Test method and test system of all-vanadium redox flow battery

Abstract

The invention relates to the field of electrochemical energy storage, and discloses a testing method and a testing system of an all-vanadium redox flow battery, and the method comprises the following steps: describing diffusion, convection and electrochemical reaction processes of electrolyte in a battery flow channel through a partial differential equation, and establishing a dynamic distribution model of concentration gradient and current density; based on the model, an adaptive control algorithm is adopted to dynamically adjust the electrolyte flow velocity and the current density, and stable control of a test environment is realized; acquiring electrochemical impedance spectrum data by applying alternating-current voltage signals with different frequencies, extracting non-linear characteristics in an impedance spectrum by using a long-short-term memory network model, and analyzing and predicting the performance degradation trend of the battery; a multivariable optimization model is established, and parameters such as temperature, flow velocity and current density are dynamically adjusted, so that a test environment is always close to a target value. According to the method, the test precision and the environmental stability can be effectively improved, and reliable technical support is provided for performance evaluation and optimization of the all-vanadium redox flow battery.

Description

Technical Field

[0001] The invention relates to the technical field of electrochemical energy storage, and in particular to a testing method and a testing system for an all-vanadium liquid flow battery. Background Art

[0002] With the development of energy storage technology, all-vanadium liquid flow batteries have gradually become an important choice for large-scale energy storage systems due to their long life, high safety and good cycle stability. However, comprehensive and accurate testing of their performance still faces many challenges.

[0003] Existing all-vanadium liquid flow battery testing methods usually use basic electrochemical testing methods, such as constant current charge and discharge tests and electrochemical impedance spectroscopy (EIS) analysis, to evaluate the electrochemical performance of the battery. However, these testing methods have many limitations. Generally, these testing methods cannot describe in detail the dynamic changes of the concentration gradient and current density inside the battery. The flow rate and current density in the test environment often change due to fluctuations in external conditions, which affects the reliability and stability of the test data. In addition, although the electrochemical impedance spectroscopy can reflect the internal characteristics of the battery, the existing methods mostly use traditional analysis methods to process impedance data, which makes it difficult to fully extract complex nonlinear characteristics, thereby affecting the accurate prediction of battery performance degradation. At the same time, the control accuracy of parameters such as temperature, flow rate and current density during the test is not high, and it is impossible to achieve dynamic adjustment and optimization of the test conditions, resulting in large errors in the test results and low test efficiency.

[0004] Therefore, existing technologies are difficult to meet the requirements of all-vanadium liquid flow battery testing for data accuracy, environmental stability and result consistency, which affects the practical application of test data in performance evaluation and battery optimization. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a testing method and a testing system for an all-vanadium liquid flow battery, which realize high-precision parameter control, real-time data analysis and dynamic environment optimization during the testing process of the all-vanadium liquid flow battery to obtain stable and reliable performance evaluation data.

[0006] To achieve the above objectives, the present invention is implemented by the following technical scheme: A method for testing an all-vanadium liquid flow battery comprises the following steps:

[0007] Establish a distribution model of electrolyte concentration and current density, and describe the dynamic changes of concentration gradient and current density inside the battery through partial differential equations;

[0008] Based on this model, the flow rate and current density of the electrolyte are adjusted in real time through a control algorithm to maintain stability under the target test conditions;

[0009] Apply an AC voltage signal to obtain electrochemical impedance spectroscopy data at multiple frequencies;

[0010] Use machine learning algorithms to analyze electrochemical impedance spectroscopy data, extract the nonlinear characteristics of the battery, and determine and predict the performance degradation trend of the battery;

[0011] A multivariable optimization model is established to dynamically adjust the test parameters to optimize the temperature, flow rate and current density variables so that the test environment is always close to the target value.

[0012] Preferably, the step of establishing a distribution model of electrolyte concentration and current density comprises:

[0013] The concentration gradient and current density distribution model of the electrolyte in the flow channel of the all-vanadium liquid flow battery is constructed by partial differential equations. The electrolyte concentration is set as C(x,t) as a function of position x and time t, and its diffusion coefficient is defined as D, flow velocity as u, current density as I, electron transfer number as n, and Faraday constant as F.

