Electronic device, recording medium, and battery degradation factor analysis method therefor

Abstract

An electronic device for analyzing a battery degradation factor is provided. The electronic device can: acquire a plurality of pieces of battery charging data for a plurality of vehicles; identify, on the basis of the plurality of pieces of battery charging data, a plurality of factors related to battery charging; generate a plurality of factor combinations on the basis of the plurality of factors; generate, for respective factor combinations, on the basis of one or more pieces of battery charging data corresponding thereto, a plurality of factor combination features indicating a charging pattern related to the factor combination; perform, on the basis of a clustering algorithm, clustering on the plurality of factor combination features indicating a charging pattern related to the same factor combination; on the basis of a deviation in data distribution between a plurality of clusters generated to correspond to respective factor combinations; extract, from among the plurality of factor combinations, one or more factor combinations corresponding to factor combinations related to battery degradation; identify, on the basis of an artificial intelligence-based model, information related to the effect of each of the one or more factor combinations on battery degradation; and provide, on the basis of the information related to the effect on battery degradation, information related to an optimal factor combination for preventing battery degradation.

Description

Method for analyzing deterioration factors of electronic devices, recording media and their batteries

[0001] The present disclosure relates to an electronic device, a recording medium, and a method for analyzing battery degradation factors thereof, and more particularly, to a technique for selecting factors affecting battery degradation based on the diverse and complex charging patterns of electric vehicle users, and analyzing the relative influence between the selected factors and an optimal method for preventing battery degradation.

[0002] This application claims the benefit of priority to Republic of Korea Patent Application No. 2024-0117680, dated August 30, 2024, and Republic of Korea Patent Application No. 2024-0146419, dated October 24, 2024, the entire contents of which are incorporated herein by reference.

[0003] AI technology can be leveraged to identify factors that influence battery degradation, based on diverse data on the charging patterns of electric vehicle users. However, conventional approaches, such as those based on AI, analyze factors related to specific situations or environments. Factor analysis is performed by fixing all other factors except for a specific factor and analyzing its impact on the environment. This precludes analysis of the interrelationships between multiple factors and the impact of these multifactorial relationships. Consequently, this approach can lead to limitations, such as the potential for analysis results that differ somewhat from actual environments.

[0004] Meanwhile, as an alternative that takes this into account, conventional correlation analysis can be performed. However, even with correlation analysis, it can only analyze the degree of linearity or monotonicity between specific factors and other factors. It is difficult to accurately analyze the complex mathematical relationships between factors. Furthermore, analyzing environmental changes over time, such as the triggering of battery degradation in the operating environment, is nearly impossible. Therefore, there is a growing need for a method that can more precisely analyze the factors affecting battery degradation by comprehensively considering multiple interrelated factors based on the complex charging behavior of EV users.

[0005] The present disclosure is proposed to solve the above-described problems, and provides a method for analyzing an electronic device and its battery degradation factors.

[0006] The technical task to be achieved by this embodiment is not limited to the task described above, and other technical tasks can be inferred from the following examples.

[0007] An electronic device according to one embodiment comprises: a transceiver; a processor; And one or more memories storing one or more instructions, wherein the one or more instructions, when executed, cause the processor to obtain a plurality of battery charging data for a plurality of vehicles, identify a plurality of factors related to battery charging based on the plurality of battery charging data, generate a plurality of factor combinations based on the plurality of factors, generate a plurality of factor combination features representing a charging pattern related to the factor combination based on one or more corresponding battery charging data for each of the plurality of factor combinations, perform clustering targeting the plurality of factor combination features representing a charging pattern related to the same factor combination based on a clustering algorithm, extract one or more factor combinations corresponding to a factor combination related to battery degradation among the plurality of factor combinations based on a deviation in data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations, and identify information on an effect of each of the one or more factor combinations on battery degradation based on an artificial intelligence-based model, and provide information on an optimal factor combination for preventing battery degradation based on the information on the effect on battery degradation.

[0008] According to one embodiment, the plurality of battery charging data is data collected for a preset period of time for each of the plurality of batteries included in the plurality of vehicles, and may include at least one of information on a state of charge (SoC) of the battery over time, information on a charging current rate (C-rate) of the battery over time, information on a charging temperature of the battery over time, and information on a charging cycle of the battery over time.

[0009] According to one embodiment, the plurality of factors may include at least one of a charge start SoC, a charge end SoC, a C-rate, a charge temperature, a charge cycle, and a charge time zone.

