Lithium battery state-of-charge and state-of-health joint estimation method

Abstract

The invention relates to a lithium battery state-of-charge and state-of-health joint estimation method, and belongs to the technical field of lithium ion batteries. The method comprises the following steps: performing a cyclic aging experiment on a battery under a cyclic working condition; an open-circuit voltage curve is extracted through data analysis, a first-order equivalent circuit model is selected to construct a circuit, and a state equation is written; carrying out parameter identification by adopting a multi-forgetting factor recursive least square algorithm, and constructing an SOC estimator of an unscented Kalman filter based on a state equation and an observation equation to carry out SOC estimation; constructing an SOH estimation observer based on an extended Kalman filter, and estimating the SOH by using the estimated SOC; carrying out segmented dynamic identification on the open-circuit voltage curve by adopting an average moving method in combination with recursive least squares; and coupling the SOC estimator and the SOH estimator, and realizing SOC and SOH joint online estimation by applying the dynamically identified open-circuit voltage curve.

Description

Technical Field

[0001] The present invention belongs to the technical field of lithium-ion batteries, and relates to a method for jointly estimating the state of charge and state of health of a lithium battery based on parameter identification by the recursive least squares method with multiple forgetting factors. Background Art

[0002] Due to its advantages such as high power density, high energy density, long cycle life, and low self-discharge rate, lithium-ion batteries have now been widely used in devices such as electric vehicles, mobile phones, airplanes, and satellites. The market has higher and higher requirements for the operation of the battery management system. Accurately estimating the state of charge (SOC) and state of health (SOH) of lithium-ion batteries is the core function of the battery management system (BMS). SOC reflects the current charge storage state of the battery, which is expressed as the ratio of the current remaining capacity to the current maximum available capacity, and plays an important role in the charging and discharging process; SOH reflects the aging state of the battery during the entire usage cycle, which is expressed as the ratio of the maximum available capacity to the initial capacity of the battery. Therefore, its accurate estimation is helpful for the health diagnosis of the battery. There is a strong coupling relationship between SOC and SOH, and battery aging will affect the accuracy of SOC estimation.

[0003] The combined estimation method of SOC and SOH has become more popular in recent years. SOC and SOH are two key parameters of the battery, which respectively reflect the state of charge and the health degree of the battery. By jointly estimating these two parameters, the overall performance of the battery can be comprehensively evaluated, including information such as the available battery power and the remaining life. At the same time, the combined estimation of SOC and SOH can provide more accurate battery state information to help take more effective measures. There are mainly two combined estimation methods: (1) The filter method combines the measurement data with the system model to provide an estimation of the system state. Usually, when using the Kalman filtering algorithm to estimate SOC, the internal parameters of the battery model are averaged through offline parameter identification. Thus, the internal parameters of the battery such as internal resistance and capacitance are regarded as a constant value. However, in the actual operation of lithium batteries, the internal parameters will vary under different working conditions. Therefore, the internal parameters of lithium batteries cannot be regarded as a fixed value for calculation. The filter method usually uses the dual Kalman filtering algorithm. This method uses two Kalman algorithms to estimate the internal parameters of the battery and the battery SOC respectively, and then establishes a dual Kalman model with mutual input to alternately update the internal parameters of the battery to achieve the combined estimation of SOC and SOH. (2) The data-driven method does not consider the dynamics and non-linear electrochemical characteristics of the battery and can be used for the estimation of battery SOC and SOH. Using neural network models such as SVM and LSTM, a mapping model of SOC and SOH is established based on the measured current and voltage to achieve the combined estimation of SOC and SOH. Both of the above methods can achieve the estimation of SOC under aging conditions, but they do not consider the influence of different working conditions on battery parameters and other influencing factors brought by battery aging on the estimation of lithium battery SOC and SOH, resulting in low estimation accuracy of SOC and SOH. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a combined estimation method of the state of charge and health state of a lithium battery based on parameter identification by the multi-forgetting factor recursive least squares method to overcome the problem that different parameters have different response speeds to different working conditions and the problem that the estimation accuracy of SOC and SOH decreases due to aging.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A combined estimation method of the state of charge and health state of a lithium battery based on parameter identification by the multi-forgetting factor recursive least squares method specifically includes the following steps:

