Wide-temperature-interval battery state-of-health estimation method based on electrochemical impedance spectroscopy

Abstract

The invention relates to a wide-temperature-interval battery health state estimation method based on an electrochemical impedance spectrum, and belongs to the technical field of battery health state monitoring. Aiming at the problem of battery state of health (SOH) estimation misalignment caused by electrochemical impedance (EIS) temperature interference, the invention provides the following scheme: collecting EIS data within 10-30 DEG C, including extracting temperature interference resistance features, principal component dimensionality reduction features and low variance phases, constructing combined features # imgabs0 # and # imgabs1 #, and realizing SOH estimation through a support vector regression (SVR) model. The method breaks through the dependence of a traditional method on a constant temperature condition or a temperature sensor, realizes SOH high-precision estimation in a wide temperature interval, and is suitable for lithium / sodium ion battery on-line health monitoring.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery health status monitoring and relates to a method for estimating the health status of a battery in a wide temperature range based on electrochemical impedance spectroscopy. Background Art

[0002] Currently, lithium-ion and sodium-ion batteries are widely used in electric vehicles, energy storage systems, and other fields. However, battery performance degradation directly affects the reliability of equipment operation and, in severe cases, can cause safety accidents or significant economic losses. Real-time monitoring of battery state of health (SoH) is crucial to ensuring operational safety, extending battery life, and optimizing operation and maintenance plans. SoH is typically defined as the ratio of current capacity to initial capacity, but it cannot be measured directly and must be estimated through indirect characteristics.

[0003] Currently, mainstream engineering methods are mainly divided into coulomb counting, model-based methods, and data-driven methods, but all have significant limitations. Although coulomb counting is intuitive and easy to implement, it is only applicable to offline constant current operation and cannot meet the needs of online application. Model-based methods are further divided into electrochemical models and equivalent circuit models. Electrochemical models simulate the electrochemical dynamics within the battery through nonlinear partial differential equations, providing a detailed mechanistic description, but the computational cost is high and it is difficult to apply on a large scale. Equivalent circuit models are combined with parameter identification algorithms (such as Kalman filters) to estimate battery parameters online. They have good engineering practicality, but they rely heavily on the accuracy of model parameters, which limits their application under complex operating conditions. Data-driven methods have attracted much attention in recent years. They establish predictive models by mining the relationship between characteristics and capacity decay in historical battery data. However, their performance is significantly affected by data quality and completeness, making them difficult to adapt to complex actual operating conditions. EIS, as a non-invasive and in-situ characterization technique, can obtain rich internal battery information, thereby describing the battery aging state in detail. By introducing EIS measurement in battery engineering applications, as a new feature for battery SoH estimation, the estimation accuracy of battery SOH can be greatly improved.

[0004] In fact, many research reports have used EIS testing technology as a solution for quickly evaluating the SoH of retired batteries. Specific examples include: "Electrochemical impedance spectroscopy power lithium battery health status assessment method based on whale optimization algorithm" (application number CN202410386258); "A SOH estimation method for sodium-ion batteries based on electrochemical impedance spectroscopy" (application number CN202411297088).

[0005] EIS technology has good prospects for online application. However, a major problem that currently restricts the EIS technology from moving from laboratory to engineering application is that the impedance spectrum is severely affected by battery temperature, and the actual temperature of the battery under complex operating conditions is difficult to determine. As a result, the degradation information in the electrochemical impedance spectrum may become invalid under temperature interference.

[0006] Currently, most work using electrochemical impedance spectroscopy for battery aging estimation fails to address this important issue. These methods require that all EIS samples be collected only at a specified temperature and that the batteries be left at a constant temperature before testing, resulting in significant limitations in practical applications. Furthermore, another group of methods chooses to perform temperature correction on the model by collecting the battery's temperature. However, most manufacturers currently do not collect the temperature of every cell, instead employing a distributed collection scheme. Furthermore, because EIS is highly sensitive to temperature changes, temperature correction inevitably introduces significant errors.

