Energy storage battery service life prediction method and device

Abstract

Embodiments of the present invention relate to computer technology, and in particular to a method and device for predicting the service life of energy storage batteries. The method comprises: obtaining early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; determining a first health factor corresponding to the current energy storage battery based on the early external characteristic data in each energy storage battery; performing principal component analysis on the early external characteristic data in all energy storage batteries to obtain a second health factor for each energy storage battery; using the first health factor and the second health factor of the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, training a preset machine learning model to obtain a prediction model; using the first health factor and the second health factor corresponding to each energy storage battery to be tested as input and inputting them into the prediction model to obtain the service life of the current energy storage battery to be tested. The solution provided by the present invention can achieve a higher accuracy prediction of the service life of energy storage batteries.

Description

Technical Field

[0001] One or more embodiments of the present specification relate to computer technology, and more particularly, to a method and device for predicting the service life of an energy storage battery. Background Art

[0002] Energy storage batteries (such as lithium batteries, lead-acid batteries, etc.) are widely used in transportation, aerospace, and defense military fields due to their advantages such as low self-discharge rate, high operating voltage, long cycle life, and high energy density. However, after long-term use, energy storage batteries will experience aging and energy decay, which may lead to catastrophic events in the electrical system. Therefore, it is very necessary to accurately predict the service life of energy storage batteries.

[0003] In the related technology, the prediction methods of the service life of energy storage batteries mainly include experience-based statistical methods and battery performance-based methods, among which: the experience-based statistical methods mainly include the cycle number method, the ampere-hour method and its weighted method and the event-oriented aging accumulation method. This method mainly uses some statistical laws to roughly estimate the service life of energy storage batteries, which is usually only achieved by professionals with rich experience in the use of energy storage batteries; the battery performance-based methods mainly include model-based methods and data-driven methods. The model-based method relies on the battery degradation mechanism, load conditions and battery material properties and combines the failure mechanism to predict the service life. The object model architecture is determined as the application premise, and the accuracy depends on the accuracy of the model establishment; the data-driven method is to complete the training model based on historical data, and usually uses particle filtering, long-term and short-term neural networks, related vector machines and other methods to predict the service life of energy storage batteries. Summary of the invention

[0004] One or more embodiments of the present specification describe a method and device for predicting the service life of an energy storage battery, which can achieve a more accurate prediction of the service life of the energy storage battery.

[0005] According to a first aspect, a method for predicting the service life of an energy storage battery is provided, comprising:

[0006] Obtain early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; wherein the early external characteristic data accounts for a preset percentage of the full-cycle external characteristic data, the external characteristic data includes voltage, capacity, temperature and internal resistance, and M and N are both positive integers greater than 1;

[0007] Determine a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery;

[0008] Performing principal component analysis on at least one of capacity, temperature and internal resistance included in early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery;

[0009] Taking the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, training a preset machine learning model to obtain a prediction model;

[0010] The first health factor and the second health factor corresponding to each of the energy storage batteries to be tested are used as input and input into the prediction model to obtain the service life of the current energy storage battery to be tested.

[0011] In one embodiment of the present specification, determining a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery includes:

[0012] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

[0013] In one embodiment of the present specification, the nth cycle process is determined by:

[0014] For each cycle process between the mth cycle process of all the retired energy storage batteries and the last cycle process of the early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all the retired energy storage batteries in the current cycle process to obtain a correlation coefficient;

[0015] The number of cycles with the highest correlation coefficient is taken as n.

[0016] In one embodiment of the present specification, determining a first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the mth and nth cycle processes in each energy storage battery includes:

[0017] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery;

[0018] The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within a preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

[0019] In one embodiment of the present specification, the principal component analysis is performed on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain the second health factor of each energy storage battery, including:

[0020] For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed:

[0021] Constructing a matrix with (M+N) rows and n columns; wherein n is the total number of cycles of the early external characteristic data;

[0022] According to the order of energy storage batteries and cycle numbers, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries;

[0023] The high-dimensional data is subjected to dimensionality reduction processing using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as a second health factor.