[0007] S1: Conduct HPPC (Hybrid Pulse Power Characteristic) test experiments on new batteries and conduct cyclic aging experiments on the batteries under cyclic working conditions;

[0008] S2: Extract the open-circuit voltage curve through data analysis, select the first-order equivalent circuit model to construct the circuit, and write the state equation;

[0009] S3: Use the multi-forgetting factor recursive least squares method for parameter identification, and construct an SOC estimator based on the unscented Kalman filter of the state equation and the observation equation to estimate SOC;

[0010] S4: Construct an SOH estimator based on the extended Kalman filter, and estimate SOH using the estimated SOC;

[0011] S5: Use the moving average method combined with the recursive least squares method for dynamic identification of the open-circuit voltage curve;

[0012] S6: Couple the SOC estimator and the SOH estimator, and apply the open-circuit voltage curve after dynamic identification to realize the joint online estimation of SOC and SOH.

[0013] Furthermore, step S1 specifically includes: Conduct HPPC tests on the new lithium battery to obtain battery parameter data; Use the strategies of full charge and full discharge to cycle-age the battery, and obtain the data of each charge-discharge cycle of the battery, capacity aging data, current I(t), voltage U(t), and capacity Q(t).

[0014] Furthermore, step S2 specifically includes the following steps:

[0015] S21: In the HPPC test experiment in step S1, use the standard current to charge the cell under test to full charge in a constant current and constant voltage manner, and let it stand for a period of time (1 h) to make it close to the equilibrium state. By loading a mixed pulse current excitation sequence, perform a 10% capacity (i.e., 10% SOC) discharge operation on the cell. After letting the cell stand to approach the equilibrium state, load the excitation sequence again, repeat the whole process, and select the open-circuit voltage values corresponding to SOC of 100%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 0%, and perform a sixth-order polynomial fitting to obtain the open-circuit voltage curve;

[0016] S22: Based on the first-order equivalent circuit model of the lithium battery, the state equation is:

[0017] U = U oc + V p + R 0 ·I(1)

[0018] Among them, U is the terminal voltage of the lithium-ion power battery, I represents the terminal current of the lithium-ion power battery, R 0 represents the ohmic internal resistance of the lithium-ion power battery, R p represents the polarization resistance of the lithium-ion power battery, C pRepresents the polarization capacitance of a lithium-ion power battery, V p Represents the polarization capacitance C of a lithium-ion power battery p of the terminal voltage, U oc Represents the open-circuit voltage;

[0019] The transfer function of the battery impedance in the S domain is obtained through Laplace transform:

[0020]

[0021] where s represents the complex variable of Laplace transform; through bilinear transform is mapped to the Z plane:

[0022]

[0023] where z represents the discretized complex frequency domain variable, a 1 , a 2 , a 3 represents the coefficient in the discretization process; finally, the expressions of a 1 , a 2 , a 3 can be obtained:

[0024]

[0025] where Δt represents the sampling time interval; from this, the expressions of R 0 , R P , C P can be deduced:

[0026]

[0027] Assume that the influence of the power consumed or absorbed by the battery within a unit sampling interval on its state of charge is approximately zero, that is, U OC,k = U OC,k-1 , and the temperature and aging state remain unchanged. Equation (1) can be expressed as:

[0028] U k = (1 - a 1 )U OC,k + a 1 U k-1 + a 2 I k + a 3 I k-1 (6)

[0029] Express Equation (6) in matrix form:

[0030] y k = φ k ·θ k(7)

[0031] Among them, y k is the terminal voltage value, φ k is the input matrix of the system, φ k = [1 U k-1 I k I k-1 ], θ k is the parameter matrix of the system, θ k = [(1 - a 1 ) UO C ,k a 1 a 2 a 3 ] T , U k-1 represents the terminal voltage value of the lithium-ion power battery at the (k - 1)th moment, I k represents the current of the lithium-ion power battery at the current moment, I k-1 represents the current at the previous moment, U OC,k represents the open-circuit voltage at the current moment.

