Storage battery intelligent management system and method based on multi-dimensional parameter acquisition

Abstract

The invention provides an intelligent storage battery management system and method based on multi-dimensional parameter acquisition, and relates to the technical field of battery management. Performing state response evaluation on the initial monitoring parameters based on a health evaluation mechanism and relevance between electrochemical impedance characteristics and polarization internal resistance characteristics to obtain monitoring confidence of the initial monitoring parameters corresponding to each storage battery monomer; determining a co-evolution relationship of the monitoring parameters between the adjacent storage battery monomers according to the historical multi-dimensional monitoring parameters, and determining state covariable characteristics of the monitoring parameters between the adjacent storage battery monomers according to the co-evolution relationship and the series topological structure of the target storage battery pack; and performing multi-target correction on the initial monitoring parameters according to the monitoring confidence and the state covariable characteristics, and further completing health state management of the target storage battery pack by the corrected monitoring parameters. According to the scheme, the reliability correction of the health monitoring parameters of the storage battery pack under complex working conditions can be realized, so that the robustness of intelligent management of the storage battery is improved.

Description

Technical Field

[0001] This application relates to the field of battery management technology, and more specifically, to a smart battery management system and method based on multi-dimensional parameter acquisition. Background Technology

[0002] With the rapid development of new energy vehicles, energy storage systems, and smart grids, the health status of battery packs directly affects the safety, lifespan, and system stability of equipment. Traditional battery management systems typically rely on single voltage or temperature monitoring, which is insufficient to comprehensively reflect electrochemical performance and internal aging characteristics. Especially under complex operating conditions, fluctuations in ambient temperature, changes in load, and differences in charge-discharge cycles can lead to deviations or anomalies in monitoring parameters, affecting the accurate assessment and intelligent control of the battery pack. Therefore, developing intelligent battery management methods based on multi-dimensional parameter acquisition is of great significance for improving the reliability and lifespan management of batteries.

[0003] In existing technologies, battery health monitoring largely relies on individual cell voltage, temperature, or simple impedance measurements, lacking comprehensive analysis of the correlation between multi-dimensional parameters and the co-evolutionary patterns of adjacent cells. Under complex operating conditions, single parameters are easily affected by sensor errors, environmental disturbances, or transient load changes, causing monitoring data to deviate from the true state, thus affecting the accuracy of health assessments and management decisions. Furthermore, traditional methods do not fully utilize the series topology characteristics of battery packs and cannot effectively correct for state constraints between adjacent cells, leading to parameter inconsistencies or abnormal accumulation during charge-discharge cycles. These problems result in insufficient robustness of existing battery management systems under multiple operating conditions and cycles. Therefore, how to achieve reliable correction of battery pack health monitoring parameters under complex operating conditions, thereby improving the robustness of intelligent battery management, has become a challenge for the industry. Summary of the Invention

[0004] This application provides a battery intelligent management system and method based on multi-dimensional parameter acquisition, which can realize the reliability correction of battery pack health monitoring parameters under complex working conditions, thereby improving the robustness of battery intelligent management.

[0005] Firstly, this application provides a battery intelligent management method based on multi-dimensional parameter acquisition, comprising: Distributed multidimensional parameter acquisition is performed on each battery cell in the target battery pack to obtain the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell, and to obtain the initial monitoring parameters. Electrochemical impedance characteristics and polarization resistance characteristics are extracted from the initial monitoring parameters, and the correlation between the characteristics is determined. Based on the health evaluation mechanism of electrochemical theory, the state response evaluation of the initial monitoring parameters is carried out in combination with the correlation between the characteristics, and the monitoring confidence level of the initial monitoring parameters corresponding to each battery cell is obtained. Historical multidimensional monitoring parameters of individual battery cells are obtained, and the co-evolution relationship of monitoring parameters between adjacent battery cells is determined based on the historical multidimensional monitoring parameters. Then, the state covariate characteristics of monitoring parameters between adjacent battery cells are determined based on the co-evolution relationship and the series topology of the target battery pack. Based on the monitoring confidence level and the state covariate characteristics, the initial monitoring parameters are corrected using a multi-objective method to obtain the corrected target monitoring parameters. Then, the health status management of the target battery pack is completed based on the target monitoring parameters.

[0006] In some embodiments, the battery cell refers to an electrochemical battery unit in a battery pack that serves as an energy storage and release unit.

[0007] In some embodiments, extracting electrochemical impedance characteristics and polarization resistance characteristics from initial monitoring parameters, and then determining the correlation between the characteristics, specifically includes: The characteristic extraction frequency bands of electrochemical impedance data are determined, including the low-frequency band reflecting ohmic resistance, the mid-frequency band reflecting charge transfer resistance, and the high-frequency band reflecting diffusion impedance. The impedance modulus, the impedance phase angle corresponding to the characteristic frequency point, and the rate of change of the impedance modulus with frequency are selected from each frequency band as electrochemical impedance characteristics. The dynamic changes, the time to reach the stable value, and the correlation with the charging and discharging current of the polarization internal resistance under different charging and discharging states are extracted as characteristics of the polarization internal resistance. The correlation coefficient between electrochemical impedance characteristics and polarization internal resistance characteristics was calculated using statistical analysis methods, and the correlation between the characteristics was determined by combining the synchronous change trend during charge-discharge cycles.

[0008] In some embodiments, the health assessment mechanism based on electrochemical theory combines the correlation between features to evaluate the state response of the initial monitoring parameters, and obtains the monitoring confidence level of the initial monitoring parameters for each battery cell, specifically including: Establish a health assessment mechanism that includes threshold ranges for electrochemical impedance characteristics and polarization internal resistance characteristics under different health levels; The extracted electrochemical impedance features, polarization resistance features, and their correlations are compared with the threshold range to mark abnormal features and abnormal correlations. A state response evaluation index is constructed based on the proportion of normal characteristics, the degree of deviation of abnormal characteristics, and the scope of influence of abnormal correlation. Then, the monitoring confidence level of the initial monitoring parameters corresponding to the battery cells is determined based on the state response evaluation index.