[0024] According to a second aspect, a device for predicting the service life of an energy storage battery is provided, comprising:

[0025] An acquisition module is used to acquire early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; wherein the early external characteristic data accounts for a preset percentage of the full-cycle external characteristic data, the external characteristic data includes voltage, capacity, temperature and internal resistance, and M and N are both positive integers greater than 1;

[0026] A first determination module, configured to determine a first health factor corresponding to a current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery;

[0027] A second determination module is used to perform principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery;

[0028] A training module, used to take the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, train a preset machine learning model to obtain a prediction model;

[0029] The prediction module is used to input the first health factor and the second health factor corresponding to each of the energy storage batteries to be tested into the prediction model to obtain the service life of the current energy storage battery to be tested.

[0030] In one embodiment of the present specification, the first determining module is used to perform the following operations:

[0031] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

[0032] In one embodiment of the present specification, the nth cycle process is determined by:

[0033] For each cycle process between the mth cycle process of all the retired energy storage batteries and the last cycle process of the early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all the retired energy storage batteries in the current cycle process to obtain a correlation coefficient;

[0034] The number of cycles with the highest correlation coefficient is taken as n.

[0035] In one embodiment of the present specification, when the first determination module determines the first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the mth and nth cycle processes of each energy storage battery, the first determination module is used to perform the following operations:

[0036] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery;

[0037] The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within a preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

[0038] In one embodiment of the present specification, the second determining module is used to perform the following operations:

[0039] For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed:

[0040] Constructing a matrix with (M+N) rows and n columns; wherein n is the total number of cycles of the early external characteristic data;

[0041] According to the order of energy storage batteries and cycle numbers, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries;

[0042] The high-dimensional data is subjected to dimensionality reduction processing using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as a second health factor.

[0043] According to the method and device for predicting the service life of an energy storage battery provided in the embodiments of this specification, by obtaining early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested, then based on the voltage and capacity included in the early external characteristic data of each energy storage battery, a first health factor corresponding to the current energy storage battery is determined, and at the same time, principal component analysis is performed on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery; then, the first health factor and the second health factor corresponding to the M retired energy storage batteries are used as inputs and the service life of the M retired energy storage batteries are used as outputs, and a preset machine learning model is trained to obtain a prediction model; finally, the first health factor and the second health factor corresponding to each energy storage battery to be tested are used as inputs and input into the prediction model to obtain the service life of the current energy storage battery to be tested. Since the first health factor of the above scheme is determined based on the voltage and capacity included in the early external characteristic data of each energy storage battery, and the second health factor is obtained by performing principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries, this can make the prediction accuracy of the model trained using the first health factor and the second health factor higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 A flow chart showing a method for predicting the service life of an energy storage battery in one embodiment of this specification;

[0046] Figure 2 A schematic diagram showing a voltage-capacity curve during the m-th and n-th cycles in one embodiment of this specification;

[0047] Figure 3 for Figure 2 A partial enlarged view of the voltage-capacity curve shown;

[0048] Figure 4 A comparison chart showing a predicted value and an actual value of the service life of an energy storage battery in one embodiment of this specification is shown;

[0049] Figure 5 A comparison chart showing a predicted value and an actual value of the service life of an energy storage battery in another embodiment of the present specification is shown;

[0050] Figure 6 A schematic diagram showing the structure of a device in which a device for predicting the service life of an energy storage battery is located in one embodiment of the present specification;

[0051] Figure 7 A schematic diagram of a device for predicting the service life of an energy storage battery in one embodiment of the present specification is shown. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the embodiments of this specification clearer, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this specification.

[0053] The solution provided in this specification is described below in conjunction with the accompanying drawings.

[0054] Figure 1 A flowchart of a method for predicting the service life of an energy storage battery according to an embodiment is shown. It can be understood that the method can be executed by any device, equipment, or platform with computing and processing capabilities. Figure 1 As shown, the method includes:

[0055] Step 100: Obtain early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested.

[0056] In some embodiments, the energy storage battery includes but is not limited to a lithium battery, a lead-acid battery, and the like.

