Battery management device, battery management method, and battery management system

Abstract

According to some embodiments, a battery management apparatus includes an interface for acquiring battery data of a battery, and a controller for extracting a physical parameter of the battery based on the battery data, classifying a state of the battery corresponding to the physical parameter by using a first model trained using a supervised learning algorithm with respect to a plurality of pieces of deterioration experimental data of the battery, and calculating a probability that the battery is in an abnormal state using a second model fitted based on a classification result of the first model.

Description

Technical Field

[0001] Cross-references to related applications

[0002] This application claims priority to Korean Patent Application No. 10-2024-0080287, filed on June 20, 2024, the disclosure of which is incorporated herein by reference. Technical Field

[0003] The embodiments disclosed herein relate to battery management devices, battery management methods, and battery management systems. Background Technology

[0004] In recent years, research and development of rechargeable batteries have been actively underway. Here, a rechargeable battery is a battery capable of being recharged and discharged, and can be interpreted as including traditional Ni / Cd batteries, Ni / MH batteries, and more recently, lithium-ion batteries. Among rechargeable batteries, lithium-ion batteries can have higher energy densities than conventional Ni / Cd and Ni / MH batteries, and can be manufactured to be smaller and lighter, thus offering high availability as a power source for mobile devices. In recent years, lithium-ion batteries have expanded their applications to power electric vehicles, making them a focus of attention as a next-generation energy storage medium.

[0005] State of Health (SOH) indicates the battery's remaining lifespan, current performance status, etc. In some cases, the battery's SOH may tend to decrease rapidly. In existing technologies, SOH values ​​are measured periodically, and the state at the current point in time can be determined based on the resulting slope. Since battery anomalies caused by a decrease in SOH can cause inconvenience and risk to users, it may be necessary to predict the timing of SOH decreases in advance; however, the methods in related technologies may present difficulties in making such predictions. Summary of the Invention

[0006] Technical issues

[0007] One of the objectives of the embodiments disclosed herein is to provide a battery management device, battery management method, and battery management system that can estimate the timing of abnormal states related to SOH reduction based on the battery's physical parameters.

[0008] The technical objectives of the embodiments disclosed herein are not limited to those mentioned above, and those skilled in the art will clearly understand other technical objectives not mentioned from the following description.

[0009] Technical solution

[0010] According to some embodiments, a battery management device includes: an interface configured to acquire battery data of a battery; and a controller configured to: extract physical parameters of the battery based on the battery data; classify the state of the battery corresponding to the physical parameters using a first model trained by a supervised learning algorithm based on multiple degradation experimental data of the battery; and calculate the probability that the battery is in an abnormal state using a second model fitted by the classification result of the first model.

[0011] According to some implementations, the first model can be configured to be trained based on the physical parameters extracted from each of the plurality of degradation experimental data and the anomaly marker data pre-added to the plurality of degradation experimental data.

[0012] According to some implementations, the physical parameters may include a first feature and a second feature for training the first model, and the first feature may include loss of active material (LAM), and the second feature may include loss of lithium source (LLI).

[0013] According to some implementations, the controller can be configured to: derive a first open-circuit voltage (OCV) value at the start of life (BOL) time point and a second OCV value at the mid-life (MOL) time point based on the battery data; and extract the physical parameters based on the first OCV value and the second OCV value.

[0014] According to some implementations, the second model may include a logistic regression model, which is trained to calculate the probability based on the state of the battery and the physical parameters.

[0015] According to some implementations, the controller can also be configured to provide notification functionality to the user of the battery based on the battery's state and the probability.

[0016] According to some implementations, the controller can be configured to provide the user with the expected time point when the battery's state is to become abnormal, based on the battery's state and the probability.

[0017] According to some embodiments, a battery management method includes the following steps: acquiring battery data of a battery; extracting physical parameters of the battery based on the battery data; classifying the state of the battery corresponding to the physical parameters using a first model trained by a supervised learning algorithm using multiple degradation experimental data of the battery; and calculating the probability that the battery is in an abnormal state using a second model fitted by the classification result of the first model.

[0018] According to some implementations, the first model can be configured to be trained based on the physical parameters extracted from each of the plurality of degradation experimental data and the anomaly marker data pre-added to the plurality of degradation experimental data.

