Lithium battery state monitoring and estimating method and system

Abstract

The invention provides a lithium battery state monitoring and estimating method and system, and relates to the technical field of lithium batteries. Performing time discretization on the charge-discharge cycle data to obtain the state monitoring granularity of each time period, and determining a plurality of lithium battery state monitoring vectors according to the state monitoring label recorded by each charge-discharge cycle; performing spatial embedding on all the lithium battery state monitoring vectors to obtain a charge-discharge cycle sequence, converting each charge-discharge cycle data block in the charge-discharge cycle sequence into a key value data pair, and determining a lithium battery attenuation coefficient corresponding to each key value data pair through the lithium battery state monitoring coefficient of each key value data pair; health estimation is carried out on the target lithium battery according to all the lithium battery attenuation coefficients, a health estimation value is obtained, and the health level of the target lithium battery is determined based on the health estimation value, so that the health estimation can be carried out by extracting the attenuation characteristics, and the health estimation accuracy of the target lithium battery is improved.

Description

Technical Field

[0001] The present application relates to the technical field of lithium batteries, and more specifically, to a method and system for monitoring and estimating the state of a lithium battery. Background Art

[0002] As an important component of modern energy storage devices, lithium batteries are widely used in electric vehicles, portable electronic devices, energy storage systems and other fields. As the use time increases, the performance of lithium batteries will gradually deteriorate, so monitoring and estimating the status of lithium batteries is the key to ensuring their safety, reliability and life. However, lithium battery status monitoring and estimation is a complex technical task, which mainly involves the accurate estimation of battery parameters such as the state of charge (SOC), state of health (SOH), and remaining useful life (RUL).

[0003] The status monitoring and estimation of lithium batteries usually rely on the real-time collection of multiple sensor data such as voltage, current, temperature, etc., and calculate the current status of the battery through a certain mathematical model or algorithm. In the prior art, the health status of lithium batteries is usually evaluated by monitoring the battery capacity decay and internal resistance increase. However, the battery capacity decay is not only related to the number of cycles, but also affected by multiple factors such as temperature, charge and discharge rate, and charge and discharge depth. Therefore, how to extract the decay characteristics of the target lithium battery to perform health estimation in order to improve the accuracy of the health estimation of the target lithium battery is a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides a lithium battery status monitoring and estimation method and system, which can extract the degradation characteristics of a target lithium battery to perform health estimation, so as to improve the accuracy of health estimation of the target lithium battery.

[0005] In a first aspect, the present application provides a lithium battery status monitoring and estimation method, the estimation method comprising the following steps:

[0006] Acquire charge and discharge cycle data of a target lithium battery in an aging cycle experiment, and assign a state monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data;

[0007] Performing time discretization on the charge and discharge cycle data to obtain a state monitoring granularity for each time period in the charge and discharge cycle data, and then determining a lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to a state monitoring tag recorded for each charge and discharge cycle;

[0008] Perform spatial embedding on all lithium battery state monitoring vectors to obtain a charge and discharge cycle sequence of a target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair;

[0009] The health of the target lithium battery is estimated according to all lithium battery degradation coefficients to obtain a health estimation value of the target lithium battery, and the health level of the target lithium battery is determined based on the health estimation value.

[0010] In this embodiment, assigning a status monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data specifically includes:

[0011] Obtain the actual rated capacity of the target lithium battery before the aging cycle experiment;

[0012] For each charge-discharge cycle record in the charge-discharge cycle data, determining a cycle experiment period corresponding to the charge-discharge cycle record in the aging cycle experiment;

[0013] Obtaining the battery measured capacity of the target lithium battery during the cycle test period;

[0014] Determine the health of the lithium battery corresponding to the charge and discharge cycle record according to the actual rated capacity and the battery measured capacity;

[0015] The health of the lithium battery is used as a status monitoring label of the charge and discharge cycle record, thereby completing the allocation of the status monitoring label to each charge and discharge cycle record in the charge and discharge cycle data.

[0016] In this embodiment, determining the health of the lithium battery corresponding to the charge and discharge cycle record according to the actual rated capacity and the battery measured capacity is to use the ratio of the battery measured capacity to the actual rated capacity as the health of the lithium battery corresponding to the charge and discharge cycle record.

