is the lithium-carbon compound content calculated from the diffraction peak intensity;

[0064] To analyze the relationship between maximum lithium content and battery degradation mode, the maximum lithium content needs to be represented as a percentage, and the calculation formula is as follows:

[0065]

[0066] In the formula, LLI XRD This indicates the maximum rate of change in lithium content diagnosed by XRD testing, while Li x C fresh and Li x C aging These represent LiC6 for new batteries and aged batteries, respectively.

[0067] Based on formula (14), the X-ray diffraction peak of the negative electrode sheet under full charge is quantitatively calculated, so as to obtain the maximum lithium content conversion rate of the battery under different aging conditions. The diagnostic results based on the intensity of the X-ray diffraction peak show that the maximum lithium content conversion rate is positively correlated with the health status of the battery.

[0068] In a preferred embodiment, S4 specifically comprises:

[0069] LAM NE Correlate with half-cell capacity:

[0070] Since the unit of negative electrode half-cell capacity is mAh, and the battery degradation mode diagnostic model diagnoses the loss of negative electrode active material LAM... NE The units are Ah, and the two cannot be directly compared; therefore, to achieve quantitative correlation diagnosis of degradation patterns, it is necessary to normalize the two sets of data; the specific calculation formula is as follows:

[0071]

[0072] In the formula, LAM NE,Cap Indicates the capacity change after a half-cell test; LAM NE,OCV Indicates LAM diagnosed based on OCV method NE C NE,Cap,fresh and C NE,Cap,aging These represent the negative electrode capacities of new and degraded batteries, respectively; LAM NE,OCV,fresh and LAM NE,OCV,aging These represent the LAM values ​​of new and degraded cells diagnosed using the OCV method, respectively. NE ;

[0073] Correlate the maximum rate of change in lithium content with LLI:

[0074] Convert the LLI results calculated based on OCV to percentage form:

[0075]

[0076] In the formula, LLI OCV This indicates LLI diagnosed using the OCV method; LLI fresh and LLI aging LLI represents new batteries and aged batteries, respectively;

[0077] The correlation between the maximum lithium content change rate and LLI shows a large deviation because the calculation of the maximum lithium content using the diffraction peak intensity of the negative electrode lithium carbon compound includes LAM; therefore, it needs to be corrected.

[0078] LAM NE The results correlated with half-cell capacity indicate that the negative electrode half-cell capacity can be used to diagnose LAM (Laminated Ambulatory Marking). NE Therefore, the formula for calculating the maximum lithium content correction is as follows:

[0079]

[0080] In the formula, Li x C' represents the modified Li x C;

[0081] Therefore, we can conclude that:

[0082]

[0083] Quantitative association of LLI OCV and LLI' XRD The results were obtained, and the accuracy of the OCV-based method for diagnosing LLI was verified.

[0084] Preferably, S5 includes:

[0085] S501: Based on the diagnostic results in S2, construct batteries with different degradation modes, plot the OCV curves of the batteries, and analyze the impact of different degradation modes on the OCV curves.

[0086] S502: The IC curves of batteries with different degradation modes are obtained by using the polynomial center smoothing method based on the OCV curve;

[0087] S503: Obtain all characteristic peaks on the IC curve, and select those corresponding to the loss LAM of the negative electrode active material. NE The characteristic peak that shows a significant correlation with the loss of active lithium ions (LLI) can be used to diagnose the degradation mode of the battery.

[0088] S504: Considering the battery capacity at C / 25 rate, the active lithium ion loss (LLI) result in S503 is corrected;

[0089] S505: To meet practical applications, a rate correction factor is introduced to optimize the correction result of S504.

[0090] In a preferred embodiment, S501 specifically includes:

[0091] Using the degradation mode diagnosis results based on the OCV method as a reference, batteries with different degradation modes were constructed using the method of controlling variables;

[0092] In the analysis of the impact of degradation modes on the OCV curve, since the capacity of the negative electrode decreases with aging, while the capacity of the positive electrode remains essentially unchanged during aging; therefore, x is used... PE The impact of degradation modes on the OCV curve is analyzed, and the specific calculation method is as follows:

[0093]

[0094] In the formula, x PE Indicates the lithium intercalation rate of the positive electrode; U p U represents the positive electrode potential; n C represents the negative electrode potential; PE Indicates positive electrode capacity; C NE Indicates the nearby capacity; LAM liPE Indicates the loss of cathode material content; LAM dePE Indicates the absence of lithium cathode material loss; LAM liNE Indicates the loss of lithium-containing anode materials; LAM deNE This indicates that there is no loss of lithium anode material;

[0095] x PE The coordinate range is as follows:

[0096]

[0097] Batteries with different degradation modes were constructed and then introduced into x. PE Using a coordinate system, analyze the impact of different degradation modes on the OCV curve;

[0098] Specifically, S502 is as follows:

[0099] Using k-th order polynomial curves By fitting the data and minimizing the sum of squares of the differences between the actual and fitted values ​​according to the least squares principle, the coefficients 'a' of each order of the polynomial can be obtained. i This transforms the solution to the IC curve into solving for the coefficient a. i Solve for;

[0100] Let p = 2m + 1, and the p data points be (x... -m ,y -m ),(x -m+1 ,y-m+1 ),...,(x0,y0),...,(x m ,y m If ), then the fitting matrix can be expressed as:

[0101]

[0102] That is:

[0103] Y (2m+1) =X (2m+1)×k ·A k×1 +E (2m+1)×1 (twenty two)

[0104] In the formula, A is the coefficient matrix; E is the deviation matrix;

[0105] The objective function is:

[0106]

[0107] The least squares solution to the coefficient matrix A is:

[0108]

[0109] In a preferred embodiment, S503 specifically includes:

[0110] Obtain all characteristic peaks on the IC curve and select those related to the loss of LAM in the negative electrode active material. NE Peak B shows a significant correlation with the loss of active lithium ions (LLI), and Peak C shows a significant correlation with the loss of active lithium ions (LLI). A significant correlation refers to the correlation between LAM and LLI. NE As LLI increases, the peak value and peak area of ​​Peak B gradually decrease; as LLI increases, the peak value and peak area of ​​Peak C gradually decrease.

[0111] Diagnosing battery degradation modes using the area of ​​characteristic peaks PeakB and PeakC:

[0112]

[0113] In the formula, Area B,fresh and Area B,aging These represent the areas of the Peak B characteristic peak obtained at C / 25 rate for brand new and aged batteries, respectively; Area C,fresh and Area C,aging These represent the areas of the Peak C characteristic peaks obtained at C / 25 rate for brand new and aged batteries, respectively.

[0114] Preferably, S504 considers the battery capacity at C / 25 rate and corrects the active lithium ion loss (LLI) result in S503, specifically as follows:

[0115]

[0116] In the formula, Capacity fresh This is the capacity of the new battery at C / 25;

[0117] To meet practical application requirements, S505 introduces a rate correction coefficient to optimize the correction result of S504, specifically as follows:

[0118] y LLI =f1 / Area C,aging (27)

[0119] In the formula, y LLI f1 represents the magnification correction factor, and f1 represents the area of ​​the Peak C characteristic peak at different magnifications.

[0120] A system for implementing online diagnosis of battery degradation modes includes an offline calibration module and an online calibration module;

[0121] The offline calibration module is used to obtain the IC curve of the battery to remove polarization effect, and to measure the active lithium ion loss (LLI) and negative electrode active material loss (LAM) at different rates. NE Make corrections and optimizations;

[0122] The online calibration module is used to collect voltage and current during battery operation in real time; it determines whether the battery has reached a state suitable for diagnosing degradation modes based on the voltage data; if it is in a suitable diagnostic state, it obtains the local IC curve based on the voltage and current, and obtains the corresponding peak area based on this. Then, it determines the rate parameter based on the current signal, and then inputs it into the offline calibration module to diagnose the battery's LLI and LAM online. NE .

