Lithium battery health state estimation model and estimation method thereof

Abstract

The invention discloses a lithium battery health state estimation model and an estimation method thereof, and the method comprises the steps: firstly, building a lithium battery health state estimation model, specifically # imgabs0 #, then substituting a battery constant-current charging voltage and corresponding charging time into the lithium battery health state estimation model, and carrying out the lithium battery health state estimation through a curve fitting method; according to the method, a relational expression between constant-current charging voltage of the battery and corresponding charging time is established, a lithium battery health state estimation model is obtained, voltage and charging time data of the battery in the constant-current charging process are collected, a relation curve is fitted by using a curve fitting method, and finally, the estimated value of the SOH is obtained by analyzing the characteristics of the fitted curve. According to the method, the estimation precision and stability can be improved, and experimental results show that the model has a good SOH estimation effect, the overall operation time is short, the maximum estimation relative error is small, and the root mean square error is small; the method is suitable for various lithium battery types, and a good SOH estimation effect is achieved; the method has the advantages of simplicity, small calculation amount, high SOH estimation accuracy and the like, and is suitable for practical engineering application.

Description

Technical Field

[0001] The present invention belongs to the field of new energy and energy-saving technology, and in particular relates to a lithium battery health status estimation model and an estimation method thereof. Background Art

[0002] The estimation of the state of health (SOH) of lithium-ion batteries is crucial to the safe and stable operation of batteries. Accurate SOH estimation can provide a data basis for the estimation of the state of charge (SOC) of the battery, and can also provide important guidance for the subsequent retirement and recycling of lithium-ion batteries. SOH can be defined by the internal resistance or capacity of the lithium-ion battery. Among them, the internal resistance of the battery will be affected by many factors. The internal resistance of batteries in different health states will be affected by the ambient temperature, discharge rate and SOC. The SOH defined by internal resistance requires that the internal resistance should be obtained under uniform conditions so that the calculated SOH is comparable, but this is difficult to achieve in actual engineering. In contrast, the SOH defined by capacity is more stable and easier to implement. Therefore, the present invention studies the SOH defined by capacity. In order to solve the problem that the state estimation algorithm in the battery management system requires high algorithm accuracy and small amount of calculation in actual vehicle applications, an improved battery state of health (SOH) estimation method based on the equivalent circuit model has been developed. The main technical problems to be solved are:

[0003] (1) Estimating battery state of health (SOH) using equivalent circuit model and curve fitting method:

[0004] This method first establishes a relationship between the battery's constant-current charging voltage and the corresponding charging time, using the parameters of the equivalent circuit model and the estimated SOH as undetermined coefficients. Specifically, the equivalent circuit model includes parameters such as the battery's internal resistance and capacitance, which affect the change in voltage during the charging process. Then, by collecting data on the battery's voltage and charging time during constant-current charging, a curve fitting method is used to construct the relationship. Finally, by analyzing the characteristics of the fitted curve, the estimated SOH value is obtained.

[0005] (2) Improving estimation accuracy and stability: The performance of the proposed model is verified based on the cycle life experimental data of 18650 lithium iron phosphate batteries. The experimental results show that the model has a good SOH estimation effect, with a total running time of 51.0739 seconds, a maximum estimation relative error of 0.011359, and a root mean square error of 0.00021236.

[0006] (3) Adaptability to various lithium battery types: The model was validated using publicly available battery experimental data from Tongji University, and similarly achieved good SOH estimation results. This model has the advantages of simplicity, low computational complexity, and high SOH estimation accuracy, making it suitable for practical engineering applications.