[0014] The concentration distribution of the electrolyte is modeled by the diffusion-convection equation, and the dynamic change of the concentration satisfies the following partial differential equation:

[0015]

[0016] Among them, the diffusion term of electrolyte represents the diffusion process of concentration along the flow channel, and the convection term represents the convection flow of electrolyte, and the electrochemical reaction term Indicates the rate of change of electrolyte concentration during battery charging and discharging.

[0017] Preferably, the partial differential equation model is discretized using a finite difference method or a finite element method, and the partial differential equation is discretized with a spatial step and a time step to obtain a numerical solution of the electrolyte concentration and the current density at each discrete node, and during the test process, iterative calculations are performed to obtain real-time concentration gradients and current density distributions.

[0018] Preferably, the control algorithm collects electrolyte concentration and current density data in real time, compares the data with preset target concentration gradient and current density, and automatically adjusts the flow rate and current density of the electrolyte according to the deviation.

[0019] Preferably, the control algorithm sets an adaptive adjustment coefficient based on the target concentration and current density, and uses a closed-loop feedback control method to update the set values ​​of the flow rate and current density in real time.

[0020] Preferably, the collection of electrochemical impedance spectroscopy data includes applying a small amplitude AC voltage signal at different frequencies, and forming impedance spectrum data by measuring the real and imaginary impedance parts of the battery at each frequency point. The impedance spectrum is used to reflect the charge transfer and electrolyte diffusion characteristics of the battery, and further extract the internal state information of the battery.

[0021] Preferably, the machine learning algorithm adopts a long short-term memory network model or a variational autoencoder model, takes multi-frequency impedance spectrum data as time series input, automatically extracts the nonlinear degradation characteristics of the battery through model training, and predicts the battery performance attenuation trend.

[0022] Preferably, the multivariable optimization model constructs constraints on test parameters to minimize the error between actual test performance and target performance, and, on the premise of satisfying the constraints, implements dynamic optimization and adjustment of the test environment to obtain test environment stability close to the target.

[0023] Preferably, the multivariable optimization model is solved by the Lagrange multiplier method, the upper and lower limits of each test parameter are set as constraints, and the optimization model is updated in real time in combination with the data collected during the test, so that the system automatically adjusts the temperature, flow rate and current density according to the real-time response of the battery.

[0024] The present invention also provides a testing system for an all-vanadium liquid flow battery, comprising:

[0025] An electrolyte storage and circulation device for achieving electrolyte flow control;

[0026] A current density regulating device, used for dynamically adjusting the current density of the battery;

[0027] Temperature control system, used to adjust the test environment temperature of the battery;

[0028] Data acquisition module, used to collect electrochemical impedance spectroscopy data and real-time data during the test;

[0029] The control processor is connected to the data acquisition module and runs the control algorithm. It automatically adjusts the flow rate, current density, and temperature test parameters by analyzing the data, realizes real-time control of the internal concentration and current density of the battery, and adjusts the test conditions based on the optimization model.

[0030] The present invention provides a testing method and testing system for an all-vanadium liquid flow battery, which has the following beneficial effects:

[0031] 1. The present invention can describe the material migration process inside the battery in detail by establishing a partial differential equation distribution model of electrolyte concentration and current density. The model accurately captures the changes in concentration gradient and current density, provides a basis for the accuracy of data collection during the test, and makes the test data more representative and credible.

[0032] 2. The present invention uses an adaptive control algorithm to adjust the electrolyte flow rate and current density in real time, ensuring that the predetermined concentration gradient and current density can always be maintained under different test conditions. This dynamic adjustment mechanism effectively responds to fluctuations in the external environment, allowing the test environment to remain stable under the target state, thereby reducing the impact of data deviation and external interference.

[0033] 3. Through the collection of electrochemical impedance spectroscopy data and the analysis of machine learning models, the present invention can identify the internal characteristics of the battery under multi-frequency conditions, including charge transfer and diffusion characteristics. This in-depth analysis helps to reveal the root cause of battery degradation, thereby providing a more accurate basis for battery life prediction and health management.

