A method and apparatus for predicting battery state of health

Abstract

The application provides a battery state of health prediction method and device, a prior distribution of a target parameter is determined, the target parameter is a parameter in a decay empirical formula, the decay empirical formula is used for predicting a state of health SOH of a battery according to working condition data of the battery; based on the decay empirical formula, the prior distribution of the SOH of the battery is obtained according to the prior distribution of the target parameter and historical working conditions of the battery; the posterior distribution of the target parameter of the battery is determined according to the prior distribution of the SOH of the battery and historical SOH of the battery, the historical SOH is the SOH of the battery actually measured; the SOH of the battery is predicted according to the posterior distribution of the target parameter of the battery and target working conditions of the battery. The application can comprehensively consider the overall law and individual difference of the decay of the battery, and more accurately predict the SOH of the battery under the target working conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of battery evaluation, in particular to a battery state of health prediction method and device. BACKGROUND

[0002] The state of health (SOH) of a battery is often used to represent the ability of the current battery to store electrical energy relative to a new battery, and the value of SOH is generally defined as the ratio of the current capacity of the battery to the rated capacity of the battery. The SOH of the battery represents the capacity degradation of the battery and is closely related to the remaining life of the battery.

[0003] Currently, the SOH value of the battery is estimated mainly by using a pre-determined battery degradation empirical formula. The battery degradation empirical formula is a formula summarized by experiments and expert experience, and the parameters of the degradation empirical formula are generally determined by fitting battery degradation experimental data under various conditions.

[0004] However, there are performance differences between different batteries, and even batteries produced by the same technology and formula will have certain differences in the degradation process. Since the traditional degradation empirical formula cannot consider the individual differences of the battery, the above method cannot accurately predict the SOH value of the battery in actual application, and thus it is difficult to accurately estimate the remaining life of the battery. SUMMARY

[0005] Therefore, the present application provides a battery state of health prediction method and device, which considers the degradation differences of individual batteries on the basis of the battery degradation empirical formula, and is beneficial to improving the accuracy of battery state of health prediction.

[0006] In one aspect, the present application provides a battery state of health prediction method, which comprises:

[0007] determining a prior distribution of a target parameter, the target parameter being a parameter in a degradation empirical formula, and the degradation empirical formula being used to predict the state of health (SOH) of a battery according to working condition data of the battery;

[0008] based on the degradation empirical formula, obtaining a prior distribution of the SOH of the battery according to the prior distribution of the target parameter and the historical working condition of the battery;

[0009] determining a posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and the historical SOH of the battery, the historical SOH being the SOH of the battery actually measured;

[0010] predicting the SOH of the battery according to the posterior distribution of the target parameter of the battery and the target working condition of the battery.

[0011] Optionally, the determining the prior distribution of the target parameter comprises:

[0012] determining a hyper-prior distribution of the target parameter, the hyper-prior distribution of the target parameter being a distribution of parameters applicable to determining the prior distribution of the target parameter;

[0013] sampling the parameters of the prior distribution of the target parameter from the hyper-prior distribution of the target parameter, and determining the prior distribution of the target parameter according to the sampled parameters of the prior distribution.

[0014] Optionally, the obtaining the prior distribution of the SOH of the battery based on the attenuation empirical formula and the prior distribution of the target parameter and the historical working conditions of the battery comprises:

[0015] sampling the prior distribution of the target parameter to obtain a sample of the prior distribution of the target parameter;

[0016] obtaining the prior distribution of the SOH of the battery based on the attenuation empirical formula and the sample of the prior distribution of the target parameter and the historical working conditions of the battery.

[0017] Optionally, the determining the posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and the historical SOH of the battery comprises:

[0018] calculating a likelihood function of the target parameter according to the prior distribution of the SOH of the battery and the historical SOH of the battery;

[0019] determining the posterior distribution of the target parameter of the battery based on a Bayesian formula and the likelihood function of the target parameter.

[0020] Optionally, the target working condition of the battery is a future working condition of the battery.

[0021] the future working condition is obtained by prediction in the following manner:

[0022] obtaining a prior distribution of the working condition of the battery;

[0023] obtaining a posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery, the historical working condition being the actually measured working condition of the battery;

[0024] taking the posterior distribution of the working condition of the battery as the future working condition of the battery.