[0010] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to identify one or more battery charge data corresponding to a first argument included in a first argument combination among a plurality of argument combinations and one or more battery charge data corresponding to a second argument included in the first argument combination, and to generate a plurality of first argument combination features representing a charging pattern associated with the first argument combination based on the one or more battery charge data corresponding to the first argument and the one or more battery charge data corresponding to the second argument.

[0011] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to cluster a plurality of first factor combination features corresponding to a first factor combination among a plurality of factor combination features into a plurality of first clusters based on a clustering algorithm, determine whether a deviation between distributions of state of health (SoH) changes of a battery corresponding to each of the plurality of first clusters satisfies a first condition, and determine that the first factor combination corresponds to a factor combination related to battery degradation if the deviation between distributions of SoH changes corresponding to each of the plurality of first clusters satisfies the first condition.

[0012] According to one embodiment, the distribution of SoH variations corresponding to each of the plurality of first clusters may be a distribution formed based on the SoH variations corresponding to each of one or more first factor combination features included in each of the plurality of first clusters.

[0013] According to one embodiment, the first condition may include at least one of a condition regarding whether there is overlap between data distributions of each of the plurality of clusters and a condition regarding a difference between data mean values ​​of each of the plurality of clusters.

[0014] According to one embodiment, the artificial intelligence-based model may be a model constructed by modeling the correlation between the learning input data set and the learning output data set, using each of a plurality of factor combination features corresponding to each of one or more factor combinations and the SoH value of the vehicle corresponding to each of the plurality of factor combination features as a learning input data set, and the SoH change amount of the vehicle corresponding to each of the plurality of factor combination features as a learning output data set.

[0015] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to extract, for each of one or more factor combinations, a representative factor combination feature corresponding to a first type or a second type, generate a plurality of input data sets including a preset SoH value and the extracted representative factor combination feature, wherein each of the plurality of input data sets is generated such that only one of the representative factor combination features included corresponds to the second type, input the plurality of input data sets into an artificial intelligence-based model, and determine an SoH variation corresponding to each of the plurality of input data sets output from the artificial intelligence-based model as information regarding an effect of each of the one or more factor combinations on battery degradation.

[0016] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to identify an input data set corresponding to a smallest value among SoH variations corresponding to each of a plurality of input data sets, and to identify a factor combination corresponding to a second type of representative factor combination features included in the identified input data set as an optimal factor combination.

[0017] In one embodiment, the artificial intelligence-based model may be a model further trained to output an optimal set of factor combinations in which at least one factor combination is changed in response to the input data set, based on the SoH variation corresponding to the input data set.

[0018] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to extract, for each of one or more factor combinations, a representative factor combination feature corresponding to a first type or a second type, generate a plurality of input data sets including preset SoH values ​​and the extracted representative factor combination features, the plurality of input data sets corresponding to all cases of extracting the representative factor combination features for each of the one or more factor combinations, input the plurality of input data sets into a trained model, identify an optimal factor combination set corresponding to each of the plurality of input data sets output from the trained model, and provide the optimal factor combination set corresponding to each of the plurality of input data sets as information regarding the optimal factor combination.

[0019] According to one embodiment, a method for analyzing a battery degradation factor may include the steps of: acquiring a plurality of battery charging data for a plurality of vehicles; identifying a plurality of factors related to battery charging based on the plurality of battery charging data; generating a plurality of factor combinations based on the plurality of factors; generating a plurality of factor combination features representing a charging pattern related to the factor combination based on one or more corresponding battery charging data for each of the plurality of factor combinations; performing clustering targeting the plurality of factor combination features representing a charging pattern related to the same factor combination based on a clustering algorithm; extracting one or more factor combinations corresponding to a factor combination related to battery degradation from among the plurality of factor combinations based on a deviation in data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations; identifying information regarding an effect of each of the one or more factor combinations on battery degradation based on an artificial intelligence-based model; and providing information regarding an optimal factor combination for preventing battery degradation based on the information regarding the effect on battery degradation.