[0019] According to some implementations, the physical parameters may include a first feature and a second feature for training the first model, and the first feature may include loss of active material (LAM), and the second feature may include loss of lithium source (LLI).

[0020] According to some implementations, the step of extracting the physical parameters includes: deriving a first open-circuit voltage (OCV) value at the start of life (BOL) time point and a second OCV value at the mid-life (MOL) time point based on the battery data; and extracting the physical parameters based on the first OCV value and the second OCV value.

[0021] According to some implementations, the second model includes a logistic regression model, which is trained to calculate the probability based on the state of the battery and the physical parameters.

[0022] According to some implementations, the battery management method may further include: providing a notification function to the user of the battery based on the state of the battery and the probability.

[0023] According to some implementations, providing the notification function may include: providing the user with the time point at which the battery's state is expected to become abnormal, based on the battery's state and the probability.

[0024] According to some embodiments, a battery management system includes: a battery; a charging / discharging device for charging and discharging the battery; and a battery management device configured to acquire battery data of the battery, extract physical parameters of the battery based on the battery data, classify the states of the battery corresponding to the physical parameters using a first model trained by a supervised learning algorithm based on multiple degradation experimental data of the battery, and calculate the probability that the state of the battery is an abnormal state using a second model fitted by the classification result of the first model.

[0025] Beneficial effects

[0026] According to the embodiments disclosed herein, a battery management device, battery management method, and battery management system can be provided that can estimate the timing of abnormal states related to SOH reduction based on the physical parameters of the battery.

[0027] The technical effects of the embodiments disclosed herein are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art based on the disclosure herein. Attached Figure Description

[0028] Figure 1 The components constituting a battery management system according to some embodiments are shown.

[0029] Figure 2 The components constituting a battery management device according to some embodiments are shown.

[0030] Figure 3 This demonstrates a conventional technique for determining battery anomalies based on the SOH slope.

[0031] Figure 4 The process of training an AI model according to some implementation methods is shown, as well as the process of using the AI ​​model to calculate battery state and anomaly probability.

[0032] Figure 5 A method for training a first model using the physical parameters of a battery according to some implementations is shown.

[0033] Figure 6 This illustrates a method for training a second model using the classification results of a first model, according to some implementations.

[0034] Figure 7 The steps constituting a battery management method according to some embodiments are shown. Detailed Implementation

[0035] In the following description, embodiments disclosed herein will be illustrated with reference to the accompanying drawings. However, this is not intended to limit the disclosure herein to the specific embodiments, and it should be construed as including various modifications, equivalents, and / or substitutions of the embodiments disclosed herein.

[0036] It should be understood that the embodiments and terminology used herein are not intended to limit the technical features stated herein to specific embodiments, and include various changes, equivalents, or substitutions to corresponding embodiments. Regarding the description of the drawings, similar or related reference numerals may be used to denote similar or related elements. It should be understood that the singular form of the noun corresponding to an item may include one or more things unless the relevant context clearly indicates otherwise.

[0037] As used herein, each of the phrases such as “A or B,” “at least one of A and B,” “at least one of A or B,” “A, B or C,” “at least one of A, B and C,” and “at least one of A, B or C” may include any or all possible combinations of the items listed together in the corresponding one of the phrases. Terms such as “first” and “second,” “first,” “second,” “A,” “B,” “(a),” or “(b)” are used simply to distinguish the corresponding component from another component and do not limit the component in any other way (e.g., in terms of importance or order) unless otherwise specifically stated.

[0038] In this specification, it should be understood that if an element (e.g., a first element) is referred to as being "connected to," "linked to," or "in contact with" another element (e.g., a second element) with or without the terms "operably" or "communically," it means that the element can be connected to the other element directly (e.g., via wired or wireless) or indirectly (e.g., via a third element).

[0039] The methods disclosed herein can be included and provided in a computer program product. The computer program product can be traded as a product between a seller and a buyer. The computer program product can be distributed in the form of a machine-readable storage medium (e.g., an optical disc read-only memory and a CD-ROM), or distributed online via an app store (e.g., downloaded or uploaded), or directly between two operator devices. If distributed online, at least a portion of the computer program product can be temporarily generated or at least temporarily stored in a machine-readable storage medium, such as the memory of a manufacturer's server, the memory of an app store's server, or the memory of a relay server.

