Method for estimating soc of aqueous sodium-ion battery

Abstract

The application discloses a kind of aqueous sodium-ion battery SOC estimation methods, comprising the following steps: step 1, establish aqueous sodium-ion battery DP equivalent circuit model and aqueous sodium-ion battery terminal voltage mathematical model;Step 2, the Laplace transform is carried out to the aqueous sodium-ion battery terminal voltage mathematical model obtained in step 1 and is discretized using bilinear transformation, then using FFRLS algorithm to calculate to be identified parameter;Step 3, based on AUKF algorithm estimates the SOC of aqueous sodium-ion battery.The application uses least square parameter estimation method with forgetting factor to realize the parameter identification of battery model, improves AUKF algorithm in combination with noise adaptive step, and can accurately estimate the real-time SOC of aqueous sodium-ion battery energy storage system.

Description

TECHNICAL FIELD

[0001] The application relates to the field of battery SOC estimation methods, and particularly relates to a water-based sodium ion battery SOC estimation method. BACKGROUND

[0002] For water-based sodium ion batteries with time-varying and nonlinear characteristics, a widely studied internal SOC state estimation method is a related algorithm based on Kalman filter, including Unscented Kalman Filter (UKF) and Extended Kalman Filter (EKF), which are used to improve the estimation accuracy of SOC. The specific description is as follows:

[0003] 1. Extended Kalman Filter algorithm: the algorithm has high accuracy in estimating the real-time SOC of water-based sodium ion batteries, and can improve the accuracy of SOC estimation by combining other filter algorithms or optimizing the battery model. Combining EKF with ampere-hour integration method can effectively suppress the influence of system noise, thereby enhancing the overall accuracy of SOC estimation.

[0004] 2. Unscented Kalman Filter algorithm (AUKF algorithm): the algorithm uses unscented transformation to avoid errors that may occur in the linearization process, and usually combines a first-order RC circuit model to apply UKF estimation algorithm to the SOC estimation of water-based sodium ion batteries, thereby improving the reliability and accuracy of the SOC estimation algorithm. Based on the PNGV equivalent circuit model, EKF and UKF are used to estimate the SOC of SIB, and UKF is superior to EKF in accuracy and stability.

[0005] However, the accuracy of the above-mentioned water-based sodium ion battery SOC estimation method based on Kalman filter depends on the accuracy of the model, and due to the complexity of the algorithm, more computing resources are required. SUMMARY

[0006] The application provides a water-based sodium ion battery SOC estimation method to solve the problem of high accuracy and computing resource requirements of the water-based sodium ion battery SOC estimation method based on Kalman filter in the prior art.

[0007] In order to achieve the above-mentioned purpose, the technical scheme adopted by the application is as follows:

[0008] The water-based sodium ion battery SOC estimation method comprises the following steps:

[0009] Step 1, establish a water-based sodium-ion battery DP equivalent circuit model, and determine a water-based sodium-ion battery terminal voltage mathematical model based on the water-based sodium-ion battery DP equivalent circuit model, wherein the water-based sodium-ion battery terminal voltage mathematical model has unknown battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant;

[0010] Step 2, Laplace transform is performed on the water-based sodium-ion battery terminal voltage mathematical model obtained in step 1, and bilinear transform discretization is adopted to obtain a discrete pulse transfer function G(z -1 );

[0011] Then, the FFRLS algorithm is used to calculate and identify the to-be-identified parameters in the discrete pulse transfer function G(z -1 ), and the unknown battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant in the battery terminal voltage mathematical model are solved through the to-be-identified parameters, so as to obtain the water-based sodium-ion battery terminal voltage mathematical model with known battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant;

[0012] Step 3, based on the water-based sodium-ion battery terminal voltage mathematical model with known battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant obtained in step 2, a discrete state space model of the water-based sodium-ion battery is established; then, the AUKF algorithm is used to estimate the SOC of the water-based sodium-ion battery based on the discrete state space model of the water-based sodium-ion battery.