[0020] A non-transitory computer-readable recording medium having recorded thereon a program for executing a battery degradation factor analysis method according to one embodiment of the present invention on a computer, the method comprising: acquiring a plurality of battery charging data for a plurality of vehicles; identifying a plurality of factors related to battery charging based on the plurality of battery charging data; generating a plurality of factor combinations based on the plurality of factors; generating a plurality of factor combination features representing a charging pattern related to the factor combination based on at least one corresponding battery charging data for each of the plurality of factor combinations; performing clustering targeting the plurality of factor combination features representing a charging pattern related to the same factor combination based on a clustering algorithm; extracting at least one factor combination corresponding to a factor combination related to battery degradation from among the plurality of factor combinations based on a deviation in data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations; identifying information regarding an effect of each of at least one factor combination on battery degradation based on an artificial intelligence-based model; and providing information regarding an optimal factor combination for preventing battery degradation based on the information regarding the effect on battery degradation.

[0021] According to the present disclosure, various combinations of factors affecting battery degradation can be analyzed to determine a combination of factors that can minimize battery degradation.

[0022] Additionally, according to the present disclosure, by analyzing the charging patterns of electric vehicle users in detail, an optimal solution for preventing battery degradation can be provided.

[0023] The effects of the invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

[0024] Figure 1 illustrates a block diagram of an electronic device according to one embodiment.

[0025] Figure 2 shows a flowchart of a battery degradation factor analysis method according to one embodiment.

[0026] Figure 3 illustrates a process for generating factor combination features according to one embodiment.

[0027] FIG. 4 illustrates a process of extracting one or more factor combinations based on a clustering algorithm according to one embodiment.

[0028] Figure 5 shows the data distribution among multiple clusters related to factor combination 1 according to one embodiment.

[0029] Figure 6 shows the data distribution among multiple clusters related to factor combination 2 according to one embodiment.

[0030] FIG. 7 illustrates a process for identifying information about the impact of each of one or more factor combinations on battery degradation based on an artificial intelligence-based model according to one embodiment.

[0031] Figure 8 illustrates a process for training an artificial intelligence-based model according to one embodiment.

[0032] Figure 9 illustrates an input data set and an output data set of an artificial intelligence-based model according to one embodiment.

[0033] Figure 10 illustrates a process for identifying an optimal set of factor combinations based on an artificial intelligence-based model according to one embodiment.

[0034] FIG. 11 illustrates an input data set and an output data set of a trained artificial intelligence-based model according to one embodiment.

[0035] The terms used in the examples have been selected from widely used, current terms, taking into account the functions of the present disclosure. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, in which case their meanings will be described in detail in the relevant description. Therefore, the terms used in this disclosure should not be defined simply as names, but rather based on the meanings of the terms and the overall content of the present disclosure.

[0036] When a part of the specification is said to "include" a component, this does not exclude other components, but rather implies the inclusion of other components, unless otherwise specifically stated. Furthermore, terms such as "part" and "module" used in the specification mean a unit that processes at least one function or operation, which may be implemented in hardware, software, or a combination of hardware and software.

[0037] The expression "at least one of a, b, and c" described throughout the specification may encompass 'a alone', 'b alone', 'c alone', 'a and b', 'a and c', 'b and c', or 'all of a, b, and c'.

[0038] Below, embodiments of the present disclosure are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the present disclosure. However, the present disclosure may be implemented in various different forms and is not limited to the embodiments described herein.

[0039]

[0040] Hereinafter, embodiments of the present disclosure relating to an electronic device for clustering data regarding a battery charging pattern are described in detail with reference to the drawings.

[0041] Figure 1 illustrates a block diagram of an electronic device according to one embodiment.

[0042] Referring to FIG. 1, an electronic device (100) may include, according to one embodiment, a transceiver (110), a processor (120), and a memory (130). The electronic device (100) illustrated in FIG. 1 only includes components related to the present embodiment. Therefore, it will be understood by those skilled in the art related to the present embodiment that other general components may be included in addition to the components illustrated in FIG. 1.

[0043] For example, the electronic device (100) may include a communication device including one or more transceivers (110), an input unit, and an output unit. The communication unit is a device for performing wired / wireless communication and may communicate with an external electronic device. The external electronic device may be a terminal or a server. In addition, communication technologies used by the communication unit may include GSM (Global System for Mobile communication), CDMA (Code Division Multi Access), LTE (Long Term Evolution), 5G, WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Bluetooth, RFID (Radio Frequency Identification), Infrared Data Association (IrDA), ZigBee, NFC (Near Field Communication), etc. The input unit may be, for example, a traditional keypad or keyboard, a mouse, a microphone for receiving voice signals, a camera, and various other input means for detecting or receiving various types of user input. The output unit may be, for example, a display that outputs images, a speaker that outputs sounds, a haptic device that generates vibrations, and various other forms of output means.