[0017] In this embodiment, the charge-discharge cycle data is time discretized to obtain the state monitoring granularity of each time period in the charge-discharge cycle data, specifically including:

[0018] Get the set discrete coefficient;

[0019] The charge-discharge cycle data is time discretized based on the discrete coefficient to obtain the state monitoring granularity of each time period in the charge-discharge cycle data.

[0020] In this embodiment, determining the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to the state monitoring tag recorded in each charge and discharge cycle specifically includes:

[0021] For each state monitoring granularity of the time period, obtain the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity;

[0022] Determine the typical characteristics of the state monitoring at the state monitoring granularity through the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity;

[0023] Determine the battery characteristics corresponding to the state monitoring granularity;

[0024] A lithium battery state monitoring vector of the target lithium battery at the state monitoring granularity is constructed according to the state monitoring typical characteristics and the battery characteristics, and then a lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity is obtained.

[0025] In this embodiment, all lithium battery state monitoring vectors are spatially embedded to obtain the charge and discharge cycle sequence of the target lithium battery, which specifically includes:

[0026] Construct a state monitoring matrix of the target lithium battery based on all lithium battery state monitoring vectors;

[0027] dividing the condition monitoring matrix into a plurality of condition monitoring blocks;

[0028] Each state monitoring block is spatially embedded to obtain multiple charge and discharge cycle data blocks of the target lithium battery;

[0029] The charge and discharge cycle sequence of the target lithium battery is constructed through all the charge and discharge cycle data blocks.

[0030] In this embodiment, converting each charge-discharge cycle data block in the charge-discharge cycle sequence into a key-value data pair specifically includes:

[0031] For each charge-discharge cycle data block in the charge-discharge cycle sequence, converting the battery feature in the charge-discharge cycle data block into a key of a key-value data pair;

[0032] The typical characteristics of state monitoring in the charge-discharge cycle data block are converted into the value of the key-value data pair, thereby completing the key-value data pair conversion of each charge-discharge cycle data block in the charge-discharge cycle sequence.

[0033] In this embodiment, the health estimation of the target lithium battery according to the degradation coefficients of all lithium batteries is to input all the degradation coefficients of the lithium batteries into a health status estimation model based on machine learning to perform health estimation.

[0034] In this embodiment, determining the health level of the target lithium battery based on the health estimation value specifically includes:

[0035] Get the preset lithium battery health level table;

[0036] The health level of the target lithium battery in the lithium battery health level table is determined according to the health estimation value, thereby obtaining the health level of the target lithium battery.

[0037] In a second aspect, the present application provides a lithium battery state monitoring and estimation system for executing a lithium battery state monitoring and estimation method, the estimation system comprising:

[0038] An initialization module, used to obtain charge and discharge cycle data of a target lithium battery in an aging cycle experiment, and to assign a state monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data;

[0039] A lithium battery state monitoring module is used to time discretize the charge and discharge cycle data to obtain the state monitoring granularity of each time period in the charge and discharge cycle data, and then determine the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to the state monitoring tag recorded for each charge and discharge cycle;

[0040] A lithium battery degradation assessment module is used to spatially embed all lithium battery state monitoring vectors to obtain a charge and discharge cycle sequence of a target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair;

[0041] The health estimation module is used to perform health estimation on the target lithium battery according to all lithium battery degradation coefficients, obtain a health estimation value of the target lithium battery, and determine the health level of the target lithium battery based on the health estimation value.

[0042] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects:

[0043] By acquiring the charge and discharge cycle data of the target lithium battery in the aging cycle experiment, and assigning a state monitoring label to each charge and discharge cycle record in the charge and discharge cycle data; time discretizing the charge and discharge cycle data to obtain the state monitoring granularity of each time period in the charge and discharge cycle data, and then determining the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to the state monitoring label of each charge and discharge cycle record; spatially embedding all lithium battery state monitoring vectors to obtain the charge and discharge cycle sequence of the target lithium battery, converting each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determining the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair; health estimation of the target lithium battery is performed based on all lithium battery degradation coefficients to obtain the health estimation value of the target lithium battery, and the health level of the target lithium battery is determined based on the health estimation value.