[0013] Further in step 1, the water-based sodium-ion battery terminal voltage mathematical model is shown in the following formula:

[0014]

[0015] Wherein: Δt is a sampling time interval; U t represents the battery terminal voltage; U OCV represents the open-circuit voltage; I represents the battery terminal current; R0 represents the battery ohmic internal resistance; R1 represents the electrochemical polarization internal resistance; R2 represents the concentration polarization internal resistance; τ1 represents the electrochemical polarization time constant; τ2 represents the concentration polarization time constant; R0, R1, R2, τ1, τ2 are all unknown quantities to be solved.

[0016] Further in step 2, the discrete pulse transfer function G(z -1 ) is shown in the following formula:

[0017]

[0018] Wherein: Z is a discretization operator; θ = [b1, b2, b3, b4, b5] is a to-be-identified parameter; G(z -1 ) represents a discrete pulse transfer function; b1, b2, b3, b4, b5 are coefficients of a complex variable z respectively; U OCV (z -1 ) represents a discrete pulse transfer function of an open circuit voltage; U t (z -1 ) represents a discrete pulse transfer function of a terminal voltage; I(z -1 ) represents a discrete pulse transfer function of a terminal current.

[0019] In further step 2, the process of calculating the to-be-identified parameters in the discrete pulse transfer function G(z -1 ) by using the FFRLS algorithm is as follows:

[0020] First, the covariance matrix, the to-be-identified parameter, the state of charge, the open circuit voltage are initialized to obtain the covariance matrix P FFRLS (0), the to-be-identified parameter θ(0), the state of charge SOC(0), the open circuit voltage U ocv (0) at the initial time, i.e., time 0.

[0021] Then, a multi-time cycle recursive iteration of the FFRLS algorithm is performed, wherein a forgetting factor φ is introduced into the recursive process in the k-time cycle recursive iteration, and thus the covariance matrix, the to-be-identified parameter, and the initial data of the gain matrix at the k+1 time are obtained based on the covariance matrix, the to-be-identified parameter at the k time. Through the multi-time cycle iteration, until the estimation error of the accuracy of the to-be-identified parameter is less than a set value, the iteration is stopped, and the identification of the to-be-identified parameter is completed.

[0022] In further step 2, the calculation formulas of the unknown battery ohmic resistance, the electrochemical polarization resistance, the concentration polarization resistance, the electrochemical polarization time constant, and the concentration polarization time constant in the battery terminal voltage mathematical model are solved by the to-be-identified parameter as follows:

[0023]

[0024] Wherein: b1, b2, b3, b4, b5 in the to-be-identified parameter θ = [b1, b2, b3, b4, b5] are obtained by the FFRLS algorithm identification; T is a sampling time constant; R0 is a battery ohmic resistance, R1 is an electrochemical polarization resistance, R2 is a concentration polarization resistance, τ1 is an electrochemical polarization time constant, and τ2 is a concentration polarization time constant.

[0025] This invention employs the FFRLS identification algorithm (i.e., least squares parameter estimation method) to improve model accuracy and reduce computational resource requirements. Then, the AUKF algorithm (i.e., unscented Kalman filter algorithm) is used to estimate the SOC. The unscented transformation avoids errors in the linearization process and improves the accuracy and real-time performance of the calculation, thus meeting the requirements of online detection.