[0009] In some embodiments, determining the co-evolution relationship of monitoring parameters between adjacent battery cells based on historical multidimensional monitoring parameters specifically includes: Historical multidimensional monitoring parameters under the same charge-discharge cycle number, ambient temperature and load conditions are selected to form a historical parameter dataset for adjacent cells. Time series analysis was performed on the historical parameter dataset to calculate the amount of change, rate of change, and fluctuation amplitude of each monitoring parameter; The difference in the changes in monitoring parameters between adjacent battery cells, the deviation rate of the change rate, and the ratio of the fluctuation amplitude are statistically analyzed to determine the stable correlation range. Then, based on the stable correlation range, the synchronous change pattern of monitoring parameters between adjacent battery cells is extracted as the co-evolution relationship.

[0010] In some embodiments, determining the state covariate characteristics of monitoring parameters between adjacent battery cells based on the co-evolution relationship and the series topology of the target battery pack specifically includes: The series topology of the target battery pack is determined by the physical connection method and electrical circuit relationship of the target battery pack; A state covariate model of multi-dimensional monitoring parameters is constructed based on the co-evolution relationship of adjacent battery cells; The series topology of the target battery pack is set as a constraint condition for the state covariate model, limiting the boundary consistency of voltage, current and temperature. The evolution process of monitoring parameters between adjacent battery cells is simulated using the state covariate model, and the state covariate characteristics of monitoring parameters between adjacent battery cells are extracted from the simulation results.

[0011] In some embodiments, the initial monitoring parameters are multi-objective corrected based on the monitoring confidence level and the state covariate characteristics to obtain the corrected target monitoring parameters, specifically including: For each battery cell, when the monitoring confidence level of the initial monitoring parameter of the battery cell is less than the preset confidence threshold, all adjacent battery cells adjacent to the battery cell are selected. The target constraint parameters of the pre-trained calibration model are updated by using the monitoring confidence level of individual battery cells and the state covariate characteristics of monitoring parameters between adjacent battery cells. The initial monitoring parameters of the battery cells are corrected based on the pre-trained correction model updated with the target constraint parameters to obtain the corrected target monitoring parameters for each battery cell, and then the corrected target monitoring parameters for each battery cell are obtained.

[0012] Secondly, this application provides a battery intelligent management system based on multi-dimensional parameter acquisition, comprising: The data acquisition module is used to collect distributed multi-dimensional parameters of each battery cell in the target battery pack, and to obtain the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell to obtain the initial monitoring parameters. The feature processing module is used to extract electrochemical impedance features and polarization internal resistance features from the initial monitoring parameters, and then determine the correlation between features. Based on the health evaluation mechanism of electrochemical theory, the correlation between features is combined to evaluate the state response of the initial monitoring parameters and obtain the monitoring confidence level of the initial monitoring parameters for each battery cell. The feature processing module is also used to acquire historical multidimensional monitoring parameters of individual battery cells, determine the co-evolution relationship of monitoring parameters between adjacent battery cells based on the historical multidimensional monitoring parameters, and then determine the state covariate characteristics of monitoring parameters between adjacent battery cells based on the co-evolution relationship and the series topology of the target battery pack. The calibration module is used to perform multi-target calibration on the initial monitoring parameters based on the monitoring confidence level and the state covariate characteristics to obtain the calibrated target monitoring parameters, and then complete the health status management of the target battery pack based on the target monitoring parameters.

[0013] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described intelligent battery management method based on multi-dimensional parameter acquisition.

[0014] Fourthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned intelligent battery management method based on multi-dimensional parameter acquisition.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In this embodiment, distributed multidimensional parameter acquisition is performed on each battery cell in the target battery pack to obtain the voltage, temperature, multi-frequency electrochemical impedance, and polarization resistance of each battery cell, thus obtaining initial monitoring parameters. Electrochemical impedance characteristics and polarization resistance characteristics are extracted from the initial monitoring parameters to determine the correlation between the characteristics. Based on the health evaluation mechanism of electrochemical theory and the correlation between the characteristics, the initial monitoring parameters are evaluated for state response to obtain the monitoring confidence level of the initial monitoring parameters for each battery cell. Historical multidimensional monitoring parameters of the battery cells are obtained, and the co-evolution relationship of monitoring parameters between adjacent battery cells is determined based on the historical multidimensional monitoring parameters. Then, based on the co-evolution relationship and the series topology of the target battery pack, the state covariate characteristics of the monitoring parameters between adjacent battery cells are determined. Based on the monitoring confidence level and the state covariate characteristics, the initial monitoring parameters are corrected for multiple objectives to obtain the corrected target monitoring parameters. Finally, the health status management of the target battery pack is completed based on the target monitoring parameters.

