Lithium ion battery SOC estimation method considering capacity attenuation factor

By combining the augmented extended Kalman filtering method and forgetting factor matrix, the model error problem caused by capacity attenuation in lithium-ion battery SOC estimation is solved, the accuracy and robustness of SOC estimation are improved, and the battery capacity attenuation error is achieved. Effective compensation and resistance to SOC initial value error.

CN119024170APending Publication Date: 2024-11-26JILIN UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411225008.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-03
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

When the existing SOC estimation method of lithium-ion batteries faces battery capacity attenuation, model errors lead to a decrease in SOC estimation accuracy. In particular, the Kalman filtering method cannot effectively compensate for the error caused by capacity attenuation after the number of battery charge and discharge increases.

Method used

Augmented extended Kalman filtering (AEKF) method is used to combine the model error caused by capacity attenuation as new state quantities and the original state quantities for synchronous estimation, and a forgetting factor matrix is ​​introduced, and a first-order RC equivalent circuit model is used to achieve a synchronous estimation. The time integral model establishes the battery state space equation, performs SOC estimation and error compensation.

Benefits of technology

It improves the accuracy and robustness of SOC estimation, can effectively compensate model errors under the attenuation of battery capacity, enhances the resistance to initial value errors, and improves the accuracy and real-time nature of SOC estimation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024170A_ABST
    Figure CN119024170A_ABST
Patent Text Reader

Abstract

The invention discloses a lithium ion battery SOC (State of Charge) estimation method considering a capacity attenuation factor, which is characterized by comprising the following steps of: establishing a battery state space equation according to a battery first-order RC equivalent circuit model and an ampere-hour integral model; taking a model error caused by the capacity attenuation factor as a newly added state quantity, combining the newly added state quantity with an original state quantity, and carrying out synchronous estimation to obtain a lithium ion battery SOC estimation state-space equation considering the capacity attenuation factor; and a battery state space equation is solved, and meanwhile, a forgetting factor matrix is introduced, so that the adaptability of the algorithm to SOC initial value deviation is improved. According to the method, the model error caused by capacity fading is added into the original state vector, and synchronous estimation and compensation of the model error are realized by using the augmented extended Kalman filtering, so that the SOC estimation precision is improved. Meanwhile, due to the introduction of the forgetting factor matrix, the robustness of the algorithm in practical application is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of power batteries, and in particular relates to a lithium-ion battery SOC estimation method taking capacity decay factors into consideration. Background Art

[0002] Lithium-ion batteries are widely used in electric vehicles. In order to ensure the safety and stability of lithium-ion batteries during operation, it is crucial to accurately estimate the battery state. The battery state of charge (SOC) is one of the most basic parameters in the battery management system, and accurate estimation of SOC is of great significance.

[0003] At present, the commonly used SOC online estimation methods mainly include the ampere-hour integration method, the data-driven method and the Kalman filter method.

[0004] The ampere-hour integration method is the simplest SOC online estimation method. This method is greatly affected by the SOC initial value error and current measurement error. Moreover, this method is an open-loop estimation method. As the running time increases, the accumulated error will cause the SOC estimation accuracy to gradually decrease.

[0005] The data-driven method ignores the internal reaction of the battery and establishes the mapping relationship between external measurable variables and SOC through training with a large amount of data. In order to ensure the accuracy of the SOC data in the training data, the data used for training this method is often obtained in the laboratory. However, the actual operation of electric vehicles depends on driving habits and environmental factors, and has strong random characteristics. Laboratory test data cannot cover all actual operating conditions. More importantly, since the battery is not modeled, in theory, the data-driven method cannot guarantee the effectiveness and accuracy of the output.

[0006] The Kalman filter method estimates the SOC based on the battery modeling. This method follows the prediction and correction principle, has a closed-loop nature, has a small amount of calculation, and can meet the real-time requirements of SOC estimation. One of the important prerequisites for the effectiveness of the Kalman filter method is the accuracy of the model. There is little research on this issue in the field of battery SOC estimation. In particular, as the number of battery charge and discharge times increases, its capacity will gradually decrease, resulting in increasingly obvious model errors, which will reduce the accuracy of SOC estimation.

