Battery parameter identification method and system

Abstract

The invention discloses a battery parameter identification method and system, belongs to the technical field of battery management, and solves the problems of high model dependency, weak anti-noise capability, difficulty in balancing real-time performance and precision, lack of aging adaptability and limited generalization capability of an existing method. The method comprises the steps of exciting a battery based on a mixed pulse power characteristic test or a dynamic working condition test, collecting dynamic test data of the battery, pre-constructing a dual-polarization model, identifying and analyzing the dynamic test data by the dual-polarization model, and outputting a parameter identification result. In the invention, the dual-polarization model comprises the ohmic internal resistance and the two RC parallel networks for representing electrochemical polarization and concentration polarization respectively, and the voltage change of the battery in the charging and discharging process can be reflected more accurately, so that the electrochemical process of the battery is described more accurately; the precision, robustness and full-life-cycle applicability of battery parameter identification are remarkably improved, and key technical support is provided for a high-safety and high-reliability battery management system.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery management, and in particular relates to a battery parameter identification method and system. Background Art

[0002] With the explosive growth of new energy vehicles, energy storage power stations and other fields, lithium-ion batteries, as core energy storage components, their safety and service life directly determine the reliability of the entire system. The Battery Management System (BMS), as the "brain" of battery operation, relies on accurate perception of battery status for its core functions, and battery parameter identification is the technical cornerstone for achieving this goal. Accurate estimation of battery parameters is a key input for the BMS to formulate charging and discharging strategies, balancing control, fault warning and life prediction. Therefore, improving the accuracy, real-time performance and robustness of parameter identification is of great significance to ensuring battery safety, extending service life and reducing system costs.

[0003] At present, battery parameter identification methods mainly include offline fitting based on equivalent circuit model (ECM) (such as HPPC test) and online identification technology (such as recursive least squares RLS, extended Kalman filter EKF), as well as intelligent optimization algorithms (such as particle swarm optimization PSO, genetic algorithm GA) and data-driven methods (such as neural network, support vector regression). In recent years, technological development has tended to integrate multiple methods, such as combining adaptive filtering (AEKF) with machine learning to improve parameter tracking capabilities under dynamic conditions, or using cloud data to collaboratively optimize model parameters. However, existing methods still have significant shortcomings:

[0004] 1) The model is highly dependent, and ECM is difficult to accurately describe the nonlinear characteristics of the battery (such as the relaxation effect), while the electrochemical model calculation is complex and difficult to apply online.

[0005] 2) Weak anti-noise capability. Online identification algorithms (such as RLS) are easily affected by sensor noise, resulting in parameter drift and affecting the reliability of BMS decision-making.

[0006] 3) It is difficult to balance real-time performance and accuracy. High-precision models are time-consuming to calculate, and lightweight models are not accurate enough to meet the BMS's requirements for both high real-time performance and high accuracy.

[0007] 4) Lack of aging adaptability: The parameter identification model cannot dynamically track the time-varying parameters caused by battery aging. Estimation errors accumulate during long-term operation and cannot support full lifecycle management.

[0008] 5) The generalization ability is limited. The data-driven method relies on large-scale labeled data, making cross-model migration difficult and having a high risk of failure in small sample scenarios.

[0009] Therefore, there is an urgent need for a battery parameter identification method and system that can break through model dependence, suppress noise interference, dynamically track aging characteristics, and have strong generalization and high real-time performance. Summary of the Invention

[0010] The purpose of the present invention is to address the shortcomings of the existing technology and provide a battery parameter identification method and system to solve the problems of the existing methods such as strong model dependence, weak noise resistance, difficulty in balancing real-time and accuracy, lack of aging adaptability and limited generalization ability.

[0011] Existing battery parameter identification methods have the problems of strong model dependence, weak noise resistance, difficulty in balancing real-time and accuracy, lack of aging adaptability and limited generalization ability. To address the above problems, we propose a battery parameter identification method and system. In short, when implementing the method, the battery is first stimulated based on a mixed pulse power characteristic test or a dynamic operating condition test, and the dynamic test data of the battery is collected. Then, a dual-polarization model is pre-constructed, and the pre-constructed dual-polarization model is optimized by identifying parameters. The dynamic test data is used as input, and the dual-polarization model is executed. The dual-polarization model identifies and analyzes the dynamic test data and outputs the parameter identification results. In an embodiment of the present invention, the dual-polarization model includes an ohmic internal resistance and two RC parallel networks that characterize electrochemical polarization and concentration polarization, respectively. It can more accurately reflect the voltage changes of the battery during the charging and discharging process, thereby more accurately describing the electrochemical process of the battery, reducing dependence on the traditional ECM model structure, and making up for its shortcomings in describing the nonlinear characteristics of the battery. It systematically solves the core problems of the existing technology models, such as insufficient nonlinear description, noise sensitivity, poor real-time performance, weak aging adaptability, and low generalization ability, and significantly improves the accuracy, robustness, and applicability of battery parameter identification throughout the entire life cycle, providing key technical support for high-safety and high-reliability battery management systems.