[0007] In order to solve the problem of temperature change interference faced when using EIS for battery SoH estimation in actual scenarios, and to make the EIS-based battery SOH estimation method applicable to a wide temperature range, a wide-temperature range battery health state estimation method based on electrochemical impedance spectroscopy is proposed. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide a method for estimating the state of health of batteries over a wide temperature range based on electrochemical impedance spectroscopy. In view of the problem that EIS is easily affected by temperature and the battery SOH estimation method based on EIS cannot be used in a wide temperature range environment, the present invention aims to propose a new idea for extracting and constructing impedance features that are resistant to temperature interference, and to establish a battery SOH estimation model suitable for a wide temperature range. By training the temperature estimation model using EIS data at any single temperature within the temperature range, high-precision estimation of the battery SoH at other temperatures can be achieved. This method is suitable for commercial lithium-ion and sodium-ion batteries, and avoids the introduction of temperature sensing parameters.

[0009] In order to achieve the above object, the present invention provides the following technical solutions:

[0010] A method for estimating the state of health of a battery over a wide temperature range based on electrochemical impedance spectroscopy includes the following steps:

[0011] Collecting electrochemical impedance spectroscopy (EIS) data of the battery in the temperature range of 10°C to 30°C, wherein the frequency range of the EIS is 2000 Hz to 0.01 Hz and the battery is in 100% state of charge (SOC);

[0012] Based on the EIS data, the real part, imaginary part, amplitude, and phase angle characteristics of the impedance are extracted, and the correlation between each characteristic and the battery state of health (SOH) and temperature is calculated. The characteristic intervals that meet the following conditions are selected:

[0013] The Pearson correlation coefficient satisfies ρ XY(SOH) >ρ XY(temp) , and the coefficient of variation satisfies the coefficient of variation Where δ is the preset threshold;

[0014] Extracting the impedance characteristics against temperature interference from the characteristic interval includes:

[0015] The real part of the impedance at 1000 Hz is taken as the ohmic impedance characteristic

[0016] The imaginary part and amplitude features are subjected to principal component analysis (PCA) dimensionality reduction, and the first principal component with a cumulative contribution rate exceeding 95% is retained as the imaginary part feature Zim PCA and amplitude characteristics Z PCA ;

[0017] Select Phase Variance The minimum frequency point phase is used as the anti-interference phase characteristic Z P ;

[0018] Construct a combined temperature-resistance feature, including:

[0019] Take the phase angle of 10Hz frequency as the basis P and calculate the ratio of P to the imaginary part of the mid-frequency band

[0020] Select the ratio of the imaginary part of the solid electrolyte interface film SEI and the charge transfer impedance semicircle vertex frequency

[0021] Based on the ohmic impedance characteristics Imaginary characteristic Zim PCA , amplitude characteristic Z PCA , phase characteristics Z P , ratio characteristics and Train the Support Vector Regression (SVR) model to output the battery health status

[0022] Furthermore, in the correlation screening, the Pearson correlation coefficient is calculated as Where X is the impedance characteristic, Y is the SOH or temperature; when calculating the correlation between the impedance characteristic and the battery health state SOH, Y is the SOH value, and When calculating the temperature dependence of the impedance characteristics, Y is the temperature value, and

[0023] Furthermore, the imaginary part of the mid-frequency band is the imaginary part value at a frequency of 22 Hz, and the ratio The temperature fluctuation variance is less than 0.05.

[0024] Furthermore, the method for determining the frequencies of the SEI and charge transfer impedance semicircle apex is to select the frequency points corresponding to the SEI impedance semicircle and the charge transfer impedance semicircle apex in the 25°C EIS data when the battery SOH decays to 90%, respectively.

[0025] Furthermore, the SVR model is trained using a Gaussian kernel function, hyperparameters are determined by Bayesian optimization, and the objective function is the root mean square error of the validation set. minimize.

[0026] Furthermore, the battery needs to be left to stand for more than 30 minutes before collecting the EIS data, and the battery is a lithium-ion battery or a sodium-ion battery.

[0027] Furthermore, the cumulative contribution rates of the imaginary part and amplitude features after PCA dimensionality reduction are both over 98%.

[0028] Furthermore, the anti-interference phase characteristic Z P The frequency is 0.065Hz, and its temperature variance is less than 0.1.