[0007] Therefore, it is of great significance to study a lithium-ion battery SOC estimation method based on Kalman filtering that can estimate and compensate for the model error caused by battery capacity decay in real time. Summary of the invention

[0008] In order to solve the problem of estimating and compensating the model error caused by capacity decay, the present invention provides a lithium-ion battery SOC estimation method taking the capacity decay factor into consideration.

[0009] The present invention provides a lithium-ion battery SOC estimation method taking into account capacity decay factors, characterized in that it comprises the following steps:

[0010] Step 1: Establish the battery state space equation based on the battery first-order RC equivalent circuit model and the ampere-hour integral model:

[0011]

[0012] In the formula, k is a non-negative integer, which is used to represent the number of iteration steps; SOC k is the battery charge state; η represents the coulomb efficiency; Δt is the sampling period; C N Indicates the rated capacity of the battery; I B,k is the current; U P,k , R P and C P are polarization voltage, polarization resistance and polarization capacitance respectively; U t,k+1 is the terminal voltage; R 0 is the ohmic internal resistance; U OCV is the open circuit voltage; w 1,k and w 2,k represents process noise; v k+1 represents the measurement noise;

[0013] Step 2: Using the Augmented Extended Kalman Filter (AEKF), the model error caused by capacity decay is used as the difference between the new state quantity and the original state quantity [SOC k U P,k ] T Combined and estimated synchronously, the state space equation of lithium-ion battery SOC estimation considering capacity attenuation is obtained:

[0014]

[0015] in:

[0016]

[0017] In the formula, x k is the system state vector, b k is the model error caused by capacity decay, A k-1 is the system state transfer matrix, u k-1 is the system input vector, w k-1 is the process noise, y k is the system observation vector, C k is the system observation matrix, SOC k|k-1 represents the prior estimate of the battery state of charge from step k-1 to step k, z kis the linearized remainder of the observation equation;

[0018] Step 3: Solve the lithium-ion battery SOC estimation state space equation described in step 2 in a non-decoupling manner, and introduce a forgetting factor matrix;

[0019] Iteration process:

[0020] (1) When k = 0, initialize:

[0021]

[0022] In the formula, is the initial value of the posterior estimate of the state quantity; x 0 is the initial value of the true value of the state quantity; is the initial value of the posterior error covariance of the state quantity;

[0023] (2) When k = 1, 2, 3, ...:

[0024] Update the prior estimate of the state quantity:

[0025] In the formula, is the prior estimate of the state quantity, is the posterior estimate of the state quantity;

[0026] Update of prior error covariance of state quantity:

[0027] In the formula, is the prior error covariance of the state quantity, is the state quantity posterior error covariance, Q k-1 is the noise w k-1 covariance of k is the forgetting factor matrix;

[0028] Innovation value update:

[0029] In the formula, r x,k is the state quantity update;

[0030] Kalman gain:

[0031] In the formula, K x,k is the Kalman gain, R k is the noise v k The covariance of

[0032] Update the posterior estimate of the state quantity:

[0033] Update of the state quantity posterior error covariance:

[0034] Where I is the unit matrix;

[0035] Forgetting factor matrix:

[0036]

[0037] In the formula, It is the element in the forgetting factor matrix that changes adaptively based on the new information;

[0038] Theoretical value of state quantity innovation covariance:

[0039]

[0040] In the formula, is the theoretical value of the innovation covariance of the state quantity;

[0041] State quantity innovation covariance estimate:

[0042] In the formula, is the estimated value of the state quantity innovation covariance, r x,i is the state quantity update;

[0043] renew:

[0044]

[0045] During the iteration process, due to the transfer matrix A k Due to the existence of , the posterior estimation result of the model error caused by capacity decay in the kth step will be substituted into the prior estimation calculation of SOC in the k+1th step, that is, the estimation and compensation of the model error caused by capacity decay are realized while SOC is estimated, thereby improving the accuracy of SOC estimation.

