Lithium ion battery life prediction method and system based on health state detection

Abstract

The invention discloses a lithium ion battery life prediction method and system based on health state detection, and relates to the technical field of batteries, and the method comprises the steps: collecting a plurality of real-time operation data of a lithium ion battery in a charge-discharge cycle process, and obtaining a battery state data set; traversing the battery state data set to extract a health characteristic parameter set for health assessment, and obtaining a health state value of the battery; performing cycle use prediction according to the health state value, obtaining a battery residual cycle prediction frequency, performing conversion according to the use duration in combination with the battery use frequency, and generating battery residual use prediction time; and calculating a battery life decline rate and combining with the battery residual use prediction time to perform battery life prediction, generating a residual use life estimation value to perform health state evaluation on the battery, and generating a battery life report. The technical problems of inaccurate battery health assessment and insufficient life prediction precision in the prior art are solved, and the technical effects of improving the battery health assessment accuracy and the life prediction reliability are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery, in particular to a lithium ion battery life prediction method and system based on health state detection. BACKGROUND

[0002] Lithium ion batteries will be affected by factors such as charging and discharging current, temperature environment and cycle number during long-term use, and their performance will gradually decline. The existing health assessment method often relies on a single parameter such as capacity or internal resistance, which is difficult to fully reflect the true state of the battery. At the same time, most life prediction is based on static data, lacking comprehensive analysis and correction of dynamic characteristics in the battery operation process, resulting in deviation of the evaluation results and difficulty in ensuring the accuracy of life prediction. SUMMARY

[0003] The present application provides a lithium ion battery life prediction method and system based on health state detection, which is used to solve the technical problems of inaccurate battery health assessment and insufficient life prediction accuracy in the prior art.

[0004] In view of the above problems, the present application provides a lithium ion battery life prediction method and system based on health state detection.

[0005] The first aspect of the present application provides a lithium ion battery life prediction method based on health state detection, which comprises: Synchronously collecting a plurality of real-time operation data of the lithium ion battery in the charging and discharging cycle process according to a sampling period to obtain a battery state data set; traversing the battery state data set to extract a health feature parameter set for health assessment to obtain a health state value of the battery; performing cycle use prediction according to the health state value to obtain a battery remaining cycle prediction number, converting the battery remaining cycle prediction number according to the use time length in combination with the battery use frequency to generate a battery remaining use prediction time; calculating a battery life decline rate in combination with the battery remaining use prediction time to perform battery life prediction, generating a remaining use life estimation value, performing health state assessment on the battery, and generating a battery life report.

[0006] The second aspect of the present application provides a lithium ion battery life prediction system based on health state detection, which comprises: The data acquisition module is used for synchronously collecting a plurality of real-time operation data of the lithium ion battery in the charging and discharging cycle process according to a sampling period, and obtaining a battery state data set; the health assessment module is used for traversing the battery state data set to extract a health feature parameter set for health assessment, and obtaining a health state value of the battery; the cycle use prediction module is used for performing cycle use prediction according to the health state value, and obtaining a battery remaining cycle prediction number, which is converted according to a use time length in combination with a battery use frequency, to generate a battery remaining use prediction time; and the battery life prediction module is used for calculating a battery life decline rate, performing battery life prediction in combination with the battery remaining use prediction time, generating a remaining use life estimation value, performing health state assessment on the battery, and generating a battery life report.

[0007] The one or more technical solutions provided in the application have at least the following technical effects or advantages: The application synchronously collects a plurality of real-time operation data of the lithium ion battery in the charging and discharging cycle process according to a sampling period, and obtains a battery state data set; traverses the battery state data set to extract a health feature parameter set for health assessment, and obtains a health state value of the battery; performs cycle use prediction according to the health state value, and obtains a battery remaining cycle prediction number, which is converted according to a use time length in combination with a battery use frequency, to generate a battery remaining use prediction time; calculates a battery life decline rate, performs battery life prediction in combination with the battery remaining use prediction time, generates a remaining use life estimation value, performs health state assessment on the battery, and generates a battery life report. The application solves the technical problems of inaccurate battery health assessment and insufficient life prediction accuracy in the prior art, and achieves the technical effects of improving the accuracy of battery health assessment and the reliability of life prediction by combining multi-source data acquisition and health feature analysis. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0009] Figure 1 A lithium ion battery life prediction method based on health state detection provided by the embodiment of the application is shown in the flowchart. Figure 2 A lithium ion battery life prediction system structure based on health state detection provided by the embodiment of the application is shown in the structural diagram.

[0010] The reference signs are explained as follows: data acquisition module 11, health assessment module 12, cycle use prediction module 13, and battery life prediction module 14. DETAILED DESCRIPTION

[0011] This application provides a method and system for predicting the lifespan of lithium-ion batteries based on health status detection. It addresses the technical problems of inaccurate battery health assessment and insufficient lifespan prediction accuracy in the prior art by combining multi-source data acquisition with health feature analysis, thereby improving the accuracy of battery health assessment and the reliability of lifespan prediction.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, this application provides a method for predicting the lifespan of lithium-ion batteries based on health status detection, the method comprising: Step S100: Collect multiple real-time operating data of the lithium-ion battery during the charge-discharge cycle according to the sampling period to obtain the battery status dataset.

[0015] In this embodiment of the application, when collecting multiple real-time operating data of a lithium-ion battery during the charge-discharge cycle according to the sampling period, the voltage sensor, current sensor and temperature sensor are first triggered according to the pre-set sampling period to collect multi-dimensional operating data such as battery voltage, current and temperature. The collected results are then time-aligned and grouped for storage to obtain a battery status dataset reflecting the battery's working state.

[0016] Furthermore, the method provided in the application embodiments, which synchronously collects multiple real-time operating data of the lithium-ion battery during the charge-discharge cycle according to the sampling period to obtain a battery state dataset, also includes: Multiple sampling periods with fixed time intervals are set, and voltage, current, and temperature sensors are synchronously triggered based on these multiple sampling periods. Based on these multiple sampling periods, the voltage sensors collect positive and negative electrode voltage data of the battery in a time sequence. Based on these multiple sampling periods, the current sensors collect charging and discharging current data of the battery in the current direction. Based on these multiple sampling periods, the temperature sensors collect multiple temperature data points on the battery surface at various battery measurement points. The positive and negative electrode voltage data, the charging and discharging current data, and the multiple temperature data are aligned to generate an original operating parameter set. The original operating parameter set is grouped and stored according to the battery cycle period to obtain the battery state dataset.