[0040] According to the embodiments disclosed herein, each component (e.g., a module or program) of the above-described components may include a single entity or multiple entities, and some of the multiple entities may be arranged separately from the other components. According to the embodiments disclosed herein, one or more components or operations of the above-described components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, multiple components (e.g., modules or programs) may be integrated into a single component. In this case, the integrated component may still perform one or more functions of each of the multiple components in the same or similar manner as the functions performed by the corresponding component of the multiple components prior to integration. According to the embodiments disclosed herein, operations performed by modules, programs, or other components may be performed sequentially, in parallel, repeatedly, or heuristically, or one or more operations may be performed in a different order or omitted, or one or more other operations may be added.

[0041] Figure 1 Examples may be made of elements constituting a battery management system according to some embodiments.

[0042] Reference Figure 1 The battery management system 100 may include a charging / discharging device 110, a battery 120, and a battery management device 130. However, the battery management system is not limited to this, and some components may be omitted from the battery management system 100, or other common components may be further included in the battery management system 100.

[0043] The charging / discharging device 110 can be configured to charge or discharge the battery 120. For example, the charging / discharging device 110 may include a power supply device, etc. The charging / discharging device 110 can apply charge / discharge cycles to the battery 120 to perform a degradation test on the battery 120. According to an embodiment, the charging / discharging device 110 may include a power supply device, and the power supply device may include a mobile device such as an electric vehicle (EV), a hybrid electric vehicle (HEV), an electric bicycle, etc.

[0044] Battery 120 may include a battery pack, etc., which is a management target of system 100. The battery pack of battery 120 may include multiple battery modules, and each battery module may include multiple battery cells. According to embodiments, battery 120 can be installed in various types of mobile devices. The charge / discharge cycle performed by charge / discharge device 110 can be applied to battery 120 to artificially induce various degradation conditions.

[0045] The battery management device 130 can perform operations for diagnosing or managing the battery 120. The battery management device 130 can acquire battery data from the battery 120 and diagnose or manage the state of the battery 120 based on the battery data. According to an embodiment, the battery management device 130 may include a battery management system (BMS) and / or an external device, wherein the BMS is configured on-board with the battery 120, and the external device is remotely located outside the vehicle. The external device may include a charger from a battery charging station, battery diagnostic equipment, a cloud computing server, etc.

[0046] The battery management system 100 may also include a management server. The management server manages the management results of the battery management device 130. The management server can exchange data with the battery management device 130 via wired / wireless communication. When a defect in the battery 120 is diagnosed or its lifespan is predicted, the results can be transmitted to the management server and recorded in a database. According to an embodiment, the battery management device 130 can perform diagnostic operations by executing battery management software, and the management server can provide updated information about the battery management software to the battery diagnostic device 130.

[0047] Figure 2 Examples may be made of elements constituting a battery management device according to some embodiments.

[0048] Reference Figure 2 The battery management device 130 may include an interface 131 and a controller 132. However, the battery management device is not limited to this, and some components may be omitted from the battery management device 130, or other general components may be further included in the battery management device 130.

[0049] Interface 131 can acquire battery data from battery 120. According to an embodiment, interface 131 may include a communication unit configured to receive battery data and / or a sensor unit configured to measure battery data. According to an embodiment, when the battery management device 130 is implemented externally to the vehicle, the communication unit may receive battery data via methods such as wired data communication or wireless data communication. Alternatively, when the battery management device 130 is implemented in an in-vehicle configuration, the sensor unit may be configured to measure values ​​associated with battery 120, such as voltage, current, temperature, and resistance.

[0050] The controller 132 may have a structure for executing commands that implement the operation of the battery management device 130. The controller 132 may be implemented as an array of multiple logic gates for handling various operations, or as a general-purpose microprocessor, and may include a single processor or multiple processors. For example, the controller 132 may be implemented as at least one of a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), and an application processor (AP).