[0016] Therefore, this application performs multi-objective correction on the initial monitoring parameters based on the monitoring confidence level and the state covariate characteristics to obtain the corrected target monitoring parameters. Firstly, by comprehensively collecting voltage, temperature, multi-frequency electrochemical impedance, and polarization resistance data for each battery cell, initial monitoring parameters are formed. This captures the internal electrochemical state and operating characteristics of the battery from the source, solving the problem of incomplete information from traditional single-parameter monitoring and providing a reliable data foundation for subsequent refined analysis. Secondly, by extracting electrochemical impedance and polarization resistance characteristics and analyzing the correlation between these characteristics, combined with the health evaluation mechanism of electrochemical theory, a monitoring confidence level is constructed. This accurately identifies the reliability level of individual cell parameters, enabling early identification and risk quantification of abnormal or deviating parameters. This aspect ensures the reliability of the monitoring data and provides a scientific basis for state response evaluation. Then, based on historical multidimensional monitoring parameters, a co-evolutionary relationship between adjacent cells is constructed, and state covariate characteristics are generated by combining the series topology of the battery pack. This allows the spatial constraints and temporal evolution laws between monitoring parameters to be systematically incorporated into the correction model, solving the parameter inconsistency problem caused by neglecting the mutual influence of adjacent cells in traditional methods. Finally, based on the monitoring confidence level and state covariate characteristics, the initial monitoring parameters are corrected using a multi-objective method to obtain the corrected target monitoring parameters. This achieves adaptive correction of abnormal deviations, ensuring the stability and consistency of monitoring data under complex environments and cyclic conditions, thereby improving the accuracy of the overall health status assessment of the battery pack and the robustness of intelligent management. In summary, the proposed solution can achieve reliable correction of battery pack health monitoring parameters under complex operating conditions, thereby improving the robustness of intelligent battery management. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of a battery intelligent management method based on multi-dimensional parameter acquisition, according to some embodiments of this application; Figure 2 This is a schematic flowchart illustrating the process of determining monitoring confidence level according to some embodiments of this application; Figure 3 This is a flowchart illustrating the determination of state covariate characteristics according to some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a battery intelligent management system based on multi-dimensional parameter acquisition, according to some embodiments of this application; Figure 5 This is a schematic diagram of the structure of a computer device that implements a battery intelligent management method based on multi-dimensional parameter acquisition, according to some embodiments of this application. Detailed Implementation

[0018] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] refer to Figure 1 The figure is an exemplary flowchart of a battery intelligent management method based on multi-dimensional parameter acquisition, according to some embodiments of this application. This battery intelligent management method based on multi-dimensional parameter acquisition mainly includes the following steps: In step 101, distributed multidimensional parameter acquisition is performed on each battery cell in the target battery pack to obtain the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell, thus obtaining the initial monitoring parameters.

[0020] It should be noted that, in this application, a single battery cell refers to an electrochemical battery unit used for energy storage and release within a battery pack. It should also be noted that, in this application, the distributed multi-dimensional parameter acquisition of each battery cell in the target battery pack refers to simultaneously acquiring multi-source data such as voltage, temperature, and electrochemical characteristics at each cell. Specifically, in implementation, an integrated sensing node is deployed on each battery cell of the target battery pack. This sensing node includes a voltage detection module, a temperature sensor, a multi-frequency electrochemical impedance spectroscopy circuit, and a polarization internal resistance measurement unit, used to simultaneously acquire the voltage and temperature of each cell. Parameters such as electrochemical impedance and polarization resistance are collected, and each sensing node is evenly arranged along the electrode connection end or shell interface of the battery cell to ensure the comprehensiveness and real-time performance of data acquisition. The collected data is directly transmitted to the host computer or battery management system via wired bus or wireless communication, realizing distributed acquisition and centralized management of multi-dimensional parameters of each cell in the target battery pack. It should be further noted that in this application, electrochemical impedance can be obtained by using a multi-frequency impedance measurement circuit, and polarization resistance can be recorded by using a polarization resistance measurement device. Other dimensional parameters can be directly obtained by corresponding sensors, which will not be elaborated here.

[0021] In step 102, electrochemical impedance characteristics and polarization resistance characteristics are extracted from the initial monitoring parameters, and the correlation between the characteristics is determined. Based on the health evaluation mechanism of electrochemical theory, the state response evaluation of the initial monitoring parameters is carried out in combination with the correlation between the characteristics, and the monitoring confidence level of the initial monitoring parameters corresponding to each battery cell is obtained.

[0022] In some embodiments, extracting electrochemical impedance characteristics and polarization resistance characteristics from initial monitoring parameters, and then determining the correlation between the characteristics, can be achieved through the following steps: The characteristic extraction frequency bands of electrochemical impedance data are determined, including the low-frequency band reflecting ohmic resistance, the mid-frequency band reflecting charge transfer resistance, and the high-frequency band reflecting diffusion impedance. The impedance modulus, the impedance phase angle corresponding to the characteristic frequency point, and the rate of change of the impedance modulus with frequency are selected from each frequency band as electrochemical impedance characteristics. The dynamic changes, the time to reach the stable value, and the correlation with the charging and discharging current of the polarization internal resistance under different charging and discharging states are extracted as characteristics of the polarization internal resistance. The correlation coefficient between electrochemical impedance characteristics and polarization internal resistance characteristics was calculated using statistical analysis methods, and the correlation between the characteristics was determined by combining the synchronous change trend during charge-discharge cycles.

[0023] It should be noted that the impedance magnitude in this application refers to the degree to which an electrode impedes current under alternating current, used to quantify the overall impedance level of an electrochemical system at a specific frequency; the characteristic frequency point in this application is a frequency position in the electrochemical impedance spectrum with significant physical or chemical significance, corresponding to the dominant response of charge transfer, electrolyte diffusion, or ohmic impedance; the impedance phase angle in this application is the angle formed by the ratio of the imaginary part to the real part of the impedance, used to reflect the relative proportion of capacitive and resistive components in the electrochemical system; the electrochemical impedance characteristics in this application are a set of key parameters characterizing the electrochemical state of the battery extracted from the impedance magnitude, impedance phase angle, and their frequency variation laws; the polarization internal resistance characteristics in this application are a set of parameters characterizing the polarization behavior of the battery, such as the dynamic change of internal resistance, stabilization time, and relationship with charge and discharge current under different charge and discharge states; the correlation coefficient in this application is a numerical index used to quantify the degree of linear or nonlinear correlation between two or more characteristics, which can reflect the synchronous change law and mutual influence strength between characteristics.