[0032] Furthermore, step S3 specifically includes the following steps:

[0033] S31: Perform parameter identification using the multi-forgetting-factor recursive least squares method:

[0034] According to the least squares method, we can obtain:

[0035]

[0036] Among them, represents the parameter to be estimated, L(k) represents the gain matrix, and φ(k) represents the input matrix;

[0037] Since there are multiple parameters in the system, and the rates of change of multiple parameters for different conditions are different, and a single forgetting factor is prone to identification errors in the case of a relatively stable system, the present invention adopts the multi-forgetting-factor recursive least squares method to identify the system parameters and obtains the criterion function:

[0038]

[0039] Among them, represents the loss function, n represents the order; λ k represents four forgetting factors corresponding to four parameters respectively, and satisfies 0 < λ i ≤1, and a new gain matrix L new (k) is obtained:

[0040]

[0041] Among them, P i (k - 1) represents the new covariance matrix corresponding to the i-th forgetting factor; P(k) is the new covariance matrix:

[0042] P(k) = A -1 [I - L(k)φ T (k)P(k - 1)]A -1 (11)

[0043] Among them, A = diag[λ 1 , λ 2 , λ 3 , λ 4 , and the expression is obtained when the criterion function criterion represented by Equation (9) is minimized:

[0044]

[0045] S32: Construct an SOC estimator of the unscented Kalman filter based on the state equation and the observation equation for SOC estimation. According to the ampere-hour integration method and combined with the first-order equivalent circuit model, the state-space equation can be obtained:

[0046]

[0047] Among them, the system parameter θ = [θ 1 , θ 2 , θ 3 T , θ 2 = R P , θ 3 = R 0 , S k+1 represents the SOC value at the current moment, S k represents the SOC value at the previous moment, C n represents the maximum available capacity of the battery at the current moment, T s represents the sampling time interval, U k represents the current current, C n is represented by the nominal capacity in the first cycle; y k represents the terminal voltage value, V oc (S k ) represents the open-circuit voltage value corresponding to the SOC, V k represents the polarization voltage;

[0048] Simplifying Equation (13) gives:

[0049]

[0050] Among them, x k = [S k , V​k T , S k represents the SOC, V k represents the polarization voltage; g(x k , u k ) = V oc (S k ) + V k , v k = θ 3 ·u k ;

[0051] Initialize the state vector x k and the error covariance matrix

[0052] Construct 2n + 1 sigma points according to the unscented transformation:

[0053]

[0054] where λ = α 2 (n + h) - n is the scaling factor, where α represents the ratio coefficient and h represents the time step, and then perform state update:

[0055]

[0056] where x i,k|k-1 represents the state estimate value of the i-th sigma point at time k, f(x i,k-1 ) represents the state transition function, represents the prior estimate of the state at time k, represents the prior covariance matrix of the state estimate at time k, represents the contribution (weight coefficient) of each sigma point to the covariance matrix, represents the weight coefficient of each sigma point, q 0 represents the initial time deviation, Q 0 represents the process noise covariance matrix;

[0057] Then perform measurement update:

[0058]

[0059] where x i,k|k-1 represents the state estimate value of the i-th sigma point at time k, represents the measurement function, represents the initial input, represents the measurement prediction value at time k, g(x i,k|k-1 , u k ​) represents the measurement function, r 0 represents the error during the measurement process, represents the covariance matrix of the measurement, represents the covariance matrix between the state and the measurement, represents the prior state estimate value at time k;

[0060] Calculate the Kalman gain, update the state estimate and the error covariance:

[0061]

[0062] where, K k represents the Kalman gain, represents the state estimate value at time k, P k represents the error covariance matrix.

