A method, apparatus, device, and storage medium for assessing the health status of energy storage batteries.

Abstract

This application discloses a method, apparatus, device, and storage medium for assessing the health status of energy storage batteries, relating to the field of battery technology. The method includes: determining a full lifecycle database of the energy storage battery based on a static impedance testing platform, a dynamic impedance testing platform, and battery aging external characteristic data; performing time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset and battery aging external characteristic data in the database, and determining a first set of time-varying dominant factors for battery aging based on the feature mining results; determining a second set of time-varying dominant factors for battery aging based on the first set of time-varying dominant factors for battery aging, the dynamic electrochemical impedance information dataset in the full lifecycle database, and grey relational analysis; and determining the target health status assessment result based on the second set of time-varying dominant factors for battery aging and preset multi-model fusion learning rules. This application can monitor the dynamic characteristics of the battery in real time under operating conditions, improving the accuracy and real-time performance of battery assessment.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method, apparatus, device, and storage medium for assessing the health status of energy storage batteries. Background Technology

[0002] Currently, electrochemical energy storage batteries are being used more and more widely, but during long-term use, they will gradually age due to various factors (such as charge and discharge cycles, temperature changes, overcharge and over-discharge, etc.), leading to a decline in their performance.

[0003] Traditional battery health assessment methods primarily rely on directly measuring parameters such as battery capacity or internal resistance, which is insufficient for real-time monitoring and fails to accurately reflect the battery's actual performance. To address this, existing methods employ electrochemical impedance spectroscopy (EIS), but this approach typically operates under static conditions, failing to reflect the battery's dynamic characteristics during actual operation, resulting in poor accuracy and real-time performance. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a method, apparatus, device, and storage medium for assessing the health status of energy storage batteries. This method enables real-time monitoring of the dynamic characteristics of batteries under operating conditions, thereby improving the accuracy, efficiency, precision, and real-time performance of battery health status assessment. This provides key technical support for the safe, reliable operation, and intelligent maintenance of electrochemical energy storage systems. The specific solution is as follows:

[0005] Firstly, this application provides a method for assessing the health status of energy storage batteries, including:

[0006] Based on static impedance testing platform, dynamic impedance testing platform and battery aging external characteristic data, a full life cycle database of energy storage batteries is established.

[0007] Based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, feature mining in the time domain and frequency domain is performed, and based on the corresponding feature mining results and Pearson correlation analysis, the first battery aging time-varying dominant factor set is determined.

[0008] Based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and the grey relational analysis method, the second battery aging time-varying dominant factor set is determined.

[0009] The battery health status is assessed based on the second battery aging time-varying dominant factor set and the preset multi-model fusion learning rules to determine the target health status assessment result corresponding to the energy storage battery.

[0010] Optionally, the step of determining the full life-cycle database of the energy storage battery based on the static impedance testing platform, the dynamic impedance testing platform, and battery aging external characteristic data includes:

[0011] During the cyclic aging process with a preset number of cycles, charge and discharge data are collected throughout the entire life cycle of the energy storage battery to determine the external characteristics data of battery aging.

[0012] During the cyclic aging process of the preset number of cycles, the static electrochemical impedance spectrum and dynamic electrochemical impedance spectrum of different charge state nodes throughout the entire life cycle are determined based on the static impedance test platform and the dynamic impedance test platform, so as to obtain the static electrochemical impedance information dataset and the dynamic electrochemical impedance information dataset.

[0013] Based on a preset data preprocessing strategy, error removal and reliability verification are performed on the battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset to determine the preprocessed battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset.

[0014] Optionally, the step of performing time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and determining the first set of time-varying dominant factors for battery aging based on the corresponding feature mining results and Pearson correlation analysis, includes:

[0015] Based on the static electrochemical impedance information dataset, relaxation time distribution algorithm and equivalent circuit model in the full life cycle database, data reflecting the external characteristics of battery aging are mined from the time domain and frequency domain to determine the battery static electrochemical impedance spectrum degradation feature set.

[0016] Based on the degradation features in the battery static electrochemical impedance spectroscopy degradation feature set and the battery health status data, Pearson correlation analysis and maximum information coefficient analysis were performed to determine the correlation analysis results; the battery health status data includes battery capacity decay data.

[0017] Based on the correlation analysis results, each degradation feature in the battery static electrochemical impedance spectroscopy degradation feature set is screened to determine the first set of battery aging time-varying dominant factors.

[0018] Optionally, based on the static electrochemical impedance information dataset, relaxation time distribution algorithm, and equivalent circuit model in the full life cycle database, data reflecting the battery's aging external characteristics are mined from both time and frequency domain perspectives to determine the battery's static electrochemical impedance spectrum degradation feature set, including:

[0019] The static electrochemical impedance information dataset in the full life cycle database is processed based on the relaxation time distribution algorithm to determine the time-domain relaxation time distribution curve.

[0020] Based on the time-domain relaxation time distribution curve and the preset peak detection rule, the position and parameter value of the peak are identified to determine the time-domain feature mining results.

[0021] By fitting the equivalent circuit model of the static electrochemical impedance information dataset and optimizing the circuit parameter values ​​using a nonlinear least squares algorithm, the frequency domain feature mining results are determined.

[0022] The time-domain feature mining results and the frequency-domain feature mining results are mapped point-by-point at different cycle rounds and different state-of-charge nodes to determine the degradation feature set of the battery static electrochemical impedance spectrum.

[0023] Optionally, determining the second set of dominant factors for battery aging based on the first set of dominant factors for battery aging, the dynamic electrochemical impedance information dataset in the full life cycle database, and grey relational analysis includes:

[0024] Frequency bands are removed from the dynamic electrochemical impedance information dataset in the full life cycle database to determine the dataset after the first removal.