[0025] Optionally, the obtaining the posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery comprises:

[0026] sampling from the prior distribution of the operating condition of the battery to obtain a sample of the prior distribution of the operating condition of the battery;

[0027] obtaining a posterior distribution of the operating condition of the battery according to the sample of the prior distribution of the operating condition of the battery and the historical operating condition of the battery.

[0028] Optionally, the obtaining the posterior distribution of the operating condition of the battery according to the prior distribution of the operating condition of the battery and the historical operating condition of the battery comprises:

[0029] calculating a likelihood function of the operating condition of the battery according to the prior distribution of the operating condition of the battery and the historical operating condition of the battery;

[0030] obtaining the posterior distribution of the operating condition of the battery according to the likelihood function of the operating condition of the battery based on a Bayesian formula.

[0031] In another aspect, the application further provides a device for predicting a state of health of a battery, the device comprising:

[0032] a target parameter prior distribution determination module configured to determine a prior distribution of a target parameter, the target parameter being a parameter in an empirical formula for attenuation, the empirical formula for attenuation being used to predict a state of health (SOH) of the battery according to operating condition data of the battery;

[0033] an SOH prior distribution determination module configured to obtain a prior distribution of the SOH of the battery based on the empirical formula for attenuation, the prior distribution of the target parameter and a historical operating condition of the battery;

[0034] a target parameter posterior distribution determination module configured to determine a posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and a historical SOH of the battery, the historical SOH being the SOH of the battery actually measured;

[0035] a prediction module configured to predict the SOH of the battery according to the posterior distribution of the target parameter of the battery and a target operating condition of the battery.

[0036] In another aspect, the application further provides a device, the device comprising: a processor and a memory;

[0037] the memory is configured to store instructions;

[0038] the processor is configured to execute the instructions in the memory, and execute the method in the above aspects.

[0039] In another aspect, the application also provides a computer readable storage medium, which stores program codes or instructions, when running on a computer, to enable the computer to perform the method of the above aspect.

[0040] Therefore, the embodiments of the application have the following beneficial effects:

[0041] The application first determines the prior distribution of the target parameter, which is the parameter in the attenuation empirical formula; based on the attenuation empirical formula, the prior distribution of the SOH of the battery is obtained according to the prior distribution of the target parameter and the historical working conditions of the battery; the posterior distribution of the target parameter of the battery is determined according to the prior distribution of the SOH of the battery and the historical SOH of the battery; and the SOH of the battery is predicted according to the posterior distribution of the target parameter of the battery and the target working condition of the battery. The application first determines the prior distribution of the SOH of the battery by combining the prior distribution of the target parameter in the attenuation empirical formula and the historical working conditions of the battery. The prior distribution of the SOH of the battery reflects the distribution of the predicted SOH of the battery based on the attenuation empirical formula, and can represent the overall law of battery attenuation. Then, the distribution law of the target parameter is adjusted according to the difference between the prior distribution of the SOH of the battery and the actually measured historical SOH of the battery, to obtain the posterior distribution of the target parameter of the attenuation empirical formula that is suitable for the specific battery attenuation process, that is, to obtain the posterior distribution of the target parameter that matches the individual situation of the specific battery. Furthermore, the SOH of the battery is predicted based on the posterior distribution of the target parameter, so as to comprehensively consider the overall law of battery attenuation and the individual difference of the battery, and to more accurately predict the SOH of the battery under the target working condition. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0043] Figure 1 A flowchart of a battery health state prediction method provided by an embodiment of the application;

[0044] Figure 2 A schematic diagram of a battery health state prediction device provided by an embodiment of the application. DETAILED DESCRIPTION

[0045] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the scope of protection of the present application.

[0046] At present, the estimation of the SOH value of a battery is mainly based on a battery attenuation empirical formula. The battery attenuation empirical formula is used to represent the relationship between various working condition data of the battery and the SOH value of the battery. Generally, the parameters in the attenuation empirical formula are determined by expert experience or experiments, and mainly reflect the attenuation law of the whole battery of a certain type or a certain model, without considering the performance difference between different battery individuals, that is, it is generally difficult to accurately reflect the attenuation law of a single battery, and accordingly, it is also difficult to accurately estimate the remaining life of the battery based on the attenuation empirical formula.