[0029] A battery health status estimation system includes:

[0030] Data acquisition module, used to collect EIS data of the battery in the temperature range of 10℃ to 30℃;

[0031] A feature extraction module, configured to execute the battery health state estimation method according to any one of claims 1 to 8;

[0032] The model calculation module outputs a battery SOH estimation value through a pre-trained SVR model based on the features.

[0033] Furthermore, the system is integrated into a battery management system (BMS) and does not require an additional temperature sensor for SOH estimation.

[0034] The beneficial effects of the present invention are:

[0035] (1) By screening impedance features that are weakly correlated with temperature (such as ohmic impedance, principal component dimensionality reduction features, and low-variance phase features) and constructing combined features, the influence of temperature fluctuations on EIS data is effectively eliminated, so that the SoH estimation model can be directly applied in a wide temperature range of 10℃ to 30℃ without constant temperature standing or temperature compensation, solving the problem that the traditional EIS method is limited by laboratory constant temperature conditions.

[0036] (2) The Pearson correlation coefficient and coefficient of variation are combined to screen features to ensure that the selected features are sensitive to battery aging but insensitive to temperature changes; by combining features (such as the ratio of the imaginary parts of the semicircle vertices), the essential changes in the shape of the EIS curve with aging are captured, which significantly improves the model's ability to characterize the internal degradation mechanism of the battery, making the SoH estimation error (RMSE) less than 2%, which is better than existing methods that rely on single features or temperature correction.

[0037] (3) Avoid errors introduced by inaccurate distributed temperature acquisition, reducing system complexity and cost; only 30 minutes of resting time is required (traditional methods require more than 3 hours), and it supports direct testing of batteries at any SOC = 100%, significantly shortening detection time and meeting the real-time monitoring needs of electric vehicles, energy storage systems, etc.; it is applicable to various types of batteries such as lithium-ion batteries and sodium-ion batteries, and has a low sample size requirement. Through the Gaussian kernel function and Bayesian hyperparameter optimization of the SVR model, it achieves fast convergence and high robustness under small sample sizes. The proposed feature extraction and modeling method can be integrated into existing battery management systems (BMS) without the need for additional hardware modification, providing a low-cost, high-reliability solution for battery health management.

[0038] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0040] Figure 1 This is a flow chart of a method for estimating the state of health of batteries over a wide temperature range based on electrochemical impedance spectroscopy.

[0041] Figure 2 EIS spectra of batteries under different SOH; Figure 2 (a) Capacity decay curve and EIS diagram of battery No. 1 during aging; Figure 2 (b) is the capacity decay curve and EIS diagram of battery No. 2 during the aging process; Figure 2 (c) is the capacity decay curve and EIS diagram of battery No. 3 during the aging process; Figure 2 (d) is the capacity decay curve and EIS diagram of battery No. 4 during aging;

[0042] Figure 3 The real part, imaginary part, amplitude and phase of the impedance of the battery at different temperatures and SOH; Figure 3 (a) is the real part of the impedance of the battery at different SOH; Figure 3 (b) is the imaginary part of the impedance of the battery at different SOH; Figure 3 (c) is the impedance amplitude of the battery at different SOH; Figure 3 (d) Impedance phase angle of the battery at different SOH; Figure 3 (e) is the real part of the impedance of the battery at different temperatures; Figure 3 (f) is the imaginary part of the impedance of the battery at different temperatures; Figure 3 (g) is the real part of the impedance of the battery at different temperatures; Figure 3 (h) is the impedance phase angle of the battery at different temperatures;

[0043] Figure 4 The anti-temperature interference characteristic diagram of the battery P / Zim characteristics under different SOH; Figure 4 (a) Curves showing the change of different phase to imaginary part ratios with temperature; Figure 4 (b) is the curve of the ratio of the extracted characteristic phase to the imaginary part changing with temperature; Figure 4 (c) The change curve of P / Zim characteristics during battery aging and its fluctuation diagram affected by temperature;

[0044] Figure 5 For different SOH battery Zim ct / Zim sei Characteristic anti-temperature interference characteristic diagram; Figure 5 (a) is the EIS of the battery collected at 10°C, 25°C and 30°C when (a) is 93%; Figure 5 (b) EIS of the battery collected at 10°C, 25°C, and 30°C when (a) was 92%; Figure 5 (c) EIS of the battery collected at 10°C, 25°C and 30°C when (a) was 90%; Figure 5 (d) is the Zim value during battery aging ct / Zim sei Characteristic variation curve and its fluctuation diagram due to temperature disturbance;