[0017] In this embodiment, technical experts first set multiple fixed time interval sampling periods according to requirements, such as 10ms for current, 100ms for voltage, and 1s for temperature, and then synchronously trigger the voltage sensor, current sensor, and temperature sensor.

[0018] Subsequently, during multiple sampling periods, voltage sensors collected positive and negative electrode voltage data of the battery in a time-series manner, obtaining positive and negative electrode voltage data reflecting the voltage change pattern during the battery charging and discharging process. Next, during multiple sampling periods, current sensors collected charging and discharging current data of the battery in the direction of current flow, obtaining charging and discharging current data that distinguishes the current characteristics under different operating conditions. Simultaneously, during multiple sampling periods, temperature sensors collected multiple temperature data points on the battery surface at various measurement points, obtaining multiple temperature data reflecting the overall thermal state and local temperature distribution characteristics of the battery.

[0019] After completing the multi-source acquisition of voltage, current and temperature, the positive and negative voltage data, charging and discharging current data and multiple temperature data are unified to the same time base. Time alignment processing is used to eliminate possible delays and offsets between samples, thereby generating a raw set of operating parameters containing multi-dimensional information of voltage, current and temperature.

[0020] Finally, the original set of operating parameters is grouped and stored according to the battery cycle, with one complete charging process and one complete discharging process as a cycle unit to obtain the battery state dataset.

[0021] Step S200: Traverse the battery state dataset to extract the health feature parameter set for health assessment and obtain the battery health state value.

[0022] In this embodiment of the application, when traversing the battery state dataset, feature analysis is performed on the positive and negative electrode voltage data, charge and discharge current data, and multiple temperature data. Voltage curve change features, current step response features, and temperature rise rate features are extracted in sequence, and a set of health feature parameters is constructed after normalization and dimensionality reduction.

[0023] Subsequently, a multi-dimensional health analysis was performed using a set of health characteristic parameters, yielding results for voltage, current, temperature, and cycle history. These results were then combined with capacity and internal resistance health assessments, and weighted according to data mapping relationships to obtain a comprehensive health status value. Finally, the comprehensive health status value was corrected through multi-level verification and calibration steps to obtain the final health status value. The health status value ranges from 0% to 100%, where 100% indicates the battery is in brand new condition, and lower values ​​indicate more severe battery performance degradation.

[0024] Furthermore, in the method provided in the application embodiments, extracting the health feature parameter set by traversing the battery state dataset further includes: The positive and negative electrode voltage data are traversed and analyzed according to voltage characteristics. The duration of voltage plateaus is extracted for health feature analysis to obtain voltage curve change characteristics. The charge and discharge current data are traversed and analyzed according to current characteristics. The charge and discharge current efficiency parameter is extracted for health feature analysis to obtain current step response characteristics. The multiple temperature data are traversed and analyzed according to temperature characteristics. The battery surface temperature gradient parameter is extracted for health feature analysis to obtain temperature rise rate characteristics. The voltage curve change characteristics, current step response characteristics, and temperature rise rate characteristics are normalized to generate an initial set of health feature parameters. Based on the initial set of health feature parameters, dimensionality reduction is performed to generate a key feature parameter set for feature importance evaluation, thus constructing the health feature parameter set.

[0025] In this embodiment, when analyzing the positive and negative voltage data based on voltage characteristics, the positive and negative voltage data are first constructed into a voltage time series. This time series is then smoothed to reduce the impact of sampling noise. Subsequently, the rate of change of the voltage time series is calculated to obtain a voltage slope sequence, and plateau intervals in the voltage curve are identified using a pre-set slope threshold. Within the identified voltage plateau intervals, the duration, start and end times, and average level of the voltage plateau are statistically analyzed to complete the health characteristic analysis and obtain the voltage curve change characteristics that characterize the stability and performance degradation of the charging and discharging process.

[0026] When analyzing and filtering charging and discharging current data based on current characteristics, the collected current data is first constructed into a current time series, and then divided into charging current data and discharging current data through sign determination. Within the charging current data interval, the actual charging current is calculated by integrating the current time series. Simultaneously, the theoretical charging current is calculated by multiplying the constant current value and its duration within the interval. Dividing the two yields the charging current efficiency parameter. Similarly, within the discharging current data interval, the actual discharging current is calculated by integrating, and the theoretical discharging current is calculated by multiplying the constant current value and its duration, thus obtaining the discharging current efficiency parameter. Subsequently, step detection is performed on the current time series. The start and end points of the current step are identified through threshold changes. The step amplitude, step response time, and steady-state error are extracted to complete the health characteristic analysis, obtaining the current step response characteristics reflecting the battery's dynamic response capability.

[0027] When analyzing multiple temperature data points based on their temperature characteristics, the temperature data from each measuring point on the battery surface are first constructed into a temperature time series and then timestamped. The temperature difference between adjacent measuring points at the same time is calculated and combined with the distance between the measuring points to obtain the battery surface temperature gradient parameters. Simultaneously, the temperature time series of a single measuring point is subjected to differencing or linear regression to obtain the rate of temperature change over time, completing the health characteristic analysis and obtaining the temperature rise rate characteristic reflecting the non-uniformity of the battery surface temperature distribution and the trend of thermal change.

[0028] Next, the voltage curve change characteristics, current step response characteristics, and temperature rise rate characteristics are normalized. The numerical standardization method is used to map the characteristics of different physical quantities to a unified scale, generating an initial parameter set for health characteristics.

[0029] Subsequently, dimensionality reduction is performed based on the initial parameter set of health features. In this process, based on the initial parameter set of health features, firstly, through relevant calculations and the construction of the covariance matrix, combined with eigenvalue decomposition, a sequence of feature vectors is obtained. Then, the first K feature vectors are extracted to construct a projection matrix, and the initial parameter set of health features is projected onto this projection matrix to achieve feature selection, ultimately generating a set of key feature parameters.

[0030] When generating a set of key feature parameters for feature importance assessment, the variance contribution rate of each feature is first calculated based on the set, and the features are then ranked according to their variance contribution rates. Subsequently, features with higher variance contribution rates are retained, while those with lower rates are removed, thus avoiding interference from redundant and invalid information. Finally, a healthy feature parameter set is constructed through this selection process.