[0034] 4. The present invention introduces a long short-term memory network model to extract time series features and predict degradation trends of the collected impedance spectrum data, which can identify the degradation mode and abnormal degradation phenomenon of the battery in advance. Through real-time monitoring and prediction of the battery status, early warning can be achieved, helping to take maintenance or adjustment measures in time, thereby extending the actual service life of the battery.

[0035] 5. Through the dynamic adjustment of the multivariable optimization model, the present invention realizes automatic optimization of key parameters such as temperature, flow rate and current density, so that the test process always remains close to the target value. This dynamic adjustment mechanism reduces the need for manual adjustment of test parameters, improves test efficiency and consistency of test data, and provides a stable and efficient operating environment for the test process. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0037] Figure 2 It is a schematic diagram of the system structure of the present invention.

[0038] Among them, 100, electrolyte storage and circulation device; 200, current density adjustment device; 300, temperature control system; 400, data acquisition module; 500, control processor. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] Please see attached Figure 1 The present invention provides a method for testing an all-vanadium redox flow battery, aiming to improve the accuracy, data analysis depth and testing efficiency of the all-vanadium redox flow battery testing process. The following is a detailed description of each step of the present invention.

[0041] S1. Establish the distribution model of electrolyte concentration and current density

[0042] During the test of the all-vanadium liquid flow battery, the distribution of electrolyte concentration and current density plays an important role in accurately reflecting the internal state of the battery. By accurately modeling the concentration gradient and current density of the electrolyte, the material migration and reaction process inside the battery can be better described, thereby providing reliable data support for subsequent real-time control, data acquisition and performance analysis. Based on this, the present invention first establishes a distribution model of electrolyte concentration and current density, and uses partial differential equations to describe the diffusion, convection and electrochemical reaction rate of the electrolyte, thereby obtaining the dynamic changes of the concentration gradient and current density inside the battery.

[0043] In this embodiment, the concentration of the electrolyte is set to C(x, t), that is, its concentration changes dynamically with the spatial position x and time t, the current density is set to J(x, t), the diffusion coefficient is set to D, the flow rate of the electrolyte is u, the number of electron transfers is n, and the Faraday constant is F. In general, the concentration distribution inside the battery is affected by diffusion, convection and electrochemical reaction rate, so the concentration distribution can be modeled by the following diffusion-convection reaction equation:

[0044]

[0045] Among them, the first represents the diffusion term of the electrolyte, which is used to describe the diffusion process of the electrolyte along the concentration gradient of position x in the flow channel over time; the second term is the convection term, which represents the convection behavior of the electrolyte in the flow channel with the flow velocity u, reflecting the change of concentration with the flow of the fluid; the third term Represents the electrochemical reaction rate term, which indicates the rate at which the electrolyte concentration changes with the current density I due to the redox reaction occurring inside the battery.

[0046] In some embodiments, specifically, the partial differential equation model can more intuitively characterize the dynamic distribution of the concentration gradient and current density of the electrolyte in the battery flow channel. Since the effects of diffusion and convection at different locations and flow rates are different, the partial differential equation of this embodiment can provide a more accurate description of concentration changes, which is helpful for analyzing the impact of different test conditions on the migration of substances inside the battery.

[0047] In order to realize the application of the model in the actual calculation process, the present invention discretizes the partial differential equation so as to calculate using numerical methods. In this embodiment, the finite difference method is used to discretize time and space, and the space step is set to Δx and the time step is set to Δt. As an option, the concentration gradient at position x can be discretized by the finite difference method, and the following difference equation is obtained:

[0048]

[0049] In the above formula, represents the electrolyte concentration at position i and time t. Through this difference equation, the concentration change of the electrolyte can be calculated at each discretized spatial point, so that the concentration value of each point is updated at the time step Δt.

[0050] In general, through the above discretization method, the dynamic distribution of electrolyte concentration and current density can be obtained at multiple discrete locations in the battery flow channel, and continuously updated over time to describe the dynamic change process of electrolyte concentration in the entire flow channel.