[0025] Based on the first battery aging time-varying dominant factor set and the first removed dataset, the second removed dataset is determined;

[0026] Based on the second removed dataset and the grey relational analysis method, the relational analysis results are determined.

[0027] Optionally, determining the correlation analysis results based on the second removed dataset and the grey relational analysis method includes:

[0028] Grey relational analysis is performed on the health features in the second removed dataset and the battery health status data to determine the correlation analysis results.

[0029] Based on the correlation analysis results, feature filtering is performed on the second excluded dataset to determine the second set of battery aging time-varying dominant factors.

[0030] Optionally, the battery health status assessment based on the second battery aging time-varying dominant factor set and preset multi-model fusion learning rules includes:

[0031] The second set of time-varying dominant factors in battery aging is used as model input information and input into multiple different preset regression models to determine multiple initial battery health status assessment results.

[0032] Based on a preset linear regression algorithm, feature fusion is performed on the initial battery health status assessment results and the second battery aging time-varying dominant factor set to determine the target health status assessment result corresponding to the energy storage battery.

[0033] Secondly, this application provides a device for assessing the health status of an energy storage battery, comprising:

[0034] The database construction module is used to determine the full life cycle database of energy storage batteries based on static impedance testing platform, dynamic impedance testing platform and battery aging external characteristic data;

[0035] The first factor set determination module is used to perform time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and to determine the first battery aging time-varying dominant factor set based on the corresponding feature mining results and Pearson correlation analysis.

[0036] The second factor set determination module is used to determine the second battery aging time-varying dominant factor set based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and the grey relational analysis method.

[0037] The evaluation result determination module is used to evaluate the battery health status based on the second battery aging time-varying dominant factor set and the preset multi-model fusion learning rules, so as to determine the target health status evaluation result corresponding to the energy storage battery.

[0038] Thirdly, this application provides an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] A processor is used to execute the computer program to implement the steps of the aforementioned energy storage battery health status assessment method.

[0041] Fourthly, this application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the steps of the aforementioned energy storage battery health status assessment method.

[0042] As can be seen, in this application, a full life cycle database of the energy storage battery is determined based on a static impedance testing platform, a dynamic impedance testing platform, and battery aging external characteristic data; time-domain and frequency-domain feature mining is performed based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and a first set of time-varying dominant factors for battery aging is determined based on the corresponding feature mining results and Pearson correlation analysis; a second set of time-varying dominant factors for battery aging is determined based on the first set of time-varying dominant factors for battery aging, the dynamic electrochemical impedance information dataset in the full life cycle database, and grey relational analysis; and battery health status is assessed based on the second set of time-varying dominant factors for battery aging and a preset multi-model fusion learning rule to determine the target health status assessment result corresponding to the energy storage battery. In other words, this application first establishes a full lifecycle database for the energy storage battery based on static and dynamic impedance testing platforms and battery aging external characteristic data. Then, it utilizes the static electrochemical impedance information dataset and battery aging external characteristic data from the database to perform time-domain and frequency-domain feature mining, and combines this with Pearson correlation analysis to determine a first set of time-varying dominant factors for battery aging. Next, it uses the first set of time-varying dominant factors for battery aging and the aforementioned dynamic electrochemical impedance information dataset from the database to determine a second set of time-varying dominant factors for battery aging. Finally, based on the second set of time-varying dominant factors for battery aging and preset multi-model fusion learning rules, it determines the target health status assessment result. This allows for real-time monitoring of the battery's dynamic characteristics under operating conditions, thereby improving the accuracy, efficiency, precision, and real-time performance of battery health status assessment, and providing key technical support for the safe, reliable operation and intelligent maintenance of electrochemical energy storage systems. Attached Figure Description

[0043] 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.

[0044] Figure 1 A flowchart of a method for assessing the health status of an energy storage battery is provided in this application;

[0045] Figure 2 A schematic diagram of the battery charging and discharging voltage and current variation curves provided in this application;

[0046] Figure 3 A schematic diagram of health features extracted based on the relaxation time distribution algorithm provided in this application;

[0047] Figure 4A schematic diagram showing the correspondence between an equivalent circuit model and impedance spectrum curves at different frequency bands provided in this application;

[0048] Figure 5 A schematic diagram of the initial parameter identification process for an equivalent circuit model provided in this application;

[0049] Figure 6 A schematic diagram of an equivalent circuit model parameter identification process based on the least squares method provided in this application;

[0050] Figure 7 A schematic diagram of a multi-model fusion algorithm based on machine learning is provided for this application;

[0051] Figure 8 A schematic diagram showing the comparison curves between the estimated and actual health status values ​​of an electrochemical energy storage system provided in this application;

[0052] Figure 9 A schematic diagram of a battery health status assessment device provided in this application;

[0053] Figure 10 This application provides a structural diagram of an electronic device. Detailed Implementation

[0054] 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.

[0055] Traditional battery health assessment methods primarily rely on directly measuring parameters such as battery capacity or internal resistance, which is insufficient for real-time monitoring and fails to accurately reflect the battery's actual performance. To address this, existing methods employ electrochemical impedance spectroscopy (EIS), but this approach typically operates under static conditions, failing to reflect the battery's dynamic characteristics during actual operation, resulting in poor accuracy and real-time performance.

[0056] Therefore, this application provides a battery health status assessment scheme that can monitor the dynamic characteristics of the battery in real time during operation, thereby improving the accuracy, efficiency, precision and real-time performance of battery health status assessment.

[0057] See Figure 1 As shown in the figure, an embodiment of the present invention discloses a method for assessing the health status of an energy storage battery, including:

[0058] Step S11: Based on the static impedance test platform, dynamic impedance test platform, and battery aging external characteristic data, determine the full life cycle database of the energy storage battery.