[0026] This invention employs a least-squares parameter estimation method with a forgetting factor to identify the parameters of the battery model. It also improves the AUKF algorithm by incorporating a noise adaptive step, enabling accurate estimation of the real-time SOC of an aqueous sodium-ion battery energy storage system. This ensures that each energy storage module operates at its optimal state, achieving dynamic power allocation, extending the lifespan of the parallel energy storage system, and guaranteeing its long-term stable operation. A simulation model is established to analyze the divided operating modes, verifying the effectiveness of the proposed strategy. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the equivalent circuit model of a water-based sodium-ion battery (DP) established in an embodiment of the present invention. Detailed Implementation

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0029] This embodiment discloses a method for estimating the state of charge (SOC) of an aqueous sodium-ion battery, including the following steps:

[0030] Step 1: Establish the equivalent circuit model of the aqueous sodium-ion battery DP, as follows: Figure 1 As shown. Figure 1 Middle U OCV U represents the open-circuit voltage of an aqueous sodium-ion battery. t R0 represents the terminal voltage of the aqueous sodium-ion battery; I represents the battery terminal current; R0 represents the battery's internal resistance in ohms; R1 represents the electrochemical polarization resistance; C1 represents the electrochemical polarization capacitance; R2 represents the concentration polarization resistance; C2 represents the concentration polarization capacitance; U1 represents the voltage across R1 and C1; U2 represents the voltage across R2 and C2. Figure 1 In the DP equivalent circuit model of the aqueous sodium-ion battery shown, the resistance generated inside the aqueous sodium-ion battery due to the Ohmic effect and polarization effect is described by two RC networks composed of resistors R1, C1, R2, and C2.

[0031] Based on such Figure 1 The equivalent circuit model of the aqueous sodium-ion battery DP shown can be used to obtain the initial estimate of the SOC state of the aqueous sodium-ion battery as shown in formula (1.1):

[0032]

[0033] In formula (1.1), SOC0 represents the initial value of the state of charge, C N represents the rated capacity, η represents the coulombic efficiency (η≈1); SOc represents the current state of charge of the battery; I d represents the battery terminal current; and t represents the charging and discharging time.

[0034] And based on the water-based sodium-ion battery DP equivalent circuit model as shown in Figure 1 , a mathematical model of the water-based sodium-ion battery terminal voltage can be obtained as shown in formula (1.2):

[0035]

[0036] In formula (1.2), Δt is the sampling time interval; U t represents the battery terminal voltage; U OCV represents the open circuit voltage; I represents the battery terminal current; R0 represents the battery ohmic internal resistance; R1 represents the electrochemical polarization internal resistance; R2 represents the concentration polarization internal resistance; τ1 represents the electrochemical polarization time constant; and τ2 represents the concentration polarization time constant.

[0037] In the above mathematical model of the water-based sodium-ion battery terminal voltage, the battery ohmic internal resistance R0, the electrochemical polarization internal resistance R1, the concentration polarization internal resistance R2, the electrochemical polarization time constant τ1, and the concentration polarization time constant τ2 are unknown quantities.

[0038] Step 2, solve the unknown battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant in the mathematical model of the battery terminal voltage, thereby obtaining the mathematical model of the water-based sodium-ion battery terminal voltage with known battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant, and concentration polarization time constant, the process is as follows:

[0039] (2.1), Laplace transform the water-based sodium-ion battery terminal voltage mathematical model as shown in formula (1.2) obtained in step 1 and use bilinear transform to discretize, to obtain a discrete impulse transfer function G(z -1 ) as shown in formula (2.1):

[0040]

[0041] In formula (2.1), Z is a discretization operator; θ=[b1, b2, b3, b4, b5] is a parameter to be identified; G(z -1 ) represents the discrete impulse transfer function; b1, b2, b3, b4, and b5 are coefficients of the complex variable z, respectively; U OCV (z -1 ) represents the discrete impulse transfer function of the open circuit voltage; U t (z -1) represents the discrete pulse transfer function of terminal voltage; I(z -1 ) represents the discrete pulse transfer function of terminal current.