[0063] Furthermore, step S4 specifically includes: constructing an SOH estimator based on the extended Kalman filter, and at the same time, SOH can be defined as:

[0064]

[0065] where, Q n is the rated capacity, Q is the current maximum available capacity, use the estimated SOC to estimate SOH, and taking the capacity as the state variable, the state space equation of SOH can be obtained:

[0066]

[0067] where, Q k is the capacity at the previous moment, Q k+1 is the capacity at the current moment, r k represents the process error, V oc (S k ) represents the open-circuit voltage value corresponding to SOC, V k represents the polarization voltage;

[0068] Initialize the state vector and the covariance matrix:

[0069]

[0070] where, represents the initial value of the error covariance matrix;

[0071] Perform state update:

[0072] Q k = Q k-1 (26)

[0073] Update the error covariance matrix:

[0074]

[0075] Among them, represents the error covariance matrix of the capacity estimate, Q Q represents the process noise covariance matrix;

[0076] Calculate the Kalman gain, update the state estimate and the error covariance:

[0077]

[0078] Among them, represents the Kalman gain, represents the Jacobian matrix, L Q represents the measurement noise covariance matrix, g(x k-1 , u k-1 , Q k-1 ) represents the measurement equation.

[0079] Furthermore, step S5 specifically includes: using the moving average method combined with the recursive least squares method for segmental dynamic identification of the open-circuit voltage curve. When the battery ages, the open-circuit voltage curve will drift, which affects the SOC estimation result. Therefore, dynamic identification is required. The identification process:

[0080]

[0081] Among them, e ocv (k) is the error of the open-circuit voltage curve, i b (k) represents the current at time k, represents the polarization voltage corresponding to time k, L represents the order, a p represents the polynomial coefficient of the open-circuit voltage curve, p represents the order with respect to SOC, represents the SOC estimation value at time k, and the recursive least squares method is applied for parameter identification:

[0082]

[0083] Among them, represents the parameter to be estimated, represents the estimated value of the polynomial coefficient of the open-circuit voltage curve;

[0084] Obtain a new open-circuit voltage expression:

[0085]

[0086] Among them, n p represents the order of SOC, a i represents the coefficient of SOC of order i, s i represents the i-th power of SOC.

[0087] The beneficial effects of the present invention are as follows: Considering the different influences of different operating conditions on parameters, the present invention uses the multi-forgetting-factor recursive least squares method for parameter identification, which can achieve more accurate parameter estimation under different environmental changes, provide accurate capacity for state-of-charge estimation when the battery ages, and reduce errors; and dynamically updates the open-circuit voltage curve through the moving average method, solves the problem that the open-circuit voltage curve drifts when the lithium battery ages, and improves the accuracy and stability of the state-of-charge and state-of-health estimation of the lithium battery.

[0088] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0090] Figure 1 is the flow chart of the method for jointly estimating the state-of-charge and state-of-health of the lithium battery of the present invention;

[0091] Figure 2 is the circuit topology schematic diagram of the first-order equivalent circuit model adopted in this embodiment;

[0092] Figure 3 is the open-circuit voltage curve graph of the power battery;

[0093] Figure 4 are the state-of-charge estimation results of multi-forgetting-factor parameter identification and single-forgetting-factor parameter identification;

[0094] Figure 5 are the state-of-health estimation results. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0095] The following uses specific specific examples to illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation manners. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present invention schematically. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0096] Among them, the attached drawings are only for illustrative purposes, showing only schematic diagrams rather than physical diagrams, and should not be construed as limiting the present invention; in order to better illustrate the embodiments of the present invention, some components in the attached drawings will be omitted, enlarged or reduced, which do not represent the dimensions of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the attached drawings may be omitted.

[0097] In the attached drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc. indicating the orientation or positional relationship, they are based on the orientation or positional relationship shown in the attached drawings. This is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the attached drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0098] Please refer to Figures 1 to 5 , considering that battery parameters have different rates of change with respect to environmental changes, the maximum available capacity decreases over time, and the open-circuit voltage curve shifts when the lithium battery ages, the present invention proposes a joint estimation method for state of charge and state of health based on multi-forgetting factor recursive least squares parameter identification. This method uses multi-forgetting factor recursive least squares parameter identification to identify the parameters in the first-order equivalent circuit model, achieving independent and accurate tracking of the parameters with each different rate of change. Then, UKF is used for SOC estimation to achieve accurate SOC estimation. MVFRLS not only achieves accurate parameter estimation but also improves the accuracy of SOC estimation. Then, the estimated SOC is used to perform SOH estimation, realizing the update of the capacity and providing an accurate capacity for SOC estimation at the same time. Furthermore, the open-circuit voltage curve is dynamically updated by combining the moving average method and recursive least squares method, achieving the correction of the open-circuit voltage curve under aging conditions. The application of the joint method for battery state of charge and state of health effectively improves the estimation accuracy, reliability, and real-time performance of the proposed algorithm.