[0031] Furthermore, in the method provided in the application embodiments, the generation of a key feature parameter set based on the initial set of health features through dimensionality reduction processing also includes: Based on the voltage curve change characteristics, the current step response characteristics, and the temperature rise rate characteristics, correlation calculations are performed to obtain multiple parameter correlation coefficients. Multiple health characteristic parameters are calculated according to these correlation coefficients to construct a covariance matrix. The covariance matrix is ​​then decomposed into eigenvalues ​​to obtain multiple eigenvalues, each containing multiple eigenvectors. These eigenvectors are then sorted in descending order according to their eigenvalues ​​to obtain a eigenvector sequence. The first K eigenvectors of this sequence are extracted to construct a projection matrix. The initial set of health characteristic parameters is projected onto this projection matrix for feature filtering to generate the key characteristic parameter set.

[0032] In this embodiment of the application, when performing correlation calculations based on voltage curve change characteristics, current step response characteristics, and temperature rise rate characteristics, the Pearson correlation coefficient method is used to calculate the linear correlation degree of any two health characteristics one by one in the sample dimension, and to complete the pairwise pairing calculations in sequence to obtain multiple parameter correlation coefficients.

[0033] Next, when calculating multiple health feature parameters based on the correlation coefficients of multiple parameters, the mean of each health feature in the initial parameter set is first removed. Then, the covariance value is calculated for each pair of health features with reference to the correlation coefficients of the parameters. All covariance values ​​are then summarized and arranged according to the row and column positions of the features to construct the covariance matrix.

[0034] Subsequently, when performing eigenvalue decomposition on the covariance matrix, the symmetric matrix eigenvalue decomposition method is used to decompose the covariance matrix into a set of numerical values ​​and vectors, resulting in multiple eigenvalues, each of which corresponds to an eigenvector.

[0035] Next, when sorting multiple eigenvectors in descending order according to multiple eigenvalues, the eigenvalues ​​are sorted from largest to smallest based on their size, and the corresponding eigenvectors are rearranged in the same order to obtain a sequence of eigenvectors arranged from highest to lowest.

[0036] Then, the first K eigenvectors of the eigenvector sequence are extracted to construct a projection matrix. The first K eigenvectors in the eigenvector sequence are combined column-wise to form a matrix, which is used as the projection matrix.

[0037] Finally, when projecting the initial set of health features onto the projection matrix for feature selection, the initial set of health features is mapped to the subspace spanned by the projection matrix through linear projection to generate a dimension-reduced feature representation. Each dimension in this dimension-reduced representation is used as a key feature to form a key feature parameter set.

[0038] Furthermore, in the method provided in the application embodiments, the health feature parameter set is used to perform a health assessment to obtain the battery's health status value, which further includes: Multidimensional analysis is performed based on the set of health feature parameters to extract multidimensional health analysis results, including voltage health analysis results, current health analysis results, temperature health analysis results, and cycle history analysis results. Battery capacity is assessed based on the current and temperature health analysis results to obtain capacity health assessment results. Battery internal resistance is assessed based on the current and temperature health analysis results to obtain internal resistance health assessment results. A mapping analysis is performed between the voltage, current, and temperature health analysis results and the battery state dataset according to the cycle history analysis results to construct a data mapping relationship. The capacity health assessment results and internal resistance health assessment results are weighted according to the data mapping relationship to calculate a comprehensive health state value. The comprehensive health state value is then subjected to hierarchical verification, and recalibrated based on the multi-level verification results to construct the battery's overall health state value.

[0039] In this embodiment, when performing multidimensional analysis based on a set of health characteristic parameters, the least squares method is used to perform linear fitting on the scatter points of voltage curve change characteristics with the number of cycles, the scatter points of current step response characteristics with the number of cycles, and the scatter points of temperature rise rate characteristics with the number of cycles. Specifically, with the number of cycles as the independent variable and the corresponding characteristic values ​​as the dependent variable, the closed-form solution of the slope and intercept is calculated to obtain three regression lines. The coefficient of determination and mean square residual of each line are calculated, thereby forming voltage health analysis results, current health analysis results, and temperature health analysis results. Simultaneously, using the number of cycles and the capacity value measured in each cycle as a set of data, the least squares method is used to calculate the slope and intercept of the regression line of capacity with the number of cycles, forming the cycle history analysis results.

[0040] When assessing battery capacity based on current and temperature health analysis results, the process begins by integrating the discharge process using the current health analysis results to obtain the battery discharge capacity. Then, the ratio of the discharge capacity to the rated capacity is calculated to obtain the capacity retention rate. Subsequently, battery capacity data from multiple historical cycle periods is retrieved and combined with the temperature health analysis results to calculate the capacity decay rate over the number of cycles. Next, the charging process is analyzed based on the capacity decay rate, and a curve showing the change in charging capacity over the number of cycles is plotted. The capacity retention rate is then used as a health assessment indicator to incrementally evaluate the charging capacity curve, ultimately generating the final capacity health assessment result.

[0041] When assessing the battery's internal resistance based on current and temperature health analysis results, the following steps are taken: First, battery state-of-charge (SOC) data is retrieved based on the current health analysis results. The ratio of battery terminal voltage drop to pulse current is calculated using the DC pulse method to obtain the DC internal resistance value. Then, the DC internal resistance value is analyzed for changes over multiple historical cycles to obtain the trend of internal resistance variation with the number of cycles. Subsequently, the rate of increase of the internal resistance variation trend is calculated according to the temperature health analysis results to determine the internal resistance growth rate. Finally, impedance spectroscopy analysis is used to decompose the frequency characteristics and extract the evolution features of the internal resistance at different frequencies, thus forming the internal resistance health assessment result.

[0042] Subsequently, a mapping analysis was performed on the voltage health analysis results, current health analysis results, temperature health analysis results, and battery state dataset based on the cycle history analysis results. In this process, linear regression and least squares methods were used to establish the data mapping relationship. Specifically, using the cycle number as an index, the voltage health analysis results, current health analysis results, and temperature health analysis results corresponding to each cycle were used as inputs, and the capacity and internal resistance values ​​recorded in the battery state dataset under the same cycle were used as outputs. The regression coefficients and intercepts for each input and output were calculated, and the correlation coefficients and mean squared residuals were provided and significance tests were performed. Finally, the coefficients, intercepts, and statistics of each input to the capacity and internal resistance values ​​were summarized to form a data mapping relationship describing the relationship between multidimensional health characteristics and cycle performance, thus obtaining the data mapping relationship.