[0042] (2.2), the FFRLS algorithm is used to calculate the to-be-identified parameters θ in the discrete pulse transfer function G(z -1 ) and then the unknown battery ohmic resistance R0, electrochemical polarization resistance R1, concentration polarization resistance R2, electrochemical polarization time constant τ1, and concentration polarization time constant τ2 in the battery terminal voltage mathematical model are solved from the to-be-identified parameters, so that the water-based sodium-ion battery terminal voltage mathematical model with known battery ohmic resistance R0, electrochemical polarization resistance R1, concentration polarization resistance R2, electrochemical polarization time constant τ1, and concentration polarization time constant τ2 is obtained. The FFRLS algorithm identification process is as follows:

[0043] 2.2a) initializing the covariance matrix, the to-be-identified parameters, the state of charge, and the open-circuit voltage to obtain the covariance matrix P FFRLS (0) at the initial time, i.e., time 0, the to-be-identified parameters θ(0), the state of charge SOC(0), and the open-circuit voltage U ocv (0).

[0044] And let P FFRLS (k), θ(k), SOC(k), and U ocv (k) represent the covariance matrix, the to-be-identified parameters, the state of charge, and the open-circuit voltage at time k, respectively, and P FFRLS (k+1), θ(k+1), SOC(k+1), and U ocv (k+1) represent the covariance matrix, the to-be-identified parameters, the state of charge, and the open-circuit voltage at time k+1, respectively.

[0045] 2.2b) then performing a multi-time loop recursion of the FFRLS algorithm, wherein a forgetting factor φ is introduced into the recursion process when the k-time loop recursion is performed, so that the covariance matrix, the to-be-identified parameters, and the initial data of the gain matrix at time k+1 are obtained based on the covariance matrix, the to-be-identified parameters at time k, and the calculation process is shown in formula (2.2):

[0046]

[0047] In formula (2.2), Φ(k+1) represents the system observation data at time k+1, which is a column vector composed of the differences between the open-circuit voltages and the terminal voltages at different times; K(k+1) is the gain matrix at time k+1, which is an intermediate parameter in the algorithm calculation process; The meanings of θ(k) and θ(k+1) are consistent, and they represent the to-be-identified parameters at time k and time k+1, respectively; φ is a forgetting factor, and its value is generally 0.95-0.99.

[0048] The obtained K(k+1), P FFRLS (k+1) as the initial data of the k+1 time recursion, and thus realize the multi-round time cycle recursion, until the estimation error of the to-be-identified parameter precision obtained is less than the set value ε, then stop the cycle, and complete the identification of the to-be-identified parameter.

[0049] 2.2c) The unknown battery ohmic internal resistance R0, electrochemical polarization internal resistance R1, concentration polarization internal resistance R2, electrochemical polarization time constant τ1, and concentration polarization time constant τ2 in the battery terminal voltage mathematical model are solved by the identified to-be-identified parameters, and the calculation formula is as follows:

[0050]

[0051] Wherein: b1, b2, b3, b4, and b5 in the to-be-identified parameter θ = [b1, b2, b3, b4, b5] are identified by the FFRLS algorithm; T is the sampling time constant; R0 is the battery ohmic internal resistance, R1 is the electrochemical polarization internal resistance, R2 is the concentration polarization internal resistance, τ1 is the electrochemical polarization time constant, and τ2 is the concentration polarization time constant.

[0052] Step 3, estimate the SOC of the aqueous sodium-ion battery based on the AUKF algorithm, and the specific process is as follows:

[0053] (3.1) Based on the known aqueous sodium-ion battery terminal voltage mathematical model of the battery ohmic internal resistance R0, electrochemical polarization internal resistance R1, concentration polarization internal resistance R2, electrochemical polarization time constant τ1, and concentration polarization time constant τ2 obtained in step 2, the discrete state space model of the aqueous sodium-ion battery is established as shown in formulas (3.1) and (3.2):

[0054]

[0055] U t (k+1) = U OCV -R0I(k)-U1(k)-U2(k) (3.2)

[0056] Wherein: k is the discrete time index, k = 0, 1, 2…n; SOC(k) is the state of charge at k time; U1(k) is the voltage across R1C1 at k time; U2(k) is the voltage across R2C2 at k time; I(k) is the terminal current at k time; U t (k+1) is the terminal voltage at k time.