[0051] The controller 132 can operate in conjunction with a memory configured to store various data, commands, mobile applications, computer programs, etc. The memory can be configured separately from or integrated with the controller 132. The controller 132 can process various operations by executing commands stored in the memory. The memory can be implemented as a non-volatile device such as read-only memory (ROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electrically erasable PROM (EEPROM), flash memory, parallel random access memory (PRAM), magnetoresistive RAM (MRAM), resistive RAM (RRAM), ferroelectric RAM (FRAM), etc., or as a volatile device such as dynamic RAM (DRAM), static RAM (SRAM), synchronous dynamic RAM (SDRAM), parallel RAM (PRAM), etc., or as a hard disk drive (HDD), solid-state drive (SSD), secure digital storage (SD), micro SD, or combinations thereof.

[0052] Interface 131 can be configured to acquire battery data of battery 120. Battery data may include voltage, current, temperature, resistance, etc., of battery 120. Battery data may be acquired in units such as cells, modules, and battery packs of battery 120. Battery data may include time-series data containing values ​​collected at regular intervals. According to embodiments, battery data may include measurement data acquired when battery 120 is charged or discharged by charging / discharging device 110.

[0053] The controller 132 can be configured to extract physical parameters of the battery 120 based on battery data. These physical parameters can refer to parameters representing the physical characteristics of the battery 120. For example, physical parameters may include loss of active material (LAM), loss of lithium source (LLI), etc. The physical parameters can be used as machine learning features to identify the state of rapid decrease in the state of equilibrium (SOH) of the battery 120.

[0054] The controller 132 can be configured to classify the states of the battery 120 corresponding to physical parameters using a first model trained using a supervised learning algorithm on multiple degradation experimental data sets of the battery 120. These multiple degradation experimental data sets can be generated by repeatedly charging and discharging the battery 120 under different degradation conditions. For example, the charging / discharging device 110 can apply different charge / discharge cycles to the battery 120, and during this period, data output from the battery 120 can be collected as multiple degradation experimental data sets. For example, the multiple degradation experimental data sets can include different combinations of active material loss (LAM) values ​​and lithium source loss (LLI) values, and each combination can be assigned a label value. The label value can indicate whether the battery 120 is abnormal according to each degradation condition. The multiple degradation experimental data sets can be used as training data to train the first model. For example, the first model can include a support vector machine (SVM) model. After training is complete, the first model can receive physical parameters as input and output the corresponding battery states.

[0055] The controller 132 can be configured to calculate the probability that the state of battery 120 is an anomalous state using a second model fitted with the classification results of the first model. When training of the first model is complete, it can be used to fit the second model. Based on this fit, the probability of battery 120 being anomalous, output by the second model, can be optimized. For example, logistic regression can be used to train the second model to output the probability of battery 120 being anomalous based on physical parameters and the output of the first model. By referring to the outputs of the first and second models, it is possible to check whether an anomalous state exists in the current battery 120, and when no anomalous state exists, the probability of an anomalous state in battery 120 can be estimated.

[0056] According to the implementation method, the first model can be configured to be trained based on physical parameters extracted from each of multiple degradation test data sets and anomaly-labeled data pre-added to the multiple degradation test data sets. The anomaly-labeled data can be pre-added to the multiple degradation test data sets, and through the anomaly-labeled data, it can be determined whether the degradation test conditions of each degradation test data set lead to an anomaly or defect in the battery 120. The state of the battery 120 classified by the anomaly-labeled data can be represented by physical parameters. The first model can learn the relationship between the physical parameters and the state of the battery 120.

[0057] According to an implementation, the physical parameters may include a first feature and a second feature for training the first model, wherein the first feature may include loss of active material (LAM) and the second feature may include loss of lithium source (LLI). The physical parameters may represent a combination of LAM and LLI values, and the first model may learn how the combination of LAM and LLI values ​​relates to the state of battery 120.

[0058] According to an implementation, the controller 132 can be configured to derive a first open-circuit voltage (OCV) value at the start of life (BOL) time point and a second OCV value at the mid-life (MOL) time point based on battery data, and extract physical parameters based on the first and second OCV values. The LAM value and LLI value can be calculated based on the OCV difference between the BOL and MOL time points.