[0043] Subsequently, when performing a weighted analysis of the capacity health assessment results and internal resistance health assessment results according to the data mapping relationship, a weighted average method based on the normalization of the absolute value of the correlation coefficient is used. The absolute values ​​of the correlation coefficients corresponding to the capacity health assessment results and internal resistance health assessment results are extracted from the data mapping relationship, normalized, and then the weight coefficients are obtained. The capacity health assessment results and internal resistance health assessment results are then linearly combined according to their weights, and the numerical range is normalized to a percentage expression when necessary. For example, when the weight of the capacity health assessment result is 0.6 and the weight of the internal resistance health assessment result is 0.4, the comprehensive health status value is equal to the weighted sum of the two values ​​at 0.6 and 0.4, yielding the comprehensive health status value.

[0044] Finally, when grading and verifying the comprehensive health status value, a health level threshold is first set. For example, 80% to 100% is considered healthy, 60% to 80% is considered slightly degraded, 40% to 60% is considered moderately degraded, and below 40% is considered severely degraded. The batteries corresponding to the comprehensive health status values ​​are then subjected to actual tests, such as constant current constant voltage charge-discharge experiments and DC pulse experiments, to obtain the actual capacity and internal resistance. These results are then compared with the predicted values ​​to calculate the deviation. If there is a systematic deviation between the predicted and actual values ​​within a certain level range, the comprehensive health status value for that range is recalibrated, eliminating the deviation through additive or proportional correction. For example, if the predicted comprehensive health status value is 75%, but the actual tested capacity shows a corresponding state of 70%, the predicted result is uniformly lowered by 5% within that range. After multiple levels of verification and calibration, a health status value reflecting the actual operating state of the battery is finally constructed.

[0045] Furthermore, in the method provided in the application embodiments, the battery capacity is assessed based on the current health analysis results and the temperature health analysis results to obtain a capacity health assessment result, which further includes: Based on the current health analysis results, battery capacity analysis is performed to obtain the battery discharge capacity; the ratio of the battery discharge capacity to the rated capacity is calculated to obtain the capacity retention rate; battery capacity data from multiple historical cycle periods are retrieved and decay calculations are performed according to the temperature health analysis results to obtain the capacity decay rate; charging analysis is performed according to the capacity decay rate, a charging capacity curve is plotted, and the capacity retention rate is used as a battery health assessment index to incrementally evaluate the charging capacity curve, generating the capacity health assessment result.

[0046] In this embodiment of the application, when performing battery capacity analysis based on current health analysis results, the charging and discharging current data is first integrated over time. Specifically, the current time series during the battery discharge phase is used as input, and the product of current and time is accumulated according to the integration calculation method to obtain the battery discharge capacity.

[0047] Next, when calculating the ratio of battery discharge capacity to rated capacity, the battery discharge capacity is compared with the rated capacity specified in the battery design parameters. Using the battery discharge capacity as the numerator and the rated capacity as the denominator, the ratio is calculated to obtain the capacity retention rate.

[0048] Subsequently, when retrieving battery capacity data from multiple historical cycle periods and calculating capacity decay based on the temperature health analysis results, the stored cycle data is first retrieved, and the discharge capacity of each cycle is paired with the corresponding temperature data. Then, a linear regression method is used, with the number of cycles as the independent variable and the discharge capacity as the dependent variable, to calculate the slope of the fitted straight line and obtain the capacity decay rate.

[0049] When performing charging analysis and plotting the charging capacity curve according to the capacity decay rate, the charging current data is first integrated over time to obtain the charging capacity for each cycle. Then, the charging capacity curve is plotted with the number of cycles on the x-axis and the charging capacity on the y-axis.

[0050] Finally, when using capacity retention rate as a battery health assessment indicator and incrementally evaluating the charging capacity curve, the initial cycle capacity retention rate is first set as 100% as the baseline. Then, the difference between the charging capacity of each cycle and the initial cycle charging capacity is calculated for multiple cycles. This difference is then compared to the rated capacity to obtain the capacity retention rate decrease for each cycle. These points showing the change in capacity retention rate with the number of cycles are then connected to form a charging capacity curve, visually reflecting the capacity decay trend. By calculating the difference and analyzing the trend of incremental changes in the charging capacity curve, the rate of capacity decay and interval characteristics at different cycle stages are identified, thereby generating a capacity health assessment result. This result is expressed as a percentage, representing the degree of capacity decay during battery use.

[0051] Furthermore, in the method provided in the application embodiments, the method for assessing the battery internal resistance based on the current health analysis results and the temperature health analysis results to obtain internal resistance health assessment results further includes: Based on the current health analysis results, the battery state of charge data is retrieved to perform DC pulse calculations and obtain the DC internal resistance value; the change of the DC internal resistance value is analyzed based on multiple historical cycle periods to construct the internal resistance change trend; the internal resistance change trend is calculated according to the temperature health analysis results to determine the internal resistance growth rate, and impedance spectrum analysis is performed to extract the internal resistance health assessment results.

[0052] In this embodiment, when retrieving battery state-of-charge data based on current health analysis results to perform DC pulse calculations, a DC pulse current of known amplitude is first applied to the battery under a constant state of charge, and the change in battery terminal voltage at the instant the pulse is applied is recorded. The DC internal resistance value is obtained by calculating the ratio of the battery terminal voltage drop to the pulse current.

[0053] When analyzing the changes in the DC internal resistance value based on multiple historical cycle periods, DC internal resistance values ​​obtained from different cycle periods are retrieved, paired with the corresponding cycle number, and arranged according to the time series. Using a linear regression method with the cycle number as the independent variable and the DC internal resistance value as the dependent variable, the slope of the fitted line is calculated to obtain the internal resistance growth rate. This internal resistance growth rate characterizes the increase in internal resistance per unit cycle number.

[0054] Subsequently, based on the temperature health analysis results, the upward trend of internal resistance change was calculated, and the trends of internal resistance change under different temperature conditions were compared. Specifically, the slopes of the internal resistance change curves under normal temperature and high temperature conditions were calculated, and the internal resistance growth rate was obtained by calculating the difference.

[0055] When performing impedance spectroscopy analysis to extract the internal resistance health assessment results, electrochemical impedance spectroscopy was used. A small-signal AC current was applied within the frequency range of 10 mHz to 100 kHz, and complex impedance data were collected. The ohmic impedance was obtained by reading the high-frequency horizontal intercept based on the Nyquist plot, the charge transfer impedance was obtained by measuring the horizontal span of the mid-frequency semicircle, and the diffusion impedance was obtained based on the slope of the low-frequency near 45-degree slope. Subsequently, the DC internal resistance value, the internal resistance growth rate, the ohmic impedance, the charge transfer impedance, and the diffusion impedance were normalized according to a unified dimension. After weighted normalization based on the absolute value of the Pearson correlation coefficient between each parameter and the cycle performance, the results were linearly synthesized. After consistency verification, the quantitative results in percentage form were output to extract the internal resistance health assessment results.

