All-vanadium redox flow battery full life cycle health state monitoring method based on capacity attenuation model

Abstract

The invention provides an all-vanadium redox flow battery full life cycle health state monitoring method based on a capacity fading model, which comprises the following steps: considering the influence of electrolyte flow velocity, bypass current, proton migration and a series of side reactions on battery capacity, and establishing an all-vanadium redox flow battery capacity fading model considering proton migration according to mass conservation and charge conservation; estimating the state of health of the battery through a capacity attenuation model of the all-vanadium redox flow battery; and correcting the estimated state of health of the battery by adopting Kalman filtering, and outputting the corrected state of health of the battery. According to the method, the capacity attenuation model is selected for modeling, factors such as flow velocity, bypass current and electrolyte volume change are considered, the battery capacity change relation is accurately reflected, the data high-dimensional calculation and set thought is combined, the battery health state is estimated through set Kalman filtering, and a new thought is provided for optimizing and monitoring the life cycle problem of the battery in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of power system energy storage, and in particular to a method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model. Background Art

[0002] Driven by the global wave of green energy technology innovation, the demand for efficient and environmentally friendly energy storage technology has become increasingly urgent. Against this background, the importance of energy storage technology continues to rise. It not only helps to reduce excessive dependence on fossil fuels and reduce environmental burden, but also optimizes energy distribution and improves the stability and reliability of power systems. Vanadium redox flow battery (VRB) has attracted widespread attention due to its excellent safety performance, flexible power and capacity design, excellent cycle life and environmental friendliness.

[0003] To ensure that all-vanadium liquid flow batteries can operate continuously and safely, the monitoring and management of battery health status is crucial. Due to the primary consideration of battery safety and service life, the current problems that need to be solved urgently include the lack of effective monitoring and early warning methods for battery field operation conditions, how to improve the safety and reliability of battery systems, and how to enable battery management systems to accurately reflect the status of the entire life cycle in complex operating environments. In order to accurately evaluate and predict the health status of all-vanadium liquid flow batteries, traditional monitoring methods may not meet this demand because they often lack accurate capture and effective early warning of the real-time operating status of the battery. Advanced capacity decay models are developed to describe the law of battery capacity changes.

[0004] Therefore, the present invention proposes a method for monitoring the health status of all-vanadium liquid flow batteries throughout their life cycle based on a capacity decay model. This method can not only effectively integrate the multi-dimensional operating data of the battery, but also accurately estimate and predict the health status of the battery through advanced filtering algorithms, thereby providing strong support for the full life cycle management of the battery. Summary of the invention

[0005] In view of the problems existing in the above-mentioned background technology, the present invention provides a method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model.

[0006] A method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model comprises the following steps:

[0007] Considering the effects of electrolyte flow rate, bypass current, proton migration and a series of side reactions on battery capacity, a capacity decay model of all-vanadium liquid flow battery considering proton migration is established based on the conservation of mass and charge.

[0008] Estimating battery health status through the capacity decay model of all-vanadium liquid flow battery;

[0009] The estimated battery health state is corrected using Kalman filtering, and the corrected battery health state is output.

[0010] Furthermore, the capacity decay model of the all-vanadium liquid flow battery considering proton migration includes a vanadium ion concentration dynamic sub-model and a proton concentration dynamic sub-model.

[0011] Furthermore, the vanadium ion concentration dynamic submodel includes the main reaction influence term, the diffusion reaction and side reaction influence term, and the electrolyte flow rate influence term, as follows:

[0012] (i) Main reaction influence term:

[0013] The expression of the main reaction influencing term is related to the charge / discharge current density, the electrode surface area, the number of transferred electrons, and the Faraday constant, and is expressed as:

[0014]

[0015] (ii) Diffusion reaction and side reaction effects:

[0016] The diffusion reaction and side reaction influence terms describe the diffusion of vanadium ions in the battery stack and the influence of possible side reactions on the vanadium ion concentration. The expression is:

[0017]

[0018]

[0019] (iii) Electrolyte flow rate influence:

[0020] The electrolyte flow rate influence term describes the effect of electrolyte flow rate on vanadium ion concentration, and the expression is:

[0021]

[0022] Combining the above items, the dynamic model of vanadium ion concentration in the battery stack is expressed as:

[0023]