[0044] It can be seen that in the present application, the degradation characteristics of the target lithium battery can be extracted for health estimation; wherein, by assigning a state monitoring tag to each charge and discharge cycle record, the health status of the battery can be reflected in real time, and the state monitoring tag can be used to analyze the trend of the health status of the target lithium battery; then, the charge and discharge cycle data is decomposed into smaller time periods through time discretization, so that the state monitoring is more detailed, and the state monitoring granularity within each time period can provide detailed information, which helps to more accurately capture the performance changes of the target battery in different time periods, thereby making the health estimation more accurate, and the lithium battery state monitoring vector can more accurately calculate the degradation rate of the target lithium battery; secondly, by converting the charge and discharge cycle data block into a key-value data pair, the complex original data can be structured, which helps to improve the efficiency and accuracy of the health estimation of the target lithium battery. After spatially embedding each charge and discharge cycle data block in the charge and discharge cycle sequence, the battery degradation characteristics can be more accurately extracted, and by calculating the lithium battery degradation coefficient, the dynamic monitoring of the lithium battery degradation state can be achieved; finally, the health estimation of the target lithium battery through all the lithium battery degradation coefficients can help determine a more accurate health level of the target lithium battery.

[0045] In summary, the technical solution adopted in this application can extract the degradation characteristics of the target lithium battery to perform health estimation, so as to improve the accuracy of health estimation of the target lithium battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only the embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 is a flow chart of a lithium battery status monitoring and estimation method provided in the present application;

[0048] Figure 2 It is a module structure diagram of the lithium battery status monitoring and estimation system provided in this application. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0050] The embodiment of the present application provides a lithium battery state monitoring estimation method and system, the core of which is to obtain the charge and discharge cycle data of the target lithium battery in the aging cycle experiment, and assign a state monitoring label to each charge and discharge cycle record in the charge and discharge cycle data; time discretize the charge and discharge cycle data to obtain the state monitoring granularity of each time period in the charge and discharge cycle data, and then determine the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to the state monitoring label of each charge and discharge cycle record; spatially embed all lithium battery state monitoring vectors to obtain the charge and discharge cycle sequence of the target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair; perform health estimation on the target lithium battery according to all lithium battery degradation coefficients to obtain the health estimation value of the target lithium battery, and determine the health level of the target lithium battery based on the health estimation value. The above scheme can be used to extract the degradation characteristics of the target lithium battery for health estimation, so as to improve the accuracy of health estimation of the target lithium battery.

[0051] Embodiment 1:

[0052] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1As shown in FIG. 1 , this figure is an exemplary flow chart of a lithium battery status monitoring and estimation method according to this embodiment of the present application, and the estimation method includes the following steps:

[0053] In step S1, the charge and discharge cycle data of the target lithium battery in the aging cycle experiment is obtained, and a state monitoring tag is assigned to each charge and discharge cycle record in the charge and discharge cycle data.

[0054] In specific implementation, an aging cycle experiment is performed on the target lithium battery. Usually, the aging cycle experiment includes multiple charge and discharge cycles, each charge and discharge cycle includes a charging process and a discharging process, and in the aging cycle experiment, 50 charge and discharge cycles can be used as a cycle experiment period to monitor the charge and discharge cycle process of the target battery in real time, and the data of the target battery is recorded once in each cycle experiment period, that is, the charge and discharge cycle record, which includes the charge and discharge voltage, charge and discharge current, temperature and time, so that the data composed of all the charge and discharge cycle records can be combined as charge and discharge cycle data.

[0055] Preferably, in this embodiment, the following method can be used to assign a status monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data, namely:

[0056] Obtain the actual rated capacity of the target lithium battery before the aging cycle experiment;

[0057] For each charge-discharge cycle record in the charge-discharge cycle data, determining a cycle experiment period corresponding to the charge-discharge cycle record in the aging cycle experiment;

[0058] Obtaining the battery measured capacity of the target lithium battery during the cycle test period;

[0059] Determine the health of the lithium battery corresponding to the charge and discharge cycle record according to the actual rated capacity and the battery measured capacity;

[0060] The health of the lithium battery is used as a status monitoring label of the charge and discharge cycle record, thereby completing the allocation of the status monitoring label to each charge and discharge cycle record in the charge and discharge cycle data.