[0057] (3.2) The discrete state space model of the aqueous sodium-ion battery is described in the form as shown in formula (3.3):

[0058]

[0059] In formula (3.3), w k is state noise, whose mean is 0 and whose covariance is Q k ; v k is observation noise, whose mean is 0 and whose covariance is R k ; x k represents state quantity, y k represents observation value U t (k); u k represents input quantity I(k); SOC(k) is the state of charge at time k; U1(k) is the voltage across R1C1 at time k; U2(k) is the voltage across R2C2 at time k; I(k) is the terminal current at time k; U t (k+1) is the terminal voltage at time k. f() represents the equation of state quantity; h() represents the equation of observation quantity.

[0060] (3.3) Then, the state quantity x k in the discretized state space model of the aqueous sodium-ion battery is estimated by using the AUKF algorithm, so as to obtain the SOC of the aqueous sodium-ion battery. The process is as follows:

[0061] 3.3a) The state quantity x T at time k is subjected to unscented transformation to obtain a sigma point set of the state quantity x and corresponding weights, as shown in formulas (3.4) and (3.5):

[0062]

[0063] In formulas (3.4) and (3.5), x is a 2-dimensional signal, so n takes the value of 2, and the superscript i takes the value of 0-4; and P represent the mean and variance of the state quantity x (i) (k|k) is the mean of the state quantity x at time k, and P(k|k) represents the variance of the state quantity x at time k. ω m is the weight of the state prediction mean; ω c is the covariance weight; the two values here have a one-to-one correspondence with x (i) (k|k), that is, is the weight corresponding to x (0) (k|k).

[0064] The parameter λ = α 2(n+K)-n, which can be controlled by the value of a to adjust the effect of high order terms on the results, is a very small normal number, the selection of K needs to ensure that the matrix (n+K)P is semi-positive definite; parameter β is a weight coefficient which is not negative, in this paper, 0≤a≤1, β=2, K=3-n.

[0065] 3.3b) The obtained 2n+1 sigma points are substituted into formula (3.1) to obtain the further predicted value of sigma as shown in formula (3.6):

[0066] x (i) (k+1|k)=f[k,x (i) (k|k)](3.6)

[0067] In formula (3.6), x (i) (k+1|k) represents the predicted value of the sigma point at k+1 time, f[] is the prediction function, and i takes 0-4.

[0068] 3.3c) The further predicted value x (i) (k+1|k) is weighted and summed to obtain the weight ω c (i) The predicted mean value of the state quantity and its covariance matrix can be obtained, and the result is shown in formula (3.7):

[0069]

[0070] In formula (3.7), x (i) (k+1|k) represents the predicted value of the sigma point at k+1 time, represents the weight of the corresponding x (i) (k+1|k), and Q is the covariance of the state noise. n takes 2.

[0071] 3.3d) The predicted mean value x obtained by formula (3.7) is subjected to unscented transformation, and the updated sigma point set is shown in formula (3.8):

[0072]

[0073] In formula (3.8): represents the mean value of the updated sigma point, n takes 2 parameters, and λ=a 2 (n+K)-n, which can be controlled by the value of a to adjust the effect of high order terms on the results, is a very small normal number, the selection of K needs to ensure that the matrix (n+K)P is semi-positive definite; parameter β is a weight coefficient which is not negative, in this paper, 0≤a≤1, β=2, K=3-n.

[0074] 3.3e) Substitute the updated sigma point set into formula (3.1) to predict the observations at time k+1, as shown in formula (3.9):

[0075] y (i) (k+1|k)=h(x i (k|k),u k (3.9)

[0076] In formula (3.9): y (i) (k+1|k) represents the sigma point set of the predicted values ​​of the observations at time k+1 at time k, and h() is the prediction function.