[0047] In order to solve the above problems, the present application provides a battery health state prediction method and device, which adjusts the parameters of the attenuation empirical formula to make it suitable for the attenuation process of a specific battery, thereby facilitating the improvement of the accuracy of SOH prediction of the specific battery.

[0048] In order to facilitate understanding, the battery health state prediction method and device provided by the embodiments of the present application will be described in detail below in conjunction with the drawings.

[0049] Reference Figure 1 As shown in the figure, the flowchart of the battery health state prediction method provided by the embodiments of the present application can include the following steps:

[0050] S101: Determine the prior distribution of the target parameter.

[0051] The target parameter is a parameter in the attenuation empirical formula, and the attenuation empirical formula is used to predict the health state SOH of the battery according to the working condition data of the battery.

[0052] In the embodiments of the present application, the attenuation empirical formula of the battery is used to represent the relationship between the health state of the battery and the working condition data. The parameter in the attenuation empirical formula is represented by θ, and the various working condition data is represented by X, and the attenuation empirical formula can be represented as SOH = f(X|θ). When applied to the SOH prediction of a specific battery, the parameter θ in the above attenuation empirical formula is regarded as the target parameter.

[0053] In the embodiment of the present application, the modeling method of the hierarchical Bayesian model is adopted, and first, the prior distribution of the target parameter needs to be determined. The prior distribution of the target parameter is the probability distribution of the target parameter determined according to prior knowledge. Specifically, the prior knowledge can be expert experience and the like, and is mainly based on the overall cognition of the characteristics of the same type or the same type of battery. For example, for different batteries produced by the same technology and formula, the process of SOH degradation of the battery has certain differences, but the overall degradation law is similar, and therefore, the probability distribution of the value of the target parameter can be estimated based on the overall degradation law and the degradation experience formula. It should be noted that since the degradation experience formula can have multiple parameters, the target parameter can also represent multiple parameters in the degradation experience formula, and the prior distribution of the target parameter can represent the prior distribution of each of the multiple parameters in the degradation experience formula. It should be noted that the prior distribution of the target parameter is irrelevant to the degradation law of a specific battery, but is based on the hypothesis and estimation of the target parameter based on prior knowledge.

[0054] When constructing the hierarchical Bayesian model of the target parameter, the hyper-prior distribution of the target parameter can be selected. According to the possible value range of the parameter and the related experience knowledge, a suitable statistical distribution can be selected as the hyper-prior distribution and the prior distribution of the target parameter. The hyper-prior distribution of the target parameter is a distribution used to determine the prior distribution of the target parameter, and specifically can be the distribution of the parameters of the prior distribution and the like. There are the following specific implementation manners:

[0055] In one possible implementation manner, the determination of the prior distribution of the target parameter includes:

[0056] determining the hyper-prior distribution of the target parameter, the hyper-prior distribution of the target parameter being a distribution suitable for determining the prior distribution of the target parameter;

[0057] sampling the parameters of the prior distribution of the target parameter from the hyper-prior distribution of the target parameter, and determining the prior distribution of the target parameter according to the sampled parameters of the prior distribution.

[0058] The hyper-prior distribution of the target parameter is also determined based on prior knowledge, and specifically, the prior knowledge can be expert experience, etc., and is mainly based on overall cognition of characteristics of the same model or the same type of battery. The hyper-prior distribution of the target parameter is matched with the prior distribution of the target parameter, and can be specifically determined by the prior distribution and the parameter type of the prior distribution; for example, if the prior distribution of the target parameter is selected as a Gaussian distribution, and the parameter type of the prior distribution is the mean and the variance, then a Gaussian distribution can be selected as the hyper-prior distribution of the mean, and a HalfCauchy distribution can be selected as the prior distribution of the variance; if the prior distribution of the target parameter is selected as a binomial distribution, and the parameter type of the prior distribution is the number of experiments n and the probability p, then a Poisson distribution can be selected as the hyper-prior distribution of n, and a Gaussian distribution can be selected as the hyper-prior distribution of p.