[0059] In this embodiment, to monitor the health status of a battery, such as a lithium-ion battery, a multi-source heterogeneous database is first constructed based on static / dynamic electrochemical impedance spectroscopy and external characteristic data. Specifically, during a preset number of cycle aging cycles, charge-discharge datasets are collected throughout the battery's entire lifespan to determine the battery's aging external characteristic data. During the preset number of cycle aging cycles, static and dynamic electrochemical impedance spectra at different state-of-charge nodes are determined using static and dynamic impedance testing platforms to obtain static and dynamic electrochemical impedance information datasets. Based on a preset data preprocessing strategy, error removal and reliability verification are performed on the battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset to determine the preprocessed battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset. The battery aging external characteristic data can be collected using charge-discharge equipment.

[0060] It is important to understand that in this embodiment, to collect electrochemical impedance information and capacity changes during the aging process of lithium-ion batteries, simultaneous acquisition of SEIS (Static Electrochemical Impedance Spectroscopy) and DEIS (Dynamic Electrochemical Impedance Spectroscopy) was designed. The acquisition was conducted in two groups: static EIS (Electrochemical Impedance Spectroscopy) test groups: C1 and C2; and dynamic EIS (Electrochemical Impedance Spectroscopy) test groups: C3 and C4. The steps were as follows: first, at 25... The initial capacity was calibrated using a constant current and constant voltage at 0.3C, and then placed at 25°C. The battery was aged in a constant temperature chamber at 1C. Specifically, during the charging phase, a constant current of 1.5A was used for charging until the voltage reached the cutoff value of 4.2V, followed by a constant voltage charging phase until the current dropped below 20mA. During the constant current discharge phase, the battery was discharged at a constant current of 1.5A until the current dropped below 20mA. The experiment ended when the discharge capacity was less than 80% of the initial capacity. During the test, the static / dynamic EIS were as follows: Static EIS test group: After 10 cycles of aging, the battery was moved to a 25°C chamber. Constant current discharge was performed under environmental conditions, and a VMP-300 electrochemical workstation (Versatile Modular Potentiostat) was connected at different SOC points. A 5mV AC voltage perturbation and a 100kHz–0.01Hz frequency sweep were used, with 10 logarithmic intervals taken every tenth harmonic, taking approximately 600 seconds to obtain the EIS at different static SOCs. Dynamic EIS test group: After static testing at 20% SOC (State of Charge) intervals during cycling, the battery was not allowed to rest. A perturbation signal composed of 21 sinusoidal currents was superimposed on the discharge current using a DEIS platform. The platform acquired voltage and current responses in real time, and after FFT (Fast Fourier Transform) analysis, the dynamic EIS was output synchronously within 10 seconds. Finally, static / dynamic electrochemical impedance spectra at different SOC points throughout the entire life cycle of the lithium battery were obtained. Figure 2 As shown, the voltage and current variation curves of a lithium-ion battery during charging and discharging cover key operational information of the battery during constant current and constant voltage charging and constant current discharging processes.

[0061] Furthermore, after obtaining the above data, multi-dimensional data preprocessing and reliability verification will be performed to remove erroneous data and ensure its reliability. Specifically, moving averages combined with polynomial fitting can be used to eliminate measurement errors, and Savitzky-Golay (least squares smoothing filter) can be used to smooth and remove environmental errors introduced by high-frequency white noise. Subsequently, the Lin-KK tool will be used to verify the causality, linearity, and stability of the impedance across the entire frequency band.

[0062] Step S12: Based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, perform time-domain and frequency-domain feature mining, and determine the first battery aging time-varying dominant factor set based on the corresponding feature mining results and Pearson correlation analysis.

[0063] In this embodiment, after constructing the database, SEIS data at different SOC points throughout the entire lifecycle are used to mine a set of battery aging characteristics from the time / frequency domain perspective. This ultimately establishes a time-varying dominant factor set for the static electrochemical impedance spectroscopy (SES) degradation characteristics of the energy storage battery, namely, the first battery aging time-varying dominant factor set. In other words, based on the SES data set, relaxation time distribution algorithm, and equivalent circuit model in the entire lifecycle database, data reflecting the external characteristics of battery aging are mined from both the time and frequency domains to determine the battery SES degradation characteristic set. Based on each degradation characteristic in the battery SES degradation characteristic set and the battery health status data, Pearson correlation analysis and maximum information coefficient analysis are performed to determine the correlation analysis results. The battery health status data includes battery capacity decay data. Based on the correlation analysis results, each degradation characteristic in the battery SES degradation characteristic set is screened to determine the selected first battery aging time-varying dominant factor set. Specifically, the determination of the battery static electrochemical impedance spectroscopy degradation feature set involves first processing the static electrochemical impedance information dataset in the full life cycle database using a relaxation time distribution algorithm to determine the time-domain relaxation time distribution curve; then, based on the time-domain relaxation time distribution curve and a preset peak detection rule, identifying the peak position and parameter values ​​to determine the time-domain feature mining result; finally, by fitting the equivalent circuit model of the static electrochemical impedance information dataset and optimizing the circuit parameter values ​​using a nonlinear least squares algorithm, determining the frequency-domain feature mining result; and then mapping the time-domain feature mining result and the frequency-domain feature mining result point-by-point at different cycle cycles and different state-of-charge nodes to determine the battery static electrochemical impedance spectroscopy degradation feature set.

[0064] Specifically, in combination Figure 3 As shown, in this embodiment, the SEIS undergoes DRT transformation (Distribution of Relaxation Times), including regularized least squares deconvolution to convert the frequency domain impedance into a time domain relaxation time distribution curve. A peak detection algorithm is used to identify the peak position, and its peak area, abscissa value, and ordinate value are calculated. Specifically, this includes:

[0065] The SEIS data (frequency domain impedance spectrum, including the real part) With the imaginary part According to the following... Functions perform DRT transformation:

[0066] ;

[0067] In the formula, The relaxation time (time constant) is the time constant. The maximum value, It is the minimum value; Let be the distribution function to be determined; The internal resistance is ohmic; This indicates the frequency value.