[0007] Currently, mainstream SOH estimation methods include data-driven methods, electrochemical models, and equivalent circuit models. While data-driven methods offer higher accuracy, they require a large amount of data to train the model, making them incapable of estimating SOH over the entire lifecycle of lithium-ion batteries. Furthermore, data-driven models suffer from weak generalization, poor interpretability, and high computational complexity, making them difficult to apply in practical engineering. While electrochemical models can delve deeper into the internal mechanisms of lithium-ion batteries and achieve high SOH estimation accuracy, their high computational complexity and high experimental equipment costs often limit them to theoretical laboratory research and hinder practical application. State of health estimation for lithium-ion batteries based on charging curves, Journal of Power Sources 249 (2014) 457-462, establishes a model for the relationship between SOH and constant-current charging voltage based on an equivalent circuit model. Using a simple curve fitting method to fit the model to the constant-current charging voltage of the battery, the SOH estimate is obtained. This model offers clear physical meaning, a low computational complexity, and high SOH estimation accuracy, making it suitable for practical engineering applications. However, this algorithm does not take into account the problem that the constant current charging voltage platform of lithium iron phosphate batteries is difficult to fit, and the data used to verify the model are all complete charging voltage data. Summary of the Invention

[0008] In order to overcome the defects of the above-mentioned prior art, the purpose of the present invention is to provide a lithium battery health status estimation model and an estimation method thereof. By adding a segmentation point in each constant current charging voltage curve, only the curve to the right of the segmentation point is used for model fitting, and the curve to the left of the segmentation point is discarded. The battery SOH estimation can be performed using the segmented charging data, which simplifies the complexity of the model and improves the calculation efficiency and accuracy.

[0009] In order to achieve the above object, the technical solution adopted by the present invention is:

[0010] A lithium battery health status estimation model, specifically:

[0011]

[0012] Among them, I is the constant current charging current of the battery, U' is the terminal voltage of the battery in the nth cycle, t' is the constant current charging time of the battery in the nth cycle, R0 is the ohmic internal resistance of the new battery equivalent circuit model, R1 and R2 are the polarization internal resistances of the new battery equivalent circuit model, τ1 and τ2 are the time constants of the new battery equivalent circuit model, τ1=R1C1, τ2=R2C2, C1 and C2 are the polarization capacitances of the new battery equivalent circuit model, k is a parameter related to the battery health status, k=1 / S OH .

[0013] The lithium battery health status estimation model meets two conditions:

[0014] (1) The RC parameters of the battery remain unchanged during the same charge and discharge cycle;

[0015] (2) During different charge and discharge cycles, the relationship between the open circuit voltage (OCV) and the SOC of the battery remains unchanged.

[0016] A method for estimating the health status of a lithium battery comprises the following steps:

[0017] Step 1: Establish a lithium battery health status estimation model;

[0018] Step 2: Substitute the battery constant current charging voltage and the corresponding charging time into the model established in step 1, and estimate the health status of the lithium battery by curve fitting method.

[0019] The specific method of step 1 is:

[0020] According to the second-order equivalent circuit model of the battery, the voltage data and corresponding time data of the battery in the constant current charging stage are substituted into the lithium battery health status estimation model, that is, formula (7), and fitted to obtain the undetermined parameters k : During the constant current charging process, the battery terminal voltage U is regarded as a zero-state response; in the first charging cycle, the relationship between the battery terminal voltage U and the charging time t is calculated by formula (1):

[0021]

[0022] Where U OC (S OC ) is the open circuit voltage of the battery, I is the constant current charging current; τ1 and τ2 are the time constants of the two RC branches, where τ1 = R1C1, τ2 = R2C2, R1 and R2 are the polarization internal resistances of the two RC branches, C1 and C2 are the polarization capacitances of the two RC branches; R0 is the ohmic internal resistance; S OC is the battery SOC, S OC The relationship with charging time t is:

[0023]

[0024] Where C ap is the actual capacity of the first cycle; the battery terminal voltage of the nth (n>1) charging cycle is calculated as follows:

[0025]

[0026] Where t' is the constant current charging time of the battery in the nth cycle, R'0, R'1, R'2, τ'1 and τ'2 are the parameters of the battery in the nth cycle, S OC The relationship with charging time t' is:

[0027]

[0028] Where C' ap is the actual capacity of the nth cycle; subtracting equation (3) from equation (1) yields:

[0029]

[0030] From formula (2) and formula (4), we can get:

[0031]

[0032] Substitute equation (6) into equation (5) and let k = C ap / C' ap ,have to:

[0033]

[0034] Formula (7) is the final lithium battery health status estimation model; according to the calculation formula of k, k = C ap / C' ap , we know that k and the health status of the battery are reciprocals of each other, that is, k = 1 / S OH .