[0024] Where, t is time; c i is the concentration of vanadium ions of various valences in the battery stack; k i is the diffusion coefficient of vanadium ions of various valences; V s represents the volume of the battery stack; d is the thickness of the proton exchange membrane; N is the number of single cells in the battery stack; S M is the area of ​​the proton exchange membrane; J represents the charge / discharge current density; S Erepresents the electrode surface area; z represents the number of electrons transferred in the redox reaction; F is the Faraday constant; Q n and Q p represent the flow rates of negative and positive electrolytes, respectively;

[0025] The dynamic model of vanadium ion concentration in the battery storage tank describes the dynamic changes of vanadium ion concentration in the storage tank under the action of flow rate and volume changes, specifically:

[0026]

[0027] Among them, V n and V p Represent the volumes of the negative and positive electrode storage tanks respectively.

[0028] Furthermore, the proton concentration dynamic sub-model includes side reaction influence terms, proton migration balance terms, bypass current influence terms, charge and discharge current influence terms, and electrolyte flow rate influence terms, as follows:

[0029] (i) Side effects

[0030]

[0031] (ii) Proton transfer equilibrium term

[0032]

[0033] (iii) Bypass current influence

[0034]

[0035] (iv) Charge and discharge current influence

[0036]

[0037] (v) Influence of electrolyte flow rate

[0038]

[0039] In summary, the dynamic model of proton concentration in the battery stack is expressed as:

[0040]

[0041] in, and represent the proton concentrations in the negative and positive electrode stacks, respectively. and represent the concentration of protons in the negative and positive electrode reservoirs, respectively;

[0042] The dynamic model of proton concentration in the liquid storage tank is expressed as:

[0043] positive electrode:

[0044]

[0045] negative electrode:

[0046]

[0047] Furthermore, the battery health status is estimated by using the all-vanadium liquid flow battery capacity decay model, specifically including:

[0048] The capacity attenuation model of the all-vanadium liquid flow battery is solved to obtain the reactant concentration in the electrolyte of the all-vanadium liquid flow battery, the battery capacity is calculated through the relationship between the battery capacity and the reactant concentration, and the battery health state SOH is estimated based on the calculated battery capacity.

[0049] Further, the reactant includes V 2+ and VO 2+ , the battery capacity is calculated based on the relationship between battery capacity and reactant concentration:

[0050]

[0051] Where C represents the capacity of the battery, Respectively represent +2 valence, +4 valence V in the battery stack 2+ The concentration of Respectively represent +2-valent and +4-valent VO in the storage tank 2+ The concentration, V s Indicates the volume of the battery stack, V n t and V p t Represent the volumes of the negative electrode and positive electrode storage tanks respectively; F is the Faraday constant; N is the number of single cells in the battery stack;

[0052] The battery health state SOH is estimated based on the calculated battery capacity, specifically:

[0053]

[0054] Among them, SOH is the SOH value of the current cycle, C is the battery capacity of the current cycle, C 0 It is the rated capacity of the battery in its new state.

[0055] Furthermore, the method of correcting the estimated battery health state by using Kalman filtering and outputting the corrected battery health state specifically includes:

[0056] The collective Kalman filter observer is designed. According to the orthogonality principle, the estimated battery health state SOH is used as the state vector of the system at time k-1 to directly predict the state of the system at time k. k-1 Perform projection decomposition to obtain the best approximation in the space spanned by past information, that is, to obtain the predicted particle set and the covariance prediction P kk-1 :

[0057]

[0058] In the formula, is the state prediction of each member at time k, is the state vector of the i-th member at time k-1, f(·) is the state transfer function of the system, u k-1 is the control input at time k-1; is the mean of the prediction set, P kk-1 is the covariance matrix estimate of the prediction error; and i = 1, 2, ... N;

[0059] In the update phase, the prediction results are corrected according to the observed values, and the state estimate and covariance matrix are updated;

[0060]

[0061] in, is the mean of the updated state set, calculated as:

[0062]

[0063] In formulas (20)-(23), y k is the observed value at time k, h(·) is the system observation function, K k is the Kalman gain; H k is the Jacobian matrix of the observation function, R k is the error matrix of observation noise; P k is the covariance matrix estimate of the prediction error; and i = 1, 2, ... N;

[0064] The prediction and update steps are repeated, and as more capacity data is input, a more accurate SOH estimate is iteratively obtained as the corrected battery health state.