[0059] The selection of the hyper-prior distribution of the target parameter will be described in detail below by taking the parameter type of the prior distribution of the target parameter as the mean and the variance as an example. For example, based on prior knowledge, it is understood that the value range of a certain target parameter is 0-1, based on experiments on the same model battery, it is obtained that the mean of the target parameter is a, the mean of the prior distribution of the target parameter is μ, and the variance of the prior distribution of the target parameter is σ, then: a Beta distribution with the mean a and the variance 1 can be selected as the prior distribution of μ, that is, μ ~ Beta(a, 1); a HalfCauchy distribution with the variance 1 can be selected as the prior distribution of σ, that is, σ ~ HalfCauchy(1). It should be noted that the above distribution form is only an example, and the selection of the specific distribution form is not limited herein.

[0060] According to the hyper-prior distribution of the target parameter, the mean and the variance of the prior distribution of the target parameter can be sampled, and specifically, the sampling can be performed based on a Markov Chain Monte Carlo (MCMC) method or a sampling method of variational inference (VI) in the sampling process. According to the sampled mean and variance, the prior distribution of the target parameter can be specifically determined. Using the example in the above, from the prior distribution of the mean μ ~ Beta(a, 1) and the prior distribution of the variance σ ~ HalfCauchy(1), the prior distribution of the target parameter can be obtained as θ ~ Beta(μ, σ). It should be noted that the above distribution form is only an example, and the selection of the specific distribution form is not limited herein.

[0061] S102: Based on the attenuation empirical formula, the prior distribution of the SOH of the battery is obtained according to the prior distribution of the target parameter and the historical working conditions of the battery.

[0062] In the embodiments of the present application, the prior distribution of the target parameter θ and the historical working conditions X of the battery are substituted into the attenuation empirical formula SOH=f(X|θ) to obtain the prior distribution of the SOH of the battery. The prior distribution of the SOH of the battery is a distribution of the SOH of the battery determined according to prior knowledge, and is irrelevant to the actually measured SOH of the battery.

[0063] In a possible implementation, the prior distribution of the SOH of the battery is obtained according to the prior distribution of the target parameter and the historical working conditions of the battery based on the attenuation empirical formula, including:

[0064] sampling from the prior distribution of the target parameter to obtain a sample of the prior distribution of the target parameter;

[0065] obtaining the prior distribution of the SOH of the battery according to the sample of the prior distribution of the target parameter and the historical working conditions of the battery based on the attenuation empirical formula.

[0066] In the embodiments of the present application, the sample of the prior distribution of the target parameter is obtained by sampling from the prior distribution of the target parameter, and specifically, the sampling can be performed based on the Markov Monte Carlo method or the sampling method of variational inference. According to the attenuation empirical formula SOH=f(X|θ), the sample is selected from the prior distribution of the target parameter θ as the value of the target parameter θ, and the historical working conditions of the battery are substituted into the attenuation empirical formula as X to obtain the SOH of the battery. Specifically, the sample of the prior distribution of the target parameter can include multiple samples, and the historical working conditions of the battery can also be organized in the form of a statistical distribution; according to the multiple samples of the prior distribution of the target parameter and the distribution of the historical working conditions of the battery, a distribution of the SOH of the battery can be obtained, and the distribution is taken as the prior distribution of the SOH of the battery.

[0067] S103: determining the posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and the historical SOH of the battery, the historical SOH being the actually measured SOH of the battery.

[0068] In the embodiments of the present application, the prior distribution of the SOH of the battery can be obtained according to the prior distribution of the target parameter of the battery, that is, the prior distribution of the SOH of the battery under the prior distribution of the target parameter is assumed in advance; based on the actually measured historical SOH of the battery, the target parameter of the battery can be adjusted to adapt to the attenuation law of the specific battery to obtain the posterior distribution of the target parameter of the battery. It should be noted that the historical SOH is obtained by measuring the SOH of the specific battery.