[0077] 3.3f) Combining weight ω m For the predicted value y of the observation (i) The weighted summation of (k+1|k) yields the predicted mean and covariance matrix of the terminal voltage predictions, as shown in formula (3.10):

[0078]

[0079] In formula (3.10): P represents the mean of the predicted values ​​of the observations at time k+1 from the observations at time k. ykyk With P xkyk Let R represent the covariance matrix of the observation itself, and the covariance matrix between the observation and the state variable, where R is the covariance of the observation noise.

[0080] 3.3g) Calculate the Kalman gain, as shown in formula (3.11):

[0081]

[0082] In formula (3.11): P ykyk With P xkyk Let K(k+1) represent the covariance matrix of the observation itself, and the covariance matrix between the observation and the state variable, where K(k+1) is the Kalman gain.

[0083] 3.3h) The process noise Q and observation noise R in equation (3.3) are adaptively updated, and the noise adaptive equation is shown in equation (3.12):

[0084]

[0085] Formula (3.12): Q(k) is the covariance of the state noise at time k, and R(k) is the covariance of the observation noise at time k.

[0086] H(k) is an approximate value of the information covariance at time k, and its calculation formula is shown in formula (3.13):

[0087]

[0088] In formula (3.13), M is the window size, and the value of M is selected as 5; y (i) (k)=y (i) (k+1|k), represents the sigma point set of the predicted value of the observation at k+1 time at k time. (k+1|k), represents the sigma point set of the predicted value of the observation at k+1 time at k time.

[0089] 3.3i) Finally, the state variable and covariance of the system are updated, as shown in formula (3.14):

[0090]

[0091] Formula (3.14): (k+1|k+1) represents the covariance matrix of the state quantity at k+1 time, which is used as the next step input of the algorithm.

[0092] The above steps 3.3a)-3.3i) are repeated in turn to complete the SOC estimation of the sodium ion battery.

[0093] The preferred embodiments of the present application are described in detail above with reference to the accompanying drawings, and the examples described in the present application are only used to describe the preferred embodiments of the present application, and do not limit the concept and scope of the present application. In the above specific embodiments, each specific technical feature described above can be combined in any appropriate manner without contradiction, and such combination should also be considered as disclosed by the present disclosure as long as it does not deviate from the technical concept of the present application. In order to avoid unnecessary repetition, the present application does not further describe various possible combinations.

[0094] The present application is not limited to the specific details described in the above embodiments, and various modifications and improvements of the technical solutions of the present application made by those skilled in the art within the scope of the technical concept of the present application and without departing from the design idea of the present application should fall within the protection scope of the present application. The technical content claimed by the present application has been fully recorded in the claims.

Claims

1. A method for SOC estimation of aqueous sodium-ion batteries, characterized in that, The method comprises the following steps: Step 1, establishing a water-based sodium-ion battery DP equivalent circuit model, and determining a water-based sodium-ion battery terminal voltage mathematical model based on the water-based sodium-ion battery DP equivalent circuit model, wherein the water-based sodium-ion battery terminal voltage mathematical model has unknown battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant and concentration polarization time constant; Step 2, Laplace transform is performed on the water-based sodium-ion battery terminal voltage mathematical model obtained in step 1, and bilinear transform discretization is adopted to obtain a discrete impulse transfer function G(z -1 ) Then, the FFRLS algorithm is used to calculate the to-be-identified parameters in the discrete pulse transfer function G(z -1 ) to be identified, and the unknown battery ohmic resistance, electrochemical polarization resistance, concentration polarization resistance, electrochemical polarization time constant, and concentration polarization time constant in the battery terminal voltage mathematical model are inversely solved by the to-be-identified parameters, so that the water-based sodium ion battery terminal voltage mathematical model with known battery ohmic resistance, electrochemical polarization resistance, concentration polarization resistance, electrochemical polarization time constant, and concentration polarization time constant is obtained. Step 3, based on the water-based sodium-ion battery terminal voltage mathematical model with known battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant and concentration polarization time constant obtained in step 2, a discrete state space model of the water-based sodium-ion battery is established; then, the AUKF algorithm is used to estimate the SOC of the water-based sodium-ion battery based on the discrete state space model of the water-based sodium-ion battery.