[0046] The present invention has the following beneficial effects:

[0047] The present invention adds the model error caused by capacity decay to the original state vector, and uses the augmented extended Kalman filter to achieve synchronous estimation and compensation of the model error, thereby improving the accuracy of SOC estimation. At the same time, the introduction of the forgetting factor matrix improves the robustness of the method described in the present invention in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flow chart of a lithium-ion battery SOC estimation method considering capacity decay factors of the present invention;

[0049] Figure 2 It is the SOC estimation error of the EKF and AEKF algorithms in the embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention will be further described below in conjunction with specific examples. The examples are used to explain the present invention but are not intended to limit the present invention.

[0051] like Figure 1 As shown, the present invention provides a lithium-ion battery SOC estimation method considering capacity decay factors, comprising the following steps:

[0052] Step 1: Establish the battery state space equation based on the battery first-order RC equivalent circuit model and the ampere-hour integral model:

[0053]

[0054] In the formula, k is a non-negative integer, which is used to represent the number of algorithm iterations; SOC k is the battery charge state; η represents the coulomb efficiency; Δt is the sampling period; C N Indicates the rated capacity of the battery; I B,k is the current; U P,k , R P and C P are polarization voltage, polarization resistance and polarization capacitance respectively; U t,k+1 is the terminal voltage; R 0 is the ohmic internal resistance; U OCV is the open circuit voltage; w 1,k and w 2,k represents process noise; v k+1 represents the measurement noise.

[0055] In this embodiment, R 0 =0.0091Ω, R P =0.0134Ω, C P =3000F,U OCV The fitting formula is:

[0056] U OCV (SOC k+1 )=83.212×SOC k 7 -302.18×SOC k 6 +445.2×SOC k 5 -341.85×SOC k 4 +

[0057] 146.86×SOC k 3 -35.578×SOC k 2 +4.8137×SOC k +2.9606;

[0058] Step 2: Using the Augmented Extended Kalman Filter (AEKF), the model error caused by capacity decay is used as the difference between the new state quantity and the original state quantity [SOC k U P,k ] T Combined and estimated synchronously, the state space equation of lithium-ion battery SOC estimation considering capacity attenuation is obtained:

[0059]

[0060] in:

[0061]

[0062] In the formula, x k is the system state vector, b k is the model error caused by capacity decay, A k-1 is the system state transfer matrix, u k-1 is the system input vector, w k-1 is the process noise, y k is the system observation vector, C k is the system observation matrix, SOC k|k-1 represents the prior estimate of the battery state of charge from step k-1 to step k, z k is the linearized remainder of the observation equation;

[0063] Step 3: Solve the lithium-ion battery SOC estimation state space equation described in step 2 in a non-decoupled manner, and introduce a forgetting factor matrix to improve the algorithm's adaptability to SOC initial value deviations.

[0064] Iteration process:

[0065] (1) When k = 0, initialize:

[0066]

[0067] In the formula, is the initial value of the posterior estimate of the state quantity; x 0 is the initial value of the true value of the state quantity; is the initial value of the posterior error covariance of the state quantity;

[0068] (2) When k = 1, 2, 3, ...:

[0069] Update the prior estimate of the state quantity:

[0070] In the formula, is the prior estimate of the state quantity, is the posterior estimate of the state quantity;

[0071] Update of prior error covariance of state quantity:

[0072] In the formula, is the prior error covariance of the state quantity, is the state quantity posterior error covariance, Q k-1 is the noise w k-1 covariance of k is the forgetting factor matrix;

[0073] Innovation value update:

[0074] In the formula, r x,k is the state quantity update;

[0075] Kalman gain:

[0076] In the formula, K x,k is the Kalman gain, R k is the noise v k The covariance of

[0077] Update the posterior estimate of the state quantity:

[0078] Update of the state quantity posterior error covariance:

[0079] Where I is the unit matrix;

[0080] Forgetting factor matrix:

[0081]

[0082] In the formula, It is the element in the forgetting factor matrix that changes adaptively based on the new information.

[0083] Theoretical value of state quantity innovation covariance:

[0084]

[0085] In the formula, is the theoretical value of the innovation covariance of the state quantity;

[0086] State quantity innovation covariance estimate:

[0087] In the formula, is the estimated value of the state quantity innovation covariance, r x,i is the state quantity update;

[0088] renew:

[0089]

[0090] During the iteration process, due to the transfer matrix A k Due to the existence of , the posterior estimation result of the model error caused by capacity decay in the kth step will be substituted into the prior estimation calculation of SOC in the k+1th step, that is, the estimation and compensation of the model error caused by capacity decay are realized while SOC is estimated, thereby improving the accuracy of SOC estimation.