[0057] In this embodiment, the early external characteristic data accounts for a preset percentage of the full cycle external characteristic data. In some embodiments, the preset percentage may be 10%-20%. For example, if the number of cycles of the full cycle is 2000, the early external characteristic data may refer to the external characteristic data of the first 200 cycles, or the external characteristic data of the first 400 cycles.

[0058] It should be noted that in existing data-driven research on energy storage battery life prediction, it is often necessary to obtain data from about 40% of the energy storage battery's life from the beginning of use to the end of its life to achieve an ideal prediction result. However, the embodiments of this specification can accurately predict the life of the energy storage battery only through its early external characteristic data, so that the future use of the energy storage battery can be judged as early as possible.

[0059] In this embodiment, the external characteristic data includes voltage, capacity, temperature and internal resistance. The capacity includes charging capacity and discharging capacity. For the sake of simplicity, they are collectively referred to as capacity. During the use of the energy storage battery, the external characteristic data such as voltage, capacity, temperature and internal resistance will be recorded. The subsequent health factor that can reflect the aging process of the energy storage battery is extracted based on these external characteristic data. For the specific extraction process, please refer to the following.

[0060] In this embodiment, both M and N are positive integers greater than 1. It can be understood that the larger the value of M, the higher the prediction accuracy of the prediction model subsequently trained using the first health factor and the second health factor corresponding to the retired energy storage battery, and the specific value of M is not specifically limited here. Similarly, the specific value of N is not specifically limited.

[0061] Step 102: Determine a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery.

[0062] It should be noted that in the process of predicting the service life of energy storage batteries, if feature extraction is not considered, the prediction accuracy may be affected, because the capacity of energy storage batteries often does not decay immediately during the initial aging process. If you want to predict the service life or remaining useful life (RUL) of energy storage batteries when the capacity has not yet decayed significantly, then extracting aging features is particularly important. After extracting the appropriate features, accurate predictions can be made using different types of machine learning algorithms. For the life prediction of energy storage batteries, the health indicator (HI) is generally used to describe indicators that are highly correlated with life. In the relevant research on health factors, commonly used health factors include charging time under constant current (CC) or constant voltage (CV) conditions, the slope of the voltage-capacity curve, etc. It is worth noting that the voltage-capacity curve is a rich data source, from which a lot of battery-related information can be discovered.

[0063] In existing research, most of them extract health factors by analyzing the voltage-capacity curve during the charge and discharge process, and find that as the number of battery charge and discharge cycles increases, the voltage-capacity curve will show an obvious change trend, and the charging time will gradually shorten. Therefore, the charging time required to reach the cut-off voltage can be used as a health factor.

[0064] In step 102, the first health factor is determined based on the voltage and capacity of each energy storage battery. Specifically, step 102 may include:

[0065] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

[0066] In this embodiment, in order to make the selected first health factor capable of representing more data, two cycle nodes can be first determined in the early external characteristic data. The selection of the first cycle node needs to avoid the initial use stage of the energy storage battery. This is because the energy storage battery has a capacity increase process during early use. This slow cycle or a small increase in capacity after the rest period is due to the charge storage of the negative electrode beyond the positive electrode area. Therefore, in order to reduce the impact on the extraction of the first health factor, this capacity increase process should be avoided. Therefore, the first cycle node should be selected at the first cycle node where the capacity begins to decay, that is, the mth cycle process is the first cycle process where the capacity begins to decay. Secondly, in order to make the fitness of the first health factor higher, it is necessary to ensure that the second cycle node has as large a difference as possible with the first cycle node, but the correlation between the first health factor and the service life of the second cycle node may be poor, so the second cycle node should be selected between the first cycle node and the last cycle node of the early external characteristic data, that is, the nth cycle process is the cycle process between the mth cycle process and the last cycle process of the early external characteristic data.

[0067] In order to determine the value of n more accurately, existing data of retired energy storage batteries can be used.