[0035] The specific method of step 2 is:

[0036] 2.1 Segment processing of constant current charging voltage data

[0037] The SOC value at the starting position of the battery voltage platform area is set, and the constant current charging time corresponding to the SOC value is used as the segmentation point of the constant current charging voltage. The constant current charging voltage is divided into two segments, namely, the constant current charging voltage corresponding to the SOC segmentation point less than the constant current charging voltage corresponding to the SOC segmentation point and the constant current charging voltage corresponding to the SOC segmentation point greater than the SOC segmentation point. The data of the constant current charging voltage corresponding to the SOC segmentation point less than the SOC segmentation point is discarded, and only the constant current charging voltage data corresponding to the SOC segmentation point greater than the SOC segmentation point is used to fit the lithium battery health status estimation model.

[0038] 2.2 For the constant current charging voltage data corresponding to the SOC segmentation point, the measured battery terminal voltage and the corresponding time are substituted into the lithium battery health status estimation model obtained in step 1, that is, formula (7), by the curve fitting method, and the parameter k is obtained by fitting. Then, the relationship between SOH and k is S OH =1 / k to calculate the battery health state SOH.

[0039] In step 2.2, the measured battery terminal voltage and the corresponding time are substituted into the lithium battery health status estimation model obtained in step 1, that is, formula (7) for fitting. In formula (7), R0, R1, R2, τ1 and τ2 are new battery parameters, which are measured by hybrid pulse power (HPPC) experiments.

[0040] The beneficial effects of the present invention are:

[0041] 1. This invention adds SOC segmentation points to the constant current charging voltage curve, discards curves with constant current charging voltages below the SOC segmentation points, and only uses curves with constant current charging voltages above the SOC segmentation points for model fitting. By using only fragmented charging data for SOH estimation, the problem of high data integrity requirements is resolved, making the model more flexible and adaptable in practical applications. This simplifies the model complexity and improves computational efficiency.

[0042] 2. Using a curve fitting method, the segmented constant current charging voltage data and the corresponding time are substituted into the final lithium-ion battery SOH estimation model for fitting. Compared with data-driven methods, the fitting method has a smaller computational load.

[0043] 3. The optimization model is based on the characteristic that lithium iron phosphate batteries have a large voltage platform area. Only part of the constant current charging voltage data is used to estimate the battery SOH, making it more suitable for practical engineering applications. It solves the problem that traditional SOH estimation models have difficulty in handling charging voltage platforms.

[0044] In summary, the model of the present invention is simple and improves the calculation efficiency; with the help of the curve fitting method, the calculation amount is less and it has the advantages of rapidity, stability, adaptability and high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is the second-order equivalent circuit model of the battery;

[0046] Figure 2 This is the constant current charging voltage curve of the LP1 battery in a partial cycle;

[0047] Figure 3 This is the right section of the constant current charging voltage curve of the LP1 battery in a partial cycle;

[0048] Figure 4 shows the relationship between the second-order equivalent circuit model parameters of the LP1 battery and the SOC under different health conditions, among which Figure 4(a) shows the change of parameter R0 with SOC; Figure 4(b) shows the change of parameter R1 with SOC; Figure 4(c) shows the change of parameter R2 with SOC; Figure 4(d) shows the change of parameter C1 with SOC, and Figure 4(e) shows the change of parameter C2 with SOC;

[0049] Figure 5 This is the OCV-SOC relationship of LP1 battery in different health states;

[0050] Figure 6 shows the SOH estimation results and relative errors of battery LP1. Figure 6(a) shows the SOH estimation results of battery LP1. Figure 6(b) shows the relative error of SOH estimation of battery LP1.