[0059] According to an implementation, the second model may include a logistic regression model trained to calculate probabilities based on physical parameters and the state of battery 120. When the training of the first model is complete, the first model can output the state of battery 120. In addition to the state, by considering the physical parameters together, the second model can output the probability of anomalies in battery 120. Alternatively, other suitable models besides the logistic regression model may be applied to the second model.

[0060] According to an implementation, the controller 132 can also be configured to provide notification functionality to a user associated with the battery 120 based on the state and probability of the battery 120. For example, notification functionality can be provided to a user's mobile terminal and / or a vehicle equipped with the battery 120. The notification content may include whether an anomaly currently exists in the battery 120 and the probability of such an anomaly.

[0061] According to one implementation, controller 132 can be configured to provide the user with the expected time point when the state of battery 120 is to become abnormal, based on the state and probability of battery 120. Although the current state of battery 120 is not abnormal, it may become abnormal in the near future when the probability of battery 120 becoming abnormal is high. Taking into account the battery's physical parameters and the probability of abnormality, the expected time point of the abnormal state can be calculated. According to another implementation, a second model can be trained to output the expected time point and the probability of battery 120 becoming abnormal. Alternatively, controller 132 can be configured to calculate the expected time of the abnormal state by introducing a third model separate from the first and second models. For example, the third model can be configured to calculate an earlier expected time point when the probability of abnormality is high.

[0062] Figure 3 This demonstrates a conventional technique for determining battery anomalies based on the SOH slope.

[0063] Reference Figure 3 A graph 300 can be shown illustrating a conventional technique for determining battery anomalies based on the SOH slope.

[0064] In conventional techniques, the method of checking the fluctuation trend of the SOH value during battery discharge is used, and when the magnitude of the SOH slope exceeds a certain value and / or a certain ratio, an abnormal state is determined to have occurred.

[0065] However, because traditional techniques require accumulated past data to identify trends in the SOH slope, it can be difficult to identify anomalies when only data from the current point in time is processed. Furthermore, even when considering past data, only the current state can be determined, and it is difficult to predict the state at future points in time.

[0066] Figure 4 The process of training an AI model according to some implementation methods is shown, as well as the process of using the AI ​​model to calculate battery state and anomaly probability.

[0067] Reference Figure 4 The process of training the AI ​​model 410 and the process of using the AI ​​model to calculate the battery state and the probability of anomalies 420 can be shown.

[0068] In step 411, various designs of experiments (DOEs) can be used to collect multiple degradation experiment data. In step 412, physical parameters such as LLI and LAM can be extracted for each experiment. These can be used as features for AI model learning. In step 413, a labeling task can be performed for each experiment to distinguish between normal and abnormal states. The labeling task can be prepared in advance before the model training process.

[0069] In step 414, training of the classification AI model can be performed. For multiple DOEs, the classification AI model can learn how the combination of LLI and LAM values ​​is associated with the labeled battery state. In step 415, a probabilistic AI model providing anomaly probabilities can be trained. When both the classification AI model and the probabilistic AI model have been trained, model training can be completed in step 416.

[0070] In step 421, physical parameters such as LLI and LAM values ​​can be extracted from the battery, which is the target for state estimation. In step 422, these physical parameters can be input into the AI ​​model. In step 423, the AI ​​model can provide the battery's state and the probability of anomalies. In step 424, a notification function can be provided to the user based on the battery's state and the probability of anomalies. The notification function can inform the user of the battery's current state and the expected time of anomalies.

[0071] Figure 5 A method for training a first model using the physical parameters of a battery according to some implementations is shown.

[0072] Reference Figure 5 Figure 500 illustrates a method for training a first model using the physical parameters of a battery.

[0073] Referring to Figure 500, the data points corresponding to the combination of LAM values ​​on the horizontal axis and LLI values ​​on the vertical axis can be plotted on a two-dimensional graph. Each of the plotted data points can correspond to multiple degradation experiment data. Data points can be labeled as normal or abnormal.