[0012] The present invention is implemented as follows: a battery parameter identification method, the battery parameter identification method comprising:

[0013] S10, stimulating the battery based on a hybrid pulse power characteristic test or a dynamic operating condition test, and collecting dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data;

[0014] S20, pre-constructing a dual-polarization model, optimizing the pre-constructed dual-polarization model by identifying parameters, and outputting the optimized dual-polarization model;

[0015] S30 , loading dynamic test data, taking the dynamic test data as input, executing a dual-polarization model, the dual-polarization model identifying and analyzing the dynamic test data, and outputting parameter identification results.

[0016] The hybrid pulse power characteristics test includes charge and discharge pulses at different SOC points and a rest phase test to stimulate the dynamic response of the battery.

[0017] The dual polarization model includes an ohmic internal resistance R O And two RC parallel networks R1C1, R2C 2， The two RC parallel networks R1C1 and R2C2 represent electrochemical polarization and concentration polarization respectively;

[0018] The dual-polarization model equation is expressed as:

[0019] U terminal =OCV(SOC)-I*R0-U1-U2

[0020] Among them, the parameters to be identified are θ = [R0, R1, C1, R2, C2].

[0021] When optimizing the pre-built dual-polarization model through identification parameters, the identification parameters are substituted into the dual-polarization model simulation, compared with the actual data, and the root mean square error (RMS) evaluation accuracy is calculated. At the same time, considering the influence of temperature, a correction relationship between the parameters and temperature is established, and the dual-polarization model aging parameters are regularly updated.

[0022] The dual-polarization model identification and analysis method for dynamic test data includes:

[0023] S301, using HPPC test data, calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification;

[0024] S302 , acquiring dynamic test data, and updating the parameters to be identified of the dual-polarization model θ=[R0, R1, C1, R2, C2] in real time based on the extended Kalman filter or recursive least squares method combined with the SOC identification dynamic test data, thereby completing online parameter identification.

[0025] On the other hand, the present invention also provides a battery parameter identification system, the system comprising:

[0026] A data acquisition module, which excites the battery based on a hybrid pulse power characteristic test or a dynamic operating condition test and collects dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data;

[0027] A model building module is used to pre-build a dual-polarization model, optimize the pre-built dual-polarization model by identifying parameters, and output the optimized dual-polarization model;

[0028] The parameter identification module is used to load dynamic test data, take the dynamic test data as input, execute the dual-polarization model, and the dual-polarization model identifies and analyzes the dynamic test data and outputs parameter identification results.

[0029] The model building module includes:

[0030] A model building unit, used for pre-building a dual-polarization model;

[0031] The model optimization unit optimizes the pre-built dual-polarization model by identifying parameters and outputs the optimized dual-polarization model.

[0032] The parameter identification module includes:

[0033] The offline identification unit uses HPPC test data to calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump. O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification;

[0034] The online identification unit is used to obtain dynamic test data, identify the dynamic test data based on the extended Kalman filter or the recursive least squares method combined with the SOC, and update the dual-polarization model parameters to be identified θ = [R0, R1, C1, R2, C2] in real time to complete the online parameter identification.

[0035] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0036] In an embodiment of the present invention, the dual-polarization model includes an ohmic internal resistance and two RC parallel networks that characterize electrochemical polarization and concentration polarization, respectively. It can more accurately reflect the voltage changes of the battery during the charging and discharging process, thereby more accurately describing the electrochemical process of the battery, reducing dependence on the traditional ECM model structure, and making up for its shortcomings in describing the nonlinear characteristics of the battery. It systematically solves the core problems of the existing technology models, such as insufficient nonlinear description, noise sensitivity, poor real-time performance, weak aging adaptability, and low generalization ability, and significantly improves the accuracy, robustness, and applicability of battery parameter identification throughout the entire life cycle, providing key technical support for high-safety and high-reliability battery management systems.