[0044] According to one embodiment, the electronic device (100) may be a server that acquires and processes data for multiple vehicles. Specifically, the data for multiple vehicles may include multiple battery charge data for the multiple vehicles. The multiple battery charge data may be acquired, for example, from at least one of an On-Board Diagnostics (OBD) device mounted on each vehicle, a battery management system (BMS), and a device (e.g., a database) in which battery charge data is previously stored, via the transceiver (110) of the electronic device (100). The manner in which the electronic device (100) acquires the multiple battery charge data is not limited to the above example, and it will be clearly understood by those skilled in the art that the electronic device (100) may acquire the multiple battery charge data from various devices with which it can communicate via the transceiver (110). The type of the electronic device (100) is not limited thereto, and various embodiments of the present disclosure may be applied to various devices capable of acquiring and processing data for vehicles.

[0045] The processor (120) can control the overall operation of the electronic device (100) and process data and signals. The processor (120) can be composed of at least one hardware unit. In addition, the processor (120) can operate by one or more software modules generated by executing program codes stored in the memory (130). The processor (120) can include a memory, and the processor (120) can control the overall operation of the electronic device (100) and process data and signals by executing the program codes stored in the memory.

[0046] The processor (120) may be implemented as a computer or similar device based on hardware, software, or a combination thereof. In terms of hardware, the processor (120) may be implemented in the form of an electronic circuit that processes electrical signals to perform control functions, and in terms of software, the processor (120) may be implemented in the form of a program that drives the hardware processor (120). Meanwhile, unless otherwise specified in the following description, the operation of the electronic device may be interpreted as being performed under the control of the processor (120). That is, when modules implemented in the clustering system for data regarding battery charging patterns are executed, the modules may be interpreted as controlling the processor (120) to perform the following operations of the electronic device (100).

[0047] The memory (130) can store various types of information. The memory (130) can store data temporarily or semi-permanently. For example, the memory (130) of the electronic device (100) can store data related to an operating program (OS: Operating System) for operating the electronic device (100). Examples of the memory (130) may include a hard disk drive (HDD: Hard Disk Drive), a solid state drive (SSD), flash memory, read-only memory (ROM: Read-Only Memory), random access memory (RAM: Random Access Memory), etc. The memory (130) may be provided as a built-in type or a detachable type.

[0048] In summary, the various embodiments may be implemented through various means. For example, the various embodiments may be implemented through hardware, firmware, software, or a combination thereof.

[0049] In the case of hardware implementation, the methods according to various embodiments may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, etc.

[0050] When implemented via firmware or software, the methods according to various embodiments may be implemented in the form of modules, procedures, or functions that perform the functions or operations described above. For example, software code may be stored in memory and executed by a processor. The memory may be located within or external to the processor and may exchange data with the processor via various known means.

[0051] Figure 2 shows a flowchart of a battery degradation factor analysis method according to one embodiment.

[0052] In step S210, the electronic device (100) may acquire a plurality of battery charging data for a plurality of vehicles. Here, the plurality of battery charging data is data collected for each of a plurality of batteries included in the plurality of vehicles over a preset period of time, and may include at least one of information on a state of charge (SoC) of the battery over time, information on a charging current rate (C-rate) of the battery over time, information on a charging temperature of the battery over time, and information on a charging cycle of the battery over time. The preset period of time may be a period during which charging and discharging of the batteries of the vehicle may be repeated multiple times, and may be, for example, one month. Accordingly, the plurality of battery charging data may include a plurality of time-series data including various information related to battery charging, such as changes in SoC, changes in C-rate, changes in charging temperature, and changes in charging time cycles over time while charging and discharging of each of the plurality of batteries included in the plurality of vehicles is repeated multiple times.

[0053] In step S220, the electronic device (100) may identify a plurality of factors related to battery charging based on the plurality of battery charging data. Here, the plurality of factors may include at least one of a charge start SoC, a charge end SoC, a C-rate, a charge temperature, a charge cycle, and a charge time zone. For example, if the plurality of battery charging data includes time series data indicating a change in SoC over time, the electronic device (100) may identify the charge start SoC and the charge end SoC as factors included in the plurality of factors. In another example, if the plurality of battery charging data includes time series data indicating a change in a charge time cycle over time, the electronic device (100) may identify the charge time cycle and the charge time zone as factors included in the plurality of factors. Since the plurality of battery charging data may include various time series data related to battery charging, the electronic device (100) may identify one or more various factors corresponding to each battery charging data according to an embodiment, and is not necessarily limited to the examples of the factors described above.