[0123] Beneficial Effects: This invention establishes a battery degradation mode diagnostic model based on the open-circuit voltage (OCV) curve and a microscopic diagnostic method based on X-ray diffraction (XRD) by combining macroscopic and microscopic characteristic tests. It also performs quantitative verification of the macroscopic-microscopic correlation and ultimately achieves online diagnosis using incremental capacity (IC) curve characteristic peak analysis and rate correction. This improves the accuracy and reliability of battery degradation mode diagnosis, overcomes the limitations of existing macroscopic external characteristic methods that struggle to reveal degradation mechanisms, and microscopic physical disintegration methods that are highly destructive and unable to provide quantitative diagnosis. Simultaneously, it achieves non-destructive, quantitative, and real-time online diagnosis, effectively solving the problem of accurately quantifying complex nonlinear degradation mechanisms within batteries. This meets the requirements for accurate and efficient degradation mode diagnosis in practical battery health management systems. Attached Figure Description

[0124] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0125] Figure 1 This is a flowchart of the method of the present invention;

[0126] Figure 2 The results are the fitting of the potential curves for the positive and negative electrodes.

[0127] Figure 3 The influence of different degradation modes on the electrode potential coordinate system;

[0128] Figure 4 The results of the diagnostic parameters for the degradation mode;

[0129] Figure 5 The results are for battery degradation mode diagnosis based on OCV.

[0130] Figure 6 This is the result of a half-cell capacity test;

[0131] Figure 7 The results are from XRD tests.

[0132] Figure 8 The maximum rate of change in lithium content based on XRD diagnosis;

[0133] Figure 9 A comparison of the negative electrode half-cell capacity with the negative electrode usable capacity diagnosed by the OCV method;

[0134] Figure 10 Comparison of LLI results diagnosed by OCV method and XRD;

[0135] Figure 11 The impact of different degradation modes on the OCV curve;

[0136] Figure 12 This shows the correspondence between the IC curve and the OCV curve.

[0137] Figure 13 The impact of different degradation modes on the IC curve;

[0138] Figure 14 The impact of different degradation modes on the characteristic peak area of ​​the IC curve;

[0139] Figure 15 The intensity and area of ​​the characteristic peaks of the IC curve under different degradation modes;

[0140] Figure 16 Degradation mode diagnostic results based on IC curve characteristic peaks;

[0141] Figure 17 A comparison of degradation pattern results based on IC curves and OCV curves;

[0142] Figure 18 IC curves at different charge / discharge rates;

[0143] Figure 19 Diagnostic results of degradation patterns at different magnifications;

[0144] Figure 20 This is an online diagnostic process for degradation patterns based on IC curves. Detailed Implementation

[0145] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0146] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0147] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0148] like Figure 1 As shown, a method for online diagnosis of battery degradation modes includes the following steps:

[0149] S1. Battery characteristic testing: Macroscopic and microscopic characteristic tests are conducted on the battery. Macroscopic characteristic tests include cycle testing, capacity calibration testing, and charge-discharge testing at different rates. Microscopic characteristic tests include half-cell testing and X-ray diffraction testing, in order to obtain the macroscopic and microscopic characteristics of the battery.

[0150] Cyclic test: The battery is charged with constant current and constant voltage in sequence, and then discharged with constant current. This cycle is repeated. After at least 50 cycles, a capacity calibration test is performed to obtain the current health status of the battery.

[0151] The purpose of battery cycle testing is to obtain battery samples in different health states to provide data support for subsequent degradation mode analysis and diagnosis. The detailed process of the cycle test is shown in Table 1. During the experiment, the battery was charged and discharged using a specific current. In addition, a capacity calibration test was performed every 50 cycles to determine the current health state of the battery.

[0152] Table 1. Detailed steps of the loop test

[0153]

[0154]

[0155] Capacity calibration test: According to the battery capacity calibration test requirements of GB / T31486, the battery is simultaneously subjected to constant current charging and constant voltage charging, followed by constant voltage discharging. This cycle is repeated, and the average discharge amount during the cycle is taken as the capacity calibration value of the battery in its current state; as shown in Table 2:

[0156] Table 2 Specific steps for capacity calibration testing

[0157]

[0158] In this embodiment, 10 lithium iron phosphate batteries were selected as the research objects. The batteries with different states of health (SOH) obtained according to the test steps in Tables 1 and 2 are numbered for ease of explanation, as shown in Table 3:

[0159] Table 3 Battery Number and Health Status

[0160]

[0161] The specific parameters of the selected lithium iron phosphate batteries are shown in Table 4. The battery health status is obtained by dividing the rated capacity by the capacity rating.

[0162] Table 4 Detailed parameters of lithium iron phosphate batteries

[0163]

[0164] Charge-discharge tests at different rates: The battery is charged and discharged at different rates to obtain its charge-discharge characteristics. The specific test rates include C / 25, C / 10, C / 5, C / 3, C / 2, and 1C. The specific test steps are shown in Table 5.

[0165] Table 5 Constant Current Discharge Test Procedure

[0166]

[0167]

[0168] In this embodiment, all the above-mentioned macroscopic characteristic tests are carried out in a constant temperature and humidity chamber to ensure that the temperature and humidity of the battery are constant, so as to reduce the interference caused by environmental factors.

[0169] In the microscopic characteristic test, batteries No.1-4 and No.5-8 were charged to a full charge state and discharged to a completely empty state, respectively. Then, they were disassembled to obtain complete positive and negative electrode plates for subsequent half-cell tests and XRD tests.

[0170] Half-cell testing: Batteries in different health states are charged to full capacity and discharged to empty capacity. They are then disassembled to obtain the positive and negative electrodes. Samples of the intact electrode portions are then taken to assemble the half-cell. Taking the positive electrode as an example, the specific operation process is as follows: First, lithium metal counter electrodes are evenly laid on the positive electrode substrate, and electrolyte is quantitatively injected to achieve interface wetting. The separator and positive electrode are then stacked and assembled sequentially, ensuring that the active material layers are axially symmetrically distributed. Subsequently, the spatial orientation of the electrode is precisely calibrated using positioning fixtures, and mechanical components such as insulating pads and spring sheets are assembled sequentially. Finally, the battery is sealed using a button-type encapsulation process, and allowed to stand for 12 hours to allow the electrolyte to complete the interface penetration process.

[0171] The assembled half-cell was charged and discharged using a rate of C / 25 to reduce internal polarization. It should be noted that rates lower than C / 25 can also be used; the lower the rate, the lower the impact of the current on the internal polarization. Furthermore, the test should be conducted in a constant temperature and humidity chamber to ensure environmental stability.

[0172] X-ray diffraction test: Samples are taken from the intact parts of the positive and negative electrode plates, and the electrodes in different health states are tested using an X-ray diffractometer. The scanning range is 10-80 degrees, and the scanning speed is 0.013 degrees per step.

[0173] To avoid oxidation side reactions caused by exposure of electrode materials to air, the entire process of battery disassembly, half-cell assembly, and XRD electrode sampling must be completed in a glove box protected by inert gas.