[0099] The following further elaborates on the present invention in conjunction with the attached drawings.

[0100] As Figure 1 shown, the method of the present invention mainly includes the following steps:

[0101] S1: Use a standard current to fully charge the battery cell to be tested in a constant current and constant voltage manner, and let it stand for 1 h to make it close to the equilibrium state. By applying a hybrid pulse current excitation sequence, perform 10% Q on the battery cell nFor the discharge operation at 10% SOC, after leaving the battery cell to stand until it approaches the equilibrium state, the loading excitation sequence is performed again, and the whole process is repeated. Charge at a constant current of 2.0C to 4.2V, then keep the voltage constant until the current drops to 0.05C; rest for 5 minutes; discharge at 1.0C to 2.5V; cycle aging with the strategy of resting for 5 minutes.

[0102] S2: Extract the open-circuit voltage curve from data analysis, extract the charge current curve and the discharge current curve, and then as Figure 3 Take the mean value of the two curves as the open-circuit voltage curve, and select the first-order equivalent circuit model such as Figure 2 , and write the state equation.

[0103] S3: Use the multi-forgetting factor recursive least squares algorithm for parameter identification. Design different forgetting factors for each model parameter to achieve independent tracking of different parameters. Construct an SOC estimator of the unscented Kalman filter based on the state equation and the observation equation for SOC estimation. The SOC estimation result based on the multi-forgetting factor recursive least squares parameter identification is as Figure 4 shown.

[0104] S4: To achieve accurate SOC estimation, accurate capacity must be used. However, the capacity will gradually decrease over time and with use, resulting in a gradual increase in the SOC estimation error. Therefore, it is necessary to perform a state of health estimation to calculate the accurate capacity. Thus, construct an SOH estimator based on the extended Kalman filter and estimate the SOH using the estimated SOC.

[0105] S5: As the lithium battery ages, the relationship between the open-circuit voltage (OCV) and the state of charge (SOC) of the battery will change significantly. Specifically, aging causes phenomena such as capacity decay, internal resistance increase, and interface layer change of the battery. These factors will all affect the electrochemical performance of the battery, thereby changing the mapping relationship between OCV and SOC. Therefore, accurately updating the open-circuit voltage curve is particularly important for the SOC estimation and SOH estimation of the battery. The change of the open-circuit voltage curve is slow, significantly slower than the change of SOC. Therefore, the coefficient of the open-circuit voltage curve cannot be estimated using the SOC at a single moment. Using the moving average method can avoid the influence of single-time SOC fluctuations on model update. Therefore, a moving average method combined with recursive least squares is proposed for segmental dynamic identification of the open-circuit voltage curve to achieve online update of the open-circuit voltage curve without the need for regular calibration or offline identification.

[0106] S6: Couple the SOC estimator and the SOH estimator. First, perform parameter identification using multiple forgetting factors, and then perform SOC estimation and SOH estimation. The model parameters can be accurately estimated even when the environment changes. As the battery is used, the capacity and open-circuit voltage curve change. Accurate SOH estimation and dynamic open-circuit voltage curve identification can accurately estimate the SOC, and at the same time, accurate SOC can also achieve accurate SOH estimation.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for jointly estimating the state of charge and health state of a lithium battery, characterized in that: The method specifically comprises the following steps: S1: Perform HPPC test on new batteries and perform cycle aging test on batteries under cycle conditions; S2: Data analysis to extract the open circuit voltage curve, select the first-order equivalent circuit model to build the circuit, and write the state equation; S3: The multi-forgetting factor recursive least squares method is used for parameter identification, and the SOC estimator of the unscented Kalman filter based on the state equation and the observation equation is constructed for SOC estimation; where SOC represents the state of charge and SOH represents the state of health; S4: Construct a SOH estimator based on the extended Kalman filter and use the estimated SOC to estimate the SOH; S5: Dynamic identification of open circuit voltage curve using moving average method combined with recursive least squares method; S6: Couple the SOC estimator and the SOH estimator, and use the open circuit voltage curve after dynamic identification to realize the joint online estimation of SOC and SOH.