[0061] It should be noted that in the present application, the status monitoring tag is the target lithium battery monitoring tag corresponding to the charge and discharge cycle record. The value of the status monitoring tag can indicate the health status of the target lithium battery; the lithium battery health indicates the health level of the target lithium battery; the actual rated capacity indicates the initial capacity of the target lithium battery; and the battery measured capacity indicates the capacity corresponding to the target lithium battery in the cycle experiment period.

[0062] In specific implementation, first, the actual rated capacity of the target lithium battery before the aging cycle experiment can be obtained from the official website of the battery manufacturer; then, for each charge and discharge cycle record in the charge and discharge cycle data, the cycle experiment period corresponding to the charge and discharge cycle record in the aging cycle experiment is determined, so that the battery measured capacity of the target lithium battery in the cycle experiment period can be calculated by the coulomb counting method; then, the health of the lithium battery corresponding to the charge and discharge cycle record can be determined according to the actual rated capacity and the battery measured capacity, that is, the ratio of the battery measured capacity to the actual rated capacity is used as the health of the lithium battery corresponding to the charge and discharge cycle record; finally, the health of the lithium battery can be used as a state monitoring tag for the charge and discharge cycle record, that is, the value of the state monitoring tag of the charge and discharge cycle record is the value of the lithium battery health, and the state monitoring tag allocation for each charge and discharge cycle record in the charge and discharge cycle data can be completed in the above manner.

[0063] It should be noted that by assigning a status monitoring tag to each charge and discharge cycle record, the health status of the battery can be reflected in real time, and these status monitoring tags can be used to analyze the trend of the health status of the target lithium battery.

[0064] In step S2, the charge and discharge cycle data is time discretized to obtain the state monitoring granularity of each time period in the charge and discharge cycle data, and then the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity is determined according to the state monitoring label of each charge and discharge cycle record.

[0065] In this embodiment, the charge-discharge cycle data is time discretized to obtain the state monitoring granularity of each time period in the charge-discharge cycle data, specifically including:

[0066] Get the set discrete coefficient;

[0067] The charge-discharge cycle data is time discretized based on the discrete coefficient to obtain the state monitoring granularity of each time period in the charge-discharge cycle data.

[0068] In the specific implementation, first, a discrete coefficient can be set according to historical experiments and data analysis, and the discrete coefficient is used to discretize the time attribute of the charge and discharge cycle data; then, the charge and discharge cycle data can be time discretized based on the discrete coefficient. If the discrete coefficient is K, the charge and discharge cycle data is randomly divided into K segments in time, thereby obtaining K time periods, and the length of each time period is obtained, so that the ratio of the length of the time period to the discrete coefficient can be used as the state monitoring granularity of the time period, and the state monitoring granularity represents the fine granularity of the target lithium battery state monitoring.

[0069] Preferably, in this embodiment, the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity is determined according to the state monitoring tag of each charge and discharge cycle record, which can be specifically determined in the following manner, namely:

[0070] For each state monitoring granularity of the time period, obtain the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity;

[0071] Determine the typical characteristics of the state monitoring at the state monitoring granularity through the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity;

[0072] Determine the battery characteristics corresponding to the state monitoring granularity;

[0073] A lithium battery state monitoring vector of the target lithium battery at the state monitoring granularity is constructed according to the state monitoring typical characteristics and the battery characteristics, and then a lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity is obtained.

[0074] It should be noted that in the present application, typical characteristics of state monitoring refer to typical representative state monitoring characteristics at the state monitoring granularity; battery characteristics refer to voltage, current and temperature characteristics of the target lithium battery; and the lithium battery state monitoring vector is a feature vector used to represent the state monitoring status of the target lithium battery at the state monitoring granularity.

[0075] In the specific implementation, first, for each state monitoring granularity of the time period, the state monitoring labels of all charge and discharge cycle records corresponding to the state monitoring granularity can be obtained; then, the state monitoring typical features of the state monitoring granularity are determined through the state monitoring labels of all charge and discharge cycle records corresponding to the state monitoring granularity, that is, the mean of the state monitoring labels of all charge and discharge cycle records is used as the state monitoring typical features of the state monitoring granularity; secondly, the battery features corresponding to the state monitoring granularity can be determined, that is, the voltage mean, current mean and temperature mean of the target lithium battery in the cycle experiment period corresponding to the state monitoring granularity are calculated, and the features composed of the three are used as the battery features corresponding to the state monitoring granularity; further, the feature vector composed of the state monitoring typical features and the battery features can be used as the lithium battery state monitoring vector of the target lithium battery at the state monitoring granularity, and the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity can be obtained in the above manner.