[0051] Figure 7 shows the SOH estimation results and relative errors of battery N11. Figure 7(a) shows the SOH estimation results of battery N11. Figure 7(b) shows the relative errors of SOH estimation of battery N11.

[0052] Figure 8 SOH estimation results and relative errors of battery N12. Figure 8(a) is the SOH estimation result of battery N12; Figure 8(b) is the relative error of SOH estimation of battery N12. DETAILED DESCRIPTION

[0053] The present invention will be described in further detail below with reference to the accompanying drawings.

[0054] A lithium battery health status estimation model, specifically:

[0055]

[0056] Among them, I is the constant current charging current of the battery, U' is the terminal voltage of the battery in the nth cycle, t' is the constant current charging time of the battery in the nth cycle, R0 is the ohmic internal resistance of the new battery equivalent circuit model, R1 and R2 are the polarization internal resistances of the new battery equivalent circuit model, τ1 and τ2 are the time constants of the new battery equivalent circuit model, τ1=R1C1, τ2=R2C2, C1 and C2 are the polarization capacitances of the new battery equivalent circuit model, k is a parameter related to the battery health status, k=1 / S OH .

[0057] A method for estimating the health status of a lithium battery comprises the following steps:

[0058] Step 1: Establish a lithium battery health status estimation model; the specific method is:

[0059] The health status estimation model of lithium-ion battery cells is derived based on the second-order equivalent circuit model of the battery. Figure 1 shown.

[0060] Figure 1 In the figure, R0 is the ohmic internal resistance, R1 and R2 are the polarization internal resistances, and C1 and C2 are the polarization capacitances. I is the current excitation applied to the battery, U is the terminal voltage response of the battery, and U OC Represents the open circuit voltage (OCV) of the battery, and U1 and U2 are the voltages of the two RC parallel branches respectively.

[0061] In order to make the SOH estimation model applicable to engineering practice, the application object should meet two conditions:

[0062] (1) The RC parameters of the battery remain unchanged during the same charge and discharge cycle;

[0063] (2) The relationship between the open circuit voltage (OCV) and SOC of the battery remains unchanged during different charge and discharge cycles.

[0064] According to the second-order equivalent circuit model of the battery, the voltage data and corresponding time data of the battery in the constant current charging stage are substituted into the lithium battery health status estimation model, that is, formula (7), and fitted to obtain the undetermined parameters k : During the constant current charging process, the battery terminal voltage U is regarded as a zero-state response; in the first charging cycle, the relationship between the battery terminal voltage U and the charging time t is calculated by formula (1):

[0065]

[0066] Where U OC (S OC ) is the open circuit voltage of the battery, I is the constant current charging current; τ1 and τ2 are the time constants of the two RC branches, where τ1 = R1C1, τ2 = R2C2, τ2 = R2C2, R1 and R2 are the polarization internal resistances of the two RC branches, and C1 and C2 are the polarization capacitances of the two RC branches; S OC is the battery SOC, S OC The relationship with charging time t is:

[0067]

[0068] Where C ap is the actual capacity of the first cycle; the battery terminal voltage of the nth (n>1) charging cycle is calculated as follows:

[0069]

[0070] Where t' is the constant current charging time of the battery in the nth cycle, R'0, R'1, R'2, τ'1 and τ'2 are the parameters of the battery in the nth cycle, S OC The relationship with charging time t' is:

[0071]

[0072] Where C' ap is the actual capacity of the nth cycle; subtracting equation (3) from equation (1) yields:

[0073]

[0074] From formula (2) and formula (4), we can get:

[0075]

[0076] Substitute equation (6) into equation (5) and let k = C ap / C' ap ,have to:

[0077]

[0078] Formula (7) is the final lithium battery health status estimation model; according to the calculation formula of k, k = C ap / C' ap , we know that k and the health status of the battery are reciprocals of each other, that is, k = 1 / S OH .