[0069] In a possible implementation, the posterior distribution of the target parameter of the battery is determined according to the prior distribution of the SOH of the battery and the historical SOH of the battery, including:

[0070] The likelihood function of the target parameter is calculated according to the prior distribution of the SOH of the battery and the historical SOH of the battery.

[0071] The posterior distribution of the target parameter of the battery is determined according to the likelihood function of the target parameter based on the Bayes formula.

[0072] When the prediction method based on the hierarchical Bayesian model is applied, the likelihood function of the target parameter can be calculated according to the prior distribution of the SOH of the battery and the historical SOH of the battery. The likelihood function can be used to represent the likelihood of the parameter in the statistical model; the likelihood is used to estimate the parameter of the nature of things when the result obtained by some observations is known. In the embodiments of the present application, the historical SOH of the battery can be specifically taken as the true value of the SOH, the assumed value of the SOH is selected from the prior distribution of the SOH of the battery, and the likelihood function of the target parameter calculated by the true value and the assumed value of the SOH can reflect the degree to which the SOH predicted based on the attenuation empirical formula approaches the true value when the target parameter takes different assumed values. Specifically, the assumed value of the SOH can be selected from the prior distribution of the SOH of the battery by using the sampling method, and the sampling process can be based on the Markov Monte Carlo method or the sampling method of variational inference. After the likelihood function of the target parameter is determined, the posterior distribution of the target parameter can be obtained by substituting the likelihood function of the target parameter into the Bayes formula. Thus, the adjustment of making the target parameter closer to the true value is realized, and the posterior distribution of the target parameter obtained is suitable for the attenuation law of the specific battery. Specifically, when calculating the likelihood function of the target parameter, the HalfCauchy distribution or other distribution forms can be selected as the distribution of the measurement error considering that the historical SOH has measurement error, and the selection of the distribution form is not limited herein.

[0073] S104: The SOH of the battery is predicted according to the posterior distribution of the target parameter of the battery and the target working condition of the battery.

[0074] In the embodiments of the present application, the value of the target parameter can be selected from the posterior distribution of the target parameter of the battery, and the target working condition of the battery is substituted into the attenuation empirical formula to calculate the SOH of the battery, thereby realizing the prediction of the SOH of the battery. In a possible implementation, the value of the target parameter selected from the posterior distribution of the target parameter can be obtained by using the sampling method, and the sampling process can be based on the Markov Monte Carlo method or the sampling method of variational inference. The target working condition of the battery can be determined according to the specific application scenario, and can be a pre-defined working condition or a prediction of the future working condition.

[0075] In the embodiments of the application, firstly, a prior distribution of a target parameter is determined, the target parameter being a parameter in an empirical formula of attenuation; based on the empirical formula of attenuation, the prior distribution of SOH of the battery is obtained according to the prior distribution of the target parameter and historical working conditions of the battery; the posterior distribution of the target parameter of the battery is determined according to the prior distribution of SOH of the battery and the historical SOH of the battery; and the SOH of the battery is predicted according to the posterior distribution of the target parameter of the battery and the target working condition of the battery. Based on the method of the hierarchical Bayesian model, firstly, the prior distribution of the target parameter is determined, and the prior distribution of SOH of the battery is determined in combination with the historical working conditions; and then the posterior distribution of the target parameter of the empirical formula of attenuation suitable for the specific battery attenuation process is obtained by adjusting the distribution rule of the target parameter in combination with the actually measured SOH of the battery, so that the individual differences of the battery can be considered, and the SOH of the battery under the target working condition can be more accurately predicted.

[0076] When the target working condition of the battery is a future working condition, since the working condition of the battery has uncertainty, in actual use, the influence of various factors on the working condition of the battery needs to be considered, and if the SOH of the battery under the future working condition needs to be accurately predicted, the accuracy of the prediction of the future working condition of the battery needs to be improved. In a possible implementation manner, the future working condition of the battery can be predicted based on the method of the hierarchical Bayesian model.

[0077] In a possible implementation manner, the target working condition of the battery is a future working condition of the battery.

[0078] The future working condition is obtained by prediction in the following manner:

[0079] The prior distribution of the working condition of the battery is obtained.