2. The method for jointly estimating the state of charge and health status of a lithium battery according to claim 1, characterized in that: Step S1 specifically includes: performing HPPC testing on a new lithium battery to obtain battery parameter data; performing cycle aging on the battery using a full charge and full discharge strategy to obtain data on each charge and discharge cycle of the battery, capacity aging data, current, voltage and capacity.

3. The method for jointly estimating the state of charge and health status of a lithium battery according to claim 1, characterized in that: Step S2 specifically includes the following steps: S21: According to the HPPC test experiment in step S1, the battery cell to be tested is fully charged in a constant current and constant voltage manner using a standard current, and is left to stand for a period of time to make it close to a balanced state. The battery cell is discharged at 10% SOC by loading a mixed pulse current excitation sequence. After the battery cell is left to stand to make it close to a balanced state, the excitation sequence is loaded again, and the whole process is repeated. The open circuit voltage values ​​corresponding to SOC of 100%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, and 0% are selected, and a sixth-order polynomial fitting is performed to obtain an open circuit voltage curve; S22: Modeling is based on the first-order equivalent circuit model of lithium battery, and the state equation is: U=U oc +V p +R0·I (1) Among them, U is the terminal voltage of the lithium-ion power battery, I represents the terminal current of the lithium-ion power battery, R0 represents the ohmic internal resistance of the lithium-ion power battery, and R p Represents the polarization resistance of lithium-ion battery, C p Represents the polarization capacitance of lithium-ion power batteries, V p Represents the polarization capacitance C of lithium-ion power battery p The terminal voltage, U oc Indicates open circuit voltage; After Laplace transform, the transfer function of battery impedance in S domain is obtained: Where s represents the complex variable of Laplace transform; after bilinear transformation Mapping to the Z plane: Among them, z represents the discretized complex frequency domain variable, a1, a2, a3 represent the coefficients in the discretization process; finally, the expressions of a1, a2, a3 are obtained: Among them, Δt represents the sampling time interval; thus, R0 and R P , C P The expression is: Assume that the power consumed or absorbed by the battery in a unit sampling interval has approximately zero effect on its state of charge, that is, U OC,k =U OC,k-1 , and the temperature and aging state remain unchanged, formula (1) is expressed as: And k =(1-a1)U OC,k +a1U k-1 +a2I k +a3I k-1 (6) Express formula (6) in matrix form: y k =φ k ·i k (7) Among them, y k is the terminal voltage value, φ k is the input matrix of the system, φ k =[1 U k-1 I k I k-1 ],θ k is the parameter matrix of the system, θ k =[(1-a1)U OC,k a1 a2 a3] T , U k-1 represents the terminal voltage of the lithium-ion power battery at time k-1, I k Indicates the current of the lithium-ion power battery at the current moment, I k-1 represents the current at the last moment, U OC,k Indicates the open circuit voltage at the current moment.