[0079] Therefore, once k is obtained through fitting, the estimated SOH value can be calculated. Theoretically, k can be obtained by substituting the second-order equivalent circuit model parameters of the battery at different SOHs and the charging voltage during the constant current phase of the cycle into the above equation for fitting. However, when using this model to estimate the SOH of batteries in engineering, many problems arise, including:

[0080] (1) The constant current charging voltage of the battery is steep at both ends and relatively flat in the middle, which makes it difficult to fit such a curve;

[0081] (2) When the SOC is high or low, the relationship between the battery's OCV and SOC changes greatly with the SOH, but the model assumes that this relationship remains unchanged, and in formula (5), U OC eliminated;

[0082] (3) In actual use, it is almost impossible for the battery to be completely discharged before charging, but the model theoretically requires obtaining the complete constant current charging voltage data of the battery;

[0083] (4) The actual working conditions of electric vehicles are complex. Although the battery parameters can be measured at any time by applying pulses to the battery, the accuracy is low, which is not conducive to the application of the SOH estimation model.

[0084] Step 2: Substitute the battery constant current charging voltage and the corresponding charging time into the model established in step 1, and estimate the health status of the lithium battery by curve fitting method; the specific method is:

[0085] 2.1 To solve the above problems, the constant current charging voltage data is processed in segments

[0086] The SOC value at the starting position of the battery voltage platform area is set, and the constant current charging time corresponding to the SOC value is used as the segmentation point of the constant current charging voltage. The constant current charging voltage is divided into two segments, namely, the constant current charging voltage corresponding to the SOC segmentation point less than the constant current charging voltage corresponding to the SOC segmentation point and the constant current charging voltage corresponding to the SOC segmentation point greater than the SOC segmentation point. The data of the constant current charging voltage corresponding to the SOC segmentation point less than the SOC segmentation point is discarded, and only the constant current charging voltage data corresponding to the SOC segmentation point greater than the SOC segmentation point is used to fit the lithium battery health status estimation model.

[0087] For example, if the SOC is 0.25, the charging time corresponding to this SOC is used as the segmentation point for the constant current charging voltage, dividing it into two segments. Finally, the data to the left of this point is discarded, and only the voltage data to the right of this point is used to fit the lithium battery health status estimation model.

[0088] The reason for processing the data in this way is that no matter which battery it is, at different SOHs, as long as the starting point of the data segmentation is roughly near the same SOC, it is equivalent to reducing the actual capacity of the battery by the same proportion in each cycle. Therefore, in theory, the SOH estimation values ​​obtained by fitting the segmented voltage data and the complete voltage data are consistent within a certain error range.

[0089] After segmenting the data, the curve to the right of the segmentation point becomes easier to fit, and SOH estimation can be performed using partial constant-current charging data without having to charge the battery from a completely empty state. In addition, by deleting the low-SOC portion of the constant-current charging data for all cycles, the problem of the battery's OCV-SOC relationship varying significantly at low SOC is solved. However, the problem of the OCV-SOC relationship varying with SOC at high SOC cannot be solved, which will cause errors in SOH estimation.

[0090] 2.2 For the constant current charging voltage data corresponding to the SOC segmentation point, the measured battery terminal voltage and the corresponding time are substituted into the lithium battery health state estimation model obtained in step 1, that is, formula (7), by the curve fitting method, and then the parameter k is obtained by fitting. Then, the relationship between SOH and k is SOH =1 / k to calculate the battery health state SOH.