[0174] S2. Macroscopic diagnosis of battery degradation mode: Based on the macroscopic characteristics of the battery obtained in S1, a battery degradation mode diagnosis model is established based on the open circuit voltage (OCV) method. Diagnostic parameters are determined and identified, and the diagnostic parameters are decoupled to make quantitative diagnoses of active material loss (LAM) and active lithium ion loss (LLI).

[0175] Battery degradation modes include loss of active lithium ions (LLI) and loss of positive electrode active material (LAM). PE and loss of LAM of negative electrode active material NE The positive electrode active material loses LAM PE Divided into lithium-containing cathode material loss LAM liPE and loss of LAM without lithium cathode material dePE The negative electrode active material loses LAM NE Lithium-containing anode material loss LAM liNE and loss of LAM without lithium anode material deNE ;

[0176] The establishment of the battery degradation mode diagnostic model specifically involves:

[0177] The electrode potential curves of the positive and negative electrodes of the battery are derived from different coordinate systems, namely, the positive electrode lithium intercalation rate-positive electrode potential coordinate system x. PE -y PE Negative electrode lithium insertion rate - negative electrode potential coordinate system x NE -y NE The polar coordinate system x can be used. APE -y APE and the available negative polar coordinate system x ANE -y ANE ;

[0178] Accurately fitting the electrode potential curve is a prerequisite for constructing a diagnostic model for battery OCV degradation modes. Since this invention uses a lithium iron phosphate battery, its mathematical expression needs to be improved based on the specific characteristics of the electrode potential curve for this type of battery. Fitted electrode potential curve:

[0179] For a half-cell made of a positive electrode, the potential curve mainly consists of two single-phase regions and a two-phase coexistence region sandwiched in between. To improve the accuracy of the potential curve fitting, two sets of hyperbolic tangent functions are introduced to fit the electrode potential curve. The specific fitting expression is as follows:

[0180]

[0181] In the formula, U p This represents the positive electrode potential, and the letter af represents the fitting parameters.

[0182] For a half-cell made of a negative electrode, there are many phase transition regions. To improve the accuracy of the potential curve fitting, three sets of hyperbolic tangent functions are introduced to fit the characteristics of the phase transition regions. The specific fitting expression is as follows:

[0183]

[0184] In the formula, U n Indicates the negative electrode potential;

[0185] Figure 2 A comparison between the fitting results of the positive and negative electrode potential curves and the measured results is given. It can be seen that the fitting results are not only consistent with the measured results in terms of trend, but also the errors are all controlled within 0.05V, which can be used for subsequent OCV-based degradation mode diagnosis.

[0186] Since the OCV of a battery is determined by the potentials of its positive and negative electrodes and the matching relationship between them, it is necessary to unify the positive and negative coordinate systems into a single coordinate system through coordinate transformation. To more intuitively and clearly illustrate the influence and transformation relationship of each degradation mode on the electrode coordinate transformation, this is uniformly expressed in... Figure 3 middle.

[0187] The lithium insertion rate of its positive and negative electrodes, i.e., x p x n Transform into x ANE By subtracting the magnitudes of the potentials of the positive and negative electrodes, the full cell OCV can be obtained, which can then be used to construct a battery degradation mode diagnostic model. The battery degradation mode diagnostic model is shown below:

[0188]

[0189] In the formula, U p U represents the positive electrode potential; n Indicates negative electrode potential; C PE Indicates positive electrode capacity; C NE Indicates the negative electrode capacity;

[0190] The determination and identification of diagnostic parameters specifically refers to:

[0191] For the positive electrode, its usable capacity is affected by C. PE LAM liPE LAM dePE The combined effect; therefore, the available capacity of the positive electrode is defined as the first diagnostic parameter X1, as follows:

[0192] X1 = C APE =C PE -LAM liPE -LAM dePE (4)

[0193] For the positive electrode, its usable capacity is affected by C. NE LAM liNE LAM deNE The combined effect; therefore, the available capacity of the positive electrode is defined as the second diagnostic parameter X2, as follows:

[0194] X2 = C ANE =C NE -LAM liNE -LAM deNE (5)

[0195] The position where the available negative electrode lithium insertion rate is zero is affected by the matching relationship between the positive and negative electrodes; therefore, the offset of the coordinate system is defined as the third diagnostic parameter X3, as follows:

[0196] X3 = LAM liNE -LAM dePE +LLI (6)

[0197] The combined formulas (3)-(6) simplify the battery degradation mode diagnostic model as follows:

[0198]

[0199] Since the lithium intercalation rate of the electrode cannot be directly measured experimentally, it is necessary to establish the relationship between the lithium intercalation rate of the battery electrode and the battery capacity. When the battery reaches the discharge cutoff voltage, the lithium intercalation rate of the negative electrode can be defined as the fourth diagnostic parameter X4, resulting in the following equation:

[0200]

[0201] In the formula, LR neg,100 This indicates the lithium intercalation rate of the available negative electrode when the battery reaches the charging cutoff voltage; Q aging Indicates the capacity of the degraded battery;

[0202] The relationship between the available lithium intercalation rate of the negative electrode and the battery capacity is as follows:

[0203]

[0204] In the formula, Capacity(i) represents the battery's charging capacity; x ANE,i This represents the corresponding lithium insertion rate, and i represents the charging time.

[0205] By integrating equations (8)-(9), we obtain:

[0206]

[0207] The optimal diagnostic parameters are determined based on the minimum root mean square error between the constructed OCV and the measured OCV; the optimization objective is as follows:

[0208]

[0209] In the formula, OCV(i) is obtained through low-rate discharge experiments; OCV(x) ANE,i The data is constructed using electrode potential curves; n is the total amount of data.

[0210] Particle Swarm Optimization (PSO) algorithms can maintain strong global search capabilities during the optimization process and are applicable to continuous, discrete, and multi-objective optimization problems. Figure 4 The results of diagnostic parameter identification using the PSO algorithm are presented. It can be seen that the available negative electrode capacity has declined significantly, while the available positive electrode capacity has changed little.

[0211] Assuming the probability of breakage of the active material is equal under different lithium insertion rates, then the lithium insertion rate of the isolated portion in one cycle should be equal to the average lithium insertion rate of the usable negative electrode. Therefore, the loss of the lithium-containing active material can be obtained:

[0212]

[0213] In the formula, x crk,p,fresh Indicates the lithium content when the positive electrode of a brand-new battery is broken; x crk,p,aging Indicates the lithium content when the positive electrode of an aged battery breaks; x crk,n,fresh Indicates the lithium content when the negative electrode of a brand-new battery is broken; x crk,n,aging Indicates the lithium content when the negative electrode of an aged battery breaks; X 1,fresh Indicates the usable capacity of a brand new battery positive terminal; X 1,aging Indicates the usable capacity of the positive electrode of an aged battery; X 2,fresh Indicates the usable capacity of a brand new battery negative terminal; X 2,aging Indicates the usable capacity of the negative electrode of an aged battery;

[0214] The simultaneous equations (6)-(7), (10), and (12) provide a quantitative diagnosis of active material loss (LAM) and active lithium ion loss (LLI). Figure 5 The results of the degradation pattern diagnosis are presented, showing that LLI and LAM... liNE With LAM deNE It increases in size as the battery ages, while LAM liPE and LAM dePE The size remains essentially unchanged. Among them, LLI and LAM... NE The maximum capacity losses were 3.40 Ah and 2.83 Ah, respectively. Based on these results, it can be inferred that the battery degradation is mainly caused by LLI and LAM. NE cause.

[0215] S3. Microscopic diagnosis of battery degradation mode: Based on the microscopic characteristics obtained in S1, the half-cell capacity change and the rate of change of the maximum lithium content of the battery are diagnosed.