[0024] In practice, firstly, a conventional electrochemical impedance spectroscopy (EIS) instrument can be used to perform full-frequency impedance testing on individual battery cells, collecting impedance data from 0.01 Hz to 10 kHz, including magnitude and phase angle. Then, based on industry-standard frequency band division, the impedance data is divided into three characteristic frequency bands: a low-frequency band reflecting ohmic resistance, a mid-frequency band reflecting charge transfer resistance, and a high-frequency band reflecting diffusion impedance. The low-frequency impedance primarily reflects the ohmic losses between the electrolyte and electrodes, the mid-frequency impedance reflects the electrochemical reaction resistance on the electrode surface, and the high-frequency impedance reflects the ion diffusion resistance at the electrode surface. The internal diffusion process, which requires no complex algorithms and relies solely on the testing instrument's functionality and basic industry knowledge to perform data segmentation, will not be elaborated upon here. Secondly, when screening electrochemical impedance characteristics, the impedance magnitude, phase angle at characteristic frequency points, and the rate of change of the impedance magnitude with frequency can be directly extracted from each frequency band as characteristic parameters. The impedance magnitude can be obtained by calculating the average value of all frequency points within each band. The phase angle at characteristic frequency points is read by plotting the corresponding Bode phase diagram and identifying the inflection point of the phase change. The rate of change of the impedance magnitude with frequency can be obtained by dividing the difference in magnitude between adjacent frequency points by the frequency difference. The calculations are as follows: Then, in the process of extracting polarization resistance characteristics, a conventional charge / discharge tester can be used to apply typical charge / discharge states to the battery cells. Simultaneously, a high-precision internal resistance tester is used to collect polarization resistance data in real time. The extracted characteristics include three aspects: the dynamic change value is directly calculated from the difference in internal resistance under different states; the time to reach a stable value is obtained by recording the time it takes for the internal resistance change amplitude to reach a stable condition; and the correlation value with the charge / discharge current is calculated using a linear regression correlation coefficient. The entire process relies entirely on conventional equipment data recording and basic statistical processing, which will not be elaborated upon here. Finally, when determining the correlation between characteristics, the Pearson correlation coefficient can be used to calculate the correlation between electrochemical impedance characteristics and polarization resistance characteristics. Multiple sets of test data are input into commonly used data processing tools for calculation. Simultaneously, the synchronous change trend of the charge / discharge cycle process is verified. A biaxial curve is plotted with the number of cycles, impedance magnitude, and dynamic change value of polarization resistance to observe the change trend and determine whether the two types of characteristics change synchronously with the number of cycles. Finally, the correlation between characteristics is determined by combining the correlation coefficient and trend consistency. The product of the correlation coefficient and trend consistency indicators can then be used to describe the correlation between the characteristics.

[0025] In some embodiments, reference Figure 2 As shown in the figure, this is a flowchart illustrating the process of determining monitoring confidence in some embodiments of this application. In this embodiment, the health evaluation mechanism based on electrochemical theory, combined with the correlation between features, evaluates the state response of the initial monitoring parameters. The monitoring confidence of the initial monitoring parameters corresponding to each battery cell can be obtained by the following steps: In step 1021, a health evaluation mechanism is established that includes threshold ranges for electrochemical impedance characteristics and polarization internal resistance characteristics under different health levels; In step 1022, the extracted electrochemical impedance features, polarization internal resistance features and their correlations are compared with the threshold range, and abnormal features and abnormal correlations are marked. In step 1023, a state response evaluation index is constructed based on the proportion of normal characteristics, the degree of deviation of abnormal characteristics, and the scope of influence of abnormal correlation. Then, the monitoring confidence level of the initial monitoring parameters corresponding to the battery cell is determined based on the state response evaluation index.

[0026] It should be noted that the health evaluation mechanism in this application is a rule system for classifying the electrochemical state of battery cells, distinguishing different health levels through threshold ranges in order to identify the operational reliability and potential risks of the battery; the state response evaluation index in this application is a comprehensive index used to quantify the performance and degree of abnormality of various monitoring characteristics of battery cells; and the monitoring confidence level in this application is a quantitative value reflecting the reliability and accuracy of the initial monitoring parameters of battery cells.

[0027] In practice, firstly, electrochemical impedance and polarization resistance characteristics of batteries with different health levels are collected through routine experiments. These characteristics are then analyzed using electrochemical theory to determine the features of each level. The upper and lower limits of these characteristics are compiled into a table to form a health evaluation mechanism. For example, healthy batteries correspond to lower mid-frequency impedance and polarization resistance, while failed batteries correspond to higher impedance values ​​and abnormal correlation characteristics. Secondly, the extracted features of each battery cell to be evaluated are compared one by one with the threshold table. A simple logical judgment is used to mark the health level and abnormal features of each characteristic. For example, if impedance and polarization resistance fall within the middle range, the battery is considered sub-healthy; if the correlation between features is below the threshold, it is marked as abnormal. This process relies entirely on the matching of basic data. The method combines conventional judgment methods without the need for complex modeling or algorithms. Then, it constructs state response evaluation indicators and determines monitoring confidence levels. These can be quantified using three basic indicators: the proportion of normal characteristics reflects the percentage of characteristics meeting health thresholds; the degree of deviation of abnormal characteristics reflects the extent to which characteristics deviate from normal thresholds; and the scope of influence of abnormal correlation reflects the degree of influence of abnormal characteristics on the overall assessment. The monitoring confidence levels of the initial monitoring parameters for each individual are then calculated by weighting each indicator. It should be further noted that the weights between indicators can be set based on historical data and industry experience. Furthermore, before weighting, the indicators undergo dimension elimination processing. Generally, normalization can be used to eliminate dimensions, which will not be elaborated upon here.