4. The method for jointly estimating the state of charge and health status of a lithium battery according to claim 3, characterized in that: Step S3 specifically includes the following steps: S31: Parameter identification using multi-forgetting factor recursive least squares method: According to the least squares method, we get: in, represents the parameter to be estimated, L(k) represents the gain matrix, and φ(k) represents the input matrix; The multi-forgetting factor recursive least squares method is used to identify the system parameters and obtain the criterion function: in, represents the loss function, n represents the order; λ k Represents the four forgetting factors corresponding to the four parameters, and satisfies 0<λ i ≤1, and get the new gain matrix L new (k): Among them, P i (k-1) represents the new covariance matrix corresponding to the i-th forgetting factor; P(k) is the new covariance matrix: P(k)=A -1 [I-L(k)φ T (k)P(k-1)]A -1 (11) Among them, A = diag[λ1,λ2,λ3,λ4], when the criterion function criterion represented by equation (9) is satisfied and the criterion is minimized, the expression is obtained: S32: Construct an SOC estimator of an unscented Kalman filter based on the state equation and the observation equation to estimate the SOC. The state space equation is obtained according to the ampere-hour integration method and combined with the first-order equivalent circuit model: Among them, the system parameter θ=[θ 1 ,θ 2 ,θ 3 ] Y , θ 2 =R P ,θ 3 =R0,S k+1 Indicates the current SOC value, S k Indicates the SOC value at the previous moment, C n Indicates the maximum available capacity of the battery at the current moment, T s represents the sampling time interval, U k Indicates the current, C n In the first cycle, the nominal capacity is used; y k Indicates the terminal voltage value, V oc (S k ) represents the open circuit voltage value corresponding to SOC, V k represents polarization voltage; Simplifying formula (13) yields: Among them, x k =[S k ,V k ] T , S k Indicates SOC, V k represents polarization voltage; g(x k ,u k )=V oc (S k )+V k , v k =θ 3 ·u k ; Initialize the state vector x k and the error covariance matrix Construct 2n+1 sigma points based on traceless change: Where λ = α 2 (n+h)-n is the scaling factor, where α represents the ratio coefficient and h represents the time step, and then the state is updated: Among them, x i,k|k-1 represents the state estimate of the i-th sigma point at time k, f(x i,k-1 ) represents the state transition function, represents the prior estimate of the state at time k, represents the prior covariance matrix of the state estimate at time k, Represents the contribution of each sigma point to the covariance matrix, represents the weight coefficient of each sigma point, q0 represents the initial moment deviation, and Q0 represents the process noise covariance matrix; Then perform the measurement update: Among them, x i,k|k-1 represents the state estimate of the i-th sigma point at time k, represents the measurement function, represents the initial input, represents the measured predicted value at time k, g(x i,k|k-1 ,u k ) represents the measurement function, r0 represents the error in the measurement process, represents the covariance matrix of the measurements, represents the covariance matrix between states and measurements, represents the prior state estimate at time k; Compute the Kalman gain and update the state estimate and error covariance: Among them, K k represents the Kalman gain, represents the estimated state value at time k, P k represents the error covariance matrix.

5. The method for jointly estimating the state of charge and health status of a lithium battery according to claim 4, characterized in that: Step S4 specifically includes: constructing a SOH estimator based on an extended Kalman filter, and SOH is defined as: Among them, Q n is the rated capacity, Q is the current maximum available capacity, the estimated SOC is used to estimate the SOH, and the capacity is used as the state variable to obtain the state space equation of the SOH: Among them, Q k is the capacity at the previous moment, Q k+1 is the capacity at the current moment, r k Represents the process error, V oc (S k ) represents the open circuit voltage value corresponding to SOC, V k represents polarization voltage; Initialize the state vector and covariance matrix: in, represents the initial value of the error covariance matrix; To make a status update: Q k =Q k-1 (26) Update the error covariance matrix: in, represents the error covariance matrix of capacity estimation, Q Q represents the process noise covariance matrix; Calculate the Kalman gain and update the state estimate and error covariance: in, represents the Kalman gain, represents the Jacobian matrix, L Q represents the measurement noise covariance matrix, g(x k-1 ,u k-1 ,Q k-1 ) represents the measurement equation.

6. The method for jointly estimating the state of charge and health status of a lithium battery according to claim 5, characterized in that: Step S5 specifically includes: using a moving average method combined with a recursive least squares method to perform segmented dynamic identification of the open circuit voltage curve, the identification process: Among them, e ocv (k) is the open circuit voltage curve error, i b (k) represents the current at time k, represents the polarization voltage corresponding to time k, L represents the order, a p represents the polynomial coefficient of the open circuit voltage curve, p represents the order of SOC, It represents the estimated SOC value at time k, and the recursive least squares method is used for parameter identification: in, represents the parameters to be estimated, represents the estimated value of the polynomial coefficient of the open circuit voltage curve; The new open circuit voltage expression is obtained: Among them, n p Indicates the order of SOC, a i The coefficient representing the SOC of order i, s i Represents the i-th power of SOC.

Images