[0068] In some embodiments, the nth cycle process is determined as follows:

[0069] For each cycle process between the mth cycle process of all retired energy storage batteries and the last cycle process of early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all retired energy storage batteries in the current cycle process to obtain a correlation coefficient;

[0070] The number of cycles with the highest correlation coefficient is taken as n.

[0071] In this embodiment, by performing a correlation analysis between the first health factor and the service life in each cycle process of all retired energy storage batteries, the cycle process with the highest correlation coefficient can be determined, and thus the value of n can be determined, which is conducive to obtaining a first health factor that is more accurate in predicting the service life of the energy storage battery.

[0072] In some embodiments, the Spearman or Pearson correlation analysis method may be used to analyze the correlation between the first health factor and the service life in each cycle of all retired energy storage batteries. The correlation analysis method is not described in detail herein.

[0073] After the voltage and capacity respectively included in the m-th and n-th cycle processes that can accurately characterize the first health factor are determined through the above scheme, in order to provide an accurate determination value of the first health factor, in some embodiments, the step of "determining the first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery" may include:

[0074] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery;

[0075] The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within the preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

[0076] In this embodiment, the voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within the preset capacity range is used as the first health factor corresponding to the current energy storage battery. In this way, the definite value of the first health factor can be determined, thereby facilitating subsequent model training and improving the prediction accuracy of the prediction model.

[0077] For example, see Figure 2 and Figure 3 According to the existing battery data, the voltage-capacity curves (y1 and y2, respectively) during the m-th and n-th cycles are drawn. There is an obvious interval between y1 and y2, so the integral difference or interval difference of the two curves can be calculated as the first health factor. The integral difference formula of the two curves is as follows:

[0078]

[0079] Wherein, b represents the preset capacity, y1 represents the voltage-capacity curve corresponding to the nth cycle process, y2 represents the voltage-capacity curve corresponding to the mth cycle process, and HIs is the first health factor.

[0080] Furthermore, the Spearman or Pearson correlation analysis method may be used to analyze the correlation between the first health factor and the service life to verify whether the first health factor is appropriately selected.

[0081] Step 104: Perform principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery.

[0082] It is understandable that the second health factor can be one (i.e. extracted from any one of capacity, temperature and internal resistance), two (i.e. extracted from any two of capacity, temperature and internal resistance), or three (i.e. extracted from capacity, temperature and internal resistance). The number of types of the second health factor is not specifically limited. In order to improve the prediction accuracy of the prediction model, the more types of the second health factor, the better, that is, there are three types of the second health factor.

[0083] Since there are many retired energy storage batteries, energy storage batteries to be tested, and cycle times, the dimensions of the data used to extract the second health factor are large, which is not conducive to improving the prediction accuracy of the prediction model. Therefore, it is possible to consider performing principal component analysis (PCA) on at least one of the early external characteristic data of all energy storage batteries, including capacity, temperature, and internal resistance. Figure 4 and Figure 5 , Figure 4 This is a comparison chart of the predicted value of cycle life (i.e. service life) obtained without principal component analysis and the actual value (the machine learning models used are BP neural network, SVM and GPR respectively). Figure 5 The comparison chart of the predicted value and actual value of cycle life obtained by principal component analysis (the machine learning models used are BP neural network, SVM and GPR respectively). Figure 4 and Figure 5 It can be seen that: Figure 5 The RMSE (Root Mean Square Error) and MAPE (Mean Absolute Percentage Error) in Figure 4This shows that the prediction model trained by principal component analysis has higher prediction accuracy.

[0084] In some implementations, step 104 may include:

[0085] For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed:

[0086] Construct a matrix with (M+N) rows and n columns; where n is the total number of cycles of the early external characteristic data;

[0087] According to the order of energy storage batteries and cycle times, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries;

[0088] The high-dimensional data is reduced in dimensionality using the preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as the second health factor.

[0089] In this embodiment, the high-dimensional data is reduced in dimensionality by using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as the second health factor. This can reduce the dimension of the data used to extract the second health factor, which is beneficial to improving the prediction accuracy of the prediction model.

