Method for constructing power battery pack combined life model and battery life evaluation method

Abstract

This invention discloses a method for constructing a power battery pack life model and a method for evaluating battery life. The construction method includes: constructing a preliminary cycle life model; the preliminary cycle life model is used to calculate the energy decay rate of the power battery, including a polynomial semi-empirical model with capacity as a life characteristic and an exponential semi-empirical model with internal resistance as a life characteristic, and battery energy as the dependent variable; constructing a power battery pack life model through the preliminary cycle life model and the calendar life model; the calendar life model is a power function semi-empirical model with time as the base, used to calculate the percentage of calendar capacity loss of the power battery. This invention uses capacity and internal resistance as characteristic quantities of the preliminary cycle life model, and uses battery energy as its dependent variable, while reflecting battery life degradation from both electrical performance and chemical characteristics; using two semi-empirical models and the calendar life model together to describe battery life degradation improves the accuracy of model prediction.

Description

Technical Field

[0001] This invention belongs to the field of battery life prediction technology, and in particular relates to a method for constructing a power battery pack life model and a battery life assessment method. Background Technology

[0002] As one of the most crucial components of an electric vehicle, the performance of the power battery is closely related to the overall vehicle performance, and its lifespan determines the vehicle's lifespan as well. The capacity of an onboard power battery gradually decreases during use. According to industry standards, the power battery needs to be replaced when its capacity reaches 80% of its initial capacity or its internal resistance increases to 150% of its initial internal resistance. However, replacing the battery too early will reduce the vehicle's overall economic performance, while replacing it too late will affect the vehicle's power and safety. If the lifespan degradation pattern of the power battery could be accurately predicted, the optimal balance between economy and safety / reliability could undoubtedly be found.

[0003] Currently, there are three main types of modeling methods for lithium-ion battery lifetime: electrochemical-based modeling methods, empirical formula-based modeling methods, and data-driven modeling methods. Among them, empirical formula-based modeling methods are widely used in predicting the lifetime of power batteries. Specifically, based on a semi-empirical model of battery lifetime degradation, factors affecting battery lifetime degradation, such as charge / discharge rate, operating temperature, and depth of discharge, are used as independent variables, while lifetime characteristics such as capacity, internal resistance, and power are used as dependent variables. Based on battery lifetime test results, a regression algorithm is used to fit the parameters in the semi-empirical model.

[0004] Current patents related to battery life models, such as "A Battery Life Assessment Method" (publication number CN112731164A), first establish calendar life models and cycle life models based on constant current charge-discharge cycle life test data under laboratory conditions. Then, they determine the operating simulation parameters based on user demand curves. By combining the calendar life model and cycle life model with the operating simulation parameters, they calculate the total capacity loss rate at the end of each of the set multiple prediction time units. When calculating the total capacity loss rate for each prediction time unit, they correct the current total capacity loss rate using the total capacity retention rate of the previous prediction time unit. This method uses a coupled calculation of calendar life and cycle life to comprehensively predict and assess the battery's lifespan. This invention comprehensively considers lifespan degradation caused by cycling and storage, and the model parameters fit the operating conditions as closely as possible to real-vehicle conditions. The model has good practicality and fills the gap in current lifespan model patents, which mostly do not consider the actual vehicle operating environment and storage lifespan loss. However, there are still two shortcomings: First, the model only uses capacity as a lifetime characteristic; second, there is only one semi-empirical model, and the single lifetime characteristic and empirical model formula cannot accurately describe the complex chemical system of the power battery, so the model accuracy is not high. Summary of the Invention

[0005] To address the shortcomings of the prior art, this invention provides a method for constructing a power battery pack life model and a battery life assessment method. This method uses capacity and internal resistance as characteristic quantities of a preliminary cycle life pack model, and uses the amount of electricity related to capacity and internal resistance as the dependent variable of the preliminary cycle life pack model. At the same time, it reflects battery life degradation from electrical performance and internal chemical characteristics. It uses two semi-empirical models and a calendar life model to describe battery life degradation, thereby improving the accuracy of model prediction.

[0006] The first objective of this invention is to provide a method for constructing a life model of a power battery pack.

[0007] The second objective of this invention is to provide a method for estimating the lifespan of a power battery.

[0008] The first objective of this invention can be achieved by adopting the following technical solution:

[0009] A preliminary cycle life model is constructed; the preliminary cycle life model is used to calculate the energy decay rate of the power battery, including a polynomial semi-empirical model with capacity as the life characteristic quantity and an exponential semi-empirical model with internal resistance as the life characteristic quantity, and battery energy is used as the dependent variable.

[0010] A power battery portfolio life model is constructed using a preliminary cycle life model and a calendar life model. The power battery portfolio life model is used to calculate the energy decay rate of the power battery. The calendar life model is a power function semi-empirical model with time as the base, used to calculate the percentage of calendar capacity loss of the power battery.

[0011] Furthermore, the lifespan model of the power battery pack is as follows:

[0012] Q loss,％，energy =e1*Q loss,％,cyc,energy +e2*Q loss,％,calendar +e3 (1)

[0013] Where e1, e2, e3 are constant coefficients, Q loss,％，energy Q represents the rate of energy decay of a power battery during its initial full charge and discharge cycle. loss,％,cyc,energy Q represents the rate of energy decay of the power battery. loss,％,calendar This indicates the percentage of calendar capacity loss in the power battery.

[0014] Furthermore, the preliminary cycle life model is as follows:

[0015] Q loss,％,cyc,energy =d1*Q loss,％,cyc,cap +d2*Q loss,％,cyc,R +d3 (2)

[0016] Where d1, d2, d3 are constants; Q loss,％,cy,ccap Q represents the cycle life decay rate characterized by capacity. loss,％,cyc,R This represents the cycle life decay rate, characterized by internal resistance.

[0017] Furthermore, the polynomial semi-empirical model with capacity as a lifetime characteristic is as follows:

[0018] Q loss,％,cyc,cap =b1k 2 +b2k+b3 (3)

[0019] Where b1, b2, and b3 are constant coefficients; k is the number of charge-discharge cycles.

[0020] The polynomial semi-empirical model with capacity as a lifetime characteristic is determined through the following process:

[0021] Cyclic charge-discharge life tests are conducted on the power battery to obtain the capacity value of the power battery at the corresponding cycle life stage.

[0022] Regression analysis is performed on the polynomial semi-empirical model with capacity as a lifetime characteristic using the capacity value, including:

[0023] Q loss,％,cyc,capThe value is calculated based on the capacity value; k is the number of charge-discharge cycles in the cyclic charge-discharge life test.

[0024] Connect k and the corresponding Q loss,％,cyc,cap Substituting the values ​​into formula (3), and fitting formula (3) using the least squares method, we obtain b1, b2, and b3.

[0025] Furthermore, the exponential semi-empirical model with internal resistance as a lifetime characteristic is as follows:

[0026]

[0027] Where a1 is a constant coefficient, a2 is a constant related to depth of discharge (DOD), ambient temperature (T), and charge / discharge ratio (Ratio); Ah is the total charge throughput in Ah; and the ambient temperature (T) is in °C.

[0028] The exponential semi-empirical model with internal resistance as a lifetime characteristic is determined through the following process:

[0029] A cycle charge-discharge life test is conducted on the power battery, and the capacity value of the power battery at the corresponding cycle life stage is obtained based on the cycle life test data; the cycle life test data also includes the ambient temperature, charge-discharge rate and depth of discharge during the cycle charge-discharge.

[0030] Using the cycle life test data and corresponding internal resistance values, a regression analysis is performed on the exponential semi-empirical model with internal resistance as a life characteristic, including:

[0031] The cycle life test data and the corresponding Q loss,％,cyc,R Substitute the value of Q into formula (4), take the natural logarithm of both sides of formula (4), and then use the least squares method to fit the constant coefficients a1 and a2; where Q loss,％,cyc,R The value is calculated based on the internal resistance value.

[0032] Furthermore, the preliminary cycle life model is determined through the following process:

[0033] After determining the polynomial semi-empirical model with capacity as a lifetime characteristic and the exponential semi-empirical model with internal resistance as a lifetime characteristic, regression analysis is performed on the preliminary cycle life model using the battery energy from the cyclic charge-discharge life test as the dependent variable, including:

[0034] Q loss,％,cyc，energy The value is calculated based on the battery energy value;

[0035] d1, d2, d3 according to Q loss,％,cyc，energy Q loss,％,cyc,R and Q loss,％,cyc,cap The value was obtained through regression analysis.

[0036] Furthermore, the semi-empirical calendar model is as follows:

[0037]

[0038] Among them, Q loss,％,calendar The values ​​represent the percentage of calendar capacity loss of the battery; R represents the universal gas constant; T represents the absolute temperature of the environment, in K; Ea represents the activation energy of the power battery, in J / mol; c1 represents the pre-load factor; and t represents the resting time, in days.

[0039] Furthermore, the semi-empirical calendar model is determined through the following process:

[0040] A storage test is conducted on the power battery, and the calendar capacity value of the power battery at the corresponding storage stage is obtained based on the storage test data.

[0041] Regression analysis was performed on the semi-empirical calendar model using shelving test data and calendar capacity values, including:

[0042] The Q loss,％,calendar The value is calculated based on the calendar capacity value;

[0043] Using the data from the shelved test and the corresponding Q loss,％,calendar The value is fitted to formula (5), including:

[0044] First, based on the data from the storage test, the relationship between battery capacity loss and storage time was analyzed, including:

[0045] Taking the natural logarithm of both sides of equation (5), we get:

[0046]

[0047] Then, using MATLAB, the slope of the straight line is fitted, which is c2.

[0048] Based on the calculation method of c2, the relationship between the logarithm of the battery capacity loss value and 1 / T under different resting times is analyzed: the slope is equal at different resting times, that is, -Ea / R is a constant; Ea is obtained by fitting the slope of the straight line, and c1 is obtained by fitting the intercept of the straight line.

[0049] Furthermore, the lifespan model of the power battery pack is determined through the following process:

[0050] After determining the preliminary cycle life model and the calendar semi-empirical model, regression analysis was performed on the power battery pack life model using the initial full-charge and discharge energy of the power battery in the electric vehicle durability test as the dependent variable, including:

[0051] Q loss,％，energyThe value is calculated based on the initial full charge and discharge energy.

[0052] Q loss,％,cyc,energy and Q loss,％,calendar The value is calculated based on experimental data from the durability test of the electric vehicle.

[0053] Q loss,％，energy Q loss,％,cyc,energy and Q loss,％,calendar Substituting the values ​​into formula (1), e1, e2, e3 are obtained by fitting using the linear least squares method;

[0054] Substituting e1, e2, and e3 into formula (1), we obtain the final power battery pack life model.

[0055] The second objective of this invention can be achieved by adopting the following technical solution:

[0056] A method for estimating the lifespan of a power battery, implemented based on the above-described construction method, the method comprising:

[0057] Based on the measured data of the power battery to be estimated, the energy decay rate of the power battery to be estimated is calculated using the preliminary cycle life model, and the calendar capacity loss percentage of the power battery to be estimated is calculated using the calendar life model.

[0058] Based on the degradation rate and calendar capacity loss percentage, the energy degradation rate of the power battery to be estimated is calculated using the power battery combination life model.

[0059] The present invention has the following advantages over the prior art:

[0060] The method for constructing a power battery pack life model provided by this invention comprehensively considers life degradation caused by cycling and storage. The model parameters are fitted from real vehicle conditions, making the model highly practical. Capacity and internal resistance are used together as feature quantities in the initial cycle life model, and the electrical quantity, which is closely related to capacity and internal resistance, is used as the dependent variable in the initial cycle life model. Simultaneously, battery life degradation is reflected from both electrical performance and internal chemical characteristics. Two semi-empirical models and one calendar life model are used together to describe battery life degradation, avoiding the one-sidedness of a single model describing the complex battery system and improving the accuracy of model predictions. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating the method for constructing a power battery pack life model according to an embodiment of the present invention.

[0063] Figure 2 The charts are used to verify the accuracy of the combined life model based on some real-vehicle durability test data from embodiments of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be understood that the specific embodiments described are merely used to explain this application and are not intended to limit this application.

[0065] Example:

[0066] The method for constructing a power battery pack life model provided in this embodiment uses battery energy as the dependent variable, and simultaneously reflects the impact of capacity decay and internal resistance growth on the battery's external characteristics, thus improving the accuracy of battery life prediction. The pack life model includes cycle life and calendar life. The dependent variable in the cycle life pack model is the power battery capacity decay rate. The cycle life pack model includes a polynomial semi-empirical model with capacity as the life characteristic and an exponential semi-empirical model with internal resistance as the life characteristic. The calendar life model is a power function semi-empirical model with time as the base.

[0067] like Figure 1As shown, the method for constructing the power battery pack life model provided in this embodiment includes: (i) referring to the electric vehicle operating environment, formulating a cycle charge-discharge life test and storage test matrix, and conducting reference performance tests during the cycle and storage processes, wherein: the capacity test results in the cycle charge-discharge and reference performance tests are used to perform regression analysis on the capacity semi-empirical model; the HPPC (hybrid power pulse characteristic) test results in the cycle charge-discharge and reference performance tests are used to perform regression analysis on the internal resistance semi-empirical model; the battery energy value of the capacity test in the cycle charge-discharge and reference performance tests is used as the dependent variable to perform regression analysis on the cycle life combined model; and the calendar life database is used to perform regression analysis on the calendar semi-empirical model; (ii) obtaining vehicle durability test data, which includes vehicle operating data and storage data. Part of the durability test data is used to perform regression analysis on the life model to obtain the combined life model; and part of the data is used to verify the accuracy of the model.

[0068] Specifically, the implementation steps of the method for constructing the power battery pack life model provided in this embodiment are as follows:

[0069] (1) Establish a preliminary cycle life model.

[0070] Current literature indicates that temperature, charge / discharge rate, and depth of discharge are the most significant factors affecting battery cycle life. Therefore, this study combines the actual operating environment of electric vehicles, using private car operating conditions as the application scenario and an NCM523 power battery manufactured by a domestic battery manufacturer as the test subject, to design a cycle life test scheme, as shown in Table 1.

[0071] Table 1. Life test scheme designed for NCM523 power batteries.

[0072]

[0073] (1-1) An exponential semi-empirical model with internal resistance as a lifetime characteristic.

[0074] After organizing the test data, the data 01-09 were used to fit an exponential semi-empirical model with internal resistance as a lifetime characteristic, i.e.:

[0075]

[0076] Among them, Q loss,％,cyc,R The cycle life decay rate, characterized by internal resistance, is calculated based on the internal resistance value obtained from cyclic charge-discharge reference performance tests. Specifically, it is the ratio of the increase in internal resistance relative to the initial internal resistance value to the initial internal resistance, i.e., Q. loss,％,cyc,R= Increase in internal resistance relative to initial internal resistance / initial internal resistance; a1 is a constant coefficient, a2 is a constant related to depth of discharge (DOD), ambient temperature (T), and charge / discharge ratio (Ratio), with T in °C; Ah represents the total charge throughput, in Ah.

[0077] First, take the natural logarithm ln on both sides of the formula, i.e. Then, the least squares method was used to fit the coefficients a1 and a2 in the model.

[0078] Substituting the obtained coefficients a1 and a2 into the right side of formula (1), we obtain an exponential semi-empirical model with internal resistance as a lifetime characteristic.

[0079] (1-2) A polynomial semi-empirical model with capacity as a lifetime characteristic.

[0080] Fitting a polynomial semi-empirical model with capacity as a lifetime characteristic, i.e.:

[0081] Q loss,％,cy,ccap =b1k 2 +b2k+b3 (2)

[0082] Among them, Q loss,％,cy,ccap The cycle life decay rate, characterized by capacity, is calculated based on the capacity value obtained from cyclic charge-discharge reference performance tests. Specifically, it is the ratio of the change in capacity value relative to the initial capacity value in each cycle to the initial capacity, i.e., Q. loss,％,cyc,R = The change in capacity value relative to the initial capacity value each time / the initial capacity; b1,b 2, b3 is a constant coefficient of the model, and k is the number of charge-discharge cycles. Based on the characteristics of the model, the least squares method is used to fit the model to obtain b1, b2, and b3.

[0083] (1-3) Preliminary model of cycle life.

[0084] Based on the exponential semi-empirical model with internal resistance as the lifetime characteristic and the polynomial semi-empirical model with capacity as the lifetime characteristic, a preliminary cycle life model is determined, namely:

[0085] Q loss,％,cyc，energy =d1*Q loss,％,cyc,cap +d2 * Q loss,％,cyc,R +d3 (3)

[0086] Among them, Q loss,％,cyc，Energy This represents the battery capacity decay rate. The data comes from the battery energy values ​​of capacity tests in the cyclic charge-discharge reference performance test. d1, d2, and d3 are constants. The data is obtained by regression analysis using the battery energy decay rate of capacity tests in the cyclic charge-discharge and reference performance tests as the dependent variable, combined with the cycle life test database.

[0087] (2) Establish a calendar life model.

[0088] During battery storage, time and temperature are two important parameters affecting the calendar life of power batteries, and there is an exponential relationship between time and calendar life characteristics. Therefore, an Arrhenius equation is used to establish a calendar life model, namely:

[0089]

[0090] In the formula, Q loss,％,calendar This represents the percentage of calendar capacity loss in the battery, calculated based on the calendar capacity values ​​obtained from reference performance tests during the storage period. Specifically, it is the ratio of the change in calendar capacity relative to the initial calendar capacity in each reference performance test to the initial calendar capacity; R represents the universal gas constant, i.e., 8.31 J·mol⁻¹. -1 ·K -1 T represents absolute temperature in K, and its relationship with the ambient temperature in the database is: T(K) = ambient temperature (°C) + 273.15; Ea represents the activation energy of the power battery in J / mol; c1 represents the pre-coefficient; t represents the shelving time in days.

[0091] Data from 10 to 12 were used to fit an exponential semi-empirical model with calendar capacity as a lifespan characteristic. This included: firstly, analyzing the relationship between battery capacity loss and the number of days in storage based on data from storage tests. To more intuitively understand the relationship, the natural logarithm of both sides of the model was taken, resulting in the equation... It can be seen that there is a linear relationship between the natural logarithm of battery capacity loss and the logarithm of time in days. The slope of the straight line fitted using MATLAB is c2.

[0092] Referring to the calculation method of c2, the relationship between the logarithm of the battery capacity loss value and 1 / T under different storage time days is analyzed, as shown in the figure. loss,％,calendar Proportional to 1 / T, the slope is the same for different resting times, that is, -Ea / R is a constant. Ea is obtained by fitting the slope of the straight line, and c1 is obtained by fitting the intercept of the straight line.

[0093] (3) Establish a battery life model based on the preliminary cycle life model and calendar life model.

[0094] Based on the preliminary cycle life model in step (1) and the calendar life model in step (2), the battery life model is established as follows:

[0095] Q loss,％，energy =e1*Q loss,％,cy,cenergy+ e2*Q loss,％,caelndar +e3 (5)

[0096] In the formula, e1, e2, and e3 are constant coefficients, obtained as follows:

[0097] First, the actual vehicle durability test data is divided into fixed time intervals, t0 to t1. 10 There are a total of 11 time periods. The initial full charge and discharge energy, total charge throughput Ah, static duration and ambient temperature of the battery are statistically analyzed in each time period. Using data from 5 time periods from t0 to t4, e1, e2 and e3 are obtained by linear least squares fitting.

[0098] Use t5~t 10 The model accuracy was validated using a total of 6 data sets, and the results are as follows: Figure 2 As shown in Table 2, the relative error analysis between model prediction and actual vehicle statistics shows a maximum relative error of 0.26%, indicating high model accuracy and the ability to accurately predict degradation under actual vehicle operating conditions. The model calculation result represents the lifespan degradation rate. In this example, the energy of the power battery increases during the initial use, resulting in a negative degradation rate, which is normal. However, to provide a more detailed and specific representation of the data... Figure 2 Battery degradation is described using discharge energy retention rate. The relationship between energy degradation rate and energy retention rate is: energy degradation rate + energy retention rate = 100%.

[0099] Table 2. Analysis of Relative Errors in Actual Vehicle Statistics and Model Prediction

[0100] Time period Actual vehicle statistics Model prediction relative error t5 97.83％ 97.96％ 0.13％ t6 98.08％ 98.32％ 0.24％ t7 97.43％ 97.65％ 0.23％ t8 96.55％ 96.78％ 0.24％ t9 95.87％ 96.12％ 0.26％ t10 94.91％ 95.16％ 0.26％

[0101] This embodiment also provides a method for estimating the lifespan of a power battery, including:

[0102] Based on the measured data of the power battery to be estimated, the energy decay rate of the power battery to be estimated is calculated using the preliminary cycle life model, and the calendar capacity loss percentage of the power battery to be estimated is calculated using the calendar life model.

[0103] Based on the degradation rate and calendar capacity loss percentage, the energy degradation rate of the power battery to be estimated is calculated using the power battery combination life model.

[0104] Specifically, the measured data includes the number of charge-discharge cycles, static duration, ambient temperature during static conditions, and resting time.

[0105] Substituting the ambient temperature and total charge throughput into the exponential semi-empirical model with internal resistance as the lifetime characteristic (i.e., formula (1) for the coefficients a1 and a2), the cycle lifetime decay rate with internal resistance as the characteristic is obtained.

[0106] Substituting the number of charge-discharge cycles into the polynomial semi-empirical model with capacity as the lifetime characteristic (i.e., formula (2) with the coefficients b1, b2, b3 determined), the cycle lifetime decay rate with capacity as the characteristic is obtained.

[0107] Furthermore, the cycle life decay rate characterized by internal resistance and the cycle life decay rate characterized by capacity are substituted into the preliminary cycle life model (i.e., the formula (3) for determining constants d1, d2, d3) to obtain the battery capacity decay rate.

[0108] Substituting the ambient temperature and the time of rest into the calendar life model (i.e., the formula (4) for the coefficients of determination c1 and c2), the percentage of calendar capacity loss of the battery is obtained.

[0109] Furthermore, the battery capacity decay rate and the percentage of battery calendar capacity loss are substituted into the battery life model (i.e., the formula (5) for the coefficients e1, e2, e3) to obtain the energy decay rate of the power battery to be estimated.

[0110] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium. It should be noted that although the method operations of the above embodiments are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Rather, the described steps can be performed in a different order. Additionally or alternatively, certain steps can be omitted, multiple steps can be combined into one step, and / or one step can be broken down into multiple steps.

[0111] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A method for constructing a lifespan model of a power battery pack, characterized in that, The method includes: A preliminary cycle life model is constructed; the preliminary cycle life model is used to calculate the cycle life decay rate of the power battery energy, including a polynomial semi-empirical model with capacity as the life characteristic and an exponential semi-empirical model with internal resistance as the life characteristic. A power battery portfolio life model is constructed using a preliminary cycle life model and a calendar life model. The power battery portfolio life model is used to calculate the energy decay rate of the power battery. The calendar life model is a power function semi-empirical model with time as the base, used to calculate the calendar capacity loss percentage of the power battery. The lifespan model of the power battery pack is as follows: （1） in, The constant coefficient, This indicates the rate of energy decay of the power battery during its initial full charge and discharge cycle. This indicates the rate of energy decay during the cycle life of the power battery. This indicates the percentage of calendar capacity loss in the power battery. The preliminary cycle life model is as follows: （2） in, It is a constant; This represents the cycle life decay rate, characterized by capacity. This represents the cycle life decay rate, characterized by internal resistance.

2. The construction method of claim 1, wherein, The polynomial semi-empirical model with capacity as a lifetime characteristic is as follows: （3） in, is a constant coefficient; k is the number of charge-discharge cycles; The polynomial semi-empirical model with capacity as a lifetime characteristic is determined through the following process: Cyclic charge-discharge life tests are conducted on the power battery to obtain the capacity value of the power battery at the corresponding cycle life stage. Regression analysis is performed on the polynomial semi-empirical model with capacity as a lifetime characteristic using the capacity value, including: The value is calculated based on the capacity value; k is the number of charge-discharge cycles in the cyclic charge-discharge life test. Substitute the values of k and corresponding into equation (3), and fit equation (3) using the least squares method to obtain .

3. The construction method of claim 1, wherein, The exponential semi-empirical model that uses internal resistance as a lifetime characteristic is as follows: （4） in, The constant coefficient, is a constant related to depth of discharge (DOD), ambient temperature (T), and charge / discharge ratio (Ratio); Ah is the total charge throughput in Ah; ambient temperature (T) is in °C. The exponential semi-empirical model with internal resistance as a lifetime characteristic is determined through the following process: A cycle charge-discharge life test is conducted on the power battery, and the capacity value of the power battery at the corresponding cycle life stage is obtained based on the cycle life test data; the cycle life test data also includes the ambient temperature, charge-discharge rate and depth of discharge during the cycle charge-discharge. Using the cycle life test data and corresponding internal resistance values, a regression analysis is performed on the exponential semi-empirical model with internal resistance as a life characteristic, including: The cycle life test data and corresponding Substitute the value into formula (4), take the natural logarithm of both sides of formula (4), and then use the least squares method to transform the constant coefficients. , Perform fitting; where, The value is calculated based on the internal resistance value.

4. The construction method of claim 1, wherein, The preliminary cycle life model was determined through the following process: After determining the polynomial semi-empirical model with capacity as a lifetime characteristic and the exponential semi-empirical model with internal resistance as a lifetime characteristic, regression analysis is performed on the preliminary cycle life model using the battery energy from the cyclic charge-discharge life test as the dependent variable, including: a value of the battery energy value is calculated from the battery energy value; Regression analysis was performed according to the values of , and .

5. The construction method of claim 1, wherein, The semi-empirical calendar model is as follows: （5） in, The value represents the percentage of calendar capacity loss of the power battery, R represents the universal gas constant, and T represents the absolute temperature of the ambient temperature, in K. The activation energy of a power battery is expressed in J / mol. represents the pre-conversion factor; t represents the shelving time, in days.

6. The construction method according to claim 5, characterized in that, The semi-empirical calendar model is determined through the following process: A storage test is conducted on the power battery, and the calendar capacity value of the power battery at the corresponding storage stage is obtained based on the storage test data. Regression analysis was performed on the semi-empirical calendar model using shelving test data and calendar capacity values, including: the value is calculated from the calendar capacity value; Using the shelf test data and corresponding values for Equation (5), a fit was made including: First, based on the data from the storage test, the relationship between battery capacity loss and storage time was analyzed, including: Taking the natural logarithm of both sides of formula (5), we get: ； Then use MATLAB to fit the straight line slope, that is ; based on The calculation method analyzes the relationship between the logarithm of battery capacity loss and 1 / T under different resting times: the slope is equal at different resting times, that is... It is a constant; Ea is obtained by fitting the slope of the line, and Ea is obtained by fitting the intercept of the line. .

7. The construction method of claim 1, wherein, The lifespan model of the power battery pack is determined through the following process: After determining the preliminary cycle life model and the calendar semi-empirical model, regression analysis was performed on the power battery pack life model using the initial full-charge and discharge energy of the power battery in the electric vehicle durability test as the dependent variable, including: The value is calculated based on the initial full charge and discharge energy. and values are calculated from experimental data obtained during the vehicle durability test of the electric vehicle; Substituting the values of , and into equation (1), a linear least squares fit gives ; Will Substituting into formula (1), we obtain the final power battery pack life model.

8. A method for estimating the lifetime of a power battery, based on the construction method according to any one of claims 1 to 7, characterized in that, The method includes: Based on the measured data of the power battery to be estimated, the energy decay rate of the power battery to be estimated is calculated using the preliminary cycle life model, and the calendar capacity loss percentage of the power battery to be estimated is calculated using the calendar life model. According to the attenuation rate and the calendar capacity loss percentage, the energy attenuation rate of the power battery to be estimated is calculated by using the power battery combined life model.

Images