[0051] In some embodiments, in order to ensure the stability and accuracy of the model calculation, the present embodiment further sets the initial boundary conditions and the initial values ​​of the flow rate and current density. As an option, the initial concentration can be set to a reasonable starting value according to the electrolyte type and the battery charge and discharge state; the boundary conditions include the concentration boundary values ​​at the flow channel inlet and outlet to ensure that the concentration of the model at different spatial positions has a consistent boundary reference.

[0052] S2, model-based control algorithm to achieve real-time regulation

[0053] During the test of the all-vanadium liquid flow battery, real-time adjustment of the electrolyte flow rate and current density to maintain stability under the target test conditions is a key link. Based on the electrolyte concentration and current density distribution model established in the aforementioned step S1, step S2 further uses a control algorithm to achieve dynamic adjustment of the test conditions to cope with test deviations caused by changes in the external environment or internal fluctuations in the system. In general, in order to ensure the accuracy and consistency of the test results, the system needs to collect electrolyte concentration and current density data in real time, and perform feedback control based on the deviation value on this basis. Therefore, real-time dynamic adjustment can be achieved through an adaptive control algorithm so that the test process can continue to stabilize at the predetermined target state.

[0054] In this embodiment, the electrolyte flow rate u and current density I are dynamically adjusted by an adaptive control algorithm, aiming to make the concentration gradient and current density of the system close to the set target value to ensure the stability of the test conditions. Specifically, the system first compares the real-time collected electrolyte concentration and current density data with the set target value C target and J target Compare and calculate the deviation value. In general, the system determines the adjustment amount according to the deviation value, so the real-time adjustment of flow rate and current density can be achieved through the following feedback control formula:

[0055] u(t+1)=u(t)+ku(C target -C(x,t)

[0056] I(t+1)=I(t)+k I (J target -J(t)

[0057] Among them, ku and k I It is an adaptive adjustment coefficient, which is used to control the response speed of the adjustment to adapt to different system states and test requirements. In general, these adjustment coefficients can be obtained through experimental calibration to ensure that the adjustment process can quickly respond to deviation changes and will not cause system oscillation or data instability due to excessive adjustment. Therefore, the control algorithm has strong stability and adaptability, and is suitable for the dynamic test process of all-vanadium liquid flow batteries.

[0058] In some embodiments, specifically, when the concentration gradient deviation collected by the system is large, the control algorithm will increase the adjustment range of the flow rate accordingly, so that the concentration gradient quickly approaches the target value to reduce the impact of external environmental changes on the test conditions; when the concentration gradient deviation is small, the control algorithm will appropriately reduce the adjustment range to ensure the stability of the system adjustment process. As an option, the current density is adjusted by detecting the deviation between the real-time current value and the target current density, and adjusting the set value of the current density according to the deviation to keep the test conditions stable.

[0059] In some embodiments, in order to improve the flexibility of adaptive control, this embodiment further introduces a dynamic adjustment mechanism, that is, the system dynamically modifies the adjustment coefficients ku and k according to the concentration gradient and current density changes under different working conditions during the test process. I , thus ensuring the adjustment accuracy under different test conditions. For example, when the ambient temperature or electrolyte characteristics change, the system automatically recalibrates the adjustment coefficient to adapt to the new test requirements and ensure the control effect.

[0060] Through the above-mentioned adaptive control algorithm, the system can effectively adjust the electrolyte flow rate and current density so that the concentration gradient and current density inside the battery remain in a stable state close to the target.

[0061] S3. Apply AC voltage signal and obtain electrochemical impedance spectroscopy data

[0062] During the performance test of the all-vanadium liquid flow battery, applying an AC voltage signal and collecting electrochemical impedance spectroscopy data at multiple frequencies can provide key information about the internal state of the battery. In general, electrochemical impedance spectroscopy (EIS) data can characterize processes such as charge transfer and electrolyte diffusion within the battery, thereby facilitating in-depth analysis of the dynamic characteristics of the battery and its reaction mechanism. Therefore, in step S3, by applying AC voltage signals of different frequencies and obtaining electrochemical impedance spectroscopy data, high-precision raw data can be provided for subsequent machine learning analysis and performance degradation prediction.