[0090] High-dimensional data has a great impact on the training and calculation process of the model, so principal component analysis is used to reduce the dimension of these high-dimensional data to obtain a new matrix. For example, the appropriate dimension can be selected based on the requirement that the cumulative contribution rate of the principal component analysis method is greater than 80%. Assuming that the cumulative contribution rate of the first two dimensions reaches more than 80%, each matrix element in the matrix of the first two dimensions can be used as the second health factor.

[0091] It should be noted that when principal component analysis is used for data dimensionality reduction, the data of all energy storage batteries need to be organized into a matrix, which is conducive to improving the prediction accuracy of the prediction model. On the contrary, if the data of retired energy storage batteries and energy storage batteries to be tested are organized into matrices separately, this will reduce the prediction accuracy of the prediction model.

[0092] In some implementations, the second health factor may be extracted based on the discharge capacity, the maximum temperature, and the average internal resistance.

[0093] Step 106: Taking the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, a preset machine learning model is trained to obtain a prediction model.

[0094] In this embodiment, the data of the first health factor corresponding to the M retired energy storage batteries is an M*1 matrix vector, and the data of the second health factor corresponding to the M retired energy storage batteries is an M*Y matrix vector (Y is the number of dimensions obtained after principal component analysis).

[0095] In some implementations, the machine learning model may be a BP model, and the specific type of the machine learning model is not limited herein.

[0096] Step 108: taking the first health factor and the second health factor corresponding to each energy storage battery to be tested as input and inputting them into the prediction model to obtain the service life of the current energy storage battery to be tested.

[0097] In this embodiment, the data of the first health factor corresponding to the N energy storage batteries to be tested is an N*1 matrix vector, and the data of the second health factor corresponding to the N energy storage batteries to be tested is an N*Y matrix vector (Y is the number of dimensions obtained after principal component analysis).

[0098] In this way, each row vector in the N*(1+Y) matrix can be input into the prediction model to obtain the corresponding service life of the energy storage battery to be tested, and then the remaining service life of the energy storage battery to be tested can be determined.

[0099] In summary, the above method obtains early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested, and then determines the first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery, and at the same time performs principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain the second health factor of each energy storage battery; then, the first health factor and the second health factor corresponding to the M retired energy storage batteries are used as inputs and the service life of the M retired energy storage batteries are used as outputs to train the preset machine learning model to obtain a prediction model; finally, the first health factor and the second health factor corresponding to each energy storage battery to be tested are used as inputs and input into the prediction model to obtain the service life of the current energy storage battery to be tested. Since the first health factor of the above scheme is determined based on the voltage and capacity included in the early external characteristic data of each energy storage battery, and the second health factor is obtained by performing principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries, the prediction accuracy of the model trained using the first health factor and the second health factor can be higher.

[0100] like Figure 6 and Figure 7As shown, the embodiments of this specification provide a device where a device for predicting the service life of an energy storage battery is located and a device for predicting the service life of an energy storage battery. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, such as Figure 6 As shown, it is a hardware structure diagram of the device where the energy storage battery service life prediction device provided in the embodiment of this specification is located, except Figure 6 In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 7 As shown, as a device in a logical sense, the CPU of the device in which it is located reads the corresponding computer program instructions in the non-volatile memory into the internal memory and runs them.

[0101] like Figure 7 As shown, in one embodiment of the present specification, a device for predicting the service life of an energy storage battery is proposed, comprising:

[0102] The acquisition module 700 is used to acquire early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; wherein the early external characteristic data accounts for a preset percentage of the full-cycle external characteristic data, and the external characteristic data includes voltage, capacity, temperature and internal resistance, and M and N are both positive integers greater than 1;

[0103] A first determination module 702 is used to determine a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery;

[0104] The second determination module 704 is used to perform principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery;

[0105] A training module 706 is used to train a preset machine learning model by taking the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output to obtain a prediction model;

[0106] The prediction module 708 is used to input the first health factor and the second health factor corresponding to each of the energy storage batteries to be tested into the prediction model to obtain the service life of the current energy storage battery to be tested.