[0068] Meanwhile, the 'find_peaks' function in MATLAB can be used to identify peak positions. A threshold of 50% above the mean peak value can be set, thus obtaining the peak position as follows: The height value is denoted as The peak integral area can be obtained using the formula shown below, denoted as: :

[0069] ;

[0070] In the formula, This is the peak position; This is the time window. Following the steps above, feature mining from a temporal perspective can be completed, yielding the temporal feature matrix. .

[0071] Simultaneously, an equivalent circuit model of the SEIS is fitted, and the circuit parameter values ​​are optimized using a nonlinear least squares algorithm. The outputs ohmic impedance, charge transfer impedance, SEI film impedance, CPE-T (Constant Phase Element, time constant part) pseudo-capacitance, and CPE-P (Constant Phase Element-Polarization, phase angle exponent part) exponent factor are calculated. The rates of change of ohmic impedance and contact impedance during cycling are calculated to obtain conduction loss and lithium-ion loss values. Specifically, this includes:

[0072] The equivalent circuit model and its structure, selected based on the correspondence with the impedance spectrum curves at each frequency band, are determined according to... Figure 4 The equivalent circuit model structure shown can be used to obtain its impedance as follows:

[0073] ;

[0074] In the formula, , and This indicates a constant-phase element. and Parameters representing constant-phase elements; and Represents Weber impedance element Parameters; For parameters related to Weber impedance elements; The resistance of a constant-phase element; These are the parameters of a constant-phase element; Here is another parameter representing the Weber impedance element; L is the inductance; This is another parameter related to Weber elements; ω is the angular frequency.

[0075] In the impedance spectrum fitting process, local fitting can be performed first on the parts with circular and straight line morphological features. Then, the final values ​​of the piecewise local fitting are used as the initial values ​​for the overall fitting, thereby achieving parameter identification of the impedance spectrum. The specific steps for obtaining the initial values ​​of the piecewise local fitting are as follows:

[0076] (1) Interpolation is used to obtain the intersection values ​​of the impedance spectrum and the real axis as the initial values ​​for fitting. Since the magnitude of stray inductance in the impedance spectrum is relatively fixed and its order of magnitude is very small, the initial value is set to... ;

[0077] (2) Since the frequency points on the left arc of the impedance spectrum are densely distributed, four frequency points are selected at intervals starting from the point closest to the real axis to perform local fitting of the left arc, thus obtaining... and Initial value, where, These are parameters for constant-phase elements, used to describe the phase characteristics within a specific frequency range of the impedance spectrum; The resistance parameter of a constant-phase element represents the contribution of the constant-phase element to the impedance within a specific frequency range.

[0078] (3) Find the highest point of the arc section, and select data from 4 frequency points near the highest point to fit the load transfer resistance circuit to obtain the result. and Initial values ​​for components, where, These are parameters for constant-phase elements, used to describe the phase characteristics within a specific frequency range of the impedance spectrum;

[0079] (4) Find the last four points of the impedance spectrum and fit the straight line segment to obtain the impedance spectrum. The initial value of .

[0080] Based on the obtained initial parameter values, the electrochemical impedance spectroscopy of lithium-ion batteries can be solved by overall fitting, such as... Figure 5 As shown, the values ​​of each impedance component can be obtained, thereby identifying and fitting the parameters of the equivalent circuit model. The fitting method employs a nonlinear least squares approach, and the formula for calculating the iteration step size during fitting is:

[0081] ;

[0082] In the formula, The iteration step size is x; x is the dependent variable. For fitting residuals; Let x be the Jacobian matrix; It is the identity matrix; Here, represents the damping adjustment factor; T denotes the transpose of the matrix. The residual e during the fitting process is defined as:

[0083] ;

[0084] In the formula, n represents the sampling point location; This represents the angular frequency of the i-th sampling point; This represents the calculated imaginary part of the impedance at the i-th sampling point; This represents the calculated real part of the impedance at the i-th sampling point; This represents the real part value obtained from the fitting; This represents the imaginary part obtained from the fitting; This represents the real part of the impedance obtained from actual measurement; This represents the imaginary part of the impedance obtained from actual measurements; the goal of the fitting is to minimize the sum of errors at each frequency point represented by this formula. The entire parameter identification process is as follows: Figure 6 As shown, the frequency domain feature mining can be completed by following the above steps, resulting in a frequency domain feature matrix. Among them, Figure 6 In the equation -IM(Z), the imaginary part of the impedance Z is represented; Re(Z) is represented by the real part of the impedance Z; L represents inductance; and R represents resistance.

[0085] Furthermore, after extracting the time-domain and frequency-domain features, the two are mapped point-by-point along with the cycle number and state of charge (SOC) to establish a static electrochemical impedance spectroscopy (SIP) degradation feature set covering the entire life cycle and the entire charge window of the energy storage battery. Specifically, the time-domain features (peak area, abscissa value, ordinate value) and frequency-domain features (ohmic impedance, charge transfer impedance, SEI film impedance, CPE-T, CPE-P exponent factor) extracted in the previous steps are mapped point-by-point along with the cycle number and SOC. In particular, the time-domain and frequency-domain feature values ​​corresponding to different SOC points at each cycle number are mapped one-to-one to form a complete feature matrix, thereby constructing a static electrochemical impedance spectroscopy (SIP) degradation feature set covering the entire life cycle and the entire charge window of the energy storage battery.