[0063] In this embodiment, a small-amplitude sinusoidal voltage source is used in the process of applying the AC voltage signal to ensure that the working state of the battery does not change significantly during the test. Specifically, the frequency range of the AC voltage signal covers low frequency to high frequency bands to capture the characteristic responses of various electrochemical processes inside the battery. In some embodiments, the setting of the signal frequency may include a frequency range from 0.01 Hz to 10 kHz, so that the test data can cover the dynamic response of charge transfer, double layer effect and diffusion process.

[0064] While applying the AC voltage signal, the system collects impedance data at multiple frequencies. In general, the impedance value is expressed in the complex form of Z(ω)=Z′(ω)+jZ″(ω), where Z′(ω) is the real impedance, representing the charge transfer impedance of the battery; Z″(ω) is the imaginary impedance, reflecting the characteristics of the diffusion process in the battery. Specifically, the impedance values ​​at different frequencies can reveal different electrochemical processes: in the low-frequency range, the impedance value mainly reflects the diffusion characteristics of the electrolyte; in the medium-frequency range, it reflects the charge transfer impedance of the battery; and in the high-frequency range, it usually manifests as the capacitance effect of the double layer.

[0065] In this embodiment, in order to fully collect the impedance spectrum data, the system measures the impedance response of the battery at each preset frequency point and records the corresponding real and imaginary data. As an option, the data in the low-frequency and high-frequency bands can be weighted sampled to improve the sampling accuracy of the diffusion process and the double-layer effect. In some embodiments, the signal amplitude during the data acquisition process can be appropriately adjusted to avoid battery polarization caused by excessive signal strength, while ensuring that the impedance response has good linear characteristics.

[0066] The collected impedance spectrum data will form the Nyquist diagram of the battery, that is, an impedance image with Z′(ω) as the horizontal axis and Z″(ω) as the vertical axis. In general, different areas in the Nyquist diagram correspond to different electrochemical processes of the battery, for example, the high-frequency area corresponds to charge transfer, and the low-frequency area corresponds to the diffusion process. Therefore, this image can intuitively display the dynamic electrochemical characteristics of the battery and provide a reference for subsequent analysis.

[0067] The impedance spectrum data collected in some embodiments will be further used for equivalent circuit fitting analysis to construct an equivalent circuit model of the battery. Specifically, the system can equate the reaction process inside the battery to a combination of a group of resistors, capacitors and inductors based on the impedance characteristics at different frequencies, and obtain the parameter values ​​of each component through a fitting algorithm, thereby more accurately describing the internal electrochemical process of the battery.

[0068] S4. Performance degradation analysis and prediction based on machine learning

[0069] During the performance test of the all-vanadium liquid flow battery, analyzing the electrochemical impedance spectroscopy (EIS) data can reflect the state changes inside the battery, and then judge and predict the performance degradation trend of the battery. In general, the nonlinear characteristics of the battery internal charge transfer, electrolyte diffusion, etc. contained in the EIS data can be extracted and modeled by machine learning algorithms. Therefore, in step S4, the present invention analyzes the electrochemical impedance spectroscopy data by a machine learning algorithm to achieve state evaluation of battery performance and prediction of degradation trends, thereby improving the reliability and depth of the test results.

[0070] In this embodiment, the long short-term memory network (LSTM) model is used to process and analyze the electrochemical impedance spectroscopy data to fully capture the nonlinear degradation characteristics of the battery. Specifically, the LSTM model is suitable for processing time series data and can extract long-term and short-term correlation features from the time series. In general, in order to enable the model to learn the nonlinear change law of EIS data, the impedance spectrum data at multiple frequencies is first serialized into time series input, including the real part Z′(ω) and imaginary part Z″(ω) data of the impedance corresponding to each frequency.