[0065] Furthermore, it also includes: formulating the following maintenance strategies and optimization measures based on the corrected battery health status:

[0066] Preventive maintenance: When the SOH drops to a preset threshold, the warning mechanism is triggered to remind the user to perform battery maintenance, including balanced charging and replacement of electrolyte, to prevent further deterioration of battery performance;

[0067] Dynamically adjust the charge and discharge strategy: According to the changes in SOH, timely adjust the battery's charge and discharge depth, rate and temperature management strategy to slow down battery aging and extend its service life;

[0068] Performance optimization: By regularly analyzing the SOH change trend, potential problems in the battery system can be identified, including electrolyte imbalance and active material loss, and corresponding optimization measures can be taken, including improving the battery management system algorithm and optimizing the battery pack structure design to improve the overall performance of the energy storage system.

[0069] The present invention has the following beneficial effects: the present invention selects a capacity attenuation model for modeling, considers factors such as flow rate, bypass current and electrolyte volume change, accurately reflects the relationship between battery capacity changes, and combines high-dimensional data calculation and set thinking to estimate the battery health state with set Kalman filtering, providing new ideas for real-time optimization and monitoring of battery life cycle issues. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 The present invention is a flowchart of a method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model.

[0071] Figure 2 This is a comparison chart of the capacity decay results of the traditional capacity decay model and the capacity decay model considering proton migration.

[0072] Figure 3 V in the battery stack and the liquid storage tank 2+ Difference plot of concentration distribution. DETAILED DESCRIPTION

[0073] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments, but the embodiments should not be construed as limiting the present invention.

[0074] The experimental parameter settings of the all-vanadium liquid flow battery in the specific example are shown in Table 1.

[0075] Table 1 Experimental parameter settings of all-vanadium liquid flow battery

[0076]

[0077]

[0078] like Figure 1 As shown, the embodiment of the present invention proposes a method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model, comprising the following steps:

[0079] Step 1: Model Design

[0080] The traditional capacity decay model does not take into account the proton migration phenomenon, which affects the volume of the electrolyte, causing an imbalance in the vanadium ion concentrations at the positive and negative electrodes, and ultimately leading to an imbalance in the battery capacity.

[0081] Due to the unique capacity recoverability of all-vanadium liquid flow batteries, the present invention considers the impact of electrolyte flow rate, bypass current, proton migration and a series of side reactions on battery capacity, and establishes a capacity decay model of all-vanadium liquid flow batteries considering proton migration based on mass conservation and charge conservation, and combines it with the ensemble Kalman filter technology to iteratively update the results and improve the accuracy of the model. This technical method can be more in line with the operating mechanism of all-vanadium liquid flow batteries in structure, and realize the estimation of capacity, monitoring of its capacity decay and monitoring of battery health status.

[0082] Step 2: Establish a capacity decay model for all-vanadium liquid flow batteries considering proton migration

[0083] The capacity decay model of the all-vanadium liquid flow battery considering proton migration constructed by the present invention comprises: a vanadium ion concentration dynamic sub-model and a proton concentration dynamic sub-model.

[0084] (1) The dynamic sub-model of vanadium ion concentration in the battery stack and the liquid storage tank consists of the following main parts:

[0085] (i) Main reaction influence term:

[0086] This part represents the effect of the main redox reaction in the battery stack on the vanadium ion concentration. Its expression is usually related to the charge / discharge current density, electrode surface area, number of transferred electrons and Faraday constant, and can be expressed as:

[0087]

[0088] (ii) Diffusion reaction and side reaction effects:

[0089] This section describes the diffusion of vanadium ions in the battery stack and the effect of possible side reactions on the vanadium ion concentration. The expression can be expressed as:

[0090]

[0091] (iii) Electrolyte flow rate influence:

[0092] This section describes the effect of electrolyte flow rate on vanadium ion concentration. Since the flow rates of the negative and positive electrodes may be different, they need to be considered separately. The expression can be expressed as:

[0093]

[0094] Combining the above items, the dynamic model of vanadium ion concentration in the complete battery stack can be expressed as:

[0095]