[0086] Subsequently, based on the degradation feature set of the battery's static electrochemical impedance spectroscopy, Pearson correlation analysis and the correlation strength between the quantification characteristics of the maximum information coefficient (MIC) and capacity decay were used to screen out the time-varying dominant factor set of the static electrochemical impedance spectroscopy degradation feature set of the energy storage battery. The specific process is as follows:

[0087] (1) Perform Pearson correlation analysis on each feature in the static electrochemical impedance spectroscopy degradation feature set of the energy storage battery and the battery capacity decay data. Then, map the obtained static electrochemical impedance spectroscopy degradation feature set of the energy storage battery to the battery state of health (SOH) data to form a matrix. ,in It is the i-th eigenvalue. This is the SOH value corresponding to the i-th eigenvalue. The Pearson correlation coefficient is calculated using the formula shown below. :

[0088] ;

[0089] In the formula, The mean value of the features in the degradation feature set of the battery static electrochemical impedance spectroscopy; is the mean of the SOH values ​​corresponding to the features in the feature set; n is the number of features in the feature set.

[0090] (2) The maximum information coefficient analysis is performed on each feature in the degradation feature set of the static electrochemical impedance spectroscopy of the energy storage battery and the battery capacity decay data. Specifically, the matrix... Substitute into the formula shown below:

[0091] ;

[0092] In the formula, M is the maximum number of grid cells; B is the grid division method; For the mutual information of X and Y under partition B; This means finding the binning method that maximizes mutual information among all possible binning methods.

[0093] (3) Through Pearson correlation analysis and MIC analysis, the linear and nonlinear correlation values ​​between each feature and capacity decay were obtained. These two values ​​were combined to screen features with a high correlation to capacity decay. The specific screening method was as follows:

[0094] 1) Set threshold The threshold for the Pearson correlation coefficient, The threshold for MIC;

[0095] 2) Select the option that satisfies and The characteristics are used as the time-varying dominant factor set to determine the first battery aging time-varying dominant factor set.

[0096] Step S13: Based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and the grey relational analysis method, determine the second battery aging time-varying dominant factor set.

[0097] In this embodiment, the obtained first battery aging time-varying dominant factor set is used to identify sensitive frequency bands / points of DEIS, eliminate parts unsuitable for practical engineering applications, obtain the remaining frequency bands mapped to DEIS impedance information, and use grey relational analysis for feature selection to establish the second battery aging time-varying dominant factor set for DEIS. Specifically, frequency bands are removed from the dynamic electrochemical impedance information dataset in the full life cycle database to determine the first removed dataset; based on the first battery aging time-varying dominant factor set and the first removed dataset, the second removed dataset is determined; based on the second removed dataset and grey relational analysis, the correlation analysis result is determined. Specifically, regarding the determination of the correlation analysis result, grey relational analysis is first performed on the health features in the second removed dataset and the battery health status data to determine the correlation analysis result; then, feature selection is performed on the second removed dataset based on the correlation analysis result to determine the selected second battery aging time-varying dominant factor set.

[0098] It's important to understand that to improve the feasibility of DEIS data applications, frequencies below 1 Hz are removed based on the engineering application characteristics of dynamic electrochemical impedance spectroscopy. The reasons are as follows: First, measurement efficiency: low-frequency signals have long response times, and removing them shortens the measurement time. Second, noise interference: the low-frequency region has high noise levels, and removing it reduces noise impact and improves the signal-to-noise ratio. Third, engineering practicality: high-frequency impedance data better reflects the dynamic performance of the battery and is significant for real-time monitoring of battery health and prediction of remaining lifespan, while low-frequency data has relatively lower practical application value. In summary, given the time-consuming and noisy nature of DEIS in engineering applications, frequencies below 1 Hz are removed.

[0099] Subsequently, based on the sensitive frequency bands / points corresponding to the first set of dominant time-varying factors for battery aging, and after removing frequencies less than 1Hz, the remaining frequency bands are mapped to DEIS impedance information. Grey relational analysis is then used for feature selection, ultimately establishing the DEIS set of dominant time-varying factors for energy storage battery aging. The specific process is as follows:

[0100] Based on the first set of dominant factors for battery aging, sensitive frequency bands / points of SEIS are identified, and parts unsuitable for DEIS engineering applications are eliminated to obtain the remaining frequency bands mapped to DEIS impedance information. Next, grey relational analysis is used to quantify the correlation strength between impedance data within the sensitive frequency bands / points of the dynamic electrochemical impedance spectroscopy and battery capacity decay, thus identifying the dominant factors. First, the absolute difference is calculated:

[0101] ;

[0102] In the formula, Let t be the health status value of the t-th health status value in the i-th cycle; Let be the actual value of the t-th DEIS health characteristic in the i-th cycle. This represents the actual health status of the sample point in the i-th cycle.

[0103] Then calculate the number of samples that need to be traversed first. Calculate and :

[0104] ;

[0105] In the formula, This represents the maximum absolute difference between the DEIS health characteristics and volume decay across all sample points. This represents the minimum absolute difference between the DEIS health characteristic and the volume decline among all sample points. Let be the correlation coefficient between the t-th DEIS health feature and the sample point in the i-th cycle, and capacity decay. The larger the value, the stronger the correlation between this feature and capacity decay at this sample point. This is used to adjust the sensitivity of the correlation coefficient to the degree of deviation. The correlation coefficient is then calculated according to the following formula:

[0106] ;

[0107] In the formula, Let be the correlation between the t-th comparison sequence and the reference sequence, which is the average correlation coefficient of all sample points in the sequence. The comparison sequence refers to the DEIS feature sequence, and the reference sequence refers to the battery capacity degradation sequence. n is the total number of sample points. Let be the correlation coefficient of the t-th comparison sequence at the i-th sample point. Using the above formula, DEIS features with high correlation to lithium battery capacity degradation can be screened out, ultimately constructing a set of time-varying dominant factors for dynamic electrochemical impedance spectroscopy degradation characteristics.

[0108] Step S14: Based on the second battery aging time-varying dominant factor set and the preset multi-model fusion learning rules, perform battery health status assessment to determine the target health status assessment result corresponding to the energy storage battery.