[0080] The posterior distribution of the working condition of the battery is obtained according to the prior distribution of the working condition of the battery and the historical working condition of the battery, the historical working condition being the actually measured working condition of the battery.

[0081] The posterior distribution of the working condition of the battery is taken as the future working condition of the battery.

[0082] In an embodiment of the present application, the prior distribution of the working condition of the battery can be an assumption of the working condition distribution of the battery according to experience. The working condition of the battery can include various data such as Ah (ampere-hour), t (usage time), I (current), T (temperature), SOC (State Of Charge), DOD (Depth of discharge), and the like. Specifically, a suitable statistical distribution can be selected as the prior distribution of the working condition by the value range of the working condition and the like. For example, for the SOC of the battery, since the value is 0% to 100%, a gamma distribution or a beta distribution can be used, and the distribution of the temperature is generally a Gaussian distribution, and a Gaussian distribution can be selected as the prior distribution of the temperature, for example, a Gaussian distribution with a mean of 25 and a variance of 10 is selected as the prior distribution of the temperature.

[0083] In a possible implementation, the posterior distribution of the working condition of the battery is obtained according to the prior distribution of the working condition of the battery and the historical working condition of the battery, and includes:

[0084] sampling from the prior distribution of the working condition of the battery to obtain a sample of the prior distribution of the working condition of the battery;

[0085] obtaining the posterior distribution of the working condition of the battery according to the sample of the prior distribution of the working condition of the battery and the historical working condition of the battery.

[0086] In an embodiment of the present application, the sampling process can be based on the Markov Monte Carlo method or the sampling method of variational inference.

[0087] In a possible implementation, the posterior distribution of the working condition of the battery is obtained according to the prior distribution of the working condition of the battery and the historical working condition of the battery, and includes:

[0088] calculating a likelihood function of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery;

[0089] obtaining the posterior distribution of the working condition of the battery according to the likelihood function of the working condition of the battery based on the Bayes formula.

[0090] In the embodiments of the present application, the likelihood function of the working condition can be calculated according to the working condition sampled from the prior distribution and the historical working condition actually measured; the calculated likelihood function of the working condition can reflect the degree to which the working condition is close to the true value of the working condition at different values. Specifically, when calculating the likelihood function of the working condition, considering that the historical working condition has measurement error, a HalfCauchy distribution or other distribution form can be selected as the distribution of the measurement error, and the selection of the distribution form is not limited herein. After the likelihood function of the working condition is determined, the likelihood function of the working condition is substituted into the Bayes formula, the posterior distribution of the working condition can be obtained, which is closer to the true distribution law of the working condition, thereby improving the accuracy of the prediction of the future working condition.

[0091] Based on the above battery state of health prediction method, the embodiments of the present application also provide a battery state of health prediction device, referring to Figure 2 the figure is a schematic diagram of a battery state of health prediction device provided by the embodiments of the present application, which can include:

[0092] The target parameter prior distribution determination module 201 is configured to determine the prior distribution of the target parameter, the target parameter being a parameter in the attenuation empirical formula, and the attenuation empirical formula being used to predict the state of health SOH of the battery according to the working condition data of the battery;

[0093] The SOH prior distribution determination module 202 is configured to obtain the prior distribution of the SOH of the battery based on the attenuation empirical formula, the prior distribution of the target parameter and the historical working condition of the battery;

[0094] The target parameter posterior distribution determination module 203 is configured to determine the posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and the historical SOH of the battery, the historical SOH being the SOH of the battery actually measured;

[0095] The prediction module 204 is configured to predict the SOH of the battery according to the posterior distribution of the target parameter of the battery and the target working condition of the battery.

[0096] In a possible implementation manner, the target parameter prior distribution determination module is specifically configured to determine the super-prior distribution of the target parameter, the super-prior distribution of the target parameter being a distribution of parameters suitable for determining the prior distribution of the target parameter;

[0097] sample the parameters of the prior distribution of the target parameter from the super-prior distribution of the target parameter, and determine the prior distribution of the target parameter according to the sampled parameters of the prior distribution.

[0098] In a possible implementation manner, the SOH prior distribution determination module is specifically configured to: sample from the prior distribution of the target parameter to obtain a sample of the prior distribution of the target parameter.