[0028] It should be noted that this application's solution establishes a health evaluation mechanism based on electrochemical theory and combines the correlation between electrochemical impedance characteristics and polarization internal resistance characteristics to evaluate the state response of initial monitoring parameters. This addresses the problem of insufficient reliability of monitoring data caused by single-feature evaluation or experience-based judgment in existing technologies. Traditional methods typically rely solely on single parameter thresholds or historical experience to judge battery health, failing to fully consider the synergistic changes among various electrochemical parameters and easily overlooking potential anomalies or misjudging health status. By extracting multidimensional electrochemical features, calculating correlation coefficients between features, and constructing state response evaluation indicators, this solution can quantify the reliability of individual cell parameters, identify abnormal features and abnormal correlations, and generate specific monitoring confidence values. This provides a quantifiable basis for subsequent parameter calibration and maintenance decisions. The resulting technical effects include: first, improving the reliability of individual cell monitoring parameters and reducing judgment bias caused by sensor errors or short-term fluctuations; second, achieving multidimensional feature synergistic analysis, enabling timely detection of abnormal states; and third, providing quantifiable and operable evaluation indicators for overall battery pack health management, improving the scientific nature and accuracy of maintenance decisions.

[0029] In step 103, historical multidimensional monitoring parameters of individual battery cells are obtained, and the co-evolution relationship of monitoring parameters between adjacent battery cells is determined based on the historical multidimensional monitoring parameters. Then, the state covariate characteristics of monitoring parameters between adjacent battery cells are determined based on the co-evolution relationship and the series topology of the target battery pack.

[0030] It should be noted that the acquisition of historical multidimensional monitoring parameters of battery cells in this application refers to the collection of multidimensional characteristic data such as voltage, temperature, electrochemical impedance and polarization resistance recorded by each cell in past charge and discharge cycles.

[0031] In some embodiments, determining the co-evolution relationship of monitoring parameters between adjacent battery cells based on historical multidimensional monitoring parameters can be achieved through the following steps: Historical multidimensional monitoring parameters under the same charge-discharge cycle number, ambient temperature and load conditions are selected to form a historical parameter dataset for adjacent cells. Time series analysis was performed on the historical parameter dataset to calculate the amount of change, rate of change, and fluctuation amplitude of each monitoring parameter; The difference in the changes in monitoring parameters between adjacent battery cells, the deviation rate of the change rate, and the ratio of the fluctuation amplitude are statistically analyzed to determine the stable correlation range. Then, based on the stable correlation range, the synchronous change pattern of monitoring parameters between adjacent battery cells is extracted as the co-evolution relationship.

[0032] It should be noted that the stable correlation range in this application refers to the range in which the differences in the changes of monitoring parameters between adjacent battery cells remain consistent, and is used to identify reliable synchronous relationships between parameters; the co-evolutionary relationship in this application refers to the law that the monitoring parameters of adjacent battery cells exhibit synchronous changes during the charging and discharging process.

[0033] In practice, firstly, historical multi-dimensional monitoring parameters of all individual cells in the target battery pack are extracted from the battery operation and maintenance history database, including voltage, temperature, multi-frequency impedance, and polarization resistance. Then, these parameters are uniformly filtered according to charge / discharge cycle count, ambient temperature, and load conditions, retaining only data from adjacent cells under the same cycle count, temperature, and load conditions. For example, parameter data from adjacent cells such as cells 1 and 2, or cells 2 and 3, are selected and organized into a structured dataset of adjacent cell pairs, monitoring time, and corresponding parameter values. This achieves standardized data preprocessing. The entire process relies on basic filtering functions and structured data organization, requiring no complex operations. Secondly, the filtered historical parameter data is arranged chronologically to form a time series. For each... The parameters of each individual cell are calculated for their changes, rates of change, and amplitudes. The change is the difference between parameters at consecutive time points, the rate of change is the ratio of the change to the time interval, and the amplitude is the difference between the maximum and minimum values ​​of the parameter over a period of time. Then, the differences in changes, the deviation rate of change, and the amplitude ratio between adjacent cells are statistically calculated, and the mean and standard deviation are used to determine the stable correlation range. By observing the change patterns of these indicators within the stable range, the synchronous change trends of adjacent cells under different health states are summarized. For example, the voltage change difference remains within a certain range, the rate of change deviation remains small, and the amplitude ratio is close to consistent. Finally, this pattern is used as the co-evolution relationship of monitoring parameters between adjacent battery cells.

[0034] In some embodiments, reference Figure 3 As shown in the figure, this is a flowchart illustrating the determination of state covariate characteristics in some embodiments of this application. In this embodiment, the determination of state covariate characteristics of monitoring parameters between adjacent battery cells based on the co-evolutionary relationship and the series topology of the target battery pack can be achieved through the following steps: In step 1031, the series topology of the target battery pack is determined by the physical connection method and electrical circuit relationship of the target battery pack; In step 1032, a state covariate model of multi-dimensional monitoring parameters is constructed based on the cooperative evolution relationship of adjacent battery cells; In step 1033, the series topology of the target battery pack is set as a constraint condition of the state covariate model, limiting the boundary consistency of voltage, current and temperature; In step 1034, the evolution process of monitoring parameters between adjacent battery cells is simulated using the state covariate model, and then the state covariate features of monitoring parameters between adjacent battery cells are extracted from the simulation results.

[0035] It should be noted that the series topology in this application refers to the circuit layout formed by arranging the individual cells in the battery pack according to the electrical connection sequence, which is used to clarify the voltage and current transmission paths between cells and the overall loop relationship; the state covariate model in this application is a mathematical model used to describe the interdependence and synchronous evolution relationship of the multidimensional monitoring parameters of adjacent battery cells; the state covariate characteristics in this application are characteristics that reflect the synchronous change law of the monitoring parameters of adjacent battery cells.