[0107] In an embodiment of the present invention, the acquisition module 700 can be used to execute step 100 in the above method embodiment, the first determination module 702 can be used to execute step 102 in the above method embodiment, the second determination module 704 can be used to execute step 104 in the above method embodiment, the training module 706 can be used to execute step 106 in the above method embodiment, and the prediction module 708 can be used to execute step 108 in the above method embodiment.

[0108] In one embodiment of the device provided in this specification, the first determining module 702 is configured to perform the following operations:

[0109] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

[0110] In one embodiment of the device proposed in this specification, the nth cycle process is determined by:

[0111] For each cycle process between the mth cycle process of all the retired energy storage batteries and the last cycle process of the early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all the retired energy storage batteries in the current cycle process to obtain a correlation coefficient;

[0112] The number of cycles with the highest correlation coefficient is taken as n.

[0113] In one embodiment of the device proposed in this specification, the first determination module 702 is used to perform the following operations when determining the first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the mth and nth cycle processes in each energy storage battery:

[0114] Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery;

[0115] The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within a preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

[0116] In one embodiment of the device provided in this specification, the second determining module 704 is configured to perform the following operations:

[0117] For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed:

[0118] Constructing a matrix with (M+N) rows and n columns; wherein n is the total number of cycles of the early external characteristic data;

[0119] According to the order of energy storage batteries and cycle numbers, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries;

[0120] The high-dimensional data is subjected to dimensionality reduction processing using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as a second health factor.

[0121] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0122] According to another aspect of the embodiment, there is also provided a device for predicting the service life of an energy storage battery, comprising: at least one memory and at least one processor;

[0123] The at least one memory is used to store a machine-readable program;

[0124] The at least one processor is used to call the machine-readable program to execute the method for predicting the service life of the energy storage battery in any embodiment of this specification.

[0125] According to another embodiment, a computer-readable medium is also provided, storing instructions for causing a computer to execute the method for predicting the service life of an energy storage battery as described herein. Specifically, a method or device equipped with a storage medium may be provided, on which a software program code for implementing the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the method or device is caused to read and execute the program code stored in the storage medium.

[0126] In this case, the program code itself read from the storage medium can realize the function of any one of the above embodiments, and thus the program code and the storage medium storing the program code constitute part of this specification.

[0127] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0128] In addition, it should be clear that the functions of any of the above embodiments can be realized not only by executing the program code read by the computer, but also by completing part or all of the actual operations by operating methods operated on the computer based on the instructions of the program code.

[0129] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or the expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0130] Those skilled in the art should be aware that in one or more of the above examples, the functions described in this specification can be implemented by hardware, software, firmware, or any combination thereof. When implemented by software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium.

[0131] It is to be understood that the structure illustrated in the embodiments of this specification does not constitute a specific limitation on the device for predicting the service life of an energy storage battery. In other embodiments of this specification, the device for predicting the service life of an energy storage battery may include more or fewer components than those illustrated, or combine certain components, or split certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0132] Since the information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the method embodiments of this specification, the specific contents can be found in the description of the method embodiments of this specification and will not be repeated here.

[0133] The above specific implementation methods further explain the purpose, technical solutions and beneficial effects of this specification in detail. It should be understood that the above are only specific implementation methods of this specification and are not used to limit the protection scope of this specification. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of this specification should be included in the protection scope of this specification.

Claims

1. A method for predicting the service life of an energy storage battery, characterized in that: include: Obtain early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; wherein the early external characteristic data accounts for a preset percentage of the full-cycle external characteristic data, the external characteristic data includes voltage, capacity, temperature and internal resistance, and M and N are both positive integers greater than 1; Determine a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery; Performing principal component analysis on at least one of capacity, temperature and internal resistance included in early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery; Taking the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, training a preset machine learning model to obtain a prediction model; The first health factor and the second health factor corresponding to each of the energy storage batteries to be tested are used as input and input into the prediction model to obtain the service life of the current energy storage battery to be tested.