[0216] Changes in the active materials of the positive and negative electrodes are closely related to battery degradation. However, battery capacity loss is the result of the combined effects of LLI and LAM, making it difficult to separate and test LLI and LAM. Since pure lithium sheets can be used as the counter electrode in a half-cell, which is considered to have sufficient cyclic lithium content, changes in half-cell capacity can be used to quantitatively analyze LAM.

[0217] Figure 6 The results of the half-cell capacity tests for both positive and negative electrodes are presented. As can be seen from the figure, the capacity of the positive electrode half-cell showed almost no significant change, indicating that LAM (Layered Amplitude) occurred almost entirely. PE This phenomenon is consistent with the open-circuit voltage diagnosis. Furthermore, the negative electrode half-cell capacity gradually decreases with battery degradation, indicating a significant LAM (Last Ampere) phenomenon. NE This trend is consistent with the open-circuit voltage diagnostic results, indicating that the two results are highly consistent in the qualitative diagnosis of degradation modes.

[0218] It should be noted that batteries No. 1-4 were fully charged during disassembly. At this point, the initial state of the assembled negative electrode half-cell was lithium-carbon compound. However, lithium-carbon compounds are highly unstable, leading to unreliable capacity test results. Therefore, Figure 6 (b) Only the results of the negative electrode capacity test for batteries No.5-8 are given.

[0219] Figure 7 The XRD test results for different electrodes are presented. It can be seen that the diffraction peaks of the positive electrode material are not only relatively dispersed but also overlap. In contrast, the diffraction peaks of the negative electrode are more concentrated and easier to distinguish between different component types. Therefore, the diffraction peaks of the negative electrode can be used to characterize the battery degradation mode.

[0220] from Figure 7 (c) It can be seen that, under a fully charged state, LiC 12 The peak values ​​of the LiC6 diffraction peaks show significant changes; for example, the intensity of the LiC6 diffraction peak decreases with battery degradation. In other words, the number of lithium ions that can be intercalated decreases with increasing degradation. Figure 7 (d) In the negative electrode sheet disassembled under empty state, it can be seen that as the battery degrades, the intensity of the carbon diffraction peak gradually weakens, indicating that the loss of negative electrode active material is continuously increasing.

[0221] contrast Figure 7 (c) and 7(d) show that the total amount of lithium ions intercalated in a single full charge state can approximately represent the maximum lithium content.

[0222] By utilizing the diffraction peak intensity of the lithium-carbon compound in the negative electrode and normalizing the carbon content, the maximum lithium content corresponding to the battery's transition from a fully charged state to a fully discharged state is calculated. The calculation formula is as follows:

[0223]

[0224] In the formula, x1 is the diffraction intensity of the LiC6 peak; x2 is the diffraction intensity of the LiC6 peak. 12 The diffraction intensity of the peak; Li x C is the lithium-carbon compound content calculated from the diffraction peak intensity;

[0225] To analyze the relationship between maximum lithium content and battery degradation mode, the maximum lithium content needs to be represented as a percentage, and the calculation formula is as follows:

[0226]

[0227] In the formula, LLI XRD This indicates the maximum rate of change in lithium content diagnosed by XRD testing, while Li x C fresh and Li x C aging These represent LiC6 for new batteries and aged batteries, respectively.

[0228] Based on formula (14), the XRD diffraction peaks of the negative electrode under full charge were quantitatively calculated, thereby obtaining the maximum lithium content conversion rate of the battery under different aging conditions. The results are as follows: Figure 8 As shown in the figure, the rate of change of maximum lithium content generally exhibits a positive correlation with battery degradation.

[0229] S4. Based on the diagnostic results of S2 and S3, perform a quantitative diagnosis of the macro-micro correlation of battery degradation modes to verify the correctness of the quantitative diagnostic results made by the battery degradation mode diagnostic model.

[0230] Figure 9 The LAM (Label-Ampere) method for half-cell capacity and OCV (Optical Characteristic Value) diagnosis is given. NE The results comparison shows that the half-cell capacity loss is related to LAM. NE The consistent trend indirectly validates the effectiveness of LAM based on OCV diagnostics. NE The accuracy.

[0231] LAM NE Correlate with half-cell capacity:

[0232] Since the unit of negative electrode half-cell capacity is mAh, and the battery degradation mode diagnostic model diagnoses the loss of negative electrode active material LAM...NE The units are Ah, and the two cannot be directly compared; therefore, to achieve quantitative correlation diagnosis of degradation patterns, it is necessary to normalize the two sets of data; the specific calculation formula is as follows:

[0233]

[0234] In the formula, LAM NE,Cap Indicates the capacity change after a half-cell test; LAM NE,OCV Indicates LAM diagnosed based on OCV method NE C NE,Cap,fresh and C NE,Cap,aging These represent the negative electrode capacities of new and degraded batteries, respectively; LAM NE,OCV,fresh and LAM NE,OCV,aging These represent the LAM values ​​of new and degraded cells diagnosed using the OCV method, respectively. NE ;

[0235] Substituting the results of OCV-based diagnostics and half-cell capacity testing into equation (15), the specific results are shown in Table 6. As can be seen from the table, the values ​​are almost identical, with a maximum deviation of only 1.35%. This result indicates that the battery can diagnose LAM in degradation mode. NE In this regard, micro- and macro-level quantitative correlations were achieved, further validating the diagnostic efficacy of LAM based on the OCV method. NE The feasibility and correctness of this.

[0236] Table 6. LAM Diagnosis in Micro and Macroscopic Methods NE Results Comparison

[0237]

[0238]

[0239] Correlate the maximum rate of change in lithium content with LLI:

[0240] There is also the problem of inconsistent dimensions; therefore, it is necessary to convert the LLI results calculated based on OCV into percentage form:

[0241]

[0242] In the formula, LLI OCV This indicates LLI diagnosed using the OCV method; LLI fresh and LLI aging LLI represents new batteries and aged batteries, respectively;

[0243] further, Figure 10(a) Comparison of LLI results based on OCV and XRD diagnostics is presented. As can be seen from the figure, the calculation results from the two methods show a consistent trend, but there is a significant difference in numerical values, with maximum differences of 17.18% and 24.96%, respectively, a difference of 7.78%. The main reason for this is that the XRD-based diagnostic method uses carbon normalization when calculating the maximum lithium content. However, the half-cell capacity test results for the negative electrode show a significant LAM (Lithium Ampere) during battery degradation. NE In other words, LLI results diagnosed based on XRD include LAM. Therefore, they need to be corrected to obtain accurate results.

[0244] The results in Table 6 show that the negative electrode half-cell capacity can be used to characterize LAM. NE Meanwhile, the capacities of batteries No.1-4 and No.5-8 before disassembly were similar, making the test results of the two sets of batteries highly comparable. Therefore, the LAM of battery No.5-8 was obtained using a linear interpolation method based on the capacity of battery No.1-4. NE This leads to the correction calculation of the maximum lithium content change rate, and the calculation method is as follows:

[0245]

[0246] In the formula, Li x C' represents the modified Li x C;

[0247] Therefore, we can conclude that:

[0248]

[0249] LLI results based on OCV and XRD diagnosis are as follows: Figure 10 As shown in (b), the two corrected results are not only consistent in trend but also roughly the same in value, with a maximum error of only 1.68%. This result demonstrates that the battery achieves a quantitative correlation between micro and macroscopic parameters in diagnosing LLI (Limited Intake) degradation modes, further validating the accuracy of LLI diagnosis based on the OCV method. Thus, the correctness of LLI and LAM calculated quantitatively based on the OCV method has been verified by half-cell capacity and XRD diffraction peak values, respectively, laying a theoretical foundation for subsequent diagnosis of battery degradation modes based on IC curves.