[0071] In some embodiments, the input sequence of the LSTM model consists of multiple sets of impedance data of different frequencies, in the form of {Z(ω1), Z(ω2), …, Z(ω n )}, where each set of data includes the real and imaginary parts of the impedance. By formatting the input data into time steps, the model is able to gradually extract the dynamic characteristics of the EIS data. As an option, the data of different frequency bands can be processed in segments to highlight the characteristic changes in the key frequency interval.

[0072] Specifically, the core of the LSTM model consists of gated units, each of which has a forget gate, an input gate, and an output gate. Through these gating mechanisms, the model can filter and memorize the input impedance data, so that important feature information can be gradually retained in multi-layer loops. The state update formula of the LSTM unit is:

[0073] Forget Gate t =σ(W f ·[h t-1 ,x t ]+b f )

[0074] Input Gate i t =σ(W i ·[h t-1 ,x t ]+b i )

[0075] Candidate memory cells

[0076] Memory status update

[0077] Output gate o t =σ(W o ·[h t-1 ,x t ]+b o )

[0078] Hidden state update ht=o t *tanh(C t )

[0079] Among them, x t is the impedance data currently input, h t-1 is the hidden state of the previous time step, f t 、i t and t They are forget gate, input gate and output gate respectively, and the parameter W f , W i , W C , W o is the corresponding weight matrix, b f、b i 、b C 、b o is the bias vector.

[0080] Through the above formula, the LSTM model processes the input impedance spectrum data layer by layer to generate a prediction output. In general, the prediction output of the model is the performance degradation index of the battery to quantify the change in the degradation trend. Therefore, the model can predict the performance change trend of the battery based on the full learning of the impedance spectrum data and provide intuitive degradation information.

[0081] In some embodiments, the training process of the LSTM model uses historical EIS data, optimizes the model parameters based on actual degradation data, and uses the mean square error loss function as the optimization target, that is:

[0082]

[0083] Among them, y i is the true degradation value, is the model prediction value, and N is the number of samples. Through iterative optimization of the training process, the model has good prediction accuracy and can identify key features from the input data.

[0084] In some embodiments, the model not only predicts degradation, but also combines the characteristic change trend of impedance at multiple frequencies to determine abnormal degradation that may occur in the battery, such as nonlinear impedance changes caused by drastic changes in the external environment. By comparing with the standard impedance curve, the model can automatically identify and mark abnormal situations, making it easier to take timely measures to optimize the battery's performance.

[0085] S5. Establish a multivariable optimization model and dynamically adjust test parameters

[0086] During the test of the all-vanadium liquid flow battery, test parameters such as temperature, flow rate and current density directly affect the stability of the test environment and the accuracy of the data. In general, in order to ensure that the test process can be carried out in a state close to the target value and avoid test data errors caused by environmental condition fluctuations or external disturbances, the present invention establishes a multivariable optimization model in step S5 to achieve dynamic optimization and adjustment of temperature, flow rate and current density, thereby ensuring the stability of the test environment. Therefore, by adjusting the test parameters in real time through the optimization model, the accuracy and consistency of the test process can be significantly improved.

[0087] In this embodiment, a multivariable optimization model including temperature, flow rate and current density variables is first constructed to minimize the performance error during the test. Specifically, the goal of the optimization model is to make the actual test performance close to the target performance, which is defined as the minimization problem of the target error E(x), where the optimization variables x = [T, u, I] represent temperature, flow rate and current density respectively. In general, the target error E(x)E(x)E(x) can be expressed as the actual test performance parameter y i and the target performance parameter y target The sum of squares of :

[0088]

[0089] Where m is the number of test performance parameters. Therefore, by minimizing the target error E(x), the actual test parameters can be made close to the set target values ​​while satisfying the test conditions.

[0090] In some embodiments, temperature, flow rate and current density constraints are further introduced into the optimization model to ensure that the system is tested within a safe and stable range. Specifically, the constraints are set as the upper limit of temperature, the lower limit of flow rate and the range boundary value of current density to ensure that the parameter changes during the optimization process do not exceed the set physical limitations. As an option, the upper limit of temperature can be determined based on the tolerance temperature of the electrolyte, while the lower limit of flow rate is set according to the minimum flow requirement of the electrolyte, and the range of current density refers to the design current density of the battery to ensure the controllability of the electrochemical reaction.