[0056] Step S300: Based on the health status value, perform cycle usage prediction, obtain the remaining cycle prediction number of the battery, combine the battery usage frequency with the usage duration, and generate the remaining battery usage prediction time.

[0057] In this embodiment, when predicting recyclability based on health status values, a training sample database is first constructed using historical aging data of lithium-ion batteries, and multiple health status intervals are defined on this database. The training samples are then segmented and mapped according to different health status intervals to establish a correspondence between health status and cycle count, forming a health status-cycle count mapping network. Next, the currently obtained health status value is input into this mapping network to obtain the recyclability prediction result. Finally, operational feature analysis is performed using the battery status dataset to correct the prediction result in real time and determine the remaining cycle count of the battery.

[0058] After calculating the remaining cycle predictions, the result is combined with the battery usage frequency for duration conversion. Specifically, when the battery completes one full charge-discharge cycle per day, the remaining cycle predictions are directly equivalent to the remaining usage days; when the battery completes two full cycles per day, the remaining cycle predictions need to be divided by two to obtain the corresponding remaining usage days; if the battery operates under irregular usage scenarios, the average daily cycle count within a certain time window is used as the usage frequency parameter, and the remaining cycle predictions are then divided by this usage frequency parameter to obtain the remaining usage days. The remaining usage days can also be converted to weekly, monthly, or yearly units to generate the battery's remaining usage prediction time.

[0059] Furthermore, in the method provided in the application embodiments, the remaining battery cycle prediction is obtained by predicting the remaining battery cycle count based on the health status value, and the method further includes: A training sample database is constructed by introducing a historical aging dataset of lithium-ion batteries. Multiple health state intervals are defined, and a health state-cycle number mapping network is constructed based on the training sample database by segmenting and mapping the multiple health state intervals. The health state value is then used cyclically according to the health state-cycle number mapping network to obtain the prediction result. Based on the battery state dataset, operational analysis is performed to obtain battery operational characteristics. The prediction result is then corrected in real time according to the battery operational characteristics to determine the remaining predicted number of cycles for the battery.

[0060] In this embodiment, when introducing a pre-stored historical aging dataset of lithium-ion batteries and constructing a training sample database, data cleaning and time alignment methods are used to deduplicate, remove outliers, and unify timestamps on the original records of capacity retention, DC internal resistance, operating temperature, charge / discharge rate, and depth of discharge at different cycle counts. Subsequently, numerical standardization is used to map each physical quantity to a comparable dimensional range, and the data is stored in the database according to a field structure of cycle count—health status—operating conditions, thus obtaining the training sample database.

[0061] When defining multiple health state intervals and performing piecewise mapping, the health state range is divided into several continuous intervals based on the health state value or capacity retention rate, such as 100%~80%, 80%~60%, 60%~40%, and below 40%. Within each interval, the least squares method is used to fit the relationship between health state and cycle count, outputting the slope, intercept, and coefficient of determination. Adjacent intervals are connected using linear interpolation to maintain the continuity and smoothness of the overall relationship, resulting in a piecewise mapping relationship between health state and cycle count. When constructing the health state-cycle count mapping network, the fitted line obtained for each interval is considered a node in the network, with the health state value as input and the cycle count as output. Interval connection and interpolation methods ensure the continuity of the global mapping. Simultaneously, the coefficient of determination and residuals of each node are examined, outliers are eliminated, and recalculation is performed when necessary, thus obtaining the health state-cycle count mapping network.

[0062] When predicting cyclic use using the health state-cycle number mapping network, the current health state value of the battery is located in the corresponding interval. A piecewise function is used for interval lookup and linear interpolation calculation. In case of exceeding the boundary, a limiting extrapolation method is used to avoid abnormal results. Finally, the number of cycles from the current state to the end-of-life threshold is obtained, and the initial result of the remaining cycle prediction is obtained.

[0063] When performing operational analysis and real-time corrections based on battery state datasets, a feature statistics and weighted correction method is used. Specifically, battery operational characteristics are statistically analyzed within a recent window period to obtain feature values ​​such as average daily cycle count, average depth of discharge, charge / discharge rate, and operating temperature. Then, rate correction coefficients, temperature correction coefficients, and depth of discharge correction coefficients are calculated. The rate correction coefficient is calculated by comparing the current lifespan with a reference rate (e.g., 1C) to obtain the lifespan reduction ratio and converting it into a rate correction coefficient. The temperature correction coefficient is calculated by comparing the current operating temperature with a reference temperature (e.g., 25 degrees Celsius) to obtain the lifespan degradation ratio and converting it into a temperature correction coefficient. The depth of discharge correction coefficient is calculated by comparing the current depth of discharge with a reference depth of discharge (e.g., 100%) to obtain the lifespan degradation ratio and converting it into a depth of discharge correction coefficient. After obtaining the rate correction coefficient, temperature correction coefficient, and depth of discharge correction coefficient, they are normalized and combined into a total correction coefficient. Finally, the initial value of the remaining cycle prediction is multiplied by the total correction coefficient to obtain the real-time corrected result, determining the remaining cycle prediction number of the battery. For example, when the initial value of the remaining cycle prediction is 500, and the rate correction factor is 0.90, the temperature correction factor is 0.95, and the depth of discharge correction factor is 0.85, the total correction factor is 0.726, and the real-time corrected remaining cycle prediction is approximately 364.

[0064] Step S400: Calculate the battery life degradation rate and combine it with the predicted remaining battery usage time to predict battery life, generate an estimated remaining lifespan, assess the battery's health status, and generate a battery life report.

[0065] In this embodiment, when calculating the battery life degradation rate and predicting the remaining battery life, the process first involves collecting historical health status data sequences and performing adjacent difference calculations to obtain adjacent change quantities reflecting the gradual changes in health status. Then, based on the health status values, the battery's state evolution is performed, forming a health status change trend that varies with time or cycle count. Next, the adjacent change quantities are used to fit and regress the health status change trend to calculate the life degradation rate, and fluctuation analysis is used to identify potential abnormal aging acceleration information. For the identified abnormal aging acceleration information, the degradation rate is backtracked and optimized to obtain a more accurate battery life degradation rate. Finally, the battery life degradation rate is combined with the predicted remaining battery life to perform failure prediction and generate an estimated remaining service life.