[0250] S5. Online diagnosis of battery degradation mode: Construct batteries with different degradation levels, establish OCV curves, and obtain IC curves based on OCV curves. Diagnose degradation modes by analyzing the area of ​​their characteristic peaks, and use capacity and rate correction coefficients to correct and optimize the diagnostic results.

[0251] S501: Based on the diagnostic results in S2, construct batteries with different degradation modes, plot the OCV curves of the batteries, and analyze the impact of different degradation modes on the OCV curves.

[0252] In battery construction with varying degrees of degradation, uncertainties during the aging process and variations in manufacturing techniques can lead to overall shifts or local slope changes in the OCV curve, hindering quantitative analysis. Furthermore, isolating degradation modes to analyze their impact on the OCV curve is a challenging task. To address this, using the degradation mode diagnosis results based on the OCV method as a reference, batteries with different degradation modes were constructed using a controlled variable approach. The results are shown in Table 7.

[0253] Table 7. Specific degradation mode values ​​for the constructed battery

[0254]

[0255] In the analysis of the impact of degradation modes on the OCV curve, since the capacity of the negative electrode decreases with aging, while the capacity of the positive electrode remains essentially unchanged during aging; therefore, x is used... PE The impact of degradation modes on the OCV curve is analyzed, and the specific calculation method is as follows:

[0256]

[0257] In the formula, x PE Indicates the lithium intercalation rate of the positive electrode; U p U represents the positive electrode potential; n C represents the negative electrode potential; PE Indicates positive electrode capacity; C NE Indicates the nearby capacity; LAM liPE Indicates the loss of cathode material content; LAM dePE Indicates the absence of lithium cathode material loss; LAM liNE Indicates the loss of lithium-containing anode materials; LAM deNE This indicates that there is no loss of lithium-ion anode material.

[0258] x PE The coordinate range is as follows:

[0259]

[0260] Batteries with different degrees of degradation were constructed and then brought into x. PE The coordinate system allows for the analysis of the impact of different degradation modes on the OCV curve. Figure 11 The effects of different degradation modes on the OCV curve are presented. It can be seen that as LLI increases, the lithium insertion rate range gradually shrinks; this phenomenon also occurs in LAM. liNEThe evolution of the OCV curve, however, LAM deNE The range of lithium insertion rate does not change significantly as it increases.

[0261] In actual battery degradation, it is usually the result of multiple degradation modes working together. Figure 11 (d) illustrates the coupling effect of various degradation modes on the OCV curve. As can be seen from the figure, the lithium insertion rate range gradually shrinks with increasing battery degradation, and a significant compression phenomenon occurs near the voltage phase transition point. Compared to a single degradation mode, the lithium insertion rate range resulting from the superposition of multiple degradation modes is even smaller, and this superposition effect exacerbates the loss of battery capacity.

[0262] The diagnostic process based on OCV is relatively cumbersome and difficult to implement in practice. Since the IC curve is the derivative of the OCV curve and contains rich aging characteristic information, while also being computationally efficient and cost-effective, it can be applied to online diagnosis of battery degradation modes.

[0263] In the process of obtaining IC curves, noise is unavoidable during the experiment. Using traditional numerical differentiation (dQ / dV) methods to obtain IC curves would not only hinder the extraction of battery degradation characteristics but also negatively impact the quantitative diagnosis of degradation modes. Therefore, a polynomial central smoothing method will be employed to obtain IC curves.

[0264] S502: The IC curves of batteries with different degradation modes are obtained by using the polynomial center smoothing method based on the OCV curve;

[0265] Specifically, S502 is as follows:

[0266] Using k-th order polynomial curves By fitting the data and minimizing the sum of squares of the differences between the actual and fitted values ​​according to the least squares principle, the coefficients 'a' of each order of the polynomial can be obtained. i This transforms the solution to the IC curve into solving for the coefficient a. i Solve for it.

[0267] Let p = 2m + 1, and the p data points be (x... -m ,y -m ),(x -m+1 ,y -m+1 ),...,(x0,y0),...,(x m ,y m If ), then the fitting matrix can be expressed as:

[0268]

[0269] That is:

[0270] Y(2m+1) =X (2m+1)×k ·A k×1 +E (2m+1)×1 (twenty two)

[0271] In the formula, A is the coefficient matrix; E is the deviation matrix.

[0272] The objective function is:

[0273]

[0274] The least squares solution to the coefficient matrix A is:

[0275]

[0276] S503: Obtain all characteristic peaks on the IC curve, and select those corresponding to the loss LAM of the negative electrode active material. NE The characteristic peak that shows a significant correlation with the loss of active lithium ions (LLI) can be used to diagnose the degradation mode of the battery.

[0277] Figure 12 The correspondence between the OCV curve and the IC curve is presented. The three relatively clear phase transition processes on the OCV curve can be observed as corresponding characteristic peaks on the IC curve. Based on their order of appearance during charging, they can be divided into Peak A, Peak B, and Peak C from left to right. These characteristic peaks are closely related to the electrochemical reaction process of the battery.

[0278] Figure 13 The effects of different degradation modes on the IC curves are presented. It can be seen that with the increase of LLI, the intensity of the Peak C characteristic peak gradually weakens, while the shapes of other characteristic peaks remain basically stable. This is mainly because some lithium ions are consumed during SEI film formation, resulting in a reduction in the total amount of recyclable lithium. Furthermore, according to... Figure 3 Coordinate transformation reveals that LLI's effect on the battery open-circuit voltage curve manifests as a shift in the overall curve, and also causes changes in the aforementioned characteristic peaks. (Comparison) Figure 13 As shown in (b) and (c), with LAM deNE and LAM liNE The intensity of Peak B decreased with the increase of Peak B, but the intensity of Peak C showed an increasing and decreasing trend, respectively.

[0279] Figure 14 Given Figure 13 Peak areas of Peak B and Peak C are shown in (a) and (d). Considering the effect of the single-decay model on the IC curve, it can be seen from the figure that the effect of LLI on the peak area of ​​Peak B remains essentially constant. However, the peak area of ​​Peak B increases with LAM... NEThe value decreases as the capacity loss increases, and this phenomenon is corroborated by the decrease in the Peak B peak of the IC curve for multifactor capacity loss. Therefore, changes in Peak B can be used to perform LAM analysis. NE Quantitative calculations were performed. Furthermore, the peak area of ​​PeakC was the same under the influence of multiple factors and LLI alone, because LAM... deNE and LAM liNE The effects on Peak C can be offset, so that the peak area variation of Peak C is mainly dominated by LLI.

[0280] In the characteristic analysis of the IC curve, a high correlation was found between changes in the characteristic peak and the decay pattern. Therefore, Figure 15 The peak height and peak area of ​​Peak B under different degradation modes were statistically analyzed. The figures show that the peak height and peak area of ​​the characteristic peaks exhibit a high degree of consistency in their impact on the degradation mode. Therefore, this invention attempts to characterize the battery degradation mode from the changes in peak intensity and peak area, and proposes the following formulas to calculate LLI and LAM respectively. NE :

[0281]

[0282] In the formula, Area B,fresh and Area B,aging These represent the areas of the Peak B characteristic peak obtained at C / 25 rate for brand new and aged batteries, respectively; Area C,fresh and Area C,aging These represent the areas of the Peak C characteristic peaks obtained at C / 25 rate for brand new and aged batteries, respectively.