[0045] Figure 6 This is the cross-validation estimation effect diagram of the SVR model; Figure 6 (a) is the response distribution diagram of the SVR model; Figure 6(b) is the SOH true value-predicted value distribution diagram; Figure 6 (c) is the cross-validation residual graph of the SVR model;

[0046] Figure 7 The SoH estimation results of 2 under multiple temperature conditions based on the SVR model; Figure 7 (a) is test set #1; Figure 7 (b) is test set #2;

[0047] Figure 8 Box plot distribution of prediction errors at different temperatures. DETAILED DESCRIPTION

[0048] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0049] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0050] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0051] like Figure 1 As shown, the method for estimating the state of health of a battery in a wide temperature range based on electrochemical impedance spectroscopy includes the following steps:

[0052] In S1, a battery with a nominal capacity of 3000mAh was aged by charging and discharging under CCCV conditions at a 1C rate, and EIS data was collected every 20 cycles. The EIS frequency test range was 2000-0.01Hz, and the capacity decay curves of four batteries and EIS data at different SOH were obtained, such as Figure 2 shown. Figure 2 EIS spectra of batteries under different SOH; Figure 2 (a) Capacity decay curve and EIS diagram of battery No. 1 during aging; Figure 2 (b) is the capacity decay curve and EIS diagram of battery No. 2 during the aging process; Figure 2 (c) is the capacity decay curve and EIS diagram of battery No. 3 during the aging process; Figure 2 (d) is the capacity decay curve and EIS diagram of battery No. 4 during the aging process.

[0053] In S2, the process of extracting temperature interference resistance features from the real part, imaginary part, amplitude and phase angle features of the collected EIS data is shown by taking battery No. 1 among the four batteries as an example. The distribution curves of the four features at different frequencies with temperature and SOH are as follows: Figure 3 shown. Figure 3 The real part, imaginary part, amplitude and phase of the impedance of the battery at different temperatures and SOH; Figure 3 (a) is the real part of the impedance of the battery at different SOH; Figure 3 (b) is the imaginary part of the impedance of the battery at different SOH; Figure 3 (c) is the impedance amplitude of the battery at different SOH; Figure 3 (d) Impedance phase angle of the battery at different SOH; Figure 3 (e) is the real part of the impedance of the battery at different temperatures; Figure 3 (f) is the imaginary part of the impedance of the battery at different temperatures; Figure 3 (g) is the real part of the impedance of the battery at different temperatures; Figure 3 (h) is the battery impedance phase angle at different temperatures. Combining Pearson correlation coefficient and coefficient of variation analysis, it is found that the real and imaginary impedance parts and amplitude within 100Hz to 2500Hz effectively reflect changes in battery SOH and are insensitive to temperature interference. The intersection of the frequency ranges for the phase angle characteristics is around 0.065Hz.

[0054] In S3, the real part of the impedance at 1000 Hz is used as the first health factor Zre ohmThe 18-dimensional impedance imaginary part and impedance amplitude data within 100Hz to 2500Hz retain the first 1-dimensional principal component eigenvalue after dimensionality reduction. The cumulative contribution rate of the first principal component eigenvalue obtained by PCA dimensionality reduction of the imaginary part data is as high as 98.96%, while that of the amplitude data is 98.85%, indicating that the 1-dimensional principal component eigenvalue can reflect the main information of the imaginary part and amplitude, so they are used as Zim PCA 、Z PCA Two health factors; the 0.065Hz phase with the strongest resistance to temperature interference is selected as another health factor Z of the battery P .