[0109] In this embodiment, after determining the second battery aging time-varying dominant factor set, the battery health status is assessed using the determined second battery aging time-varying dominant factor set and a preset multi-model fusion learning rule. That is, the second battery aging time-varying dominant factor set is used as model input information and input into multiple different preset regression models to determine multiple initial battery health status assessment results. Based on a preset linear regression algorithm, feature fusion is performed on each of the initial battery health status assessment results and the second battery aging time-varying dominant factor set to determine the target health status assessment result corresponding to the energy storage battery.

[0110] It's important to understand that an initial prediction by a primary model is performed first. In this embodiment, the second set of time-varying dominant factors related to battery aging is used as the model input, serving as the input for the random forest regression algorithm, extreme random tree algorithm, K-nearest neighbor regression algorithm, and linear regression algorithm, respectively, to obtain the initial battery health prediction results for each model. Based on the initial prediction results, a Bayesian linear regression algorithm is used to fuse the initial prediction results to obtain a second-corrected health status estimate. Figure 7 As shown, the specific related process can be described as follows:

[0111] (1) Data preparation. Based on the time-varying dominant factor set of the second battery aging, the time-varying dominant factor set matrix of the dynamic electrochemical impedance spectroscopy degradation characteristics was obtained by analysis, denoted as D-HFs;

[0112] (2) Single-model prediction. D-HFs were used as inputs to KNNR (K-Nearest Neighbors Regression), ERTR (Extreme Random Tree), LR (Linear Regression), and RFR (Random Forest Regression), respectively. The prediction results were defined as follows: , , and .

[0113] (3) Multi-model and multi-feature fusion based on feature fusion and reuse. A matrix composed of all single-model prediction results and D-HFs is formed. As input to BLR (Bayesian Linear Regression), it completes multi-model, multi-feature fusion based on feature reuse. Represented as The predicted results are as follows: Figure 8 As shown in Table 1, the error information of the prediction results is as follows.

[0114] Table 1 Error Information Table of Model Prediction Results

[0115]

[0116] Wherein, RMSE stands for Root Mean Square Error; MAE stands for Mean Absolute Error; C3 and C4 are different battery labels; Fusion means fusing the prediction results of all single models.

[0117] Compared to traditional frequentist regression methods, Bayesian linear regression (BLR) can better utilize prior knowledge, improving prediction accuracy and reliability. Because BLR is highly adaptable to data and has no hyperparameters, it offers high prediction efficiency. Based on this feature and comparisons with other models, the BLR algorithm was chosen as the fusion model. The formula for the BLR algorithm is... The specific details are as follows:

[0118] ;

[0119] In the formula, These are the linear regression coefficients; Here are the regression disturbance coefficients, with a mean of 0 and a variance of . ; Let be the observed value of the independent variable (explanatory variable) at time t. In the BLR algorithm... and Let be random variables. In short, the probability distribution of the parameters is updated using the likelihood function and the prior distribution. The likelihood function for multiple linear regression of random samples is expressed as follows:

[0120] ;

[0121] In the formula, for The conditional probability density function; Represents the response variable; Let represent the predictor variable. If the disturbance term is independent, identically distributed Gaussian white noise with the same variance, then we can conclude that:

[0122] ;

[0123] In the formula, To use the mean ,variance Gaussian probability density function in The value at that location, Let y be the value of the observed value at the r-th data point. Let be the prior distribution function. and The prior distribution of the parameters can be obtained using Bayes' theorem:

[0124] ;

[0125] In the formula, the denominator is a constant, which is the response distribution under a given predicted value; Let represent the likelihood function, which represents the likelihood given parameters. and The probability of observing data y in the case of squared values; This indicates a direct proportional relationship between the left and right sides. It is easy to obtain... Expected value for:

[0126] ;

[0127] In the formula, E() is the expected function. It can be calculated according to the following formula. It is expressed as follows:

[0128] ;

[0129] In the formula, for The conditional expectation of the probability distribution of the parameter with respect to the posterior distribution of the parameter; For the new explanatory variable values; This represents the probability distribution of the predicted values.

[0130] In summary, the target health status assessment result for the energy storage battery can be determined.

[0131] It is understandable that this embodiment addresses the issues of insufficient accuracy and low efficiency in monitoring the health status of lithium-ion batteries in electrochemical energy storage systems. It proposes a rapid battery health status estimation scheme based on dynamic electrochemical impedance spectroscopy, which has significant technical advantages and application value. Firstly, in terms of perception, unlike traditional monitoring methods that rely solely on single electrical characteristics such as charge / discharge voltage and current, this invention comprehensively incorporates dynamic electrochemical impedance information to form a battery aging feature set, effectively enhancing the ability to identify early aging characteristics of the battery. Secondly, in the model building part, an ensemble learning model integrating random forest regression, extreme random tree, K-nearest neighbor regression, and linear regression algorithms is constructed. This model fully utilizes the advantages of multiple machine learning algorithms, overcoming the problems of insufficient accuracy and poor generalization of single algorithms under complex operating conditions. Based on a time-varying factor set, this model can quickly and accurately estimate the health status of electrochemical energy storage batteries and adapt to changes in battery aging characteristics under different operating conditions, improving the real-time performance and accuracy of health status estimation. Finally, in terms of application value, this invention starts from individual battery cells and constructs a health status estimation model that extends to the system level, possessing good hierarchical scalability. This invention significantly improves the accuracy and efficiency of battery health status estimation, providing key technical support for the safe, reliable operation and intelligent maintenance of electrochemical energy storage systems. It is applicable to health monitoring and operation and maintenance management in various energy storage scenarios and has good engineering practicality and promotion prospects.