[0076] It should be noted that by decomposing the charge and discharge cycle data into smaller time periods through time discretization, the state monitoring is more detailed. The state monitoring granularity in each time period can provide detailed information, which helps to more accurately capture the performance changes of the target battery in different time periods, thereby making the health estimation more accurate, and the lithium battery state monitoring vector can more accurately calculate the degradation rate of the target lithium battery.

[0077] In step S3, all lithium battery status monitoring vectors are spatially embedded to obtain the charge and discharge cycle sequence of the target lithium battery, and each charge and discharge cycle data block in the charge and discharge cycle sequence is converted into a key-value data pair, and then the lithium battery degradation coefficient corresponding to each key-value data pair is determined respectively through the lithium battery status monitoring coefficient of each key-value data pair.

[0078] In this embodiment, all lithium battery state monitoring vectors are spatially embedded to obtain the charge and discharge cycle sequence of the target lithium battery in the following manner, namely:

[0079] Construct a state monitoring matrix of the target lithium battery based on all lithium battery state monitoring vectors;

[0080] dividing the condition monitoring matrix into a plurality of condition monitoring blocks;

[0081] Each state monitoring block is spatially embedded to obtain multiple charge and discharge cycle data blocks of the target lithium battery;

[0082] The charge and discharge cycle sequence of the target lithium battery is constructed through all the charge and discharge cycle data blocks.

[0083] In the specific implementation, first, the state monitoring vectors of the lithium batteries are arranged in chronological order to form a large matrix, and the matrix is ​​used as the state monitoring matrix of the target lithium battery. The rows of the matrix can represent time periods, and the columns can represent different state monitoring features in the lithium battery state monitoring vectors; then, the state monitoring matrix can be divided into multiple state monitoring blocks, each of which corresponds to a different time period of the target lithium battery. The state monitoring matrix can be divided according to the set data block size to obtain multiple state monitoring blocks, wherein each state monitoring block contains a certain number of time period data and state monitoring features; then, each state monitoring block is spatially embedded and converted into a representation in the embedding space for further analysis and processing. A deep learning algorithm, such as an autoencoder, can be used to spatially embed each state monitoring block, that is, to represent it as a new data block in the embedding space, that is, a charge and discharge cycle data block, which retains the main features of each state monitoring block and reduces the dimension at the same time, that is, the charge and discharge cycle data block is the representation of the corresponding state monitoring block in the low-dimensional space; finally, all the charge and discharge cycle data blocks (that is, the state monitoring blocks after spatial position embedding) are arranged in chronological order to construct the charge and discharge cycle sequence of the target lithium battery.

[0084] In this embodiment, each charge-discharge cycle data block in the charge-discharge cycle sequence is converted into a key-value data pair in the following manner, namely:

[0085] For each charge-discharge cycle data block in the charge-discharge cycle sequence, converting the battery feature in the charge-discharge cycle data block into a key of a key-value data pair;

[0086] The typical characteristics of state monitoring in the charge-discharge cycle data block are converted into the value of the key-value data pair, thereby completing the key-value data pair conversion of each charge-discharge cycle data block in the charge-discharge cycle sequence.

[0087] In the specific implementation, first, it is necessary to select the battery characteristics in the charge and discharge cycle data block as the key. The battery characteristics are used to uniquely identify the data block. For example, the voltage, current or temperature in the battery characteristics can be selected as the key of the key-value data pair; then, the typical characteristics of the state monitoring in the charge and discharge cycle data block can be selected as the value of the key-value data pair. After obtaining the key and value of the key-value data pair, the key-value data pair corresponding to the charge and discharge cycle data block can be obtained. The key-value data pair corresponding to each charge and discharge cycle data block can be obtained in the above manner, that is, the key-value data pair conversion of each charge and discharge cycle data block in the charge and discharge cycle sequence is completed.