[0055] In S4, the 10Hz impedance phase angle does not change significantly during battery aging, but it changes significantly with temperature. By calculating the ratio of the 10Hz impedance to the imaginary part characteristics at different frequencies, we can find a characteristic that is affected by temperature changes similar to the 10Hz phase angle, such as Figure 4 As shown in (a), based on the variance analysis of P and Z at each frequency point im As for the fluctuation of the ratio with temperature, we found that the ratio of the 10Hz phase to the imaginary part of 22Hz is the slightest with temperature change. The fluctuation of the normalized value of the combined feature with temperature change is shown as follows: Figure 4 As shown in (b), it can be seen that this feature in healthy batteries shows good stability at different temperatures. Figure 4 (c) shows how the ratio of the 10Hz phase to the 22Hz imaginary part is affected by temperature during battery aging. First, at 25°C, this feature decreases linearly with increasing battery cycle number. This trend is similar to the variation of the inverse of the 22Hz imaginary part of the impedance with cycle number, indicating a strong correlation between this feature and battery aging. Second, the red area represents the standard deviation of this feature within the 10-30°C temperature range during battery aging. It can be seen that temperature changes cause some interference with this feature, but the overall perturbation is small and relatively stable throughout the aging process. Therefore, the ratio of the 10Hz phase to the 22Hz imaginary part (P / Zim) is used as another battery health factor.

[0056] In S5, Figure 5 (a)~ Figure 5As shown in (c), in a fresh battery, Rct is very small, and EIS basically presents a semicircular shape. As the number of cycles increases, Rct continues to increase, and the second semicircle gradually bulges, and then exceeds the first semicircle as the battery continues to age. Therefore, the present invention regards the shape of the EIS curve as a feature with good resistance to temperature perturbations. It is only necessary to extract the change information of the EIS curve shape to obtain the charge transfer impedance aging information under a dynamic temperature environment. Based on the EIS image collected at 25°C under 90% SOH of battery sample No. 1, the present invention extracts the frequency points of the arc tops of SEI and the charge transfer impedance semicircle, thereby obtaining the frequencies of the two characteristic points as 0.26 and 8.5 Hz, respectively. The ratio of the imaginary eigenvalues of the two extreme points is recorded as Zim ct / Zim sei , as another health factor of the battery. Figure 5 (d) Given Zim ct / Zim sei The changes in the characteristics during the battery aging process and its resistance to temperature interference. First, it can be seen that as the battery ages, the value of this characteristic continues to increase, which is consistent with the fact that the second semicircle in the EIS curve gradually increases; secondly, this characteristic is less affected by temperature interference during the entire aging cycle of the battery.

[0057] In S6, based on the above steps, this embodiment collects EIS data of 12 sodium ion batteries during the aging process and extracts health factors. Ten of the 12 sodium ion batteries are used as training sets to train the SVR regression model to establish a battery SOH estimation model. The extracted and constructed Zre ohm 、Zim PCA 、Z PCA 、P / Z im and Zim ct / Zim sei The five health factors such as ΔH and ΔS are used as input, and the SOH corresponding to the battery is used as output. The SVR model is trained by calling the fitrsvm function of MATLAB, and the SVR model is trained by five-fold cross validation. The constraint condition is that the RMSE between the estimated value of the SOH of the validation set and the true value is minimized. All feature data are standardized to eliminate the errors caused by different dimensions. The Gaussian kernel is selected as the kernel function of the model, and the Bayesian optimizer is used to obtain the optimal hyperparameters of the SVR model. The kernel scale of the model is 4.99, the box constraint is 2.12, and the Epsilon is 0.0029. The results of cross validation show that the estimation accuracy of the model is good, and the vast majority of the predicted values are distributed near the measured values. Figure 6 (a) shows the distribution of estimated and actual values of all samples. Figure 6 (b) shows the deviation of the observed value from the perfect prediction at each aging stage, Figure 6 (c) shows the estimated residuals under each aging state. The estimation errors of all samples are less than 4%, and the errors of all estimates except the four abnormal samples are less than 2%.

[0058] In S7, based on the estimation model established in S6, the battery SOH is verified by collecting EIS test set data at 10, 25 and 30°C. Figure 7 (a) and Figure 7 (b) shows the SOH estimation results for the test set of batteries collected across different temperature ranges. This shows that the average SoH estimated based on impedance data maintains a small error relative to the true value throughout the entire cycling process, with the MAE of the prediction set near 1%. Furthermore, the pink standard deviation ranges due to temperature effects in the two prediction samples are generally small, demonstrating that the unique health factor extraction and construction method proposed in this paper effectively suppresses temperature interference, and the SoH estimation results remain stable within the temperature range of 10-30°C.