[0283] The calculations for characterizing each decay mode using peak intensity changes are as follows:

[0284]

[0285] In the formula, Peak B,fresh and Peak B,aging These are the Peak B strength values ​​of a new battery and an aged battery at C / 25, respectively; Peak C,fresh and Peak C,aging These are the Peak C strengths of a new battery and an aged battery at C / 25, respectively.

[0286] Figure 16 The calculation results of equations (25)-(26) are presented, along with the diagnostic results based on the OCV curve. It can be seen from the figure that although the trends of the diagnostic results based on the peak area and peak height of the IC curve are consistent with those based on the OCV curve, the error when using the peak area for calculation is significantly smaller than that when using the peak height, especially for LAM. NETherefore, the calculation results based on peak area will be used as a reference in subsequent analyses.

[0287] However, the figure also shows that there are significant deviations in the LLI calculation results for both characteristics. The main reason is that the effect of LLI on the OCV curve is primarily an overall shift, determined by the entire stoichiometric window, rather than changes in local peaks or peak areas. However, the methods described above normalize the LLI calculations according to the characteristic peaks of the new battery, thus failing to fully reflect the effect of LLI on the OCV curve shift, leading to a large error compared to the actual situation.

[0288] S504: Based on the problems in calculating LLI in S503, and considering the battery capacity at C / 25 rate, the LLI result of active lithium ion loss in S503 is corrected.

[0289] S504 considers the battery capacity at C / 25 rate and corrects the active lithium ion loss (LLI) result in S503, specifically as follows:

[0290]

[0291] In the formula, Capacity fresh This is the capacity of the new battery at C / 25;

[0292] Based on equation (27), the experimental data under laboratory C / 25 were input, and the calculation results based on IC were compared with the diagnostic results based on OCV curves, such as... Figure 17 As shown in the figure, the calculated cyclic lithium-ion loss and negative electrode active material loss from the two methods not only show a high degree of consistency in their trends but also exhibit good numerical agreement. Specifically, the maximum difference for LLI is only 1.79%, while for LAM... NE The calculated maximum difference is 1.62%.

[0293] S505: To meet practical applications, a rate correction factor is introduced to optimize the correction result of S504;

[0294] To minimize the interference of battery polarization on diagnostic results, the aforementioned analysis was primarily based on discharge curves at a low rate of C / 25. However, in actual electric vehicle use, it is rare for batteries to reach a rate of C / 25. Therefore, Figure 18The IC curves of this battery at different charge / discharge rates are presented. As can be seen from the figures, with increasing rate, the characteristic peak of the IC curve shifts towards higher voltages during the charging phase and towards lower voltages during the discharging phase. This is mainly due to battery polarization. This phenomenon is more pronounced at high charge / discharge rates, resulting in more significant peak shifts; peak sticking even occurs at C / 2 rate. Furthermore, the peak value of the IC curve decreases with increasing rate. Therefore, it is necessary to further analyze the applicability of degradation mode diagnosis based on the characteristic peaks of the IC curve under different rate conditions.

[0295] Considering that the characteristic peak of the IC curve of a battery changes at different rates, the aforementioned LLI and LAM... NE The calculation formula has been modified as follows:

[0296]

[0297] In the formula, the subscript rate represents the multiplier;

[0298] By substituting the characteristic peak data at different magnifications into the above formula, the online diagnostic results at different magnifications can be obtained, as shown in the following example. Figure 19 As shown. It should be noted that the rate study here only extends to C / 3. The main reason is that at higher rates, the battery electrochemistry and concentration polarization increase, leading to the merging or disappearance of intermediate characteristic peaks with surrounding characteristic peaks. LAM was diagnosed at C / 10, C / 5, and C / 3 rates. NE LAM diagnosed with C / 25 NE The results were largely consistent, but the diagnostic results fluctuated at different magnifications. The cause was attributed to LAM (Laminated Angiography). NE It is calculated based on Peak B, and its variation is relatively small. Furthermore, Peak B is located between Peak A and Peak C, as... Figure 18 As shown in the circular frame, this peak is an asymmetric peak, and the calculation method of obtaining the IC curve using polynomial center smoothing may have affected its calculation results.

[0299] Furthermore, such as Figure 19 As shown in (b), the diagnostic jitter for LLI is almost non-existent at different magnifications, but shows an overall decreasing trend. The maximum absolute errors at C / 10, C / 5, and C / 3 magnifications are 3.25%, 4.77%, and 6.07%, respectively. Comparing the LLI diagnostic results at different magnifications reveals a negative correlation between the decreasing trend of LLI and magnification, with a relatively obvious linearity. Therefore, future research will explore magnification-corrected LLI diagnostic methods to improve diagnostic accuracy at different magnifications.

[0300] Figure 20A degradation mode diagnosis process based on IC curves is presented for practical applications. First, the battery voltage and current data are acquired online in real time. Next, the characteristic peak Peak B or Peak C is determined based on the acquired voltage data. Then, LAM is calculated using equations (28) and (29), respectively. NE And LLI. Further, the charge / discharge rate of the battery is calculated using the collected current data, and this rate is then input into an offline calibration function for LLI rate correction. Alternatively, an online rate correction function calibration can be performed based on actual collected data, such as... Figure 20 The dashed box in the image is shown. For ease of calculation, [the text is incomplete]. Figure 19 The relationship between LLI and SOH was converted into the relationship between capacity loss rate and LLI, and the corresponding rate correction coefficient was obtained. The calculation method is shown in Equation 30, and the results are shown in Table 8. The sum of capacity loss rate and health status is 100%.

[0301] y LLI =f1 / Area C,aging (30)

[0302] In the formula, y LLI f1 represents the magnification correction factor, and f1 represents the area of ​​the Peak C characteristic peak at different magnifications;

[0303] Table 8 LLI Correction Factors at Different Magnification Ratios

[0304]

[0305] Substitute the corresponding magnification correction factor into Figure 18 The calculation results yielded the corrected LLI results, as shown in Table 9. The table shows that the corrected results not only maintain the same trend as the results at the baseline magnification C / 25, but also that the maximum deviation of the calculation error decreased from 6.07% to 1.68% compared to the uncorrected version, demonstrating the feasibility of the magnification-based correction method.

[0306] Table 9. Corrected LLI diagnostic results at different magnification ratios.

[0307]

[0308] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0309] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of diagnosing a battery degradation pattern online, characterized by, The method comprises the following steps: S1, battery characteristic test: macroscopic characteristic test and microscopic characteristic test are performed on the battery, the macroscopic characteristic test comprises cycle test, capacity calibration test, different rate charge and discharge test, the microscopic characteristic test comprises half-cell test and X-ray diffraction test, so as to obtain the macroscopic characteristics and the microscopic characteristics of the battery; S2, macroscopic diagnosis of battery degradation mode: based on the battery macroscopic characteristics obtained in S1, a battery degradation mode diagnosis model is established based on an open circuit voltage (OCV) method, diagnosis parameters are determined and identified, the diagnosis parameters are decoupled, and quantitative diagnosis is made on active material loss (LAM) and active lithium ion loss (LLI); S3, microscopic diagnosis of battery degradation mode: based on the microscopic characteristics obtained in S1, the half-cell capacity change and the maximum lithium content change rate of the battery are diagnosed; S4, macro-micro correlation quantitative diagnosis of the battery degradation mode is performed based on the diagnosis results of S2 and S3, and the correctness of the quantitative diagnosis result made by the battery degradation mode diagnosis model is verified; S5, online diagnosis of the battery degradation mode: batteries with different degradation degrees are constructed, an OCV curve is established, an IC curve is obtained based on the OCV curve, the degradation mode is diagnosed by analyzing the area of the characteristic peak, and the diagnosis result is corrected and optimized by using the capacity and rate correction coefficients.