2. The aqueous sodium-ion battery SOC estimation method according to claim 1, characterized in that, In step 1, the water-based sodium-ion battery terminal voltage mathematical model is as shown in the following formula: where: Δt is the sampling time interval; U t represents the open-circuit voltage; I represents the battery terminal current; R0represents the battery ohmic internal resistance; R1represents the electrochemical polarization internal resistance; R2represents the concentration polarization internal resistance; τ1represents the electrochemical polarization time constant; τ2represents the concentration polarization time constant; R0, R1, R2, τ1, τ2are all unknown quantities to be solved. OCV represents the open-circuit voltage; I represents the battery terminal current; R0represents the battery ohmic internal resistance; R1represents the electrochemical polarization internal resistance; R2represents the concentration polarization internal resistance; τ1represents the electrochemical polarization time constant; τ2represents the concentration polarization time constant; R0, R1, R2, τ1, τ2are all unknown quantities to be solved.

3. The aqueous sodium-ion battery SOC estimation method according to claim 1, characterized in that, In step 2, the discrete pulse transfer function G(z -1 ) is given by the following equation: wherein: Z is a discretization operator; θ = [b1, b2, b3, b4, b5] is a parameter to be identified; G(z -1 ) represents a discrete pulse transfer function; b1, b2, b3, b4, b5 are coefficients of a complex variable z, respectively; U OCV (z -1 ) represents a discrete pulse transfer function of an open circuit voltage; U t (z -1 ) represents a discrete pulse transfer function of a terminal voltage; I(z -1 ) represents a discrete pulse transfer function of a terminal current.

4. The aqueous sodium-ion battery SOC estimation method according to claim 1, characterized in that, In step 2, the FFRLS algorithm is used to calculate the discrete pulse transfer function G(z) to be identified in step 1 -1 The process of identifying the parameters in the discrete pulse transfer function G(z) is as follows: First, the covariance matrix, the parameter to be identified, the state of charge, the open circuit voltage are initialized to obtain the covariance matrix P at the initial time, i.e., time 0 FFRLS (0), the parameter to be identified θ(0), the state of charge SOC(0), the open circuit voltage U ocv (0); Then, a multi-round time cycle recursive propagation of the FFRLS algorithm is performed, wherein a forgetting factor φ is introduced into the recursive propagation process in the kth time cycle recursive propagation, and thus the covariance matrix, the to-be-identified parameter and the gain matrix initial data in the k+1th time cycle are obtained based on the covariance matrix, the to-be-identified parameter in the kth time cycle; When the estimation error of the accuracy of the to-be-identified parameter obtained through the multi-round time cycle is less than a set value, the cycle is stopped, and the identification of the to-be-identified parameter is completed.

5. The aqueous sodium-ion battery SOC estimation method according to claim 1, characterized in that, In step 2, the calculation formula of the unknown battery ohmic internal resistance, electrochemical polarization internal resistance, concentration polarization internal resistance, electrochemical polarization time constant and concentration polarization time constant in the battery terminal voltage mathematical model is solved by the to-be-identified parameter as follows: Wherein: b1, b2, b3, b4 and b5 in the to-be-identified parameter θ=[b1, b2, b3, b4, b5] are identified by the FFRLS algorithm; T is a sampling time constant; R0 is the battery ohmic internal resistance, R1 is the electrochemical polarization internal resistance, R2 is the concentration polarization internal resistance, τ1 is the electrochemical polarization time constant, and τ2 is the concentration polarization time constant.

Images