[0132] Therefore, this application first establishes a full lifecycle database for the energy storage battery based on static and dynamic impedance testing platforms and battery aging external characteristic data. Then, it utilizes the static electrochemical impedance information dataset and battery aging external characteristic data from the database to perform time-domain and frequency-domain feature mining, and combines this with Pearson correlation analysis to determine a first set of time-varying dominant factors for battery aging. Next, it uses the first set of time-varying dominant factors for battery aging and the aforementioned dynamic electrochemical impedance information dataset from the database to determine a second set of time-varying dominant factors for battery aging. Finally, based on the second set of time-varying dominant factors for battery aging and preset multi-model fusion learning rules, it determines the target health status assessment result. This allows for real-time monitoring of the battery's dynamic characteristics under operating conditions, thereby improving the accuracy, efficiency, precision, and real-time performance of battery health status assessment, and providing key technical support for the safe, reliable operation and intelligent maintenance of electrochemical energy storage systems.

[0133] See Figure 9 As shown in the figure, this application also discloses a device for assessing the health status of an energy storage battery, including:

[0134] Database construction module 11 is used to determine the full life cycle database of energy storage batteries based on static impedance test platform, dynamic impedance test platform and battery aging external characteristic data.

[0135] The first factor set determination module 12 is used to perform time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and to determine the first battery aging time-varying dominant factor set based on the corresponding feature mining results and Pearson correlation analysis.

[0136] The second factor set determination module 13 is used to determine the second battery aging time-varying dominant factor set based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database and the grey relational analysis method.

[0137] The evaluation result determination module 14 is used to evaluate the battery health status based on the second battery aging time-varying dominant factor set and the preset multi-model fusion learning rules, so as to determine the target health status evaluation result corresponding to the energy storage battery.

[0138] In some specific embodiments, the database construction module 11 can be used to: collect charge and discharge datasets of the energy storage battery throughout its entire life cycle during the cyclic aging process of a preset number of cycles, in order to determine the battery aging external characteristic data; during the cyclic aging process of the preset number of cycles, determine the static electrochemical impedance spectrum and dynamic electrochemical impedance spectrum of different state of charge nodes throughout the entire life cycle based on the static impedance testing platform and the dynamic impedance testing platform, so as to obtain the static electrochemical impedance information dataset and the dynamic electrochemical impedance information dataset; and perform error elimination and reliability verification on the battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset based on a preset data preprocessing strategy, so as to determine the preprocessed battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset.

[0139] In some specific embodiments, the first factor set determination module 12 can be specifically used to: based on the static electrochemical impedance information dataset, relaxation time distribution algorithm, and equivalent circuit model in the full life cycle database, mine data reflecting the external characteristics of battery aging from the time domain and frequency domain perspectives respectively to determine the battery static electrochemical impedance spectrum degradation feature set; based on each degradation feature in the battery static electrochemical impedance spectrum degradation feature set and the battery health status data, perform Pearson correlation analysis and maximum information coefficient analysis respectively to determine the correlation analysis results; the battery health status data includes battery capacity decay data; based on the correlation analysis results, screen each degradation feature in the battery static electrochemical impedance spectrum degradation feature set to determine the screened first battery aging time-varying dominant factor set.

[0140] In some specific embodiments, the first factor set determination module 12 can be specifically used to: process the static electrochemical impedance information dataset in the full life cycle database based on the relaxation time distribution algorithm to determine the time-domain relaxation time distribution curve; identify the position and parameter value of the peak based on the time-domain relaxation time distribution curve and the preset peak detection rule to determine the time-domain feature mining result; optimize the circuit parameter value by fitting the equivalent circuit model of the static electrochemical impedance information dataset and combining it with the nonlinear least squares algorithm to determine the frequency-domain feature mining result; and map the time-domain feature mining result and the frequency-domain feature mining result point-by-point at different cycle cycles and different state of charge nodes to determine the battery static electrochemical impedance spectrum degradation feature set.

[0141] In some specific embodiments, the second factor set determination module 13 can be used to: remove frequency bands from the dynamic electrochemical impedance information dataset in the full life cycle database to determine a first removed dataset; determine a second removed dataset based on the first battery aging time-varying dominant factor set and the first removed dataset; and determine the correlation analysis result based on the second removed dataset and the grey relational analysis method.

[0142] In some specific embodiments, the second factor set determination module 13 can be used to: perform grey relational analysis on the health features in the second removed dataset and the battery health status data to determine the relational analysis result; and perform feature filtering on the second removed dataset based on the relational analysis result to determine the selected second battery aging time-varying dominant factor set.

[0143] In some specific embodiments, the evaluation result determination module 14 can be used to: use the second battery aging time-varying dominant factor set as model input information and input it into multiple different preset regression models to determine multiple initial battery health status evaluation results; perform feature fusion on each of the initial battery health status evaluation results and the second battery aging time-varying dominant factor set based on a preset linear regression algorithm to determine the target health status evaluation result corresponding to the energy storage battery.

[0144] Furthermore, embodiments of this application also disclose an electronic device, Figure 10 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application.

[0145] Figure 10This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the energy storage battery health status assessment method disclosed in any of the foregoing embodiments. Alternatively, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0146] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0147] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0148] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the energy storage battery health status assessment method executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.

[0149] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for assessing the health status of energy storage batteries. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0150] 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 it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0151] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0152] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0153] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0154] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for assessing the health status of an energy storage battery, characterized in that, include: Based on static impedance testing platform, dynamic impedance testing platform and battery aging external characteristic data, a full life cycle database of energy storage batteries is established. Based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, feature mining in the time domain and frequency domain is performed, and based on the corresponding feature mining results and Pearson correlation analysis, the first battery aging time-varying dominant factor set is determined. Based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and the grey relational analysis method, the second battery aging time-varying dominant factor set is determined. Based on the second battery aging time-varying dominant factor set and the preset multi-model fusion learning rule, the battery health status is assessed to determine the target health status assessment result of the energy storage battery. Specifically, the step of performing time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and determining the first set of time-varying dominant factors for battery aging based on the corresponding feature mining results and Pearson correlation analysis, includes: Based on the static electrochemical impedance information dataset, relaxation time distribution algorithm and equivalent circuit model in the full life cycle database, data reflecting the external characteristics of battery aging are mined from the time domain and frequency domain to determine the battery static electrochemical impedance spectrum degradation feature set. Based on the degradation features in the battery static electrochemical impedance spectroscopy degradation feature set and the battery health status data, Pearson correlation analysis and maximum information coefficient analysis were performed to determine the correlation analysis results; the battery health status data includes battery capacity decay data. Based on the correlation analysis results, each degradation feature in the battery static electrochemical impedance spectroscopy degradation feature set is screened to determine the first set of battery aging time-varying dominant factors.