2. The method of online diagnosing battery degradation mode of claim 1, wherein, The S1 is specifically: Cycle test: the battery is sequentially subjected to constant current and constant voltage charging, and then subjected to constant current discharging, so as to cycle, at least 50 cycles are performed each time, and then a capacity calibration test is performed to obtain the health status of the current battery; Capacity calibration test: the battery is subjected to constant current charging and constant voltage charging at the same time, and then subjected to constant voltage discharging, so as to cycle, and the average discharge capacity in the cycle process is taken as the capacity calibration value of the current state of the battery; Different rate charge and discharge test: the battery is subjected to different rate charge and discharge to obtain the charge and discharge characteristics of the battery, and the test rate specifically includes C / 25, C / 10, C / 5, C / 3, C / 2 and 1C; Half-cell test: the battery with different health statuses is charged to a full charge state and discharged to an empty charge state, then the positive and negative pole pieces of the battery are obtained by disassembling, then the intact parts of the pole pieces are sampled to assemble half-cells, and then C / 25 rate charging or discharging test is performed to obtain the positive and negative pole piece capacity under different health statuses; X-ray diffraction test: the intact parts of the positive and negative pole pieces are sampled, and the X-ray diffractometer is used to test the pole pieces with different health statuses, the scanning range is 10-80 degrees, and the scanning speed is 0.013 degrees per step.

3. The method of online diagnosing battery degradation modes of claim 1, wherein, The S2 is specifically: The degradation modes of the battery include loss of active lithium ions LLI, loss of positive active material LAM PE and loss of negative active material LAM NE , the loss of positive active material LAM PE is divided into loss of lithium-containing positive material LAM liPE and loss of lithium-free positive material LAM dePE , the loss of negative active material LAM NE is divided into loss of lithium-containing negative material LAM liNE and loss of lithium-free negative material LAM deNE ; The establishment of the battery degradation mode diagnosis model is specifically: The electrode potential curves of the positive and negative electrodes of the battery are from different coordinate systems, i.e. the positive electrode lithium intercalation rate-positive electrode potential coordinate system x PE -y PE , the negative electrode lithium intercalation rate-negative electrode potential coordinate system x NE -y NE , the available positive electrode coordinate system x APE -y APE , and the available negative electrode coordinate system x ANE -y ANE ; Fitting of electrode potential curve: For the half-cell made of the positive pole piece, the potential curve is mainly composed of two single-phase regions and one two-phase coexistence region in the middle, in order to improve the fitting accuracy of the potential curve, two sets of hyperbolic tangent functions are introduced to fit the electrode potential curve, and the specific fitting expression is: where U p represents the positive electrode potential, and the letters a-f represent the fitting parameters; For the half-cell made of the negative pole piece, there are more phase transition regions, in order to improve the fitting accuracy of the potential curve, three sets of hyperbolic tangent functions are introduced to fit the phase transition region characteristics, and the specific fitting expression is: In the formula, U n represents the potential of the negative electrode Since the OCV of the battery is determined by the potential of the positive and negative electrodes and the matching relationship therebetween, the positive and negative coordinate systems need to be unified into the same coordinate system through coordinate transformation, The lithium intercalation rate of the positive and negative electrodes, i.e., x p , x n is converted to x ANE , and the full battery OCV is obtained by subtracting the positive electrode potential from the negative electrode potential, which is used to construct a battery degradation mode diagnosis model as follows: wherein U p represents the positive electrode potential; U n represents the negative electrode potential; C PE represents the positive electrode capacity; C NE represents the negative electrode capacity; The specific determination and identification of the diagnosis parameter are: For the positive electrode, its available capacity is subject to C PE , LAM liPE , LAM dePE the influence of the co-action; thus, the available capacity of the positive electrode is defined as the first diagnostic parameter X1, as follows: X1= C APE = C PE -LAM liPE -LAM dePE (4) For the positive electrode, its available capacity is subject to C NE , LAM liNE , LAM deNE the influence of the co-action; thus, the positive electrode available capacity is defined as the second diagnostic parameter X2, as follows: X2= C ANE = C NE -LAM liNE -LAM deNE (5) The available negative electrode lithium intercalation rate is affected by the matching relationship between the positive and negative electrodes; therefore, the offset of the coordinate system is defined as the third diagnosis parameter X3, and the specific definition is as follows: X3 = LAM liNE - LAM dePE + LLI (6) According to formulas (3)-(6), the battery degradation mode diagnosis model is simplified as: Since the electrode lithium intercalation rate cannot be directly measured by experimental methods, the relationship between the electrode lithium intercalation rate and the battery capacity needs to be established; the available negative electrode lithium intercalation rate when the battery reaches the discharge cut-off voltage is defined as the fourth diagnosis parameter X4, and the following equation is obtained: wherein LR neg,100 represents the lithium intercalation rate of the available negative electrode when the battery reaches the charge cut-off voltage; Q aging represents the capacity of the degraded battery; The relationship between the available negative electrode lithium intercalation rate and the battery capacity is: In the formula, Capacity(i) represents the charge capacity of the battery; x ANE,i is the corresponding lithium intercalation rate, i represents the charging time; Through integration of formulas (8)-(9), the following is obtained: According to the minimum root mean square error between the constructed OCV and the measured OCV, the optimal diagnosis parameters are determined; the optimization target is as follows: In the formula, OCV(i) is obtained through a low rate discharge experiment; OCV(x ANE,i ) is constructed through an electrode potential curve; n is the total amount of data; The decoupling diagnosis parameter makes quantitative diagnosis of the active material loss LAM and the active lithium ion loss LLI, and the specific diagnosis is as follows: Assuming that the probability of active material crushing under different lithium intercalation rates is equal, the lithium intercalation rate of the isolated part in one cycle should be equal to the average lithium intercalation rate of the available negative electrode; therefore, the active material loss of the lithium-containing part is obtained as follows: wherein x crk,p,fresh represents the lithium content when the fresh battery cathode is broken; x crk,p,aging represents the lithium content when the aged battery cathode is broken; x crk,n,fresh represents the lithium content when the fresh battery anode is broken; x crk,n,aging represents the lithium content when the aged battery anode is broken; x 1,fresh represents the available capacity of the fresh battery cathode; x 1,aging represents the available capacity of the aged battery cathode; x 2,fresh represents the available capacity of the fresh battery anode; x 2,aging represents the available capacity of the aged battery anode; The simultaneous equations (6)-(7), (10), (12) realize quantitative diagnosis of different degradation modes; according to the diagnosis result, active lithium ion loss LLI and negative active material loss LAM are obtained NE is the main degradation mode of the battery.