[0066] When assessing the battery's health status, the estimated remaining lifespan is used as the core indicator and analyzed in conjunction with the capacity and internal resistance health assessment results. Specifically, capacity retention is used as an indicator of the battery's energy storage capacity, while DC internal resistance and impedance spectrum characteristics are used as indicators of the battery's transmission performance. These are then combined with the lifespan degradation rate and predicted remaining usage time to comprehensively determine the battery's operating status, resulting in a comprehensive health status value covering capacity performance, transmission performance, and lifespan degradation. When generating the battery life report, the health status value is first categorized according to threshold levels: 80%–100% is considered healthy, 60%–80% is mild degradation, 40%–60% is moderate degradation, and below 40% is severe degradation. The report then displays key parameters such as positive and negative electrode voltage curve changes, charge / discharge current efficiency parameters, temperature rise rate characteristics, capacity retention rate, internal resistance growth rate, and lifespan degradation rate in the form of charts and data, along with a dynamic evolution curve of the health status value. Next, the battery life report presents the predicted number of remaining cycles and the predicted remaining usage time, along with an estimated remaining lifespan. It also compares historical operating data with predicted trends, illustrating potential future risks and maintenance windows for the battery. Finally, based on the above comprehensive analysis, a complete battery life report is generated.

[0067] Furthermore, in the method provided in the application embodiments, calculating the battery life degradation rate and combining it with the predicted remaining battery usage time to predict battery life and generate an estimated remaining lifespan, further includes: Historical health status data sequences are collected and adjacent difference calculations are performed to obtain adjacent changes. Based on the battery's health status value, the battery's state evolution is performed to construct a health status change trend. Based on the adjacent changes, the health status change trend is fitted and regressed to calculate the degradation rate and perform fluctuation analysis to identify abnormal aging acceleration information. The degradation rate is backtracked and optimized based on the abnormal aging acceleration information to obtain the battery life degradation rate. Based on the battery life degradation rate and the battery's remaining usage prediction time, battery failure is extrapolated and predicted to generate the estimated remaining service life.

[0068] In this embodiment of the application, the pre-stored historical health status data sequence is first extracted and adjacent difference calculation is performed. The health status values ​​obtained by the battery under different cycles are arranged in chronological order, and then the difference between two adjacent cycles is calculated. These differences are the adjacent changes.

[0069] Next, the state evolution is performed based on the battery's health state value. The health state value under multiple cycles is plotted as a curve that changes with the number of cycles. This curve reflects the trend of battery performance over time. By smoothing, some random fluctuations are eliminated, thus obtaining the trend of battery health state change.

[0070] Subsequently, when fitting and regressing the trend of health status changes based on adjacent changes, a linear fitting method was used to obtain the overall decline slope, which represents the battery degradation rate. Then, fluctuation analysis was performed in conjunction with the trend. If a sudden increase in the slope was found in certain stages, it was identified as an abnormal acceleration of aging. For example, if the overall decline rate was 0.02% per cycle, and suddenly increased to 0.08% in a certain stage, it was determined that abnormal acceleration of aging had occurred in that stage, thus obtaining preliminary information on the degradation rate and abnormal acceleration of aging.

[0071] Next, when backtracking and optimizing the degradation rate based on the abnormal aging acceleration information, these abnormal stages are marked and the slope of the overall curve is recalculated to avoid deviations caused by fluctuations in individual stages, thereby obtaining a battery life degradation rate that better represents the true degradation level of the battery.

[0072] Finally, when predicting battery failure based on the battery life degradation rate and the predicted remaining battery life, a battery failure threshold is first set, for example, a health status of 80%. Then, the number of cycles required to reduce the current health status to this threshold is calculated based on the degradation rate. This number of cycles is then converted into time based on the battery usage frequency to obtain an estimate of the remaining lifespan. For example, if the current health status is 90%, the degradation rate is 0.02% per cycle, and the average number of cycles per day is 1, it would take approximately 500 cycles to reduce the health status to 80%, or approximately 500 days. This process ultimately yields the estimated remaining lifespan.

[0073] In summary, the embodiments of this application have at least the following technical effects: This application synchronously collects multiple real-time operating data of lithium-ion batteries during charge-discharge cycles according to a sampling period to obtain a battery status dataset; it then extracts a set of health feature parameters from the battery status dataset for health assessment to obtain the battery's health status value; based on the health status value, it performs cycle usage prediction to obtain the remaining cycle prediction number of the battery, combines the battery usage frequency with the usage duration, and generates the remaining usage prediction time of the battery; it calculates the battery life degradation rate and combines it with the remaining usage prediction time to predict the battery life, generating an estimated remaining lifespan value to assess the battery's health status and generate a battery life report. This invention solves the technical problems of inaccurate battery health assessment and insufficient lifespan prediction accuracy in the prior art by combining multi-source data acquisition with health feature analysis, achieving the technical effect of improving the accuracy of battery health assessment and the reliability of lifespan prediction.

[0074] Example 2, based on the same inventive concept as the lithium-ion battery life prediction method based on health status detection in the previous examples, such as... Figure 2As shown, this application provides a lithium-ion battery life prediction system based on health status detection. The system and method embodiments in this application are based on the same inventive concept. The system includes: The data acquisition module 11 is used to synchronously acquire multiple real-time operating data of the lithium-ion battery during the charge-discharge cycle according to the sampling period to obtain a battery status dataset; the health assessment module 12 is used to traverse the battery status dataset to extract a set of health feature parameters for health assessment and obtain the battery's health status value; the cycle usage prediction module 13 is used to predict the cycle usage based on the health status value, obtain the remaining cycle prediction number of the battery, and convert it according to the usage frequency and usage time to generate the remaining usage prediction time of the battery; the battery life prediction module 14 is used to calculate the battery life degradation rate and combine it with the remaining usage prediction time of the battery to predict the battery life, generate an estimated remaining service life value, assess the battery's health status, and generate a battery life report.

[0075] Furthermore, the system is also used to implement the following functions: Multiple sampling periods with fixed time intervals are set, and voltage, current, and temperature sensors are synchronously triggered based on these multiple sampling periods. Based on these multiple sampling periods, the voltage sensors collect positive and negative electrode voltage data of the battery in a time sequence. Based on these multiple sampling periods, the current sensors collect charging and discharging current data of the battery in the current direction. Based on these multiple sampling periods, the temperature sensors collect multiple temperature data points on the battery surface at various battery measurement points. The positive and negative electrode voltage data, the charging and discharging current data, and the multiple temperature data are aligned to generate an original operating parameter set. The original operating parameter set is grouped and stored according to the battery cycle period to obtain the battery state dataset.