[0091] like Figure 2 As shown, the effectiveness of the method of the present invention is verified by simulation. Assuming that the battery capacity has decayed by 1Ah during use, the actual initial value of SOC is 0.9. In order to examine the resistance of the method of the present invention to the initial value error, the initial value of SOC when the algorithm is started is set to 1. The battery is discharged according to the NYCC (New York City Cycle) cycle condition until it is in an empty state. The SOC estimation error of the method of the present invention (abbreviated as AEKF algorithm) is compared with the commonly used SOC estimation algorithm based on EKF (Extended Kalman Filter), as shown in FIG. Figure 2 As shown. Figure 2 It can be seen that in the initial stage of discharge, the SOC estimation error of the AEKF algorithm can converge to near 0 more quickly, indicating that the introduction of the forgetting factor improves the response performance of the AEKF algorithm and its ability to resist the interference of initial value errors. Figure 2 It can also be seen that during the entire discharge process, the SOC estimation error of the AEKF algorithm is always smaller than that of the EKF algorithm, indicating that the estimation and compensation of the model error caused by capacity decay effectively improves the estimation accuracy of SOC.

[0092] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements can be made without departing from the principle of the present invention, and these improvements should also be considered as the protection scope of the present invention.

Claims

1. A lithium-ion battery SOC estimation method considering capacity decay factors, characterized in that: The steps include: Step 1: Establish the battery state space equation based on the battery first-order RC equivalent circuit model and the ampere-hour integral model: In the formula, k is a non-negative integer, which is used to represent the number of iteration steps; SOC k is the battery state of charge; η represents the coulomb efficiency; Δt is the sampling period; C N Indicates the rated capacity of the battery; I B,k is the current; U P,k , R P and C P are polarization voltage, polarization resistance and polarization capacitance respectively; U t,k+1 is the terminal voltage; R0 is the ohmic internal resistance; U OCV is the open circuit voltage; w 1,k and w 2,k represents process noise; v k+1 represents the measurement noise; Step 2: Using the augmented extended Kalman filter AEKF, the model error caused by capacity decay is used as the new state quantity and the original state quantity [SOC k U P,k ] T Combined and estimated synchronously, the state space equation of lithium-ion battery SOC estimation considering capacity attenuation is obtained: in: In the formula, x k is the system state vector, b k is the model error caused by capacity decay, A k-1 is the system state transfer matrix, u k-1 is the system input vector, w k-1 is the process noise, y k is the system observation vector, C k is the system observation matrix, SOC k|k-1 represents the prior estimate of the battery state of charge from step k-1 to step k, z k is the linearized remainder of the observation equation; Step 3: Solve the lithium-ion battery SOC estimation state space equation described in step 2 in a non-decoupling manner, and introduce a forgetting factor matrix; Iteration process: (1) When k = 0, initialize: In the formula, is the initial value of the posterior estimate of the state quantity; x0 is the initial value of the true value of the state quantity; is the initial value of the posterior error covariance of the state quantity; (2) When k = 1, 2, 3, ...: Update the prior estimate of the state quantity: In the formula, is the prior estimate of the state quantity, is the posterior estimate of the state quantity; Update of prior error covariance of state quantity: In the formula, is the prior error covariance of the state quantity, is the state quantity posterior error covariance, Q k-1 is the noise w k-1 The covariance of k is the forgetting factor matrix; Innovation value update: In the formula, r x,k is the state quantity update; Kalman gain: In the formula, K x,k is the Kalman gain, R k is the noise v k The covariance of Update the posterior estimate of the state quantity: Update of the state quantity posterior error covariance: Where I is the unit matrix; Forgetting factor matrix: In the formula, It is the element in the forgetting factor matrix that changes adaptively based on the new information; the theoretical value of the state quantity new information covariance is: In the formula, is the theoretical value of the innovation covariance of the state quantity; State quantity innovation covariance estimate: In the formula, is the estimated value of the state quantity innovation covariance, r x,i is the state quantity update; renew: