Battery remaining service life prediction method, device, equipment and readable storage medium

Abstract

A battery remaining useful life prediction method, device and equipment and readable storage medium are related to the field of battery, which comprises obtaining a target energy retention rate and a target cycle test number corresponding to the last cycle test of a battery to be predicted at a target temperature level; calculating a target drift rate through a target Eyring model for describing the functional relationship between the drift rate and the temperature and the cycle test number, the target cycle test number and the target temperature level; based on the remaining useful life probability density function, the target cycle test number, the target drift rate, the target energy retention rate and the target diffusion coefficient value are substituted to obtain the remaining useful life of the battery to be predicted; the model parameter value of the target Eyring model and the target diffusion coefficient value are determined by solving the target battery degradation model composed of the initial Eyring model and the Wiener process model corresponding to the drift rate change based on the parameter estimation method. The accuracy of the battery remaining useful life prediction can be effectively improved through the present application.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of batteries, in particular to a battery remaining service life prediction method, device, equipment and readable storage medium. BACKGROUND

[0002] Lithium ion batteries often experience complex and variable operating environments during actual operation, especially different working temperature conditions. The multi-temperature environment not only increases the working load of the battery, but also puts higher requirements on its materials and structure. Among them, under different working temperature conditions, the internal chemical reaction rate of the battery often differs, leading to significant differences in performance. Moreover, the degradation process of the battery is generally nonlinear, which brings great challenges to predicting the battery life and evaluating the battery reliability indicators.

[0003] In related technologies, the degradation process of the battery is usually described using a Wiener process, and the relationship between the drift rate (i.e. the degradation rate) and the temperature level is constructed by combining the Arrhenius model, and then a battery degradation model based on a nonlinear Wiener process is constructed to evaluate the remaining service life of the battery.

[0004] However, the degradation process of the battery is not only nonlinear, but also the degradation rate is not only related to the environmental temperature, but also changes with the battery aging process. When using the Arrhenius model to construct the relationship between the drift rate and the stress level, the factor that the drift rate changes with the battery aging process is ignored, resulting in poor fitting effect and large error of the degradation model to the data, so as to be unable to accurately predict the remaining service life of the battery. SUMMARY

[0005] The present application provides a battery remaining service life prediction method, device, equipment and readable storage medium, which can effectively improve the prediction accuracy of the remaining service life of the battery.

[0006] In a first aspect, the present application provides a battery remaining service life prediction method, which comprises:

[0007] obtaining a target energy retention rate corresponding to the last cycle of the cycle test of the battery to be predicted at a target temperature level and a target cycle test number corresponding to the last cycle of the cycle test;

[0008] calculating a target drift rate by a target Arrhenius model pre-set for describing the functional relationship between the drift rate and the temperature, the cycle test number, the target cycle test number and the target temperature level;

[0009] Based on a preset remaining service life probability density function, a target number of cycle test laps, a target drift rate, a target energy retention rate, and a preset target diffusion coefficient value, a life prediction is performed to obtain a target remaining service life of the battery to be predicted;

[0010] Among them, the model parameter values ​​and the target diffusion coefficient values ​​in the target Eyring model are determined by solving the preset target battery degradation model based on a preset parameter estimation method. The target battery degradation model consists of an initial Eyring model and a Wiener process model corresponding to the drift rate change.

[0011] In conjunction with the first aspect, in one embodiment, the target Eyring model is:

[0012]

[0013] In the formula, μ1 represents the drift rate, k represents the number of cycle tests, and T represents the temperature level. and are the model parameter values ​​obtained by the parameter estimation method.

[0014] In combination with the first aspect, in one embodiment, the remaining useful life probability density function for:

[0015]

[0016] Where s represents the remaining service life, T i Indicates the target temperature level, μ1(k ′ ,T i ) indicates that at the target temperature level, the last cycle test k end The corresponding target cycle test number k ′ The target drift rate corresponding to the condition of i (k end ) represents the target energy retention rate corresponding to the last cycle test of the battery to be predicted at the target temperature level, ω represents the energy retention failure threshold, Represents the target diffusion coefficient value.

[0017] In combination with the first aspect, in one embodiment, the target battery degradation model is:

[0018]

[0019] Where, Y1(k+1) represents the energy retention rate corresponding to the k+1th cycle test, Y1(k) represents the energy retention rate corresponding to the kth cycle test, T represents the temperature level, k represents the number of cycle tests, represents an initial Eyring model, a and b are model parameters to be solved by parameter estimation method, σ1 represents a diffusion coefficient value to be solved by parameter estimation method, and B(1) represents a standard Brownian motion.

[0020] In combination with the first aspect, in an implementation, the target Eyring model is:

[0021]

[0022] wherein μ2 represents a drift rate, k represents a cycle number of a cycle test, and T represents a temperature level, and are model parameter values solved by parameter estimation method.

[0023] In combination with the first aspect, in an implementation, the probability density function of the remaining useful life is

[0024]

[0025] wherein s represents a remaining useful life, T i represents a target temperature level, μ2(k ′ ,T i ) represents a target drift rate corresponding to a target cycle number k end of a cycle test under the condition of a target temperature level T ′ , Y i (k end ) represents a target energy retention rate corresponding to a last cycle of a cycle test of the battery to be predicted under a target temperature level, and ω represents an energy retention rate failure threshold, represents a target diffusion coefficient value.

[0026] In combination with the first aspect, in an implementation, the target battery degradation model is:

[0027]

[0028] wherein Y2(k+1) represents an energy retention rate corresponding to a (k+1)th cycle of a cycle test, Y2(k) represents an energy retention rate corresponding to a kth cycle of a cycle test, T represents a temperature level, and k represents a cycle number of a cycle test, represents an initial Eyring model, c, d, and e are model parameters to be solved by parameter estimation method, σ2 represents a diffusion coefficient value to be solved by parameter estimation method, and B(1) represents a standard Brownian motion.

[0029] In a second aspect, an embodiment of the present application provides a battery remaining useful life prediction device, and the battery remaining useful life prediction device comprises:

[0030] ​a parameter obtaining module configured to obtain a target energy retention rate corresponding to a last cycle of a cycle test at a target temperature level of a battery to be predicted and a target cycle number corresponding to the last cycle of the cycle test;

[0031] a parameter calculating module configured to calculate a target drift rate based on a target Eyring model preset for describing a functional relationship between the drift rate and the temperature and the cycle number, the target cycle number, and the target temperature level;

[0032] a life prediction module configured to perform life prediction based on a preset residual service life probability density function, the target cycle number, the target drift rate, the target energy retention rate, and a preset target diffusion coefficient value, to obtain a target residual service life of the battery to be predicted.

[0033] In the target Eyring model, a model parameter value and the target diffusion coefficient value are determined by solving a preset target battery degradation model based on a preset parameter estimation method, and the target battery degradation model is composed of an initial Eyring model and a Wiener process model corresponding to a drift rate change.

[0034] In combination with the second aspect, in an implementation, the target Eyring model is:

[0035]

[0036] wherein μ1 represents the drift rate, k represents the cycle number, and T represents the temperature level, and is a model parameter value obtained by solving the parameter estimation method.

[0037] In combination with the second aspect, in an implementation, the residual service life probability density function is

[0038]

[0039] wherein s represents the residual service life, T i represents the target temperature level, μ1(k ′ ,T i ) represents the target drift rate corresponding to the target cycle number k end under the condition of the target temperature level, the last cycle of the cycle test k ′ , Y i (k end ) represents the target energy retention rate corresponding to the last cycle of the cycle test at the target temperature level of the battery to be predicted, and ω represents an energy retention rate failure threshold, represents the target diffusion coefficient value.

[0040] ​In combination with the second aspect, in an implementation, the target battery degradation model is:

[0041]

[0042] where Y1(k+1) represents the energy retention rate corresponding to the (k+1)th cycle of the cycle test, Y1(k) represents the energy retention rate corresponding to the kth cycle of the cycle test, T represents the temperature level, and k represents the cycle number of the cycle test, represents an initial Eyring model, a and b are model parameters to be solved by parameter estimation, σ1 represents a diffusion coefficient value to be solved by parameter estimation, and B(1) represents a standard Brownian motion.

[0043] In combination with the second aspect, in an implementation, the target Eyring model is:

[0044]

[0045] where μ2 represents a drift rate, k represents the cycle number of the cycle test, and T represents the temperature level, and are model parameter values solved by parameter estimation.

[0046] In combination with the second aspect, in an implementation, the probability density function of the remaining useful life is

[0047]

[0048] where s represents the remaining useful life, T i represents a target temperature level, μ2(k ′ ,T i ) represents a target drift rate corresponding to the k end th cycle of the cycle test k ′ at the target temperature level, Y i (k end ) represents a target energy retention rate corresponding to the last cycle of the cycle test at the target temperature level, and ω represents an energy retention rate failure threshold, represents a target diffusion coefficient value.

[0049] In combination with the second aspect, in an implementation, the target battery degradation model is:

[0050]

[0051] where Y2(k+1) represents the energy retention rate corresponding to the (k+1)th cycle of the cycle test, Y2(k) represents the energy retention rate corresponding to the kth cycle of the cycle test, T represents the temperature level, and k represents the cycle number of the cycle test,​ represents the initial Eyring model, c, d and e are model parameters to be solved by parameter estimation, σ2represents the diffusion coefficient value to be solved by parameter estimation, and B(1) represents the standard Brownian motion.

[0052] In a third aspect, an embodiment of the present application provides a battery remaining useful life prediction device, the battery remaining useful life prediction device comprising a processor, a memory, and a battery remaining useful life prediction program stored in the memory and executable by the processor, wherein the battery remaining useful life prediction program, when executed by the processor, implements the steps of the battery remaining useful life prediction method as described above.

[0053] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium storing a battery remaining useful life prediction program, wherein the battery remaining useful life prediction program, when executed by a processor, implements the steps of the battery remaining useful life prediction method as described above.

[0054] The technical scheme provided by the embodiments of the present application has the following beneficial effects:

[0055] The temperature and the cycle test number of turns are taken as factors affecting the degradation process of lithium ion batteries and the like, that is, the temperature and the cycle test number of turns are introduced to describe the change of the drift rate with the battery aging process, so as to accurately realize the prediction of the remaining useful life of the battery; the target Eyring model is used to construct the functional relationship between the drift rate and the temperature and the cycle test number of turns, so as to substitute the target cycle test number of turns corresponding to the last cycle test of the battery to be predicted at a target temperature level and the target temperature level into the target Eyring model, to determine the target drift rate of the battery to be predicted; wherein, the parameter estimation is performed on the target battery degradation model composed of the initial Eyring model and the Wiener process model corresponding to the change of the drift rate, to determine the model parameter value of the target Eyring model and the target diffusion coefficient value; and then, based on the remaining useful life probability density function, and by substituting the target energy retention rate corresponding to the last cycle test of the battery to be predicted at the target temperature level, the target cycle test number of turns, the target drift rate and the target diffusion coefficient value, the life prediction is performed, so as to accurately predict the remaining useful life of the battery to be predicted. It can be seen that the prediction accuracy of the remaining useful life of the battery can be effectively improved by the present application. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 FIG. 1 is a flowchart of the battery remaining useful life prediction method according to an embodiment of the present application;

[0057] Figure 2 FIG. 2 is a battery RUL probability density function curve diagram of the target battery degradation model 1 based on parameterization at 25°C according to an embodiment of the present application;

[0058] Figure 3 Battery RUL probability density function curve at 25℃ based on the target battery degradation model 1 involved in the embodiment of the present application;

[0059] Figure 4 Battery RUL probability density function curve at 45℃ based on the target battery degradation model 1 involved in the embodiment of the present application;

[0060] Figure 5 Battery RUL probability density function curve at 25℃ based on the target battery degradation model 2 involved in the embodiment of the present application;

[0061] Figure 6 Battery RUL probability density function curve at 35℃ based on the target battery degradation model 2 involved in the embodiment of the present application;

[0062] Figure 7 Battery RUL probability density function curve at 45℃ based on the target battery degradation model 2 involved in the embodiment of the present application;

[0063] Figure 8 Hardware structure schematic diagram of the battery remaining useful life prediction device involved in the embodiment of the present application. DETAILED DESCRIPTION

[0064] In order for those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0065] In order to make the purpose, technical solutions and advantages of the present application clearer, the embodiments of the present application will be described in further detail below in conjunction with the drawings.

[0066] In a first aspect, the embodiments of the present application provide a battery remaining useful life prediction method.

[0067] In an embodiment, with reference to Figure 1 , Figure 1 Flowchart of the battery remaining useful life prediction method embodiment of the present application. As shown in Figure 1 , the battery remaining useful life prediction method comprises:

[0068] Step S10: obtaining a target energy retention rate corresponding to a last cycle of the cycle test at a target temperature level of the battery to be predicted and a target cycle number of the cycle test corresponding to the last cycle.

[0069] For example, in this embodiment, when predicting the remaining useful life of the battery to be predicted, first, the battery to be predicted needs to be controlled to perform a preset number of constant power charge-discharge accelerated degradation cycle tests at a target temperature level to determine the target energy retention rate of the battery to be predicted at the last cycle of the cycle test. It should be noted that the target temperature level refers to the temperature level to be tested, and the specific value thereof can be determined according to the test requirements of the battery to be tested. For example, if a user expects to predict the remaining useful life of the battery to be predicted at 25°C, the target temperature level is 25°C. For example, if a user expects to predict the remaining useful life of the battery to be predicted at 35°C, the target temperature level is 35°C. In addition, the specific value of the preset number of cycles can be determined according to actual requirements, which is not limited herein. For example, the preset number of cycles is set to 100, the last cycle is the 100th cycle, and the target cycle number of the cycle test is 100.

[0070] It should be understood that if the remaining useful life of multiple batteries to be predicted of the same model is predicted at the same time, after obtaining the target energy retention rate of each battery to be predicted at the target temperature level, the average energy retention rate of all target energy retention rates needs to be obtained, and the target average energy retention rate is substituted into the remaining useful life probability density function to predict the remaining useful life of the battery to be predicted. The energy retention rate failure threshold in the remaining useful life probability density function needs to be adaptively adjusted to the average energy retention rate failure threshold.

[0071] Step S20: obtaining a target drift rate by a preset target Eyring model for describing the functional relationship between the drift rate and the temperature, the cycle number of the cycle test, the target cycle number of the cycle test, and the target temperature level; wherein the model parameter value in the target Eyring model and the target diffusion coefficient value are determined by parameter estimation on a target battery degradation model composed of an initial Eyring model and a Wiener process model corresponding to the change of the drift rate.

[0072] It should be noted that the target battery degradation model is:

[0073]

[0074] In the formula, Y1(k+1) represents the energy retention rate corresponding to the (k+1)th cycle of the cycle test, Y1(k) represents the energy retention rate corresponding to the kth cycle of the cycle test, T represents the temperature level, k represents the cycle number of the cycle test, represents an initial Eyring model, a and b are model parameters to be solved by parameter estimation method, σ1 represents a diffusion coefficient value to be solved by parameter estimation method, and B(1) represents a standard Brownian motion.

[0075] Alternatively, the target battery degradation model is:

[0076]

[0077] wherein Y2(k+1) represents an energy retention corresponding to the k+1th cycle of the cycle test, Y2(k) represents an energy retention corresponding to the kth cycle of the cycle test, T represents a temperature level, and k represents a cycle number of the cycle test, represents an initial Eyring model, c, d and e are model parameters to be solved by parameter estimation method, σ2 represents a diffusion coefficient value to be solved by parameter estimation method, and B(1) represents a standard Brownian motion.

[0078] Exemplarily, in the embodiment, before calculating the drift rate of the battery to be predicted, an Eyring model (i.e., a target Eyring model, hereinafter referred to as a target Eyring model for the sake of uniformity of description) for describing the functional relationship between the drift rate and the temperature and the cycle number of the cycle test needs to be constructed.

[0079] Specifically, first, the lithium ion battery and the like is subjected to a constant power charge-discharge accelerated degradation cycle test under different environmental temperatures (such as 25°C, 35°C, 45°C, etc.), until the performance characteristic quantity of all batteries reaches the failure threshold, and the charge-discharge energy data of the battery under each cycle test is recorded and saved, so as to calculate the performance characteristic quantity according to the charge-discharge energy data. The performance characteristic quantity is the energy retention, which can be obtained by the following formula:

[0080]

[0081] wherein Energy retention represents the energy retention, E i is the discharge energy of the ith cycle of the cycle test, and E0 is the initial discharge energy.

[0082] Secondly, the environmental temperature is taken as a factor affecting the battery degradation process, and the change trend of the battery energy retention is simulated based on the Wiener process of the drift rate change. The expression of the Wiener process model is as follows:

[0083] Y(t0+Δt)=Y(t0)-μΔt+σB(Δt) (2)

[0084] Wherein, t0 represents the initial cycle test number, Δt represents the interval of the cycle test number, Y(t0+Δt) represents the energy retention rate of the battery corresponding to the (t0+Δt)th cycle test number; Y(t0) represents the energy retention rate of the battery corresponding to the t0th cycle test number; B(Δt) represents the standard Brownian motion; μ represents the drift rate; σ represents the diffusion coefficient.

[0085] It should be noted that, in this embodiment, the diffusion coefficient σ of the Wiener process is regarded as a constant value, and the drift rate μ is set as a function of the temperature and the number of cycle test cycles. In addition, it should be understood that the increments Y(t4)-Y(t3) and Y(t2)-Y(t1) of the Wiener process (where, ) are independent of each other, and the increment ΔY=Y(t2)-Y(t1) obeys the Gaussian distribution ΔY~N(μΔt,σ 2 Δt), where Δt=t2-t1.

[0086] Then, in order to fit the relationship between drift rate, temperature, and number of cycle tests, this embodiment uses the generalized Eyring model shown in formula (3) to describe the above relationship:

[0087]

[0088] Where μ(k,T) represents the drift rate; k represents the number of test cycles; T represents the temperature level; γ0, γ1, γ2, γ3, m, and n are model parameters.

[0089] By substituting different parameter combinations into formula (3) for fitting experimental calculations, the following two initial Eyring models can be obtained:

[0090] Initial Eyring Model 1:

[0091] Initial Eyring Model 2:

[0092] It should be understood that under the initial Eyring model 1, temperature and the number of cycle test cycles have a synergistic effect on the drift rate, while under the initial Eyring model 2, temperature and the number of cycle test cycles will have independent effects on the drift rate respectively; where μ1(k, T) in formula (4) represents the drift rate of the initial Eyring model 1; μ2(k, T) in formula (5) represents the drift rate of the initial Eyring model 2; a, b, c, d, and e are the parameters to be estimated for the model.

[0093] Then, the drift rate change of the Eyring model-based Wiener process expression is obtained according to the initial Eyring model 1 and the initial Eyring model 2 in combination with the corresponding Wiener process model of formula (2), that is, the improved battery degradation model corresponding to the two different initial Eyring models is generated:

[0094] Target battery degradation model 1:

[0095] Target battery degradation model 2:

[0096] In the formula, t0 is assigned as k, Δt is assigned as 1, a, b, c, d and e are all model parameters to be solved, and σ1 and σ2 are diffusion coefficient values to be solved.

[0097] It should be understood that after obtaining the accelerated test data of the battery, the embodiment can obtain the estimated values of the to-be-solved parameters in the target battery degradation model 1 and the target battery degradation model 2 through a parameter estimation method, wherein the parameter estimation method is a method for parameter estimation such as maximum likelihood estimation; for example, the maximum likelihood estimation is preferably used to construct the likelihood functions corresponding to the target battery degradation model 1 and the target battery degradation model 2 respectively:

[0098]

[0099] In the formula, L1(a, b, σ1) is the likelihood function of the target battery degradation model 1; L2(c, d, e, σ2) is the likelihood function of the target battery degradation model 2; a, b, c, d and e are to-be-estimated parameters of the model, σ1 is a to-be-estimated diffusion coefficient of the target battery degradation model 1, and σ2 is a to-be-estimated diffusion coefficient of the target battery degradation model 2. i represents the temperature corresponding to the i-th temperature level (for example, i = 1, 2, 3 represent 25℃, 35℃ and 45℃ respectively); j represents the serial number of the battery sample corresponding to each temperature level (i.e. the j-th battery); k represents the cycle number of the test; k start represents the input data from the start of the k start cycle test; k end represents the input data to the end of the k end cycle test; ΔY i,j,k represents the energy retention rate increment of the j-th battery at the i-th temperature level in the k-th cycle test.

[0100] The embodiment can preferably use the minimize function of the scipy library of Python to solve formula (8) and formula (9) respectively, that is, the parameters corresponding to the maximum value of formula (8) and formula (9) are taken as the parameter estimation values of the target battery degradation model 1 and the target battery degradation model 2 It should be noted that the calculation method and principle of the parameter estimation method such as the maximum likelihood estimation method are common knowledge in the art, and therefore will not be described herein for the sake of brevity.

[0101] Finally, the are substituted into the corresponding initial Eyring model, and the are generated as the target diffusion coefficient values corresponding to different target Eyring models.

[0102] Therefore, the target drift rate of the battery to be predicted can be calculated by substituting the target cycle test number and the target temperature level into the target Eyring model. It can be understood that the target drift rate takes into account the change thereof with the battery aging process.

[0103] Further, in an embodiment, the target Eyring model is:

[0104]

[0105] In the formula, μ1 represents the drift rate, k represents the cycle test number, T represents the temperature level, and are model parameter values calculated by the parameter estimation method.

[0106] For example, in the implementation of the prediction of the remaining service life of the battery to be predicted, if the synergistic effect of the temperature and the cycle test number on the drift rate is to be considered, the model parameter values calculated by the parameter estimation method such as the maximum likelihood estimation method are substituted into the initial Eyring model 1, and the target Eyring model 1 is generated: and

[0107]

[0108] In the formula (10), μ1(k, T) represents the corresponding drift rate under the synergistic effect of the temperature and the cycle test number.

[0109] Further, in an embodiment, the target Eyring model is:

[0110]

[0111] In the formula, μ2 represents the drift rate, k represents the cycle test number, T represents the temperature level, and are model parameter values calculated by the parameter estimation method.

[0112] ​Exemplarily, in the embodiment, when implementing the prediction of the remaining useful life of the battery to be predicted, if the independent effects of the temperature and the cycle test number on the drift rate are to be considered, the model parameter values calculated by the parameter estimation method such as the maximum likelihood estimation method are substituted into the initial Eyring model 2, and the target Eyring model 2 is generated: The model parameter values are substituted into the initial Eyring model 2, and the target Eyring model 2 is generated:

[0113]

[0114] μ2(k, T) in formula (11) represents the corresponding drift rate under the independent effects of the temperature and the cycle test number.

[0115] Step S30: Life prediction is performed based on the preset remaining useful life probability density function, the target cycle test number, the target drift rate, the target energy retention rate, and the preset target diffusion coefficient value, and the target remaining useful life of the battery to be predicted is obtained.

[0116] Exemplarily, it can be understood that, after the energy retention rate failure threshold ω of the given battery is given, the time at which the energy retention rate failure threshold is first reached can be used to define the failure time S of the battery; wherein, according to the important characteristics of the Wiener process, the time S at which the energy retention rate failure threshold is first reached obeys the inverse Gaussian distribution of transformation:

[0117] S ~ tIG(((1-ω) / μ),((1-ω) 2 / σ 2 )) (12)

[0118] It can be seen that the remaining useful life s (i.e. Remaining useful life, RUL) of the battery also obeys the inverse Gaussian distribution, and based on this, the inverse Gaussian distribution expression of the remaining useful life s based on the target battery degradation model can be derived, that is, the probability density function (i.e. remaining useful life probability density function) to which the remaining useful life s obeys.

[0119] Therefore, when evaluating the remaining useful life s of the battery to be predicted, only the target cycle test number, the target drift rate, the target energy retention rate, and the target diffusion coefficient value need to be substituted into the remaining useful life probability density function for life prediction, and the time metric index of the battery reliability can be calculated; it should be noted that, as shown in Table 1, the commonly used time metric indexes include: average RUL, mode RUL, median RUL, and 95% confidence RUL, wherein, Table 1 also lists the calculation methods of each index.

[0120] Table 1: Calculation method of reliability time metric index

[0121]

[0122] Therefore, the remaining useful life can be determined according to the selected specific reliability time metric indicator. For example, if the mode RUL is selected as the reliability time metric indicator, the k corresponding to the maximum of the remaining useful life probability density function is selected as the target remaining useful life of the battery to be predicted. It can be seen that the prediction accuracy of the remaining useful life of the battery can be effectively improved by the embodiment.

[0123] Further, in an embodiment, the remaining useful life probability density function is:

[0124]

[0125] In the formula, s represents the remaining useful life, T i represents the target temperature level, μ1(k ′ , T i ) represents the target drift rate corresponding to the last cycle test k end of the battery to be predicted under the condition of the target temperature level i, the target cycle test number k ′ , Y i (k end ) represents the target energy retention rate of the battery to be predicted at the last cycle test under the target temperature level, ω represents the energy retention rate failure threshold, represents the target diffusion coefficient value.

[0126] For example, in the embodiment, when evaluating the remaining useful life of the battery to be predicted, if the synergistic effect of temperature and cycle test number on drift rate is considered, the remaining useful life probability density function is constructed based on the energy retention rate Y i (k end ) of the battery to be predicted at the last cycle test under the target temperature level i, the target Eyring model 1, and the target diffusion coefficient value . Therefore, the target drift rate corresponding to the last cycle test k end of the battery to be predicted under the condition of the target temperature level i, the target cycle test number k ′ , the energy retention rate Y i (k end ), and the target diffusion coefficient value calculated by the target Eyring model 1 are substituted into the remaining useful life probability density function , and the remaining useful life of the battery to be predicted can be predicted.

[0127] Further, in an embodiment, the remaining useful life probability density function is:

[0128]

[0129] wherein s represents the remaining useful life, T i represents the target temperature level, μ2(k ′ represents the target temperature level, T i represents the corresponding target drift rate under the condition of the target temperature level, the last cycle of the cycle test k end corresponding target cycle number of the cycle test k ′ , Y i (k end ) represents the corresponding target energy retention rate of the last cycle of the cycle test of the battery to be predicted under the target temperature level, and ω represents the energy retention rate failure threshold, represents the target diffusion coefficient value.

[0130] Exemplarily, in the embodiment, when evaluating the remaining useful life of the battery to be predicted, if the individual effects of temperature and cycle number of the cycle test on the drift rate are considered, the remaining useful life probability density function i (k end ) of the battery to be predicted at the last cycle of the cycle test under the target temperature level i, the target Eyring model 2, and the target diffusion coefficient value are used to construct the remaining useful life probability density function

[0131] Therefore, the corresponding target drift rate under the condition of the target temperature level i, the last cycle of the cycle test k end corresponding target cycle number of the cycle test k ′ , the energy retention rate Y i (k end ), and the target diffusion coefficient value calculated by the target Eyring model 2 are substituted into the remaining useful life probability density function , so that the prediction of the remaining useful life of the battery to be predicted can be performed.

[0132] The following explains the principle of the prediction of the remaining useful life of the lithium ion battery under three environmental temperatures of 25°C, 35°C, and 45°C.

[0133] Firstly, constant power charge and discharge accelerated degradation cycle tests are carried out on lithium ion batteries at three ambient temperatures of 25°C, 35°C and 45°C respectively, until the energy retention rate of all batteries decays to 0.8 (i.e. the energy retention rate threshold is 0.8), that is, the battery energy retention rate degradation amount is 0.2, and the test is stopped, and the energy retention rate data of the battery in each cycle test is recorded; then the target battery degradation model 1 and the target battery degradation model 2 based on the drift rate change of the nonlinear Wiener process are constructed; then the maximum likelihood estimation method is used to estimate and solve the parameters in the target battery degradation model 1 and the target battery degradation model 2, and the results are shown in the following table:

[0134] Table 2 Parameter estimation values of target battery degradation model 1

[0135]

[0136] It should be understood that, in order to verify the fitting effect of the target battery degradation model 1 and the target battery degradation model 2 and calculate the error, the Monte Carlo method based on the Wiener process of the drift rate change can be preferably used to fit and generate the degradation curve, and the error is calculated based on the degradation curve. Among them, the parameter values estimated above are brought into the formula (6) and the formula (7) to obtain the parameterized target battery degradation model 1 and the target battery degradation model 2:

[0137] Parameterized target battery degradation model 1:

[0138] Parameterized target battery degradation model 2:

[0139] Then the Monte Carlo method is used to simulate and generate a plurality of Wiener process curves of the formula (13) and the formula (14), and the average values of these curves are calculated, and the average values are compared with the average values of the energy retention rate data of each battery at each temperature to calculate the error; wherein the error index can be selected as the mean absolute error (Mean absolute error, MAE), the mean absolute percentage error (Mean absolute percentage error, MAPE) and the root mean square error (Root mean square error, RMSE), and the calculation formulas are shown in the formula (15), the formula (16) and the formula (17):

[0140]

[0141] Wherein, Z i,k represents the average value of the energy retention rate of a plurality of simulated curves at the kth cycle test at the ith temperature level; Y i,k ​The average value of the energy retention rate of each battery at the kth cycle of the test at the ith temperature level; wherein the error results of the two models are as shown in the following table, respectively.

[0142] Table 4 Fitting error of parameterized target battery degradation model 1

[0143]

[0144] Table 5 Fitting error of parameterized target battery degradation model 2

[0145]

[0146] From the fitting error calculation results, it can be seen that the fitting error values of the parameterized target battery degradation model 1 and the parameterized target battery degradation model 2 are at a low level, which indicates that the models have excellent fitting accuracy.

[0147] Then, the energy retention rate data of the battery whose remaining useful life needs to be predicted at the last cycle of the test and the target cycle number corresponding to the last cycle of the test are substituted into the parameterized target battery degradation model 1 and the parameterized target battery degradation model 2, respectively, so as to draw the battery RUL probability density function curves at three temperature levels based on the parameterized target battery degradation model 1 and the parameterized target battery degradation model 2, respectively, as shown in Figures 2-7

[0148] Finally, according to the battery RUL probability density function curve, the average RUL, the median RUL, the mode RUL and the 95% confidence RUL of the battery at different temperatures can be obtained, and the results are shown in Tables 6 and 7.

[0149] Table 6 Battery RUL values predicted under the parameterized target battery degradation model 1

[0150]

[0151] Table 7 Battery RUL values predicted under the parameterized target battery degradation model 2

[0152]

[0153]

[0154] ​In summary, the embodiment takes temperature and cycle test number as factors affecting the battery degradation process, uses the Eyring model to construct the functional relationship between drift rate and temperature and cycle test number, and then improves the parameter estimation of the battery degradation model by the maximum likelihood estimation method to construct the RUL probability density function satisfying the inverse Gaussian distribution to realize the accurate prediction of the remaining useful life of the battery under various working temperatures. Compared with the method using the Arrhenius model, the embodiment considers that the drift rate will change with the battery aging process, and then introduces two variables of temperature and cycle test number to construct the functional relationship of the drift rate in the Wiener process, which can greatly reduce the fitting error of the model and significantly improve the accuracy of the predicted value.

[0155] In a second aspect, the embodiment of the present application further provides a battery remaining useful life prediction device.

[0156] In an embodiment, the battery remaining useful life prediction device comprises:

[0157] A parameter acquisition module is configured to acquire a target energy retention rate corresponding to the last cycle test and a target cycle test number corresponding to the last cycle test of the battery to be predicted under a target temperature level;

[0158] A parameter calculation module is configured to calculate a target drift rate by using a preset target Eyring model describing the functional relationship between the drift rate, temperature and cycle test number, the target cycle test number and the target temperature level;

[0159] A life prediction module is configured to perform life prediction based on a preset remaining useful life probability density function, the target cycle test number, the target drift rate, the target energy retention rate and a preset target diffusion coefficient value to obtain a target remaining useful life of the battery to be predicted.

[0160] The model parameter value in the target Eyring model and the target diffusion coefficient value are determined by solving a preset target battery degradation model based on a preset parameter estimation method, and the target battery degradation model is composed of an initial Eyring model and a Wiener process model corresponding to the change of the drift rate.

[0161] Further, in an embodiment, the target Eyring model is:

[0162]

[0163] In the formula, μ1 represents the drift rate, k represents the cycle test number, T represents the temperature level, and are model parameter values obtained by the parameter estimation method.

[0164] Further, in an embodiment, the remaining useful life probability density function is:

[0165]

[0166] where s represents the remaining useful life, T i represents the target temperature level, μ1(k ′ , T i ) represents the corresponding target drift rate under the condition of the target temperature level, the last cycle of the cycle test k end , and the corresponding target cycle of the cycle test k ′ . Y i (k end ) represents the corresponding target energy retention rate of the last cycle of the cycle test of the battery to be predicted under the target temperature level, and ω represents the energy retention rate failure threshold, represents the target diffusion coefficient value.

[0167] Further, in an embodiment, the target battery degradation model is:

[0168]

[0169] where Y1(k+1) represents the energy retention rate corresponding to the (k+1)th cycle of the cycle test, Y1(k) represents the energy retention rate corresponding to the kth cycle of the cycle test, T represents the temperature level, k represents the cycle number of the cycle test, represents the initial Eyring model, a and b are model parameters to be solved by the parameter estimation method, σ1 represents the diffusion coefficient value to be solved by the parameter estimation method, and B(1) represents the standard Brownian motion.

[0170] Further, in an embodiment, the target Eyring model is:

[0171]

[0172] where μ2 represents the drift rate, k represents the cycle number of the cycle test, T represents the temperature level, and are model parameter values solved by the parameter estimation method.

[0173] Further, in an embodiment, the remaining useful life probability density function is:

[0174]

[0175] where s represents the remaining useful life, T i represents the target temperature level, μ2(k ′ , T irepresents the target temperature level, k represents the last cycle of the cycle test, and Y represents the target drift rate corresponding to the condition of the target temperature level, k end represents the target temperature level, k represents the last cycle of the cycle test, and Y represents the target drift rate corresponding to the condition of the target temperature level, k ′ represents the target temperature level, k represents the last cycle of the cycle test, and Y represents the target drift rate corresponding to the condition of the target temperature level, k i (k end represents the target energy retention rate corresponding to the last cycle of the cycle test at the target temperature level, and ω represents the energy retention rate failure threshold, represents the target diffusion coefficient value.

[0176] Further, in an embodiment, the target battery degradation model is:

[0177]

[0178] In the formula, Y2(k+1) represents the energy retention rate corresponding to the k+1 cycle of the cycle test, Y2(k) represents the energy retention rate corresponding to the k cycle of the cycle test, T represents the temperature level, and k represents the cycle number of the cycle test, represents the initial Eling model, c, d, and e are model parameters to be solved by the parameter estimation method, σ2 represents the diffusion coefficient value to be solved by the parameter estimation method, and B(1) represents the standard Brownian motion.

[0179] The functions of each module in the battery remaining service life prediction device correspond to the steps in the battery remaining service life prediction method, and the functions and implementation processes will not be repeated here.

[0180] In a third aspect, the embodiments of the present application provide a battery remaining service life prediction device. The battery remaining service life prediction device can be a personal computer (PC), a notebook computer, a server, or other devices with data processing functions.

[0181] Referring to Figure 8 , Figure 8 FIG. 1 is a schematic diagram of a hardware structure of a battery remaining service life prediction device according to an embodiment of the present application. In the embodiments of the present application, the battery remaining service life prediction device can include a processor, a memory, a communication interface, and a communication bus.

[0182] The communication bus can be of any type, used to interconnect the processor, the memory, and the communication interface.

[0183] The communication interface includes an input / output (I / O) interface, a physical interface, and a logical interface, etc. for realizing interconnection of devices inside the battery remaining usage life prediction device, and an interface for realizing interconnection of the battery remaining usage life prediction device with other devices (e.g. other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber interface, an ATM interface, etc.; the user device can be a display, a keyboard, etc.

[0184] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0185] The processor can be a general-purpose processor, which can invoke the battery remaining usage life prediction program stored in the memory and execute the battery remaining usage life prediction method provided by the embodiments of the present application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed by the battery remaining usage life prediction program when invoked can refer to the various embodiments of the battery remaining usage life prediction method of the present application, which will not be repeated here.

[0186] Those skilled in the art can understand that the hardware structure shown in the above-mentioned embodiments is not a limitation of the present application, and can include more or less components than the illustrated components, or combine certain components, or different component arrangements. Figure 8

[0187] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium.

[0188] The battery remaining usage life prediction program is stored on the computer readable storage medium of the present application, and when the battery remaining usage life prediction program is executed by the processor, the steps of the battery remaining usage life prediction method as described above are implemented.

[0189] The method implemented by the battery remaining usage life prediction program when executed can refer to the various embodiments of the battery remaining usage life prediction method of the present application, which will not be repeated here.

[0190] ​It should be noted that the above-mentioned sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages or disadvantages of the embodiments.

[0191] The terms "comprise", "comprising", "include", "including", "have" and "having" in the specification and claims of the present application and the above-described drawings are intended to cover a non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed, or can optionally include other steps or units inherent to the process, method, product or device. The terms "first", "second" and "third" and the like descriptions are used to distinguish different objects, and do not represent the order or limit the types of "first", "second" and "third".

[0192] In the description of the embodiments of the present application, "exemplary", "for example", "for instance" or "such as" are used to represent an example, illustration or description. Any embodiment or design scheme described as "exemplary", "for example" or "for instance" in the embodiments of the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. In fact, the words "exemplary", "for example", "for instance" or the like are intended to present the relevant concept in a specific manner.

[0193] In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in the text only describes the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone, in addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0194] In some of the processes described in the embodiments of the present application, a plurality of operations or steps are included in a specific order, but it should be understood that these operations or steps can be executed or executed in parallel without the order in which they appear in the embodiments of the present application, and the sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes can include more or fewer operations, and these operations or steps can be executed in sequence or in parallel, and these operations or steps can be combined.

[0195] Those skilled in the art can clearly understand the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a plurality of instructions for making a terminal device execute the method described in each embodiment of the present application.

[0196] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method of predicting the remaining useful life of a battery, the method comprising: The battery remaining useful life prediction method comprises: obtaining a target energy retention rate corresponding to a last cycle of a cycle test at a target temperature level of a battery to be predicted and a target cycle test number corresponding to the last cycle of the cycle test; calculating a target drift rate by using a target Eyring model preset for describing a functional relationship between the drift rate and the temperature and the cycle test number, the target cycle test number and the target temperature level; performing life prediction based on a preset remaining useful life probability density function, the target cycle test number, the target drift rate, the target energy retention rate and a preset target diffusion coefficient value to obtain a target remaining useful life of the battery to be predicted; wherein the model parameter value in the target Eyring model and the target diffusion coefficient value are determined by solving a preset target battery degradation model based on a preset parameter estimation method, and the target battery degradation model is composed of an initial Eyring model and a Wiener process model corresponding to the drift rate change.

2. The battery remaining lifetime prediction method according to claim 1, wherein The target Eyring model is: In the formula, μ1 represents a drift rate, k represents a number of test cycles, and T represents a temperature level. and are model parameter values obtained by a parameter estimation method.

3. The battery remaining lifetime prediction method according to claim 2, wherein The remaining useful lifetime probability density function Is: where s represents the remaining service life, T i represents the target temperature level, μ1(k ′ , T i ) represents the corresponding target drift rate under the condition of the target temperature level, the last cycle of the cycle test k end , the corresponding target cycle of the cycle test k ′ , Y i (k end ) represents the corresponding target energy retention rate of the last cycle of the cycle test of the battery to be predicted under the target temperature level, and ω represents the energy retention rate failure threshold, represents the target diffusion coefficient value.

4. The battery remaining lifetime prediction method according to claim 1, wherein The target battery degradation model is: In the formula, Y1(k+1) represents the energy retention rate corresponding to the k+1th cycle of the cycle test, Y1(k) represents the energy retention rate corresponding to the kth cycle of the cycle test, T represents the temperature level, and k represents the cycle number of the cycle test, represents the initial Arrhenius model, a and b are model parameters to be solved by the parameter estimation method, σ1 represents a diffusion coefficient value to be solved by the parameter estimation method, and B(1) represents a standard Brownian motion.

5. The battery remaining lifetime prediction method according to claim 1, wherein The target Eyring model is: In the formula, μ2 represents a drift rate, k represents a number of test cycles, and T represents a temperature level. and are model parameter values obtained by a parameter estimation method.

6. The battery remaining lifetime prediction method according to claim 5, wherein The remaining useful lifetime probability density function is: Where s represents the remaining service life, T i Indicates the target temperature level, μ2(k ′ ,T i ) indicates that at the target temperature level, the last cycle test k end The corresponding target cycle test number k ′ The target drift rate corresponding to the condition of i (k end ) represents the target energy retention rate corresponding to the last cycle test of the battery to be predicted at the target temperature level, ω represents the energy retention failure threshold, Represents the target diffusion coefficient value.

7. The battery remaining lifetime prediction method according to claim 1, wherein The target battery degradation model is: In the formula, Y2(k+1) represents the energy retention rate corresponding to the k+1th cycle of the cyclic test, Y2(k) represents the energy retention rate corresponding to the kth cycle of the cyclic test, T represents the temperature level, and k represents the cycle number of the cyclic test, represents the initial Arrhenius model, c, d and e are model parameters to be solved by parameter estimation, σ2 represents a diffusion coefficient value to be solved by parameter estimation, and B(1) represents a standard Brownian motion.

8. A battery remaining life prediction device, characterized by, The battery remaining useful life prediction device comprises: a parameter acquisition module configured to obtain a target energy retention rate corresponding to a last cycle of a cycle test at a target temperature level of a battery to be predicted and a target cycle test number corresponding to the last cycle of the cycle test; a parameter calculation module configured to calculate a target drift rate by using a target Eyring model preset for describing a functional relationship between the drift rate and the temperature and the cycle test number, the target cycle test number and the target temperature level; a life prediction module configured to perform life prediction based on a preset remaining useful life probability density function, the target cycle test number, the target drift rate, the target energy retention rate and a preset target diffusion coefficient value to obtain a target remaining useful life of the battery to be predicted; wherein the model parameter value in the target Eyring model and the target diffusion coefficient value are determined by solving a preset target battery degradation model based on a preset parameter estimation method, and the target battery degradation model is composed of an initial Eyring model and a Wiener process model corresponding to the drift rate change.

9. A battery remaining life prediction device characterized by comprising: The battery remaining useful life prediction device comprises a processor, a memory and a battery remaining useful life prediction program stored on the memory and executable by the processor, wherein the battery remaining useful life prediction program, when executed by the processor, implements the steps of the battery remaining useful life prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a battery remaining useful life prediction program, wherein the battery remaining useful life prediction program, when executed by the processor, implements the steps of the battery remaining useful life prediction method according to any one of claims 1 to 7.

Images