[0091] In order to solve the multivariable optimization problem, this embodiment uses the Lagrange multiplier method to solve the optimization model. By introducing the Lagrange multiplier λ, the Lagrange function is constructed. Its form is:

[0092]

[0093] Among them, g j (x) represents the constraint function of the test condition, and λ is the Lagrange multiplier. In general, the solution process needs to satisfy the zero gradient condition of the Lagrange equation, so the update process of the optimization variable can be obtained by the following two conditions:

[0094]

[0095] As an option, the system iteratively calculates and gradually adjusts the values ​​of the optimization variable x and the Lagrange multiplier λ to obtain the optimal test parameters within the constraints. Specifically, each step in the optimization process updates the temperature, flow rate, and current density based on real-time data to ensure that the test conditions are always stable near the target value. Generally speaking, this iterative solution method can obtain the optimal solution in a shorter time, thus meeting the needs of real-time adjustment during the test process.

[0096] In some embodiments, in order to improve the response speed and stability of the optimization model, this embodiment further introduces a dynamic adjustment mechanism. Specifically, the system dynamically adjusts the Lagrange multiplier and step size according to the error change rate during the actual test process to accelerate the convergence speed of the optimization model. As an extension, when the test data suddenly changes, the system automatically increases the iteration step size so that the optimization process can quickly respond to data changes, thereby ensuring the stability and consistency of the test environment.

[0097] Through the dynamic adjustment of the above-mentioned optimization model, the present invention can maintain the temperature, flow rate and current density of the test environment in an optimal state, significantly reduce the impact of external environmental fluctuations on the test data, and provide stable and reliable original data for subsequent data analysis.

[0098] The present invention realizes dynamic regulation and in-depth analysis of the test process by establishing a distribution model of concentration and current density, a real-time control algorithm, electrochemical impedance spectrum analysis, machine learning prediction and a multivariable optimization model. The method includes constructing the electrolyte concentration distribution using partial differential equations, adjusting the flow rate and current density through adaptive control, collecting multi-frequency impedance spectrum data and applying machine learning to analyze degradation trends, and then dynamically adjusting the temperature, flow rate and current density through multivariable optimization. This method can effectively improve the accuracy of the test process, the stability of the environment and the reliability of data analysis, and provides accurate and efficient technical support for the performance evaluation and optimization of all-vanadium liquid flow batteries.

[0099] The testing system of the all-vanadium redox flow battery described below and the testing method of the all-vanadium redox flow battery described above can be referred to each other.

[0100] Please see attached Figure 2 The present invention also provides a testing system for an all-vanadium liquid flow battery, comprising:

[0101] An electrolyte storage and circulation device 100 for achieving electrolyte flow control;

[0102] A current density adjustment device 200, used to dynamically adjust the current density of the battery;

[0103] A temperature control system 300 is used to adjust the test environment temperature of the battery;

[0104] The data acquisition module 400 is used to collect electrochemical impedance spectroscopy data and real-time data during the test process;

[0105] The control processor 500 is connected to the data acquisition module and runs the control algorithm to automatically adjust the flow rate, current density, and temperature test parameters by analyzing the data, thereby achieving real-time control of the internal concentration and current density of the battery, as well as adjustment of the test conditions based on the optimization model.

[0106] The system of this embodiment can be used to execute the above method embodiments, and its principles and technical effects are similar, which will not be repeated here.

[0107] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for testing an all-vanadium redox flow battery, characterized in that: The following steps are involved: Establish a distribution model of electrolyte concentration and current density, and describe the dynamic changes of concentration gradient and current density inside the battery through partial differential equations; Based on this model, the flow rate and current density of the electrolyte are adjusted in real time through a control algorithm to maintain stability under the target test conditions; Apply an AC voltage signal to obtain electrochemical impedance spectroscopy data at multiple frequencies; Use machine learning algorithms to analyze electrochemical impedance spectroscopy data, extract the nonlinear characteristics of the battery, and determine and predict the performance degradation trend of the battery; A multivariable optimization model is established to dynamically adjust the test parameters to optimize the temperature, flow rate and current density variables so that the test environment is always close to the target value.