[0037] In the embodiment of the present invention, the dual-polarization model, through the combination of a fast-slow dual-time-scale RC network, not only retains the computational efficiency of the equivalent circuit model, but also significantly improves the model's accuracy in depicting the nonlinear characteristics of the battery by describing the polarization process in stages. A battery management system based on the dual-polarization model can more accurately predict the battery's charge and discharge behavior, thereby optimizing charging strategies, discharge control, and thermal management strategies. This helps extend the battery's service life and improve the overall performance and safety of the battery system. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the implementation flow of the battery parameter identification method provided by the present invention.

[0039] Figure 2 It is a structural diagram of the battery parameter identification system provided by the present invention. DETAILED DESCRIPTION

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit this application. The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of this application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0041] Existing battery parameter identification methods have the problems of strong model dependence, weak noise resistance, difficulty in balancing real-time and accuracy, lack of aging adaptability and limited generalization ability. To address the above problems, we propose a battery parameter identification method and system. In short, when implementing the method, the battery is first stimulated based on a mixed pulse power characteristic test or a dynamic operating condition test, and the dynamic test data of the battery is collected. Then, a dual-polarization model is pre-constructed, and the pre-constructed dual-polarization model is optimized by identifying parameters. The dynamic test data is used as input, and the dual-polarization model is executed. The dual-polarization model identifies and analyzes the dynamic test data and outputs the parameter identification results. In an embodiment of the present invention, the dual-polarization model includes an ohmic internal resistance and two RC parallel networks that characterize electrochemical polarization and concentration polarization, respectively. It can more accurately reflect the voltage changes of the battery during the charging and discharging process, thereby more accurately describing the electrochemical process of the battery, reducing dependence on the traditional ECM model structure, and making up for its shortcomings in describing the nonlinear characteristics of the battery. It systematically solves the core problems of the existing technology models, such as insufficient nonlinear description, noise sensitivity, poor real-time performance, weak aging adaptability, and low generalization ability, and significantly improves the accuracy, robustness, and applicability of battery parameter identification throughout the entire life cycle, providing key technical support for high-safety and high-reliability battery management systems.

[0042] The embodiment of the present invention provides a battery parameter identification method. Figure 1 The figure shows a flow chart of the battery parameter identification method. The battery parameter identification method specifically includes:

[0043] S10, based on the hybrid pulse power characteristic test (HPPC) or dynamic operating condition test to stimulate the battery, collect dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data, and the dynamic operating condition test includes but is not limited to FUDS and DST tests; the hybrid pulse power characteristic test includes charge and discharge pulses at different SOC points and static stage tests to stimulate the dynamic response of the battery.

[0044] S20, pre-constructing a dual-polarization model, optimizing the pre-constructed dual-polarization model by identifying parameters, and outputting the optimized dual-polarization model;

[0045] It should be noted that the dual-polarization model includes the ohmic internal resistance R O And two RC parallel networks R1C1, R2C 2， The two RC parallel networks R1C1 and R2C2 represent electrochemical polarization and concentration polarization respectively;

[0046] The dual-polarization model equation is expressed as:

[0047] U terminal =OCV(SOC)-I*R0-U1-U2

[0048] The parameters to be identified are θ = [R0, R1, C1, R2, C2], OCV(·) represents the open circuit voltage, SOC is the state of charge, and I is the operating current of the battery.

[0049] In an embodiment of the present invention, the dual-polarization model, through the combination of a fast-slow dual-time-scale RC network, retains the computational efficiency of the equivalent circuit model while significantly improving the model's accuracy in depicting the nonlinear characteristics of the battery by describing the polarization process in stages. A battery management system based on the dual-polarization model can more accurately predict the battery's charge and discharge behavior, thereby optimizing charging strategies, discharge control, and thermal management strategies. This helps extend the battery's service life and improve the overall performance and safety of the battery system. Furthermore, the dual-polarization model can consider the dynamic response of the battery under different operating conditions, including the impact of factors such as temperature and current rate on battery performance. This enables the battery management system to more flexibly respond to various complex operating conditions and enhance the robustness and reliability of the battery system. Furthermore, the dual-polarization model can also capture changes in the battery's internal state, such as increases in internal resistance and capacity decay. This helps the battery management system promptly detect potential faults, provide early warnings, and address them, thereby avoiding sudden failures and safety incidents in the battery system.

[0050] It should be noted that when optimizing the pre-built dual-polarization model through identification parameters, the identification parameters are substituted into the dual-polarization model simulation, compared with actual data, and the root mean square error (RMS) evaluation accuracy is calculated. At the same time, the influence of temperature is considered, a correction relationship between the parameters and temperature is established, and the dual-polarization model aging parameters are regularly updated.