[0076] Furthermore, the system is also used to implement the following functions: The positive and negative electrode voltage data are traversed and analyzed according to voltage characteristics. The duration of voltage plateaus is extracted for health feature analysis to obtain voltage curve change characteristics. The charge and discharge current data are traversed and analyzed according to current characteristics. The charge and discharge current efficiency parameter is extracted for health feature analysis to obtain current step response characteristics. The multiple temperature data are traversed and analyzed according to temperature characteristics. The battery surface temperature gradient parameter is extracted for health feature analysis to obtain temperature rise rate characteristics. The voltage curve change characteristics, current step response characteristics, and temperature rise rate characteristics are normalized to generate an initial set of health feature parameters. Based on the initial set of health feature parameters, dimensionality reduction is performed to generate a key feature parameter set for feature importance evaluation, thus constructing the health feature parameter set.

[0077] Furthermore, the system is also used to implement the following functions: Based on the voltage curve change characteristics, the current step response characteristics, and the temperature rise rate characteristics, correlation calculations are performed to obtain multiple parameter correlation coefficients. Multiple health characteristic parameters are calculated according to these correlation coefficients to construct a covariance matrix. The covariance matrix is ​​then decomposed into eigenvalues ​​to obtain multiple eigenvalues, each containing multiple eigenvectors. These eigenvectors are then sorted in descending order according to their eigenvalues ​​to obtain a eigenvector sequence. The first K eigenvectors of this sequence are extracted to construct a projection matrix. The initial set of health characteristic parameters is projected onto this projection matrix for feature filtering to generate the key characteristic parameter set.

[0078] Furthermore, the system is also used to implement the following functions: Multidimensional analysis is performed based on the set of health feature parameters to extract multidimensional health analysis results, including voltage health analysis results, current health analysis results, temperature health analysis results, and cycle history analysis results. Battery capacity is assessed based on the current and temperature health analysis results to obtain capacity health assessment results. Battery internal resistance is assessed based on the current and temperature health analysis results to obtain internal resistance health assessment results. A mapping analysis is performed between the voltage, current, and temperature health analysis results and the battery state dataset according to the cycle history analysis results to construct a data mapping relationship. The capacity health assessment results and internal resistance health assessment results are weighted according to the data mapping relationship to calculate a comprehensive health state value. The comprehensive health state value is then subjected to hierarchical verification, and recalibrated based on the multi-level verification results to construct the battery's overall health state value.

[0079] Furthermore, the system is also used to implement the following functions: Based on the current health analysis results, battery capacity analysis is performed to obtain the battery discharge capacity; the ratio of the battery discharge capacity to the rated capacity is calculated to obtain the capacity retention rate; battery capacity data from multiple historical cycle periods are retrieved and decay calculations are performed according to the temperature health analysis results to obtain the capacity decay rate; charging analysis is performed according to the capacity decay rate, a charging capacity curve is plotted, and the capacity retention rate is used as a battery health assessment index to incrementally evaluate the charging capacity curve, generating the capacity health assessment result.

[0080] Furthermore, the system is also used to implement the following functions: Based on the current health analysis results, the battery state of charge data is retrieved to perform DC pulse calculations and obtain the DC internal resistance value; the change of the DC internal resistance value is analyzed based on multiple historical cycle periods to construct the internal resistance change trend; the internal resistance change trend is calculated according to the temperature health analysis results to determine the internal resistance growth rate, and impedance spectrum analysis is performed to extract the internal resistance health assessment results.

[0081] Furthermore, the system is also used to implement the following functions: A training sample database is constructed by introducing a historical aging dataset of lithium-ion batteries. Multiple health state intervals are defined, and a health state-cycle number mapping network is constructed based on the training sample database by segmenting and mapping the multiple health state intervals. The health state value is then used cyclically according to the health state-cycle number mapping network to obtain the prediction result. Based on the battery state dataset, operational analysis is performed to obtain battery operational characteristics. The prediction result is then corrected in real time according to the battery operational characteristics to determine the remaining predicted number of cycles for the battery.

[0082] Furthermore, the system is also used to implement the following functions: Historical health status data sequences are collected and adjacent difference calculations are performed to obtain adjacent changes. Based on the battery's health status value, the battery's state evolution is performed to construct a health status change trend. Based on the adjacent changes, the health status change trend is fitted and regressed to calculate the degradation rate and perform fluctuation analysis to identify abnormal aging acceleration information. The degradation rate is backtracked and optimized based on the abnormal aging acceleration information to obtain the battery life degradation rate. Based on the battery life degradation rate and the battery's remaining usage prediction time, battery failure is extrapolated and predicted to generate the estimated remaining service life.

[0083] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0084] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0085] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for predicting the lifespan of lithium-ion batteries based on health status detection, characterized in that, The method includes: Multiple real-time operating data of lithium-ion batteries during charge-discharge cycles are collected synchronously according to the sampling period to obtain a battery status dataset. The battery status dataset is traversed to extract a set of health feature parameters for health assessment, and the health status value of the battery is obtained. Based on the health status value, a cycle usage prediction is performed to obtain the remaining battery cycle prediction number. This number is then combined with the battery usage frequency and converted according to the usage duration to generate the remaining battery usage prediction time. The battery life degradation rate is calculated and combined with the predicted remaining battery usage time to predict battery life, generate an estimated remaining lifespan, assess the battery's health status, and generate a battery life report.

2. The lithium-ion battery life prediction method based on health status detection as described in claim 1, characterized in that, Multiple real-time operating data points of the lithium-ion battery during charge-discharge cycles are collected synchronously according to a sampling period to obtain a battery state dataset. The methods include: Set multiple fixed time interval sampling periods, and synchronously trigger voltage, current and temperature sensors based on multiple sampling periods; Based on the multiple sampling periods, the voltage sensor collects the positive and negative electrode voltage data of the battery in a time sequence. Based on the multiple sampling periods, the current sensor collects the charging and discharging current data of the battery in the direction of current. Based on the multiple sampling periods, the temperature sensor collects multiple temperature data of the battery surface according to the battery measuring points; The positive and negative electrode voltage data, the charging and discharging current data, and the multiple temperature data are aligned to generate an original set of operating parameters. The original set of operating parameters is grouped and stored according to the battery cycle period to obtain the battery status dataset.