[0099] Based on the attenuation empirical formula, the SOH prior distribution of the battery is obtained according to the sample of the prior distribution of the target parameter and the historical working condition of the battery.

[0100] In a possible implementation manner, the target parameter posterior distribution determination module is specifically configured to: calculate a likelihood function of the target parameter according to the SOH prior distribution of the battery and the historical SOH of the battery.

[0101] Based on the Bayesian formula, the posterior distribution of the target parameter of the battery is determined according to the likelihood function of the target parameter.

[0102] In a possible implementation manner, the target working condition of the battery is a future working condition of the battery.

[0103] The future working condition is obtained according to the following manner:

[0104] A prior distribution of the working condition of the battery is obtained.

[0105] A posterior distribution of the working condition of the battery is obtained according to the prior distribution of the working condition of the battery and the historical working condition of the battery, and the historical working condition is the actually measured working condition of the battery.

[0106] The posterior distribution of the working condition of the battery is taken as the future working condition of the battery.

[0107] In a possible implementation manner, the obtaining the posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery comprises:

[0108] A sample of the prior distribution of the working condition of the battery is obtained by sampling from the prior distribution of the working condition of the battery.

[0109] The posterior distribution of the working condition of the battery is obtained according to the sample of the prior distribution of the working condition of the battery and the historical working condition of the battery.

[0110] In a possible implementation manner, the obtaining the posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery comprises:

[0111] A likelihood function of the working condition of the battery is calculated according to the prior distribution of the working condition of the battery and the historical working condition of the battery.

[0112] Based on the Bayesian formula, a posterior distribution of the working condition of the battery is obtained according to a likelihood function of the working condition of the battery.

[0113] Based on the above battery health state prediction method, an embodiment of the present application further provides a device, which can comprise a processor and a memory.

[0114] The memory is configured to store instructions.

[0115] The processor is configured to execute the instructions in the memory, and execute the above battery health state prediction method.

[0116] Based on the above battery health state prediction method, an embodiment of the present application further provides a computer readable storage medium, which stores program codes or instructions, and when the program codes or instructions are executed on a computer, the computer is caused to execute the above battery health state prediction method.

[0117] It should be noted that the embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts of each embodiment can be referred to each other. For the system or device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0118] It should be understood that in the present application, "at least one" means one or more, and "multiple" means two or more. "And / or" is used to describe the association relationship of the associated objects, which means that there can be three kinds of relationships, for example, "A and / or B" can represent three cases: only A, only B, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally represents that the associated objects before and after are in an "or" relationship. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c, can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0119] It is also to be noted that, as used in the specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless otherwise indicated. Furthermore, to the extent that the terms "including," "includes," "having," "has," "with," or "contains" are used in either the detailed description and the claims, such terms are intended to be inclusive in a manner similar to the term "comprising" as an open transition term without precluding any additional or other elements.

[0120] The embodiments disclosed herein can each be implemented as a method, apparatus, or article of manufacture using programming instructions. The embodiments disclosed herein can be implemented using software, firmware, hardware, or a combination thereof. The various elements of the disclosed embodiments, as well as the embodiments themselves, can be constructed from any combination of hardware, software, and / or firmware. The software implementation can be implemented by one or more software modules using object-oriented design methodology, among other techniques. The software modules can be stored on any computer-readable medium, including RAM, ROM, EEPROM, flash memory, or a hard disk, to name a few. The software modules can include one or more routines.