2. The method for testing an all-vanadium redox flow battery according to claim 1, characterized in that: The step of establishing the distribution model of electrolyte concentration and current density comprises: The concentration gradient and current density distribution model of the electrolyte in the flow channel of the all-vanadium liquid flow battery is constructed by partial differential equations. The electrolyte concentration is set as C(x,t) as a function of position x and time t, and its diffusion coefficient is defined as D, flow velocity as u, current density as I, electron transfer number as n, and Faraday constant as F. The concentration distribution of the electrolyte is modeled by the diffusion-convection equation, and the dynamic change of the concentration satisfies the following partial differential equation: Among them, the diffusion term of electrolyte represents the diffusion process of concentration along the flow channel, and the convection term represents the convection flow of electrolyte, and the electrochemical reaction term Indicates the rate of change of electrolyte concentration during battery charging and discharging.

3. The method for testing an all-vanadium redox flow battery according to claim 2, characterized in that: The partial differential equation model is discretized by using a finite difference method or a finite element method, and the partial differential equation is discretized with a space step and a time step to obtain a numerical solution of the electrolyte concentration and the current density at each discrete node, and in the test process, it is iteratively calculated to obtain a real-time concentration gradient and current density distribution.

4. The method for testing an all-vanadium redox flow battery according to claim 1, characterized in that: The control algorithm collects electrolyte concentration and current density data in real time, compares them with preset target concentration gradient and current density, and automatically adjusts the flow rate and current density of the electrolyte according to the deviation.

5. The method for testing an all-vanadium redox flow battery according to claim 4, characterized in that: The control algorithm sets an adaptive adjustment coefficient based on the target concentration and current density, and uses a closed-loop feedback control method to update the set values ​​of the flow rate and current density in real time.

6. The method for testing an all-vanadium redox flow battery according to claim 1, characterized in that: The acquisition of electrochemical impedance spectroscopy data includes applying a small amplitude AC voltage signal at different frequencies, and forming impedance spectrum data by measuring the real and imaginary impedance parts of the battery at each frequency point. The impedance spectrum is used to reflect the charge transfer and electrolyte diffusion characteristics of the battery, and further extract the internal state information of the battery.

7. A method for testing an all-vanadium redox flow battery according to claim 6, characterized in that: The machine learning algorithm adopts a long short-term memory network model or a variational autoencoder model, takes multi-frequency impedance spectrum data as time series input, automatically extracts the nonlinear degradation characteristics of the battery through model training, and predicts the battery performance attenuation trend.

8. The method for testing an all-vanadium redox flow battery according to claim 1, characterized in that: The multivariable optimization model constructs constraints on test parameters to minimize the error between actual test performance and target performance, and implements dynamic optimization and adjustment of the test environment on the premise of satisfying the constraints to obtain test environment stability close to the target.

9. A method for testing an all-vanadium redox flow battery according to claim 8, characterized in that: The multivariable optimization model is solved by the Lagrange multiplier method, the upper and lower limits of each test parameter are set as constraints, and the optimization model is updated in real time in combination with the data collected during the test, so that the system automatically adjusts the temperature, flow rate and current density according to the real-time response of the battery.

10. A testing system for an all-vanadium redox flow battery, executing the method according to any one of claims 1 to 9, characterized in that: include: An electrolyte storage and circulation device for achieving electrolyte flow control; A current density regulating device, used for dynamically adjusting the current density of the battery; Temperature control system, used to adjust the test environment temperature of the battery; Data acquisition module, used to collect electrochemical impedance spectroscopy data and real-time data during the test; The control processor is connected to the data acquisition module and runs the control algorithm. It automatically adjusts the flow rate, current density, and temperature test parameters by analyzing the data, realizes real-time control of the internal concentration and current density of the battery, and adjusts the test conditions based on the optimization model.

Images