[0036] In practical implementation, firstly, mainstream electrical drawing software is used to draw the physical connection diagram of the battery pack, marking the arrangement order of individual cells, terminal blocks, and wire information. Then, based on Kirchhoff's laws, the electrical relationships of the series circuit are sorted out, including that the current of each cell is equal and the total voltage is the arithmetic sum of the voltages of each cell. At the same time, the electrical parameters of key nodes are marked, forming a complete topological structure diagram of physical connections and electrical logic. Secondly, when constructing the state covariate model of multidimensional monitoring parameters, the multidimensional monitoring parameters of adjacent cells are used as variables. A multiple linear equation for independent and dependent variables is constructed using conventional regression tools, incorporating historical co-evolution data. By substituting the data into the model, equations and coefficients describing the covariant relationships between parameters of individual cells are obtained. The model reflects the correlation strength between parameters such as voltage, electrochemical impedance, polarization resistance, and temperature. The construction process relies entirely on well-known multivariate modeling methods, requiring no complex algorithms or innovative models, which will not be elaborated upon here. Then, the current, voltage, and temperature conduction characteristics of adjacent cells in the series circuit are transformed into mathematical boundaries that the model can recognize, limiting the range of current deviation, voltage accumulation error, and temperature difference between adjacent cells to ensure that the simulation parameters conform to actual electrical constraints. Finally, the constructed covariate model and topological constraints are integrated into the simulation platform. A modular simulation was constructed, including modules for parameter input, covariate calculation, and constraint determination. The simulation cycle and step size were set to run the simulation, obtaining the evolution curves of each cell parameter over time. State covariate characteristics, including covariate coefficient stability, parameter deviation thresholds, and response delay time, were extracted from the simulation results. Statistical analysis methods were used to quantify these characteristics. It should be further noted that the state covariate model is a relational model trained based on the linear or approximately linear relationship between multidimensional monitoring parameters. The covariance relationship between adjacent battery cell parameters is quantified by establishing a multivariate linear equation between independent and dependent variables. During training, historical monitoring data under the same cycle number, temperature, and load conditions were first collected. The voltage, electrochemical impedance, polarization resistance, and temperature of adjacent cells were used as variables. A conventional regression method was used to fit the multivariate linear equation to obtain the coefficients of each parameter, reflecting the co-evolution law between cells. Parameter selection was based on the principles of physical meaning and statistical significance, ensuring that the independent variables could fully explain the changes in the dependent variable. Simultaneously, constraints were imposed on the regression coefficients to meet the electrical conditions of the series topology, such as equal current, voltage superposition, and temperature difference limitations, thereby constructing a reliable model that can be used for dynamic simulation and stable extraction of state covariate characteristics.

[0037] It should be noted that the proposed solution combines the co-evolution relationship of adjacent battery cells with the series topology to construct a multi-dimensional state covariate model and conduct simulations. This solves the problem of error accumulation caused by independent analysis of individual cell parameters and lack of constraints between adjacent cells in the prior art. This method can extract state covariate features that reflect the synchronous changes of voltage, current and temperature, so that the monitoring parameters of adjacent cells maintain physical and electrical consistency in the dynamic evolution process. This improves the accuracy and stability of monitoring data, enhances the ability to identify abnormal cells, and provides accurate reference for health assessment and subsequent parameter correction, thereby achieving technical optimization of battery pack operating status prediction and operation and maintenance decision-making.

[0038] In step 104, the initial monitoring parameters are corrected for multiple objectives based on the monitoring confidence level and the state covariate characteristics to obtain the corrected target monitoring parameters, and then the health status management of the target battery pack is completed based on the target monitoring parameters.

[0039] In some embodiments, the initial monitoring parameters are multi-objective corrected based on the monitoring confidence level and the state covariate characteristics to obtain the corrected target monitoring parameters, which can be achieved by the following steps: For each battery cell, when the monitoring confidence level of the initial monitoring parameter of the battery cell is less than the preset confidence threshold, all adjacent battery cells adjacent to the battery cell are selected. The target constraint parameters of the pre-trained calibration model are updated by using the monitoring confidence level of individual battery cells and the state covariate characteristics of monitoring parameters between adjacent battery cells. The initial monitoring parameters of the battery cells are corrected based on the pre-trained correction model updated with the target constraint parameters to obtain the corrected target monitoring parameters for each battery cell, and then the corrected target monitoring parameters for each battery cell are obtained.

[0040] It should be noted that the multi-objective calibration in this application refers to simultaneously optimizing and adjusting the initial monitoring parameters of a single entity by comprehensively considering the monitoring confidence level and the characteristics of the state covariates, so as to obtain more accurate and reliable target monitoring parameters; the target constraint parameters in this application refer to the reference values ​​and boundary conditions used to limit the adjustment range of the single entity monitoring parameters during the calibration process.

[0041] In practical implementation, firstly, in a spreadsheet storing the monitoring confidence levels and initial monitoring parameters of each cell in the target battery pack, an industry-standard confidence threshold of 80% is preset to identify cells with low confidence. Cells below this threshold are quickly located using the spreadsheet's filtering function, and their adjacent cell numbers and corresponding monitoring parameters are directly determined based on the series topology. The parameter data of the low-confidence cells and their adjacent cells are then summarized to form a dataset to be calibrated. Secondly, when updating the target constraint parameters of the pre-trained calibration model, deviation constraints on adjacent cell parameters are extracted based on state covariate characteristics, including voltage tolerance, impedance ratio range, and polarization internal resistance difference. The low-confidence cells are assigned a weight equal to their own monitoring confidence percentage, and the remaining percentage is allocated to adjacent cells according to their confidence ratios. The calibration target value is then calculated based on the weighting coefficients and initial parameters, and combined with the deviation constraints of the state covariate characteristics, the constraint parameters are updated. For example... For example, for voltage parameters, a preliminary correction value can be calculated. By adjusting the weights or using the adjacent mean correction method, the correction result can be made to be close to the parameters of adjacent high-confidence units while also meeting the allowable deviation range of the state covariate characteristics. Then, in the process of obtaining the target monitoring parameters by correcting the initial parameters, the preliminary correction value can be verified against the deviation constraints. If the deviation exceeds the constraint range, the weights of low-confidence units and adjacent units are adjusted according to the synchronous change law of the state covariate characteristics, and the weighted average is recalculated until the deviation requirements are met. Finally, the multi-dimensional parameters such as voltage, impedance, and polarization resistance are corrected to obtain the target monitoring parameters of each low-confidence unit. The entire process relies only on basic arithmetic operations, weighted averaging, and constraint verification methods, without the need for complex algorithms. It can be completed using conventional spreadsheets or data processing software. The specific operation steps and methods of data screening, weight allocation, weighted calculation, and constraint verification are fully disclosed and will not be elaborated here.