[0121] The above description of disclosed embodiments provides enough information to enable those skilled in the art to make and use the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of predicting a state of health of a battery, the method comprising: The method comprises: determining a prior distribution of a target parameter, the target parameter being a parameter in an empirical formula for predicting a state of health SOH of a battery according to working condition data of the battery; obtaining a prior distribution of the SOH of the battery based on the empirical formula, the prior distribution of the target parameter and historical working conditions of the battery; determining a posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and a historical SOH of the battery, the historical SOH being an actually measured SOH of the battery; predicting the SOH of the battery according to the posterior distribution of the target parameter of the battery and a target working condition of the battery; the obtaining of the prior distribution of the SOH of the battery based on the empirical formula, the prior distribution of the target parameter and the historical working conditions of the battery comprises: sampling the prior distribution of the target parameter to obtain a sample of the prior distribution of the target parameter; and obtaining the prior distribution of the SOH of the battery based on the sample of the prior distribution of the target parameter and the historical working conditions of the battery; the determination of the prior distribution of the target parameter comprises: determining a hyper-prior distribution of the target parameter, the hyper-prior distribution of the target parameter being a distribution of a parameter suitable for determining the prior distribution of the target parameter, the hyper-prior distribution being determined by the prior distribution and a type of parameter of the prior distribution; sampling the hyper-prior distribution of the target parameter to obtain the parameter of the prior distribution of the target parameter, and determining the prior distribution of the target parameter according to the sampled parameter of the prior distribution.

2. The method of claim 1, wherein, the determination of the posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and the historical SOH of the battery comprises: calculating a likelihood function of the target parameter according to the prior distribution of the SOH of the battery and the historical SOH of the battery; determining the posterior distribution of the target parameter of the battery based on the likelihood function of the target parameter according to a Bayesian formula.

3. The method of claim 1, wherein, the target working condition of the battery is a future working condition of the battery; the future working condition is predicted in the following manner: obtaining a prior distribution of a working condition of the battery; obtaining a posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and a historical working condition of the battery, the historical working condition being an actually measured working condition of the battery; taking the posterior distribution of the working condition of the battery as the future working condition of the battery.

4. The method of claim 3, wherein, the obtaining of the posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery comprises: sampling the prior distribution of the working condition of the battery to obtain a sample of the prior distribution of the working condition of the battery; obtaining the posterior distribution of the working condition of the battery according to the sample of the prior distribution of the working condition of the battery and the historical working condition of the battery.

5. The method of claim 3, wherein, the obtaining of the posterior distribution of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery comprises: calculating a likelihood function of the working condition of the battery according to the prior distribution of the working condition of the battery and the historical working condition of the battery; Based on a Bayesian formula, a posterior distribution of the working condition of the battery is obtained according to a likelihood function of the working condition of the battery.

6. A battery state of health prediction apparatus characterized by comprising: The device comprises: A target parameter prior distribution determination module is configured to determine a prior distribution of a target parameter, the target parameter being a parameter in an attenuation empirical formula, and the attenuation empirical formula being used to predict a state of health (SOH) of a battery according to working condition data of the battery. An SOH prior distribution determination module is configured to obtain a prior distribution of an SOH of the battery based on the attenuation empirical formula, the prior distribution of the target parameter, and historical working conditions of the battery. A target parameter posterior distribution determination module is configured to determine a posterior distribution of the target parameter of the battery according to the prior distribution of the SOH of the battery and a historical SOH of the battery, the historical SOH being an actually measured SOH of the battery. A prediction module is configured to predict the SOH of the battery according to the posterior distribution of the target parameter of the battery and a target working condition of the battery. The SOH prior distribution determination module is specifically configured to sample the prior distribution of the target parameter to obtain a sample of the prior distribution of the target parameter, and obtain the prior distribution of the SOH of the battery based on the sample of the prior distribution of the target parameter and the historical working conditions of the battery. The target parameter prior distribution determination module is specifically configured to determine a hyper-prior distribution of the target parameter, the hyper-prior distribution of the target parameter being a distribution of a parameter used to determine the prior distribution of the target parameter, and the hyper-prior distribution being determined by the prior distribution and a type of parameter of the prior distribution; sample the hyper-prior distribution of the target parameter to obtain the parameter of the prior distribution of the target parameter, and determine the prior distribution of the target parameter according to the sampled parameter of the prior distribution.

7. An apparatus, comprising: The device comprises a processor and a memory. The memory is configured to store instructions. The processor is configured to execute the instructions in the memory, and execute the method in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores program codes or instructions, which, when executed on a computer, cause the computer to execute the method in any one of claims 1-5. The computer readable storage medium stores program codes or instructions, which, when executed on a computer, cause the computer to execute the method in any one of claims 1-5.

Images