[0091] Regardless of which battery it is, at different SOHs, as long as the starting point of the data segmentation is roughly near the same SOC, it is equivalent to reducing the actual capacity of the battery by the same proportion in each cycle. Therefore, in theory, the SOH estimation values ​​obtained by fitting the segmented voltage data and the complete voltage data are consistent within a certain error range.

[0092] After segmenting the data, the curve to the right of the segmentation point becomes easier to fit, and SOH estimation can be performed using partial constant-current charging data without having to charge the battery from a completely empty state. In addition, by deleting the low SOC portion of the constant-current charging data for all cycles, the problem of large differences in the OCV-SOC relationship at low SOC for batteries at different SOH levels is solved.

[0093] Since the battery parameters change slowly during the accelerated aging process, only the new battery parameters, namely R0, R1, R2, τ1 and τ2, need to be measured through the hybrid pulse power (HPPC) experiment. In the actual operation of electric vehicles, there is no need to measure battery parameters such as R'0, R'1, R'2, τ'1 and τ'2. These parameters can be obtained by fitting the lithium battery health status estimation model.

[0094] Simulation experiment

[0095] The battery used in the experiment is 18650 lithium iron phosphate battery. The specific parameters are shown in Table 1.

[0096] Table 1 Parameters of the batteries used in the test

[0097]

[0098]

[0099] (1) Cyclic experimental data acquisition

[0100] The cycle life of a power battery is related to factors such as ambient temperature, battery charge / discharge depth, and charge / discharge rate. Due to limited experimental time, only the relationship with discharge rate was studied. The relevant experimental conditions are shown in Table 2. The batteries involved in the experiment were identical except for the discharge rate. One battery was used for each discharge rate cycle test: the battery discharged at 5C was designated LP1, the battery discharged at 10C was designated LP2, and the battery discharged at 15C was designated LP3.

[0101] Table 2 Cyclic test experimental conditions

[0102]

[0103] The experimental process of the cycle test is as follows:

[0104] (1) Perform battery performance tests, including capacity tests and HPPC tests;

[0105] (2) Charge the battery at a constant current of 5C (5.5A) to 3.65V, then charge at a constant voltage of 3.65V until the current decreases to 0.04C, and let it rest for 1 hour.

[0106] (3) Discharge the battery to 2.5V at a constant current of 5C (or 10C and 15C) and let it stand for 1 hour;

[0107] (4) Repeat (2)-(3) 18 times;

[0108] (5) Repeat (1)-(4) until the battery capacity decays to 80% of the initial capacity, and then end the test experiment.

[0109] Taking LP1 as an example, draw the constant current charging voltage curves of the 1st, 300th, 500th, 700th and 900th cycles, as shown in Figure 1. Figure 2 shown.

[0110] Depend on Figure 2 As can be seen, as the battery ages, the duration of its constant current charging phase becomes shorter and shorter. When the charging time is between 200 and 400 seconds, as the battery SOH increases, the voltage corresponding to the same time shows a pattern of first accelerating to drop and then slowly recovering.

[0111] (2) Constant current charging voltage data segmentation processing

[0112] Since the experiment uses lithium iron phosphate batteries, their charging voltage shows a characteristic of changing steeply at both ends and gently in the middle with the charging time. In order to apply the proposed model to the SOH estimation of lithium iron phosphate batteries, the original experimental data needs to be further processed. Here, a segmented method is adopted to divide the voltage data of the constant current charging stage of the lithium-ion battery in each cycle into two segments. The method for determining the segmentation point is shown in Part 1. Take the segmented SOC as 0.25, and draw the constant current charging voltage curve of LP1 for the 1st, 300th, 500th, 700th and 900th cycles, the part on the right side of the SOC of 0.25, as shown in Figure 1. Figure 3 As shown in the figure, the constant current charging curves of the battery at different aging stages are significantly different, which is also the basis for fitting these curves to obtain the battery SOH estimation value.