[0051] S30 , loading dynamic test data, taking the dynamic test data as input, executing a dual-polarization model, the dual-polarization model identifying and analyzing the dynamic test data, and outputting parameter identification results.

[0052] In this embodiment, the dual-polarization model identification and analysis method for dynamic test data includes:

[0053] S301, using HPPC test data, calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification;

[0054] S302 , acquiring dynamic test data, and updating the parameters to be identified of the dual-polarization model θ=[R0, R1, C1, R2, C2] in real time based on the extended Kalman filter or recursive least squares method combined with the SOC identification dynamic test data, thereby completing online parameter identification.

[0055] The offline identification process analyzes HPPC test data to determine the battery's parameter values under different operating conditions, providing rich initial information for the model. These offline parameter identification results based on dynamic test data enable the dual-polarization model to better capture the dynamic changes in the battery during the charging and discharging process, laying a good foundation for subsequent online identification and model application. At the same time, during the actual operation of the battery, dynamic test data is input into the dual-polarization model in real time. Based on algorithms such as extended Kalman filtering or recursive least squares, the model can quickly respond to changes in battery status and update the parameters to be identified in real time. This real-time parameter update mechanism enables the model to promptly reflect the dynamic characteristics of the battery under different operating conditions, enhances the model's ability to track and predict the battery's dynamic behavior, and better adapts to the battery's complex operating environment and changing operating conditions.

[0056] In an embodiment of the present invention, the dual-polarization model includes an ohmic internal resistance and two RC parallel networks that characterize electrochemical polarization and concentration polarization, respectively. It can more accurately reflect the voltage changes of the battery during the charging and discharging process, thereby more accurately describing the electrochemical process of the battery, reducing dependence on the traditional ECM model structure, and making up for its shortcomings in describing the nonlinear characteristics of the battery. It systematically solves the core problems of the existing technology models, such as insufficient nonlinear description, noise sensitivity, poor real-time performance, weak aging adaptability, and low generalization ability, and significantly improves the accuracy, robustness, and applicability of battery parameter identification throughout the entire life cycle, providing key technical support for high-safety and high-reliability battery management systems.

[0057] On the other hand, the present invention also provides a battery parameter identification system, such as Figure 2 As shown, the battery parameter identification system specifically includes:

[0058] The data acquisition module 100 excites the battery based on the hybrid pulse power characteristic test or the dynamic working condition test and collects dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data;

[0059] The model construction module 200 is used to pre-construct a dual-polarization model, optimize the pre-constructed dual-polarization model by identifying parameters, and output the optimized dual-polarization model;

[0060] The model building module 200 includes:

[0061] A model building unit 210, configured to pre-build a dual-polarization model;

[0062] The model optimization unit 220 optimizes the pre-built dual-polarization model by identifying parameters and outputs the optimized dual-polarization model.

[0063] The parameter identification module 300 is used to load dynamic test data, take the dynamic test data as input, execute the dual-polarization model, and the dual-polarization model identifies and analyzes the dynamic test data and outputs parameter identification results.

[0064] In this embodiment, the parameter identification module 300 includes:

[0065] The offline identification unit 310 uses the HPPC test data to calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump. O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification;

[0066] The online identification unit 320 is used to obtain dynamic test data, identify the dynamic test data based on the extended Kalman filter or recursive least squares method combined with the SOC, and update the dual-polarization model parameters to be identified θ = [R0, R1, C1, R2, C2] in real time to complete the online parameter identification.

[0067] In this embodiment of the present invention, the parameter identification module 300 comprises an offline identification unit 310 and an online identification unit 320. By combining these two units, the dual-polarization model can more accurately describe the battery's electrochemical processes and dynamic characteristics, providing the battery management system with more precise battery status information, such as remaining capacity and state of health. Based on this accurate information, the battery management system can more reliably predict the battery's charge and discharge behavior and performance changes, thereby formulating more optimized battery management strategies.

[0068] In summary, the present invention provides a battery parameter identification method and system. In an embodiment of the present invention, the dual-polarization model includes an ohmic internal resistance and two RC parallel networks that characterize electrochemical polarization and concentration polarization, respectively. It can more accurately reflect the voltage changes of the battery during the charging and discharging process, thereby more accurately describing the electrochemical process of the battery, reducing dependence on the traditional ECM model structure, and making up for its shortcomings in describing the nonlinear characteristics of the battery. It systematically solves the core problems of the existing technology model, such as insufficient nonlinear description, noise sensitivity, poor real-time performance, weak aging adaptability and low generalization ability, and significantly improves the accuracy, robustness and full life cycle applicability of battery parameter identification, providing key technical support for high-safety and high-reliability battery management systems.