2. The method for assessing the health status of energy storage batteries according to claim 1, characterized in that, The database for determining the entire lifecycle of energy storage batteries, based on static impedance testing platforms, dynamic impedance testing platforms, and battery aging external characteristic data, includes: During the cyclic aging process with a preset number of cycles, charge and discharge data are collected throughout the entire life cycle of the energy storage battery to determine the external characteristics data of battery aging. During the cyclic aging process of the preset number of cycles, the static electrochemical impedance spectrum and dynamic electrochemical impedance spectrum of different charge state nodes throughout the entire life cycle are determined based on the static impedance test platform and the dynamic impedance test platform, so as to obtain the static electrochemical impedance information dataset and the dynamic electrochemical impedance information dataset. Based on a preset data preprocessing strategy, error removal and reliability verification are performed on the battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset to determine the preprocessed battery aging external characteristic data, the static electrochemical impedance information dataset, and the dynamic electrochemical impedance information dataset.

3. The method for assessing the health status of energy storage batteries according to claim 1, characterized in that, The static electrochemical impedance information dataset, relaxation time distribution algorithm, and equivalent circuit model based on the full life cycle database are used to mine data reflecting the battery's aging external characteristics from both time and frequency domain perspectives to determine the battery's static electrochemical impedance spectrum degradation feature set, including: The static electrochemical impedance information dataset in the full life cycle database is processed based on the relaxation time distribution algorithm to determine the time-domain relaxation time distribution curve. Based on the time-domain relaxation time distribution curve and the preset peak detection rule, the position and parameter value of the peak are identified to determine the time-domain feature mining results. By fitting the equivalent circuit model of the static electrochemical impedance information dataset and optimizing the circuit parameter values ​​using a nonlinear least squares algorithm, the frequency domain feature mining results are determined. The time-domain feature mining results and the frequency-domain feature mining results are mapped point-by-point at different cycle rounds and different state-of-charge nodes to determine the degradation feature set of the battery's static electrochemical impedance spectrum.

4. The method for assessing the health status of energy storage batteries according to claim 1, characterized in that, The determination of the second battery aging time-varying dominant factor set based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and grey relational analysis includes: Frequency bands are removed from the dynamic electrochemical impedance information dataset in the full life cycle database to determine the dataset after the first removal. Based on the first battery aging time-varying dominant factor set and the first removed dataset, the second removed dataset is determined; Based on the second removed dataset and the grey relational analysis method, the relational analysis results are determined.

5. The method for assessing the health status of energy storage batteries according to claim 4, characterized in that, The determination of the correlation analysis results based on the second removed dataset and the grey relational analysis method includes: Grey relational analysis is performed on the health features in the second removed dataset and the battery health status data to determine the correlation analysis results. Based on the correlation analysis results, feature filtering is performed on the second excluded dataset to determine the second set of battery aging time-varying dominant factors.

6. The method for assessing the health status of an energy storage battery according to any one of claims 1 to 5, characterized in that, The battery health status assessment based on the second battery aging time-varying dominant factor set and preset multi-model fusion learning rules includes: The second set of time-varying dominant factors in battery aging is used as model input information and input into multiple different preset regression models to determine multiple initial battery health status assessment results. Based on a preset linear regression algorithm, feature fusion is performed on the initial battery health status assessment results and the second battery aging time-varying dominant factor set to determine the target health status assessment result corresponding to the energy storage battery.

7. A device for assessing the health status of an energy storage battery, characterized in that, include: The database construction module is used to determine the full life cycle database of energy storage batteries based on static impedance testing platform, dynamic impedance testing platform and battery aging external characteristic data; The first factor set determination module is used to perform time-domain and frequency-domain feature mining based on the static electrochemical impedance information dataset in the full life cycle database and the battery aging external characteristic data, and to determine the first battery aging time-varying dominant factor set based on the corresponding feature mining results and Pearson correlation analysis. The second factor set determination module is used to determine the second battery aging time-varying dominant factor set based on the first battery aging time-varying dominant factor set, the dynamic electrochemical impedance information dataset in the full life cycle database, and the grey relational analysis method. The evaluation result determination module is used to evaluate the battery health status based on the second battery aging time-varying dominant factor set and preset multi-model fusion learning rules, so as to determine the target health status evaluation result corresponding to the energy storage battery. The first factor set determination module is used to: based on the static electrochemical impedance information dataset, relaxation time distribution algorithm, and equivalent circuit model in the full life cycle database, mine data reflecting the external characteristics of battery aging from the time domain and frequency domain perspectives to determine the battery static electrochemical impedance spectroscopy degradation feature set; based on each degradation feature in the battery static electrochemical impedance spectroscopy degradation feature set and the battery health status data, perform Pearson correlation analysis and maximum information coefficient analysis to determine the correlation analysis results; the battery health status data includes battery capacity decay data; based on the correlation analysis results, screen each degradation feature in the battery static electrochemical impedance spectroscopy degradation feature set to determine the selected first battery aging time-varying dominant factor set.

8. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the energy storage battery health status assessment method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the energy storage battery health status assessment method as described in any one of claims 1 to 6.

Images