2. The method according to claim 1, characterized in that The determining of a first health factor corresponding to the current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery comprises: Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

3. The method according to claim 2, characterized in that The nth cycle process is determined by the following method: For each cycle process between the mth cycle process of all the retired energy storage batteries and the last cycle process of the early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all the retired energy storage batteries in the current cycle process to obtain a correlation coefficient; The number of cycles with the highest correlation coefficient is taken as n.

4. The method according to claim 2, characterized in that: The determining of a first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the mth and nth cycle processes in each energy storage battery comprises: Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery; The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within a preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

5. The method according to any one of claims 1 to 4, characterized in that The method of performing principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain the second health factor of each energy storage battery includes: For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed: Constructing a matrix with (M+N) rows and n columns; wherein n is the total number of cycles of the early external characteristic data; According to the order of energy storage batteries and cycle numbers, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries; The high-dimensional data is subjected to dimensionality reduction processing using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as a second health factor.

6. A device for predicting the service life of an energy storage battery, characterized in that: include: An acquisition module is used to acquire early external characteristic data of M retired energy storage batteries and N energy storage batteries to be tested; wherein the early external characteristic data accounts for a preset percentage of the full-cycle external characteristic data, the external characteristic data includes voltage, capacity, temperature and internal resistance, and M and N are both positive integers greater than 1; A first determination module, configured to determine a first health factor corresponding to a current energy storage battery based on the voltage and capacity included in the early external characteristic data of each energy storage battery; A second determination module is used to perform principal component analysis on at least one of the capacity, temperature and internal resistance included in the early external characteristic data of all energy storage batteries to obtain a second health factor of each energy storage battery; A training module, used to take the first health factor and the second health factor corresponding to the M retired energy storage batteries as input and the service life of the M retired energy storage batteries as output, train a preset machine learning model to obtain a prediction model; The prediction module is used to input the first health factor and the second health factor corresponding to each of the energy storage batteries to be tested into the prediction model to obtain the service life of the current energy storage battery to be tested.

7. The device according to claim 6, characterized in that The first determining module is used to perform the following operations: Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, a first health factor corresponding to the current energy storage battery is determined; wherein the m-th cycle process is the first cycle process in which the capacity begins to decay, and the n-th cycle process is the cycle process between the m-th cycle process and the last cycle process of the early external characteristic data.

8. The device according to claim 7, characterized in that The nth cycle process is determined by the following method: For each cycle process between the mth cycle process of all the retired energy storage batteries and the last cycle process of the early external characteristic data, a correlation analysis is performed between the first health factor and the service life of all the retired energy storage batteries in the current cycle process to obtain a correlation coefficient; The number of cycles with the highest correlation coefficient is taken as n.

9. The device according to claim 7, characterized in that When the first determination module determines the first health factor corresponding to the current energy storage battery based on the voltage and capacity respectively included in the mth and nth cycle processes in each energy storage battery, the first determination module is used to perform the following operations: Based on the voltage and capacity respectively included in the m-th and n-th cycle processes in each energy storage battery, determining the voltage-capacity curves respectively corresponding to the m-th and n-th cycle processes in the current energy storage battery; The voltage-capacity curves corresponding to the m-th and n-th cycle processes of each energy storage battery are subtracted, and the area enclosed by the obtained difference curve within a preset capacity range is used as the first health factor corresponding to the current energy storage battery; wherein the preset capacity is smaller than the charging capacity of the current energy storage battery during the m-th or n-th cycle.

10. The device according to any one of claims 6 to 9, characterized in that: The second determining module is used to perform the following operations: For each of the early external characteristic data of capacity, temperature and internal resistance of all energy storage batteries, the following are performed: Constructing a matrix with (M+N) rows and n columns; wherein n is the total number of cycles of the early external characteristic data; According to the order of energy storage batteries and cycle numbers, the current early external characteristic data of all energy storage batteries are correspondingly filled into the matrix to obtain high-dimensional data of all energy storage batteries; The high-dimensional data is subjected to dimensionality reduction processing using a preset principal component analysis algorithm to obtain low-dimensional data of all energy storage batteries, and each matrix element in the low-dimensional data is used as a second health factor.

Images