[0074] The first model can set a baseline 510 based on the distribution of data points. The baseline 510 can be used to distinguish whether a combination of LAM and LLI values ​​is normal or abnormal. The baseline 510 can be formed not only as a straight line, but also as various curved shapes or combinations of straight and curved shapes. The first model can be trained to adjust the position of the baseline 510 to optimally classify data points from multiple deteriorated experimental datasets.

[0075] Once the first model is trained, the position of baseline 510 can be determined. Then, when the combination of the LAM and LLI values ​​of the target battery to be estimated is input into the first model, the state of the target battery to be estimated can be classified by comparing the combined data points with baseline 510.

[0076] Figure 6 This illustrates a method for training a second model using the classification results of a first model, according to some implementations.

[0077] Reference Figure 6Figure 600 illustrates an example of using the classification results of the first model to train the second model.

[0078] The two-dimensional area of ​​chart 600 can be divided into different colors based on the probability of an anomaly. For example, legend 610 can represent the matching relationship between color and anomaly probability. This can be based on... Figure 5 The baseline 510 of chart 500 is used to form the two-dimensional area of ​​chart 600.

[0079] For example, when the training of the first model is complete, the position of baseline 510 can be determined. The position of baseline 510 may correspond to an anomaly probability of approximately 0.5. Therefore, a second model can be trained to align the boundary line representing the anomaly probability of 0.5 with the position of baseline 510. Given a boundary line representing an anomaly probability of 0.5, the boundary lines corresponding to anomaly probabilities of other values ​​can also be adjusted based on parallel shifts relative to that boundary line.

[0080] Once the locations of the boundary lines representing the probabilities of anomalies are determined, training of the second model can be completed. When the combination of the LAM and LLI values ​​of the target battery to be estimated is input into the second model, it can first be determined whether the corresponding combination is classified as normal or abnormal. When the corresponding combination is classified as normal, the anomaly probability value can be provided based on the boundary lines of the second model. According to the implementation, the second model can be configured to provide anomaly probability values ​​for combinations of LAM and LLI values ​​classified as anomalous.

[0081] Figure 7 The steps constituting a battery management method according to some embodiments are shown.

[0082] See Figure 7 The battery management method 700 may include steps 710 to 740. However, the battery checking method is not limited to this, and some steps may be omitted or other general steps may be added, and the steps of the battery management method 700 may be performed in a different order than that shown.

[0083] The battery management method 700 may include time-series processing steps in the battery management device 130. Therefore, even if the following description is omitted, the content described above for the battery management device 130 can be equally applied to the battery management method 700.

[0084] Steps 710 to 740 of the battery management method 700 can be executed by the interface 131 and controller 132 of the battery management device 130.

[0085] In step 710, the battery management device 130 may perform the step of acquiring battery data of the battery.

[0086] In step 720, the battery management device 130 may perform a step of extracting the physical parameters of the battery based on the battery data.

[0087] In step 730, the battery management device 130 may perform a step of classifying the state of the battery corresponding to physical parameters using a first model, which is trained using a supervised learning algorithm based on multiple degradation experimental data of the battery.

[0088] In step 740, the battery management device 130 may perform the step of calculating the probability that the battery is in an abnormal state using a second model fitted by the classification results of the first model.

[0089] According to an embodiment, the battery management method 700 can be implemented in the form of a computer program stored in a computer-readable storage medium. That is, the computer program may include commands for implementing the battery management method 700, and the commands of the program may be stored in a computer-readable storage medium. The computer program may include a mobile application.

[0090] According to the implementation, computer-readable storage media may include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floppy disks; and hardware devices specifically configured to store and execute computer program instructions, such as ROMs, RAMs, and flash memory. Computer program instructions may include machine language code generated by a compiler and high-level language code that can be executed by a computer using an interpreter.

[0091] Unless otherwise stated, terms such as “comprising,” “including,” or “having” as described above mean that the corresponding component may be present and should therefore be interpreted as potentially including, rather than excluding, other components. Unless otherwise defined, all terms including technical or scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments disclosed herein pertain. Commonly used terms, such as those defined in dictionaries, should be interpreted as having meanings consistent with their meanings in the context of the relevant art and will not be interpreted as having idealized or overly formal meanings unless expressly defined herein.