[0088] Preferably, in this embodiment, the lithium battery degradation coefficient corresponding to each key-value data pair can be determined by the lithium battery state monitoring coefficient of each key-value data pair in the following manner, namely:

[0089] Obtaining preset condition monitoring characteristics;

[0090] For each key-value data pair, determining a feature equivalence entropy between the key-value data pair and the state monitoring feature;

[0091] Determine the lithium battery status monitoring coefficient of the key-value data pair;

[0092] The lithium battery degradation coefficient of the key-value data pair is determined according to the lithium battery state monitoring coefficient and the characteristic equivalent entropy, and then the lithium battery degradation coefficient corresponding to each key-value data pair is obtained.

[0093] It should be noted that, in the present application, the lithium battery degradation coefficient represents the degree of attenuation and aging of the target lithium battery; the lithium battery state monitoring coefficient is a monitoring indicator used to represent the current state health of the target lithium battery; the state monitoring feature is a monitoring feature set for comparison based on historical experience; the feature equivalence entropy represents the amount of information equivalence between the key-value data pair and the state monitoring feature.

[0094] In specific implementation, first, it is necessary to define a set of monitoring features for comparison, which can be voltage, current or temperature. In actual implementation, it is necessary to select according to the type of key in the key-value data pair; then, for each key-value data pair, the feature equivalent entropy between the key-value data pair and the state monitoring feature can be determined, that is, the feature similarity between the key in the key-value data pair and the state monitoring feature is calculated, and then the feature similarity is multiplied by the preset adjustment coefficient, and the result can be used as the feature equivalent entropy between the key-value data pair and the state monitoring feature; secondly, the lithium battery state monitoring coefficient of the key-value data pair can be determined. In actual implementation, the value in the key-value data pair can be used as the lithium battery state monitoring coefficient of the key-value data pair; finally, the lithium battery degradation coefficient of the key-value data pair is determined according to the lithium battery state monitoring coefficient and the feature equivalent entropy, that is, the product of the lithium battery state monitoring coefficient and the feature equivalent entropy is used as the lithium battery degradation coefficient of the key-value data pair. The lithium battery degradation coefficient corresponding to each key-value data pair can be obtained in the above manner.

[0095] It should be noted that by converting the charge and discharge cycle data blocks into key-value data pairs, complex raw data can be structured, which helps to more efficiently process and analyze a large amount of monitoring information, thereby improving the efficiency and accuracy of the target lithium battery health estimation. After spatially embedding each charge and discharge cycle data block in the charge and discharge cycle sequence, the relationship between each data point is analyzed through the lithium battery status monitoring vector, which can more accurately extract the battery degradation characteristics, and by calculating the lithium battery degradation coefficient, dynamic monitoring of the lithium battery degradation status can be achieved.

[0096] In step S4, health estimation is performed on the target lithium battery according to all lithium battery degradation coefficients to obtain a health estimation value of the target lithium battery, and a health level of the target lithium battery is determined based on the health estimation value.

[0097] In this embodiment, the health estimation of the target lithium battery based on all lithium battery degradation coefficients is to input all lithium battery degradation coefficients into a health status estimation model based on machine learning for health estimation; in specific implementation, all lithium battery degradation coefficients are used as input vectors and input into a pre-trained health status estimation model based on machine learning. The health status estimation model can adopt a variety of machine learning algorithms, such as neural networks, support vector machines, random forests, etc. The health status estimation model can more accurately predict the health status of the current target lithium battery by learning patterns and trends in a large amount of historical data. The health status estimation model can capture nonlinear complex relationships and provide higher estimation accuracy than traditional linear models. Therefore, the health status estimation model calculates the health estimation value of the target lithium battery according to the input lithium battery degradation coefficient, and the health estimation value represents the health status of the current target lithium battery.

[0098] In this embodiment, the health level of the target lithium battery may be determined based on the health estimation value in the following manner, namely:

[0099] Get the preset lithium battery health level table;

[0100] The health level of the target lithium battery in the lithium battery health level table is determined according to the health estimation value, thereby obtaining the health level of the target lithium battery.