[0042] It should be noted that the proposed solution combines the monitoring confidence level of individual cells with the state covariate characteristics of adjacent cells to perform multi-objective correction of initial monitoring parameters. This solves the problem in traditional methods where low-confidence parameters easily lead to data inconsistencies between cells and the inability to automatically identify abnormal deviations. Technically, by dynamically adjusting the target constraint parameters of the pre-trained correction model, the correction values ​​of low-confidence cells are kept within the range of physical consistency and co-evolution laws, while maintaining synchronous changes with adjacent cells. The core beneficial effects include significantly improving the accuracy and reliability of individual cell monitoring parameters, reducing the impact of measurement errors and sensor drift, ensuring the consistency of voltage, current, and impedance data of the entire battery pack in time and space, and providing highly reliable basic data for accurate health status assessment and operation and maintenance decisions.

[0043] It should also be noted that health status management of a target battery pack based on target monitoring parameters refers to using highly reliable monitoring parameters of each individual cell, after calibration, to assess, analyze, and manage the health status of the entire battery pack, thereby supporting life prediction, anomaly diagnosis, and operation and maintenance decisions. The specific implementation includes: first, calculating key health indicators for each individual cell based on target monitoring parameters, such as state of charge, health status, internal resistance change rate, and capacity decay rate, to quantify individual cell performance and degradation trends; then, comparing the individual cell health indicators with preset health level thresholds to determine the individual cell health level, and combining state covariate characteristics and inter-cell consistency relationships, generating a group-level health score through statistical or model aggregation methods to reflect overall performance degradation and potential risks; finally, developing management strategies based on the group-level health score and the distribution of abnormal cells, including optimizing charging and discharging schemes, maintenance reminders, or cell replacement, and displaying individual and group-level health status trends through a visual interface to achieve refined and dynamic battery pack health management.

[0044] On the other hand, in some embodiments, this application provides a battery intelligent management system based on multi-dimensional parameter acquisition, referencing... Figure 4 The figure is a schematic diagram of the structure of a battery intelligent management system based on multi-dimensional parameter acquisition according to some embodiments of this application. The battery intelligent management system 400 based on multi-dimensional parameter acquisition includes: a data acquisition module 401, a feature processing module 402, and a correction module 403, which are described below: The data acquisition module 401 is used to perform distributed multi-dimensional parameter acquisition on each battery cell in the target battery pack, and to acquire the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell to obtain the initial monitoring parameters. The feature processing module 402 is used to extract electrochemical impedance features and polarization internal resistance features from the initial monitoring parameters, and then determine the correlation between features. Based on the health evaluation mechanism of electrochemical theory, the correlation between features is combined to evaluate the state response of the initial monitoring parameters, and obtain the monitoring confidence level of the initial monitoring parameters corresponding to each battery cell. In this application, the feature processing module 402 is also used to obtain historical multidimensional monitoring parameters of individual battery cells, determine the co-evolution relationship of monitoring parameters between adjacent battery cells based on the historical multidimensional monitoring parameters, and then determine the state covariate characteristics of monitoring parameters between adjacent battery cells based on the co-evolution relationship and the series topology of the target battery pack. The calibration module 403 is used to perform multi-target calibration on the initial monitoring parameters based on the monitoring confidence level and the state covariate characteristics to obtain the calibrated target monitoring parameters, and then complete the health status management of the target battery pack based on the target monitoring parameters.

[0045] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described intelligent battery management method based on multi-dimensional parameter acquisition.

[0046] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device implementing a battery intelligent management method based on multi-dimensional parameter acquisition, according to some embodiments of this application. The battery intelligent management method based on multi-dimensional parameter acquisition in the above embodiments can... Figure 5 The computer device shown is used to implement this, and the computer device 500 includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0047] Processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0048] The communication bus 502 can be used to transmit information between the aforementioned components.

[0049] Memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 503 may exist independently and be connected to processor 501 via communication bus 502. Memory 503 may also be integrated with processor 501.

[0050] The memory 503 stores program code for executing the solution of this application, and its execution is controlled by the processor 501. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. The intelligent battery management method based on multi-dimensional parameter acquisition in the above embodiments can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0051] Communication interface 504 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0052] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0053] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0054] In addition, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned intelligent battery management method based on multi-dimensional parameter acquisition.

[0055] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0056] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for intelligent battery management based on multi-dimensional parameter acquisition, characterized in that, include: Distributed multidimensional parameter acquisition is performed on each battery cell in the target battery pack to obtain the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell, and to obtain the initial monitoring parameters. Electrochemical impedance characteristics and polarization resistance characteristics are extracted from the initial monitoring parameters, and the correlation between the characteristics is determined. Based on the health evaluation mechanism of electrochemical theory, the state response evaluation of the initial monitoring parameters is carried out in combination with the correlation between the characteristics, and the monitoring confidence level of the initial monitoring parameters corresponding to each battery cell is obtained. Historical multidimensional monitoring parameters of individual battery cells are obtained, and the co-evolution relationship of monitoring parameters between adjacent battery cells is determined based on the historical multidimensional monitoring parameters. Then, the state covariate characteristics of monitoring parameters between adjacent battery cells are determined based on the co-evolution relationship and the series topology of the target battery pack. Based on the monitoring confidence level and the state covariate characteristics, the initial monitoring parameters are corrected using a multi-objective method to obtain the corrected target monitoring parameters. Then, the health status management of the target battery pack is completed based on the target monitoring parameters.