[0092] The above description is merely an example of the technical ideas disclosed herein, and various modifications and variations can be made by those skilled in the art to which the embodiments disclosed herein pertain without departing from the essential features of the embodiments disclosed herein. Therefore, the embodiments disclosed herein are not intended to limit the technical ideas of the embodiments disclosed herein, but rather to interpret them, and the scope of the technical ideas disclosed herein is not limited by these embodiments. The scope of protection disclosed herein should be interpreted by the appended claims, and all technical ideas within the scope equivalent to those claims should be interpreted as being included within the scope of the rights herein.

[0093] (Reference list)

[0094] 100: Battery Management System; 110: Charging / Discharging Device

[0095] 120: Battery; 130: Battery management device

[0096] 131: Interface 132: Controller

Claims

1. A battery management device, the battery management device comprising: An interface configured to acquire battery data: and The controller is configured to: The physical parameters of the battery are extracted based on the battery data; Using multiple degradation experimental data points for the battery, a first model trained via a supervised learning algorithm is used to classify the states of the battery corresponding to the physical parameters: and The probability that the battery is in an abnormal state is calculated using a second model fitted by the classification results of the first model.

2. The battery management device according to claim 1, wherein, The first model is configured to be trained based on the physical parameters extracted from each of the multiple degradation experiment data and the anomaly marker data pre-added to the multiple degradation experiment data.

3. The battery management device according to claim 1, wherein, The physical parameters include a first feature and a second feature used to train the first model, and The first feature includes loss of active material (LAM), and the second feature includes loss of lithium source (LLI).

4. The battery management device according to claim 1, wherein, The controller is configured to: Based on the battery data, derive the first open-circuit voltage (OCV) value at the beginning of the lifespan (BOL) and the second OCV value at the middle of the lifespan (MOL): The physical parameters are extracted based on the first OCV value and the second OCV value.

5. The battery management device according to claim 1, wherein, The second model includes a logistic regression model, which is trained to calculate the probability based on the state of the battery and the physical parameters.

6. The battery management device according to claim 1, wherein, The controller is also configured to provide notification functionality to the user of the battery based on the battery's state and the probability.

7. The battery management device according to claim 6, wherein, The controller is configured to provide the user with the expected time point when the battery's state is to become abnormal, based on the battery's state and the probability.

8. A battery management method, the battery management method comprising the following steps: Obtain battery data; The physical parameters of the battery are extracted based on the battery data; Using multiple degradation experimental data points for the battery, a first model trained via a supervised learning algorithm is used to classify the battery's states corresponding to the physical parameters: and The probability that the battery is in an abnormal state is calculated using a second model fitted by the classification results of the first model.

9. The battery management method according to claim 8, wherein, The first model is configured to be trained based on the physical parameters extracted from each of the multiple degradation experiment data and the anomaly marker data pre-added to the multiple degradation experiment data.

10. The battery management method according to claim 8, wherein, The physical parameters include a first feature and a second feature used to train the first model, and The first feature includes loss of active material (LAM), and the second feature includes loss of lithium source (LLI).

11. The battery management method according to claim 8, wherein, The steps for extracting the physical parameters include: Based on the battery data, the first open-circuit voltage (OCV) value at the beginning of the lifespan (BOL) and the second OCV value at the middle of the lifespan (MOL) are derived: The physical parameters are extracted based on the first OCV value and the second OCV value.

12. The battery management method according to claim 8, wherein, The second model includes a logistic regression model, which is trained to calculate the probability based on the state of the battery and the physical parameters.

13. The battery management method according to claim 8, further comprising the following steps: The system provides notifications to the user of the battery based on the battery's state and the probability.

14. The battery management method according to claim 13, wherein, The steps of providing the notification function include: providing the user with the expected time point when the battery's state is expected to become abnormal, based on the battery's state and the probability.

15. A battery management system, the battery management system comprising: Battery; A charging / discharging device configured to charge and discharge the battery: and A battery management device is configured to acquire battery data of a battery, extract physical parameters of the battery based on the battery data, classify the state of the battery corresponding to the physical parameters using a first model trained by a supervised learning algorithm using multiple degradation experimental data of the battery, and calculate the probability that the state of the battery is an abnormal state using a second model fitted by the classification result of the first model.

Images