[0059] In order to further demonstrate the stability of the SoH estimation method proposed in this invention in a wide temperature range, Figure 8 Boxplots of the error distribution of the SVR model's estimated values at various temperatures are presented. First, the absolute values of the minimum or smallest observed values at each temperature in the boxplots are all less than 2.5%, with no outliers, indicating that the model's estimates are good and stable at all temperatures. This demonstrates that the proposed temperature-resistant health factor combined with the SVM model has good adaptability to abnormal temperatures within the range of 10 to 30°C.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy, characterized by: The following steps are involved: Collecting electrochemical impedance spectroscopy (EIS) data of the battery in the temperature range of 10°C to 30°C, wherein the frequency range of the EIS is 2000 Hz to 0.01 Hz and the battery is at 100% state of charge (SOC); Based on the EIS data, the real part, imaginary part, amplitude, and phase angle characteristics of the impedance are extracted, and the correlation between each characteristic and the battery state of health (SOH) and temperature is calculated. The characteristic intervals that meet the following conditions are selected: The Pearson correlation coefficient satisfies ρ XY(SOH) >ρ XY(temp) , and the coefficient of variation satisfies the coefficient of variation Where δ is the preset threshold; Extracting the impedance characteristics against temperature interference from the characteristic interval includes: The real part of the impedance at 1000 Hz is taken as the ohmic impedance characteristic The imaginary part and amplitude features are subjected to principal component analysis (PCA) dimensionality reduction, and the first principal component with a cumulative contribution rate exceeding 95% is retained as the imaginary part feature Zim PCA and amplitude characteristics Z PCA ; Select Phase Variance The minimum frequency point phase is used as the anti-interference phase characteristic Z P ; Construct a combined temperature-resistance feature, including: Take the phase angle of 10Hz frequency as the basis P and calculate the ratio of P to the imaginary part of the mid-frequency band Select the ratio of the imaginary part of the solid electrolyte interface film SEI and the charge transfer impedance semicircle vertex frequency Based on the ohmic impedance characteristics Imaginary characteristic Zim PCA , amplitude characteristic Z PCA , phase characteristics Z P , ratio characteristics and Train a Support Vector Regression (SVR) model to output the battery health status 2. The method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, characterized in that: In the correlation screening, the Pearson correlation coefficient is calculated as the Pearson correlation coefficient Where X is the impedance characteristic, Y is the SOH or temperature; when calculating the correlation between the impedance characteristic and the battery health state SOH, Y is the SOH value, and When calculating the temperature dependence of the impedance characteristics, Y is the temperature value, and 3. The method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, wherein: The imaginary part of the mid-frequency band is the imaginary part value at a frequency of 22 Hz, and the ratio The temperature fluctuation variance is less than 0.

05.

4. The method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, wherein: The method for determining the SEI and charge transfer impedance semicircle apex frequencies is as follows: in the 25°C EIS data when the battery SOH decays to 90%, the frequency points corresponding to the SEI impedance semicircle and the charge transfer impedance semicircle arc tops are selected respectively.

5. The method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, wherein: The SVR model is trained using a Gaussian kernel function, and the hyperparameters are determined by Bayesian optimization. The objective function is the root mean square error of the validation set. minimize.

6. The method for estimating the state of health of a battery in a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, wherein: The battery needs to be left to stand for more than 30 minutes before collecting the EIS data, and the battery is a lithium-ion battery or a sodium-ion battery.

7. The method for estimating battery health status over a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, characterized in that: The cumulative contribution rates of the imaginary part and amplitude features after PCA dimensionality reduction are both over 98%.

8. The method for estimating the state of health of a battery in a wide temperature range based on electrochemical impedance spectroscopy according to claim 1, wherein: The anti-interference phase characteristic Z P The frequency is 0.065Hz, and its temperature variance is less than 0.

1.

9. A battery health status estimation system, characterized by: include: Data acquisition module, used to collect EIS data of the battery in the temperature range of 10℃ to 30℃; A feature extraction module, configured to execute the battery health state estimation method according to any one of claims 1 to 8; The model calculation module outputs a battery SOH estimation value through a pre-trained SVR model based on the features.

10. The battery health status estimation system according to claim 9, characterized in that: The system is integrated into the battery management system (BMS) and does not require an additional temperature sensor for SOH estimation.

Images