3. The lithium-ion battery life prediction method based on health status detection as described in claim 2, characterized in that, The method for extracting a set of health feature parameters by traversing the battery state dataset includes: The positive and negative voltage data are traversed and filtered according to voltage characteristics. The duration of voltage plateaus is extracted for health feature analysis to obtain voltage curve change characteristics. The charging and discharging current data is traversed and filtered according to current characteristics. The charging and discharging current efficiency parameters are extracted for health feature analysis to obtain current step response characteristics. The multiple temperature data are traversed and filtered according to temperature characteristics. The battery surface temperature gradient parameters are extracted for health feature analysis to obtain the temperature rise rate characteristics. The voltage curve variation characteristics, current step response characteristics, and temperature rise rate characteristics are normalized to generate an initial parameter set for health characteristics. The initial set of health features is used for dimensionality reduction to generate a set of key feature parameters for feature importance assessment, thus constructing the set of health feature parameters.

4. The lithium-ion battery life prediction method based on health status detection as described in claim 3, characterized in that, Based on the initial set of health features, a dimensionality reduction process is performed to generate a set of key feature parameters. The method includes: Based on the voltage curve change characteristics, the current step response characteristics, and the temperature rise rate characteristics, correlation calculations are performed to obtain correlation coefficients for multiple parameters; The covariance matrix is ​​constructed by calculating the correlation coefficients of the multiple health characteristic parameters. The covariance matrix is ​​decomposed into eigenvalues ​​to obtain multiple eigenvalues, which contain multiple eigenvectors. The feature vectors are sorted in descending order according to the multiple feature values ​​to obtain a feature vector sequence; Extract the first K feature vectors from the feature vector sequence to construct a projection matrix; The initial set of health features is projected onto the projection matrix for feature filtering to generate the key feature parameter set.

5. The lithium-ion battery life prediction method based on health status detection as described in claim 1, characterized in that, A health assessment is performed using a set of health characteristic parameters to obtain the battery's health status value. Methods include: Multidimensional analysis is performed based on the set of health feature parameters to extract multidimensional health analysis results, which include voltage health analysis results, current health analysis results, temperature health analysis results, and cycle history analysis results. Based on the current health analysis results and the temperature health analysis results, a health assessment of the battery capacity is performed to obtain the capacity health assessment results. Based on the current health analysis results and the temperature health analysis results, the battery internal resistance is assessed to obtain the internal resistance health assessment results. Based on the cycle history analysis results, a mapping analysis is performed on the voltage health analysis results, the current health analysis results, the temperature health analysis results, and the battery state dataset to construct a data mapping relationship; The capacity health assessment results and the internal resistance health assessment results are weighted and analyzed according to the data mapping relationship to calculate the comprehensive health status value; The comprehensive health status value is subjected to hierarchical verification, and the comprehensive health status value is recalibrated based on the multi-level verification results to construct the battery's health status value.

6. The lithium-ion battery life prediction method based on health status detection as described in claim 5, characterized in that, Based on the current health analysis results and the temperature health analysis results, a health assessment of the battery capacity is performed to obtain a capacity health assessment result. The method includes: Based on the current health analysis results, battery capacity analysis is performed to obtain the battery discharge capacity. Calculate the ratio of the battery's discharge capacity to its rated capacity to obtain the capacity retention rate; The battery capacity data from multiple historical cycle periods is retrieved and the capacity decay rate is calculated based on the temperature health analysis results. The charging analysis is performed according to the capacity decay rate, the charging capacity curve is plotted, and the capacity retention rate is used as the battery health assessment index to incrementally evaluate the charging capacity curve, thereby generating the capacity health assessment result.

7. The lithium-ion battery life prediction method based on health status detection as described in claim 5, characterized in that, Based on the current health analysis results and the temperature health analysis results, a health assessment of the battery internal resistance is performed to obtain the internal resistance health assessment results. The method includes: Based on the current health analysis results, the battery state of charge data is retrieved to perform DC pulse calculation and obtain the DC internal resistance value. Based on multiple historical cycles, the change trend of the DC internal resistance value is constructed by analyzing the changes in the internal resistance. Based on the temperature health analysis results, the internal resistance change trend is calculated to determine the internal resistance growth rate, and impedance spectrum analysis is performed to extract the internal resistance health assessment results.

8. The lithium-ion battery life prediction method based on health status detection as described in claim 1, characterized in that, Based on the health status value, a cycle usage prediction is performed to obtain the predicted number of remaining battery cycles. The method includes: A training sample database is constructed by introducing historical aging datasets of lithium-ion batteries. Multiple health state intervals are defined, and a health state-cycle number mapping network is constructed by segmenting and mapping the training sample database according to the multiple health state intervals. The health status value is predicted for cyclic use according to the health status-cycle number mapping network to obtain the prediction result; Based on the battery state dataset, operational analysis is performed to obtain battery operational characteristics. The prediction results are then corrected in real time according to these characteristics to determine the remaining cycle prediction number for the battery.

9. The lithium-ion battery life prediction method based on health status detection as described in claim 1, characterized in that, The method for calculating the battery life degradation rate and combining it with the predicted remaining battery usage time to predict battery life and generate an estimate of remaining lifespan includes: Collect historical health status data sequences and perform adjacent difference calculations to obtain adjacent changes; Based on the battery's health status value, the battery's state evolution is performed to construct a health status change trend; Based on the adjacent changes, the trend of health status change is fitted and regressed to calculate the rate of decline and perform fluctuation analysis to identify abnormal aging acceleration information. Based on the abnormal aging acceleration information, the degradation rate is backtracked and optimized to obtain the battery life degradation rate. Based on the battery life degradation rate and the predicted remaining battery usage time, battery failure is extrapolated and predicted to generate the estimated remaining lifespan.

10. A lithium-ion battery life prediction system based on health status detection, characterized in that, The system is used to execute the lithium-ion battery life prediction method based on health status detection as described in any one of claims 1-9, and the system comprises: The data acquisition module is used to synchronously acquire multiple real-time operating data of the lithium-ion battery during the charge and discharge cycle according to the sampling period, and obtain the battery status dataset. The health assessment module is used to traverse the battery status dataset to extract a set of health feature parameters for health assessment and obtain the battery's health status value. The cycle usage prediction module is used to predict the cycle usage based on the health status value, obtain the remaining cycle prediction number of the battery, combine it with the battery usage frequency according to the usage time, and generate the remaining battery usage prediction time. The battery life prediction module is used to calculate the battery life degradation rate and combine it with the remaining battery usage prediction time to predict the battery life, generate an estimated remaining lifespan, assess the battery's health status, and generate a battery life report.

Images