[0113] (3) Offline parameter identification of the battery second-order equivalent circuit model

[0114] Taking LP1 battery as an example, based on the data of HPPC test, the model parameters are identified using the offline parameter identification method. The results are shown in Figures 4 and Figure 5 Figure 4 shows how the equivalent circuit model parameters of the battery change with SOC under different health conditions. Figure 5 It shows the change of OCV with SOC in different health states of the battery.

[0115] From Figure 4 and Figure 5 As can be seen, the battery RC parameters and OCV vary slightly with SOC, demonstrating that the two model requirements are reasonable. Since the battery RC parameters also change relatively slowly with SOH, in actual electric vehicle operation, there is no need to measure battery parameters. Simply updating these parameters through the SOH estimation model is sufficient to accommodate these slow changes, and this approach will also facilitate the model's application in engineering.

[0116] Model validation and analysis

[0117] To verify the accuracy of the proposed battery SOH estimation model, we used the experimental data of the LP1 battery and the cycling test data of two batteries, NCM_CY45_05_1_#11 (denoted as N11) and NCM_CY45_05_1_#12 (denoted as N12), from the Tongji University lithium-ion battery public dataset to validate the model. The basic parameters of these two batteries are shown in Table 3.

[0118] Table 3 Basic parameters of two NCM batteries used for model verification

[0119]

[0120]

[0121] The model is evaluated using three performance metrics: runtime, relative error (RE), and root mean square error (RMSE). The runtime is the total time required to complete the SOH estimation for all battery cycles. The relative error and RMSE are calculated using Equations (8) and (9), respectively.

[0122]

[0123] Where i is the battery cycle test sequence, E RE is the relative error of the SOH estimation of the i-th cycle, S OH_real,i is the actual value of SOH in the i-th cycle, S OH_eva,i is the estimated SOH value of the i-th cycle, E RMSEis the root mean square error of SOH estimation, and n is the number of battery charge and discharge cycles.

[0124] Figures 6, 7, and 8 show the estimated SOH values ​​of the three batteries and their relative errors, respectively.

[0125] The comparison of the algorithm computational complexity, the maximum relative error and the root mean square error of the estimation results is shown in Table 4.

[0126] Table 4 Performance evaluation of SOH estimation based on three batteries

[0127]

[0128] Table 4 shows that when the model is used to estimate the SOH of these three batteries, the maximum runtime is 51.0739 seconds, indicating that the model runs quickly. The maximum relative error represents the stability of the model. The maximum relative error of the SOH estimation for the three batteries is 3.4314%. Compared with data-driven models, model stability is a significant advantage of this model. The root mean square error of the SOH estimation for all three batteries is particularly small, which is sufficient to demonstrate that the model achieves good SOH estimation results for all three batteries. This verifies the rapidity, stability, adaptability, and accuracy of the established SOH estimation model.

[0129] In summary, it can be seen from Table 4 and Figures 6 to 8 that, compared with the prior art, the present invention adopts a fitting method that can directly calculate the SOH of lithium batteries without being affected by the consistency of battery characteristics, and has the characteristics of fast calculation and high overall estimation accuracy; Figure 1 and Figure 2 It can be seen that the data segmentation method adopted in the present invention can solve the problem that the constant current charging voltage platform area of ​​lithium iron phosphate batteries is difficult to fit. The segmented data is easy to obtain in actual work, and has the characteristics of strong practicality, convenient operation and wide adaptability.

Claims

1. A lithium battery health status estimation model, characterized in that: Specifically: Wherein, I is the constant current charging current of the battery, U' is the terminal voltage of the battery in the nth cycle, t' is the constant current charging time of the battery in the nth cycle, R0 is the ohmic internal resistance of the new battery equivalent circuit model, R1 and R2 are both polarization internal resistances of the new battery equivalent circuit model, τ1 and τ2 are both time constants of the new battery equivalent circuit model, τ1=R1C1, τ2=R2C2, C1 and C2 are both polarization capacitors of the new battery equivalent circuit model, k is a parameter related to the battery health status, k=1 / S OH .