2. The method as described in claim 1, characterized in that, The battery cell refers to the electrochemical battery unit in a battery pack that serves as the storage and release unit for energy.

3. The method as described in claim 1, characterized in that, Extracting electrochemical impedance and polarization resistance characteristics from initial monitoring parameters, and then determining the correlation between these characteristics, specifically includes: The characteristic extraction frequency bands of electrochemical impedance data are determined, including the low-frequency band reflecting ohmic resistance, the mid-frequency band reflecting charge transfer resistance, and the high-frequency band reflecting diffusion impedance. The impedance modulus, impedance phase angle corresponding to characteristic frequency points, and rate of change of impedance modulus with frequency are selected from each frequency band as electrochemical impedance characteristics. The dynamic changes, the time to reach the stable value, and the correlation with the charging and discharging current of the polarization internal resistance under different charging and discharging states are extracted as characteristics of the polarization internal resistance. The correlation coefficient between electrochemical impedance characteristics and polarization internal resistance characteristics was calculated using statistical analysis methods, and the correlation between the characteristics was determined by combining the synchronous change trend during charge-discharge cycles.

4. The method as described in claim 1, characterized in that, The health assessment mechanism based on electrochemical theory combines the correlation between characteristics to evaluate the state response of initial monitoring parameters, and obtains the monitoring confidence level of the initial monitoring parameters for each battery cell, specifically including: Establish a health assessment mechanism that includes threshold ranges for electrochemical impedance characteristics and polarization internal resistance characteristics under different health levels; The extracted electrochemical impedance features, polarization resistance features, and their correlations are compared with the threshold range to mark abnormal features and abnormal correlations. A state response evaluation index is constructed based on the proportion of normal characteristics, the degree of deviation of abnormal characteristics, and the scope of influence of abnormal correlation. Then, the monitoring confidence level of the initial monitoring parameters corresponding to the battery cells is determined based on the state response evaluation index.

5. The method as described in claim 1, characterized in that, Determining the co-evolution relationship of monitoring parameters between adjacent battery cells based on historical multidimensional monitoring parameters specifically includes: Historical multidimensional monitoring parameters under the same charge-discharge cycle number, ambient temperature and load conditions are selected to form a historical parameter dataset for adjacent cells. Time series analysis was performed on the historical parameter dataset to calculate the amount of change, rate of change, and fluctuation amplitude of each monitoring parameter; The difference in the changes in monitoring parameters between adjacent battery cells, the deviation rate of the change rate, and the ratio of the fluctuation amplitude are statistically analyzed to determine the stable correlation range. Then, based on the stable correlation range, the synchronous change pattern of monitoring parameters between adjacent battery cells is extracted as the co-evolution relationship.

6. The method as described in claim 1, characterized in that, Based on the aforementioned co-evolutionary relationship and the series topology of the target battery pack, the state covariate characteristics of the monitoring parameters between adjacent battery cells are specifically determined as follows: The series topology of the target battery pack is determined by the physical connection method and electrical circuit relationship of the target battery pack; A state covariate model of multi-dimensional monitoring parameters is constructed based on the co-evolution relationship of adjacent battery cells; The series topology of the target battery pack is set as a constraint condition for the state covariate model, limiting the boundary consistency of voltage, current and temperature. The evolution process of monitoring parameters between adjacent battery cells is simulated using the state covariate model, and the state covariate characteristics of monitoring parameters between adjacent battery cells are extracted from the simulation results.

7. The method as described in claim 1, characterized in that, Based on the monitoring confidence level and the state covariate characteristics, the initial monitoring parameters are corrected using a multi-objective method to obtain the corrected target monitoring parameters, which specifically include: For each battery cell, when the monitoring confidence level of the initial monitoring parameter of the battery cell is less than the preset confidence threshold, all adjacent battery cells adjacent to the battery cell are selected. The target constraint parameters of the pre-trained calibration model are updated by using the monitoring confidence level of individual battery cells and the state covariate characteristics of monitoring parameters between adjacent battery cells. The initial monitoring parameters of the battery cells are corrected based on the pre-trained correction model updated with the target constraint parameters to obtain the corrected target monitoring parameters for each battery cell, and then the corrected target monitoring parameters for each battery cell are obtained.

8. A battery intelligent management system based on multi-dimensional parameter acquisition, characterized in that, include: The data acquisition module is used to collect distributed multi-dimensional parameters of each battery cell in the target battery pack, and to obtain the voltage, temperature, multi-frequency electrochemical impedance and polarization internal resistance of each battery cell to obtain the initial monitoring parameters. The feature processing module is used to extract electrochemical impedance features and polarization internal resistance features from the initial monitoring parameters, and then determine the correlation between features. Based on the health evaluation mechanism of electrochemical theory, the correlation between features is combined to evaluate the state response of the initial monitoring parameters and obtain the monitoring confidence level of the initial monitoring parameters for each battery cell. The feature processing module is also used to acquire historical multidimensional monitoring parameters of individual battery cells, determine the co-evolution relationship of monitoring parameters between adjacent battery cells based on the historical multidimensional monitoring parameters, and then determine the state covariate characteristics of monitoring parameters between adjacent battery cells based on the co-evolution relationship and the series topology of the target battery pack. The calibration module is used to perform multi-target calibration on the initial monitoring parameters based on the monitoring confidence level and the state covariate characteristics to obtain the calibrated target monitoring parameters, and then complete the health status management of the target battery pack based on the target monitoring parameters.

9. A computer device comprising a memory and a processor, the memory storing code, characterized in that, The processor is configured to acquire the code and execute the intelligent battery management method based on multi-dimensional parameter acquisition as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent battery management method based on multi-dimensional parameter acquisition as described in any one of claims 1 to 7.

Images