[0069] In the embodiment of the present invention, the dual-polarization model, through the combination of a fast-slow dual-time-scale RC network, not only retains the computational efficiency of the equivalent circuit model, but also significantly improves the model's accuracy in depicting the nonlinear characteristics of the battery by describing the polarization process in stages. A battery management system based on the dual-polarization model can more accurately predict the battery's charge and discharge behavior, thereby optimizing charging strategies, discharge control, and thermal management strategies. This helps extend the battery's service life and improve the overall performance and safety of the battery system.

[0070] It should be noted that for the aforementioned embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0071] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the scope of protection of the invention. Obviously, the embodiments described are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in this field can still combine, add, delete or make other adjustments to the features in the various embodiments of the present invention according to the circumstances without conflict, without making creative work, so as to obtain different other technical solutions that do not deviate from the concept of the present invention in essence, and these technical solutions also fall within the scope of protection of the present invention.

Claims

1. A battery parameter identification method, characterized in that: The battery parameter identification method includes: S10, stimulating the battery based on a hybrid pulse power characteristic test or a dynamic operating condition test, and collecting dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data; S20, pre-constructing a dual-polarization model, optimizing the pre-constructed dual-polarization model by identifying parameters, and outputting the optimized dual-polarization model; S30 , loading dynamic test data, taking the dynamic test data as input, executing a dual-polarization model, the dual-polarization model identifying and analyzing the dynamic test data, and outputting parameter identification results.

2. The battery parameter identification method according to claim 1, wherein: The hybrid pulse power characteristics test includes charge and discharge pulses at different SOC points and a rest phase test to stimulate the dynamic response of the battery.

3. The battery parameter identification method according to claim 1, wherein: The dual polarization model includes an ohmic internal resistance R O And two RC parallel networks R1C1, R2C 2， The two RC parallel networks R1C1 and R2C2 represent electrochemical polarization and concentration polarization respectively; The dual-polarization model equation is expressed as: The terminal =OCV(SOC)-I*R0-U1-U2 Among them, the parameters to be identified are θ = [R0, R1, C1, R2, C2].

4. The battery parameter identification method according to claim 3, wherein: When optimizing the pre-built dual-polarization model through identification parameters, the identification parameters are substituted into the dual-polarization model simulation, compared with the actual data, and the root mean square error (RMS) evaluation accuracy is calculated. At the same time, considering the influence of temperature, a correction relationship between the parameters and temperature is established, and the dual-polarization model aging parameters are regularly updated.

5. The battery parameter identification method according to claim 4, wherein: The dual-polarization model identification and analysis method for dynamic test data includes: S301, using HPPC test data, calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification; S302 , acquiring dynamic test data, and updating the parameters to be identified of the dual-polarization model θ=[R0, R1, C1, R2, C2] in real time based on the extended Kalman filter or recursive least squares method combined with the SOC identification dynamic test data, thereby completing online parameter identification.

6. A battery parameter identification system for implementing the battery parameter identification method according to any one of claims 1 to 5, characterized in that: The system comprises: A data acquisition module, which excites the battery based on a hybrid pulse power characteristic test or a dynamic operating condition test and collects dynamic test data of the battery, wherein the dynamic test data includes current, voltage, temperature data, and HPPC test data; A model building module is used to pre-build a dual-polarization model, optimize the pre-built dual-polarization model by identifying parameters, and output the optimized dual-polarization model; The parameter identification module is used to load dynamic test data, take the dynamic test data as input, execute the dual-polarization model, and the dual-polarization model identifies and analyzes the dynamic test data and outputs parameter identification results.

7. The battery parameter identification method according to claim 6, wherein: The model building module includes: A model building unit, used for pre-building a dual-polarization model; The model optimization unit optimizes the pre-built dual-polarization model by identifying parameters and outputs the optimized dual-polarization model.

8. The battery parameter identification system according to claim 7, wherein: The parameter identification module includes: The offline identification unit uses HPPC test data to calculate the ohmic internal resistance R by the current and voltage difference at the moment of voltage jump. O Through the voltage recovery curve in the relaxation stage, the electrochemical polarization parameter R1C1 and the concentration polarization parameter R2C2 are fitted by exponential fitting or recursive least squares method to complete the parameter offline identification; The online identification unit is used to obtain dynamic test data, identify the dynamic test data based on the extended Kalman filter or the recursive least squares method combined with the SOC, and update the dual-polarization model parameters to be identified θ = [R0, R1, C1, R2, C2] in real time to complete the online parameter identification.

Images