2. A lithium battery health status estimation model according to claim 1, characterized in that: The lithium battery health status estimation model meets two conditions: (1) The RC parameters of the battery remain unchanged during the same charge and discharge cycle; (2) During different charge and discharge cycles, the relationship between the open circuit voltage (OCV) and SOC of the battery remains unchanged.

3. A method for estimating the health status of a lithium battery, characterized in that: The specific steps include: Step 1, establishing a lithium battery health status estimation model; Step 2: Substitute the battery constant current charging voltage and the corresponding charging time into the model established in step 1, and estimate the health status of the lithium battery by curve fitting method.

4. A method for estimating the health status of a lithium battery according to claim 3, characterized in that: The specific method of step 1 is: According to the second-order equivalent circuit model of the battery, the voltage data and corresponding time data of the battery in the constant current charging stage are substituted into the lithium battery health state estimation model, that is, equation (7), for fitting, so as to obtain the unknown parameters: k : During the constant current charging process, the battery terminal voltage U is regarded as a zero state response; in the first charging cycle, the relationship between the battery terminal voltage U and the charging time t is calculated by formula (1): Where U OC (S OC ) is the open circuit voltage of the battery, I is the constant current charging current; τ1 and τ2 are the time constants of the two RC branches, where τ1 = R1C1, τ2 = R2C2, R1 and R2 are the polarization internal resistances of the two RC branches, C1 and C2 are the polarization capacitances of the two RC branches; R0 is the ohmic internal resistance; S OC is the battery SOC, S OC The relationship with the charging time t is: In the formula, C ap is the actual capacity of the first cycle; the battery terminal voltage of the nth (n>1) charging cycle is calculated as follows: Where t' is the constant current charging time of the battery in the nth cycle, R'0, R'1, R'2, τ'1 and τ'2 are the parameters of the battery in the nth cycle, S OC The relationship with charging time t' is: Where C' ap is the actual capacity of the nth cycle; subtracting equation (3) from equation (1) yields: From equation (2) and equation (4), we can get: Substitute equation (6) into equation (5) and let k = C ap / C' ap ,have to: Formula (7) is the final lithium battery health status estimation model; according to the calculation formula k = C ap / C' ap , we know that k and the health status of the battery are reciprocal, that is, k = 1 / S OH .

5. The method for estimating the health status of a lithium battery according to claim 3, characterized in that: The specific method of step 2 is: 2.1 Segment processing of constant current charging voltage data The SOC value at the starting position of the battery voltage platform area is set, and the constant current charging time corresponding to the SOC value is used as the segmentation point of the constant current charging voltage. The constant current charging voltage is divided into two sections, namely, the constant current charging voltage corresponding to the SOC segmentation point is less than the constant current charging voltage corresponding to the SOC segmentation point and the constant current charging voltage corresponding to the SOC segmentation point is greater than the SOC segmentation point; the data of the constant current charging voltage corresponding to the SOC segmentation point is discarded, and only the constant current charging voltage data corresponding to the SOC segmentation point is taken to fit the lithium battery health status estimation model; 2.2 For the constant current charging voltage data corresponding to the SOC segmentation point, the measured battery terminal voltage and the corresponding time are substituted into the lithium battery health state estimation model obtained in step 1, that is, equation (7), through the curve fitting method, and the parameter k is obtained by fitting. Then, the relationship between SOH and k is S OH =1 / k to calculate the battery health state SOH.

6. A method for estimating the health status of a lithium battery according to claim 5, characterized in that: In step 2.2, the measured battery terminal voltage and the corresponding time are substituted into the lithium battery health status estimation model finally obtained in step 1, i.e., equation (7) for fitting. In equation (7), R0, R1, R2, τ1 and τ2 are new battery parameters, which are measured by hybrid pulse power (HPPC) experiment.

Images