[0101] In the specific implementation, first, a preset lithium battery health level table needs to be prepared. The lithium battery health level table is usually formulated by experts based on a large amount of experimental data and industry standards. The lithium battery health level table lists the health levels corresponding to different health estimation values. The health level can be divided into multiple levels, for example, from "Level 1" to "Level N"; then, the obtained health estimation value can be compared with the various level standards in the preset lithium battery health level table, and the health level corresponding to the target lithium battery in the health level table is determined according to the range of the health estimation value. The health level can be directly used to indicate the current health status of the target lithium battery, and the health level can be used in the battery management system (BMS) to help users or systems make decisions, such as whether to replace the battery or adjust the charging strategy.

[0102] It can be seen that in the present application, the degradation characteristics of the target lithium battery can be extracted for health estimation; wherein, by assigning a state monitoring tag to each charge and discharge cycle record, the health status of the battery can be reflected in real time, and the state monitoring tag can be used to analyze the trend of the health status of the target lithium battery; then, the charge and discharge cycle data is decomposed into smaller time periods through time discretization, so that the state monitoring is more detailed, and the state monitoring granularity within each time period can provide detailed information, which helps to more accurately capture the performance changes of the target battery in different time periods, thereby making the health estimation more accurate, and the lithium battery state monitoring vector can more accurately calculate the degradation rate of the target lithium battery; secondly, by converting the charge and discharge cycle data block into a key-value data pair, the complex original data can be structured, which helps to improve the efficiency and accuracy of the health estimation of the target lithium battery. After spatially embedding each charge and discharge cycle data block in the charge and discharge cycle sequence, the battery degradation characteristics can be more accurately extracted, and by calculating the lithium battery degradation coefficient, the dynamic monitoring of the lithium battery degradation state can be achieved; finally, the health estimation of the target lithium battery through all the lithium battery degradation coefficients can help determine a more accurate health level of the target lithium battery. In summary, the technical solution adopted in this application can extract the degradation characteristics of the target lithium battery to perform health estimation, so as to improve the accuracy of health estimation of the target lithium battery.

[0103] Embodiment 2:

[0104] This application provides a lithium battery status monitoring and estimation system, referring to Figure 2 As shown, this figure is a schematic diagram of an estimation system according to this embodiment of the present application, and the estimation system includes:

[0105] The initialization module 100 is used to obtain the charge and discharge cycle data of the target lithium battery in the aging cycle experiment, and allocate a state monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data.

[0106] The lithium battery status monitoring module 200 is used to time discretize the charge and discharge cycle data to obtain the status monitoring granularity of each time period in the charge and discharge cycle data, and then determine the lithium battery status monitoring vector of the target lithium battery at each status monitoring granularity according to the status monitoring label of each charge and discharge cycle record.

[0107] The lithium battery degradation assessment module 300 is used to spatially embed all lithium battery status monitoring vectors to obtain the charge and discharge cycle sequence of the target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery status monitoring coefficient of each key-value data pair.

[0108] The health estimation module 400 is used to perform health estimation on the target lithium battery according to all lithium battery degradation coefficients, obtain a health estimation value of the target lithium battery, and determine a health level of the target lithium battery based on the health estimation value.

[0109] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0110] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data. It should be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A lithium battery status monitoring and estimation method, characterized in that: The estimation method comprises the following steps: Acquire charge and discharge cycle data of a target lithium battery in an aging cycle experiment, and assign a state monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data; Performing time discretization on the charge and discharge cycle data to obtain a state monitoring granularity for each time period in the charge and discharge cycle data, and then determining a lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to a state monitoring tag recorded for each charge and discharge cycle; Perform spatial embedding on all lithium battery state monitoring vectors to obtain a charge and discharge cycle sequence of a target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair; The health of the target lithium battery is estimated according to all lithium battery degradation coefficients to obtain a health estimation value of the target lithium battery, and the health level of the target lithium battery is determined based on the health estimation value.

2. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: Assigning a status monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data specifically includes: Obtain the actual rated capacity of the target lithium battery before the aging cycle experiment; For each charge-discharge cycle record in the charge-discharge cycle data, determining a cycle experiment period corresponding to the charge-discharge cycle record in the aging cycle experiment; Obtaining the battery measured capacity of the target lithium battery during the cycle test period; Determine the health of the lithium battery corresponding to the charge and discharge cycle record according to the actual rated capacity and the battery measured capacity; The health of the lithium battery is used as a status monitoring label of the charge and discharge cycle record, thereby completing the allocation of the status monitoring label to each charge and discharge cycle record in the charge and discharge cycle data.

3. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: Determining the health of the lithium battery corresponding to the charge and discharge cycle record according to the actual rated capacity and the battery measured capacity is to use the ratio of the battery measured capacity to the actual rated capacity as the health of the lithium battery corresponding to the charge and discharge cycle record.

4. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: The time discretization of the charge-discharge cycle data to obtain the state monitoring granularity of each time period in the charge-discharge cycle data specifically includes: Get the set discrete coefficient; The charge-discharge cycle data is time discretized based on the discrete coefficient to obtain the state monitoring granularity of each time period in the charge-discharge cycle data.

5. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: Determine the lithium battery status monitoring vector of the target lithium battery at each status monitoring granularity based on the status monitoring tag recorded in each charge and discharge cycle. Specifically include: For each state monitoring granularity of the time period, obtain the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity; Determine the typical characteristics of the state monitoring at the state monitoring granularity through the state monitoring tags of all charge and discharge cycle records corresponding to the state monitoring granularity; Determine the battery characteristics corresponding to the state monitoring granularity; A lithium battery state monitoring vector of the target lithium battery at the state monitoring granularity is constructed according to the state monitoring typical characteristics and the battery characteristics, and then a lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity is obtained.

6. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: All lithium battery status monitoring vectors are spatially embedded to obtain the charge and discharge cycle sequence of the target lithium battery, which specifically includes: Construct a state monitoring matrix of the target lithium battery based on all lithium battery state monitoring vectors; dividing the condition monitoring matrix into a plurality of condition monitoring blocks; Each state monitoring block is spatially embedded to obtain multiple charge and discharge cycle data blocks of the target lithium battery; The charge and discharge cycle sequence of the target lithium battery is constructed through all the charge and discharge cycle data blocks.

7. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: Converting each charge-discharge cycle data block in the charge-discharge cycle sequence into a key-value data pair specifically includes: For each charge-discharge cycle data block in the charge-discharge cycle sequence, converting the battery feature in the charge-discharge cycle data block into a key of a key-value data pair; The typical characteristics of state monitoring in the charge-discharge cycle data block are converted into the value of the key-value data pair, thereby completing the key-value data pair conversion of each charge-discharge cycle data block in the charge-discharge cycle sequence.

8. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: The health estimation of the target lithium battery according to all the lithium battery degradation coefficients is to input all the lithium battery degradation coefficients into the health status estimation model based on machine learning to perform health estimation.

9. A lithium battery status monitoring and estimation method as claimed in claim 1, characterized in that: Determining the health level of the target lithium battery based on the health estimation value specifically includes: Get the preset lithium battery health level table; The health level of the target lithium battery in the lithium battery health level table is determined according to the health estimation value, thereby obtaining the health level of the target lithium battery.

10. A lithium battery status monitoring and estimation system, used to execute a lithium battery status monitoring and estimation method as claimed in any one of claims 1 to 9, characterized in that: The estimation system comprises: An initialization module, used to obtain charge and discharge cycle data of a target lithium battery in an aging cycle experiment, and to assign a state monitoring tag to each charge and discharge cycle record in the charge and discharge cycle data; A lithium battery state monitoring module is used to time discretize the charge and discharge cycle data to obtain the state monitoring granularity of each time period in the charge and discharge cycle data, and then determine the lithium battery state monitoring vector of the target lithium battery at each state monitoring granularity according to the state monitoring tag recorded for each charge and discharge cycle; A lithium battery degradation assessment module is used to spatially embed all lithium battery state monitoring vectors to obtain a charge and discharge cycle sequence of a target lithium battery, convert each charge and discharge cycle data block in the charge and discharge cycle sequence into a key-value data pair, and then determine the lithium battery degradation coefficient corresponding to each key-value data pair through the lithium battery state monitoring coefficient of each key-value data pair; The health estimation module is used to perform health estimation on the target lithium battery according to all lithium battery degradation coefficients, obtain a health estimation value of the target lithium battery, and determine the health level of the target lithium battery based on the health estimation value.

Images