4. The method of online diagnosing battery degradation modes of claim 1, wherein, The S3 is specifically as follows: Based on the positive and negative electrode sheet capacities at different states of health obtained from the half-cell test in S1, it is derived that due to the presence of the negative active material loss LAM NE causes the negative half-cell capacity to gradually decrease as the battery degrades; According to the diffraction peak distribution of the positive and negative electrodes obtained by X-ray diffraction testing in S1, it is found that the diffraction peak distribution of the negative electrode is more concentrated and each component type is easy to distinguish; therefore, the diffraction peak of the negative electrode is used to characterize the battery degradation mode; The maximum lithium content of the battery from the full charge state to the full discharge state is calculated by normalizing the carbon using the diffraction peak intensity of the negative electrode lithium-carbon compound, and the calculation formula is as follows: wherein x1 is the intensity of the LiC6 peak; x2 is the intensity of the LiC 12 peak; Li x C is the lithium carbon content calculated from the intensity of the diffraction peak; In order to analyze the relationship between the maximum lithium content and the battery degradation mode, the maximum lithium content is characterized in the form of percentage, and the calculation formula is as follows: wherein LLI XRD represents the maximum lithium content change rate diagnosed through XRD testing, and Li x C fresh and Li x C aging represents LiC6of a new battery and an aged battery, respectively; Based on formula (14), the X-ray diffraction peak of the negative electrode plate in the full charge state is quantitatively calculated, thereby obtaining the maximum lithium content change rate of the battery under different aging states; the diagnosis result based on the X-ray diffraction peak intensity shows that the maximum lithium content change rate is positively correlated with the health state of the battery.

5. The method of online diagnosing battery degradation modes of claim 1, wherein, The S4 is specifically as follows: LAM NE Correlation with half-cell capacity: Since the unit of negative electrode half-cell capacity is mAh, and the battery degradation mode diagnostic model diagnoses the loss of negative electrode active material LAM... NE The units are Ah, and the two cannot be directly compared; therefore, to achieve quantitative correlation diagnosis of degradation patterns, it is necessary to normalize the two sets of data; the specific calculation formula is as follows: where LAM NE,Cap represents the capacity change by half-cell test; LAM NE,OCV represents the LAM NE based on OCV method diagnosis NE,Cap,fresh ; C NE,Cap,aging and C NE,OCV,fresh represent the anode capacity of the new battery and the degraded battery, respectively; LAM NE,OCV,aging and LAM NE represent the LAM of the new battery and the degraded battery based on OCV method diagnosis, respectively The maximum lithium content change rate is associated with LLI: The LLI result calculated based on OCV is converted into the form of percentage: where LLI OCV represents LLI diagnosed based on the OCV method; LLI fresh and LLI aging represent LLI of a new battery and an aged battery, respectively; Since there is a large deviation in the association result of the maximum lithium content change rate and LLI, the reason is that the maximum lithium content calculated using the diffraction peak intensity of the negative electrode lithium-carbon compound contains LAM; therefore, it needs to be corrected; LAM NE Results associated with half-cell capacity indicate that the negative half-cell capacity can be used to diagnose LAM NE ; therefore, the calculation formula of the maximum lithium content correction is as follows: wherein Li x C' represents a modified Li x C; Further, the following is obtained: Quantitative correlation of LLI OCV and LLI' XRD The results were compared and the accuracy of diagnosing LLI based on the OCV method was verified.

6. The method of online diagnosing battery degradation modes of claim 1, wherein, The S5 includes: S501: According to the diagnosis result in S2, a battery of different degradation modes is constructed, the OCV curve of the battery is drawn, and the influence of different degradation modes on the OCV curve is analyzed; S502: Based on the OCV curve, the IC curve of the battery of different degradation modes is obtained by using the polynomial center smoothing method; S503: Obtain all characteristic peaks on the IC curve, select the characteristic peaks respectively corresponding to the negative active material loss LAM NE There is a significant correlation between the characteristic peaks and the active lithium ion loss LLI, and the area of the characteristic peaks is used to diagnose the degradation mode of the battery; S504: Considering the battery capacity under the C / 25 rate, the active lithium ion loss LLI result in S503 is corrected; S505: In order to meet the actual application, the rate correction coefficient is introduced to optimize the correction result of S504.

7. The method of online diagnosing battery degradation modes of claim 6, wherein, The S501 is specifically: Using the OCV method to diagnose the degradation mode results as a reference, the control variable method is used to construct a battery with different degradation modes; In the analysis of the impact of degradation mode on OCV curve, since the negative electrode capacity of the battery will decrease with aging, while the positive electrode capacity does not lose substantially during the aging process; therefore, x PE is used to analyze the impact of degradation mode on OCV curve, and the specific calculation method is as follows: wherein x PE represents the lithium intercalation rate of the positive electrode; U p represents the positive electrode potential; U n represents the negative electrode potential; C PE represents the positive electrode capacity; C NE represents the nearby capacity; LAM liPE represents the content positive electrode material loss; LAM dePE represents the lithium-free positive electrode material loss; LAM liNE represents the lithium-containing negative electrode material loss; LAM deNE represents the lithium-free negative electrode material loss; x PE The coordinate ranges for the following are as follows: The built battery with different degradation modes is brought into x PE Coordinate system, analyze the impact of different degradation modes on OCV curve; The S502 is specifically: Adopt k order polynomial curve Fit the data, and make the square sum of the difference between actual value and fitting value minimum according to the principle of least square method, that is, get the coefficient a of each order of polynomial i , so as to convert the solution of IC curve into the solution of coefficient a i ; Let p = 2m + 1, p data points are (x -m ,y -m ),(x -m+1 ,y -m+1 ),...,(x0,y0),...,(x m ,y m ), the fitting matrix is represented as: That is: Y (2m+1) = X (2m+1)×k • A k×1 + E (2m+1)×1 (22) In the formula, A is the coefficient matrix; E is the bias matrix; The objective function is: The least square solution of the coefficient matrix A is:

8. The method of online diagnosing battery degradation mode of claim 7, wherein, The S503 is specifically: All characteristic peaks on the IC curve were obtained, and characteristic peaks Peak B and Peak C were selected, which were obviously correlated with the loss of active lithium ions LLI and the loss of negative active material LAM NE The characteristic peak Peak B and the characteristic peak Peak C were obviously correlated with the loss of active lithium ions LLI and the loss of negative active material LAM NE With the increase of LAM, the peak value and peak area of Peak B gradually decreased; with the increase of LLI, the peak value and peak area of Peak C gradually decreased; Using the area of the characteristic peak PeakB and the characteristic peak PeakC to diagnose the degradation mode of the battery: wherein Area B,fresh and Area B,aging represent the area of the Peak B characteristic peak obtained at C / 25 rate for a fresh battery and an aged battery, respectively; Area C,fresh and Area C,aging represent the area of the Peak C characteristic peak obtained at C / 25 rate for a fresh battery and an aged battery, respectively.

9. The method for diagnosing the degradation mode of the battery online according to claim 8, characterized in that, The S504 considers the battery capacity under the C / 25 rate to correct the active lithium ion loss LLI result in S503, and is specifically: where Capacity is the capacity of the new battery at C / 25; and fresh is the capacity of the new battery at C / 25; and The S505 introduces a rate correction coefficient to optimize the correction result of S504 in order to meet the actual application, and is specifically: y LLI = f1 / Area C,aging (27) wherein y LLI represents a correction factor for the magnification, and f1 represents the area of the Peak C characteristic peak at different magnifications.

10. System implementing the method of diagnosing the degradation mode of a battery on-line according to any one of claims 1 to 9, characterized in that, Including an offline calibration module and an online calibration module; The offline calibration module is used to obtain IC curves of battery removing polarization effect, and simultaneously obtain active lithium ion loss LLI and negative active material loss LAM under different rates NE Carry out correction optimization; The online calibration module is used to collect the voltage and current in the battery operation process in real time; according to the voltage data, it is judged whether the battery can be used to diagnose the degradation mode or not; If in the appropriate diagnostic state, the local IC curve is obtained according to the voltage and current, and the corresponding peak area is obtained based thereon, then the rate parameter is determined according to the current signal, and then it is brought into the offline calibration module to diagnose the LLI and LAM of the battery online NE .

Images