[0096] Where, t is time; c i is the concentration of vanadium ions of various valences in the battery stack; k i is the diffusion coefficient of vanadium ions of various valences; V s represents the volume of the battery stack; d is the thickness of the proton exchange membrane; N is the number of single cells in the battery stack; S M is the area of ​​the proton exchange membrane; J represents the charge / discharge current density; S E represents the electrode surface area; z represents the number of electrons transferred in the redox reaction; F is the Faraday constant; Q n and Q p represent the flow rates of negative and positive electrolytes, respectively;

[0097] The dynamic model of vanadium ion concentration in the battery tank describes the dynamic change of vanadium ion concentration in the tank under the action of flow rate and volume change. Its concentration change can be expressed as:

[0098]

[0099] Among them, V n and V p Respectively represent the volume of the negative electrode and positive electrode storage tanks, and other potential influencing factors such as temperature and pressure;

[0100] (2) The dynamic sub-model of proton concentration in the battery stack and the liquid storage tank consists of the following main parts:

[0101] Modeling the proton concentration in the all-vanadium liquid flow battery stack and the liquid storage tank can be described by the following aspects:

[0102] (iii) Side effect items

[0103]

[0104] (ii) Proton transfer equilibrium term

[0105]

[0106] (iii) Bypass current influence

[0107]

[0108] (iv) Charge and discharge current influence

[0109]

[0110] (v) Influence of electrolyte flow rate

[0111]

[0112] Therefore, combining the above items, the dynamic model of proton concentration in the complete battery stack and storage tank can be expressed as:

[0113]

[0114] in, and represent the proton concentrations in the negative and positive electrode stacks, respectively. and represent the concentration of protons in the negative and positive electrode reservoirs, respectively;

[0115] Fluid reservoir:

[0116] positive electrode:

[0117]

[0118] negative electrode:

[0119]

[0120] Step 3: Capacity imbalance characteristics analysis

[0121] After completing the establishment of the model and the setting of relevant parameters, in order to explore the battery capacity imbalance characteristics, the present invention simulates and analyzes the vanadium ion concentration imbalance characteristics in the battery stack. Figure 3 Shows the V in the stack and reservoir 2+ The difference in concentration distribution. From the figure, we can see that V 2+ The concentration change always lags behind the stack. This is because the main reaction occurs in the stack. The electrolyte first undergoes redox reaction in the stack and then flows into the storage tank. Therefore, the V 2+ The concentration change is delayed compared to the stack. This delay phenomenon is also called mass transfer delay and is caused by the electrolyte flow rate. At the end of charging, V 3+ There is a surplus, VO 2+ Almost completely reacted; at the end of discharge, VO 2 + There is a surplus, V 2+ This phenomenon shows the relationship between the imbalance of reactant concentration and battery capacity attenuation under long-term operation, V 2+ and VO 2+ The concentration of is still the factor that determines the theoretical capacity of the battery.

[0122] The present invention simulates the concentration of vanadium ions of various valences during the discharge process when SOC=50%, and approximately solves the V in the electrolyte of the all-vanadium liquid flow battery according to formula (5) (6): 2+ and VO2+ The concentration of the battery can be expressed by formula (16):

[0123]

[0124] In formula (16), C represents the capacity of the battery, Respectively represent +2 valence, +4 valence V in the battery stack 2+ The concentration of Respectively represent +2-valent and +4-valent VO in the storage tank 2+ The concentration, V s Indicates the volume of the battery stack, V n t and V p t Represents the volume of the negative electrode and positive electrode storage tanks respectively; F is the Faraday constant; N is the number of single cells in the battery stack. 2+ and VO 2+ The concentration is calculated and Figure 2 The conventional capacity decay model shown in the figure is compared with the capacity decay results of the model considering proton migration. It can be seen that the battery capacity decay rate is significantly slowed down before and after considering proton migration, and it will reach a steady state faster.

[0125] The model selected in the present invention models the battery stack and the liquid storage tank respectively, comprehensively considering the effects of electrolyte flow rate, proton migration, bypass current and a series of side reactions on the battery capacity, which is more in line with actual operating conditions.

[0126] Step 4: Design the Ensemble Kalman Filter Observer

[0127] The simulation capacity obtained in step 3 is defined as follows:

[0128]

[0129] Among them, SOH is the SOH value of the current cycle, C is the battery capacity of the current cycle, C 0 It is the rated capacity of the battery in its new state.

[0130] According to the orthogonality principle, the output result SOH is used as the state vector of the system at time k-1 to directly predict the state of the system at time k. k-1 Perform projection decomposition to obtain the best approximation in the space spanned by past information, that is, to obtain the predicted particle set and the covariance prediction P kk-1 .

[0131]

[0132] In formulas (18) and (19), is the state prediction of each member at time k, is the state vector of the i-th member at time k-1, f(·) is the state transfer function of the system, u k-1 is the control input at time k-1; is the mean of the prediction set, P kk-1 is the covariance matrix estimate of the prediction error; and i = 1, 2, …N.

[0133] In the update phase, the prediction results are corrected according to the observed values, and the state estimate and covariance matrix are updated.

[0134]

[0135]

[0136] in, is the mean of the updated state set, calculated as:

[0137]

[0138] In formulas (20)-(23), y k is the observed value at time k, h(·) is the system observation function, K k is the Kalman gain; H k is the Jacobian matrix of the observation function, R k is the error matrix of observation noise; P k is the covariance matrix estimate of the prediction error; and i = 1, 2, …N.

[0139] Step 5: Monitoring, evaluation and optimization

[0140] As the battery operates, the algorithm continues to receive new monitoring data and continuously updates the estimated SOH. This iterative correction process makes the SOH estimate closer to the actual health status of the battery, significantly improving the accuracy and reliability of the estimate.

[0141] Based on the revised SOH estimation results, the following maintenance strategies and optimization measures can be formulated:

[0142] 1. Preventive maintenance: When the SOH drops to a preset threshold, the early warning mechanism is triggered to remind the user to perform battery maintenance, such as balanced charging, electrolyte replacement, etc., to prevent further deterioration of battery performance.

[0143] 2. Dynamically adjust the charge and discharge strategy: According to the changes in SOH, timely adjust the battery's charge and discharge depth, rate and temperature management strategy to slow down battery aging and extend its service life.

[0144] 3. Performance optimization: By regularly analyzing the SOH change trend, potential problems in the battery system, such as electrolyte imbalance and active material loss, are identified, and corresponding optimization measures are taken, such as improving the battery management system algorithm and optimizing the battery pack structure design, to improve the overall performance of the energy storage system.

[0145] It will be easily understood by those skilled in the art that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model, characterized in that: The steps include: Considering the effects of electrolyte flow rate, bypass current, proton migration and a series of side reactions on battery capacity, a capacity decay model of all-vanadium liquid flow battery considering proton migration is established based on the conservation of mass and charge. Estimating battery health status through the capacity decay model of all-vanadium liquid flow battery; The estimated battery health state is corrected using Kalman filtering, and the corrected battery health state is output.

2. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 1, characterized in that: The capacity decay model of the all-vanadium liquid flow battery considering proton migration includes a vanadium ion concentration dynamic sub-model and a proton concentration dynamic sub-model.

3. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 2, characterized in that: The vanadium ion concentration dynamic submodel includes the main reaction influence term, the diffusion reaction and side reaction influence term, and the electrolyte flow rate influence term, which are as follows: (i) Main reaction influence term: The expression of the main reaction influencing term is related to the charge / discharge current density, the electrode surface area, the number of transferred electrons, and the Faraday constant, and is expressed as: (ii) Diffusion reaction and side reaction effects: The diffusion reaction and side reaction influence terms describe the diffusion of vanadium ions in the battery stack and the influence of possible side reactions on the vanadium ion concentration. The expression is: (iii) Electrolyte flow rate influence: The electrolyte flow rate influence term describes the effect of electrolyte flow rate on vanadium ion concentration, and the expression is: Combining the above items, the dynamic model of vanadium ion concentration in the battery stack is expressed as: Where, t is time; c i is the concentration of vanadium ions of various valences in the battery stack; k i is the diffusion coefficient of vanadium ions of various valences; V s represents the volume of the battery stack; d is the thickness of the proton exchange membrane; N is the number of single cells in the battery stack; S M is the area of ​​the proton exchange membrane; J represents the charge / discharge current density; S E represents the electrode surface area; z represents the number of electrons transferred in the redox reaction; F is the Faraday constant; Q n and Q p represent the flow rates of negative and positive electrolytes, respectively; The dynamic model of vanadium ion concentration in the battery storage tank describes the dynamic changes of vanadium ion concentration in the storage tank under the action of flow rate and volume changes, specifically: Among them, V n and V p Represent the volumes of the negative and positive electrode storage tanks respectively.

4. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 2, characterized in that: The proton concentration dynamic sub-model includes side reaction influence terms, proton migration balance terms, bypass current influence terms, charge and discharge current influence terms, and electrolyte flow rate influence terms, which are as follows: (i) Side effects (ii) Proton transfer equilibrium term (iii) Bypass current influence (iv) Charge and discharge current influence (v) Influence of electrolyte flow rate In summary, the dynamic model of proton concentration in the battery stack is expressed as: in, and represent the proton concentrations in the negative and positive electrode stacks, respectively. and represent the concentration of protons in the negative and positive electrode reservoirs, respectively; The dynamic model of proton concentration in the liquid storage tank is expressed as: positive electrode: negative electrode:

5. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 1, characterized in that: The method of estimating the battery health state by using the capacity decay model of the all-vanadium liquid flow battery specifically includes: The capacity attenuation model of the all-vanadium liquid flow battery is solved to obtain the reactant concentration in the electrolyte of the all-vanadium liquid flow battery, the battery capacity is calculated through the relationship between the battery capacity and the reactant concentration, and the battery health state SOH is estimated based on the calculated battery capacity.

6. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 5, characterized in that: The reactants include V 2+ and VO 2+ , the battery capacity is calculated based on the relationship between battery capacity and reactant concentration: Where C represents the capacity of the battery, Respectively represent +2 valence, +4 valence V in the battery stack 2+ The concentration of Respectively represent +2-valent and +4-valent VO in the storage tank 2+ The concentration, V s Indicates the volume of the battery stack, V n t and V p t Represent the volumes of the negative electrode and positive electrode storage tanks respectively; F is the Faraday constant; N is the number of single cells in the battery stack; The battery health state SOH is estimated based on the calculated battery capacity, specifically: Among them, SOH is the SOH value of the current cycle, C is the battery capacity of the current cycle, and C0 is the rated capacity of the battery in a brand new state.

7. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 6, characterized in that: The method of correcting the estimated battery health state by using Kalman filtering and outputting the corrected battery health state specifically includes: The collective Kalman filter observer is designed. According to the orthogonality principle, the estimated battery health state SOH is used as the state vector of the system at time k-1 to directly predict the state of the system at time k. k-1 Perform projection decomposition to obtain the best approximation in the space spanned by past information, that is, to obtain the predicted particle set and the covariance prediction P kk-1 : In the formula, is the state prediction of each member at time k, is the state vector of the i-th member at time k-1, f(·) is the state transfer function of the system, u k-1 is the control input at time k-1; is the mean of the prediction set, P k|k-1 is the covariance matrix estimate of the prediction error; and i = 1, 2, ... N; In the update phase, the prediction results are corrected according to the observed values, and the state estimate and covariance matrix are updated; in, is the mean of the updated state set, calculated as: In formulas (20)-(23), y k is the observed value at time k, h(·) is the system observation function, K k is the Kalman gain; H k is the Jacobian matrix of the observation function, R k is the error matrix of observation noise; P k is the covariance matrix estimate of the prediction error; and i = 1, 2, ... N; The prediction and update steps are repeated, and as more capacity data is input, a more accurate SOH estimate is iteratively obtained as the corrected battery health state.

8. The method for monitoring the health status of an all-vanadium liquid flow battery throughout its life cycle based on a capacity decay model according to claim 1, characterized in that: Also includes: Based on the corrected battery health status, formulate the following maintenance strategies and optimization measures: Preventive maintenance: When the SOH drops to a preset threshold, the warning mechanism is triggered to remind the user to perform battery maintenance, including balanced charging and replacement of electrolyte, to prevent further deterioration of battery performance; Dynamically adjust the charge and discharge strategy: According to the changes in SOH, timely adjust the battery's charge and discharge depth, rate and temperature management strategy to slow down battery aging and extend its service life; Performance optimization: By regularly analyzing the SOH change trend, potential problems in the battery system can be identified, including electrolyte imbalance and active material loss, and corresponding optimization measures can be taken, including improving the battery management system algorithm and optimizing the battery pack structure design to improve the overall performance of the energy storage system.

Images