[0054] In step S230, the electronic device (100) can generate a plurality of factor combinations based on a plurality of factors. The generation of the factor combinations can be performed randomly, for example. That is, the electronic device (100) can generate a plurality of factor combinations by randomly selecting two factors from among the plurality of factors. As a specific example, the electronic device (100) can generate factor combination 1 consisting of a charging start SoC and a charging end SoC, factor combination 2 consisting of a charging C-rate and a charging temperature, etc.

[0055] In step S240, the electronic device (100) may generate a plurality of factor combination features indicating a charging pattern related to the factor combination based on one or more corresponding battery charging data for each of the plurality of factor combinations. Specifically, the electronic device (100) may check one or more battery charging data corresponding to a first factor included in a first factor combination among the plurality of factor combinations and one or more battery charging data corresponding to a second factor included in the first factor combination, and may generate a plurality of first factor combination features indicating a charging pattern related to the first factor combination based on one or more battery charging data corresponding to the first factor and one or more battery charging data corresponding to the second factor.

[0056] That is, the electronic device (100) checks, for each factor combination, a plurality of battery charging data related to each of one or more vehicles corresponding to a factor combination for which a feature is to be generated among a plurality of battery charging data, and generates a plurality of factor combination features representing a charging pattern related to the corresponding factor combination based on the plurality of battery charging data, thereby generating a plurality of factor combination features for each factor combination. A more specific embodiment of generating a plurality of factor combination features corresponding to a specific factor combination will be described in detail with reference to FIG. 3 below.

[0057] Meanwhile, each of the generated multiple factor combination features may be image data representing a charging pattern related to the factor combination. Here, the charging pattern related to the factor combination may refer to a characteristic tendency that appears when multiple factors included in the factor combination interact with each other during a battery charging process, for example. As a specific example, the charging pattern related to factor combination 1 may be a pattern related to the frequency of charging sections where charging starts and ends, and as another example, the charging pattern related to factor combination 2 may be a pattern related to the frequency of combinations of C-rate and charging temperature. Each factor combination feature representing such a charging pattern may be, for example, two-dimensional image data, and the width and height of the image may each correspond to each of the factors constituting the factor combination.

[0058] In step S250, the electronic device (100) may perform clustering on a plurality of factor combination features that indicate charging patterns related to the same factor combination based on a clustering algorithm. That is, the electronic device (100) may perform clustering on each of the plurality of factor combination features for each factor combination, and thus, multiple clusters may be generated for each factor combination. Here, the clustering algorithm may be an algorithm that divides the target data into clusters having similar characteristics (e.g., image similarity, distance between data, etc.), and the electronic device (100) may utilize various types of clustering algorithms depending on the embodiment.

[0059] As a more specific example, the electronic device (100) can perform clustering targeting multiple factor combination features for each factor combination based on a hierarchical clustering algorithm that hierarchizes clustering target data among clustering algorithms that can set the number of clusters as a parameter (i.e., can be hyper-parameterized). The hierarchical clustering algorithm can have the advantage of having a very small amount of computational work since it repeatedly generates an initial similarity matrix based on the similarity of characteristics between clusters and then merges or divides clusters with high similarity into sub-clusters based on the matrix.

[0060] In step S260, the electronic device (100) may extract one or more factor combinations corresponding to a factor combination related to battery degradation from among the plurality of factor combinations based on the deviation in the data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations. In other words, the electronic device (100) may analyze the deviation in the data distribution between the plurality of clusters generated for each factor combination as a result of clustering performed on the plurality of factor combination features for each factor combination, and may extract one or more factor combinations determined to have an effect on battery degradation based on the result of the analysis. A more specific embodiment in which the electronic device (100) analyzes the clustering result for each factor combination to extract one or more factor combinations will be described in detail with reference to FIG. 4 below.

[0061] At step S270, the electronic device (100) can check information about the influence of each of one or more factor combinations on battery degradation based on the artificial intelligence-based model. The information about the influence of each of one or more factor combinations on battery degradation may be, for example, information about the amount of change in the state of health (SoH) of the battery according to each of one or more factor combinations (for convenience, hereinafter referred to as 'dSoH (Difference of SoH)'). Here, dSoH is an indicator indicating how much the SoH, which is the state of health of the battery, has changed over a specific time period, and can be calculated as, for example, the difference between the SoH at a point in time at which dSoH is to be measured and the amount of change in SoH at a previous specific point in time. Since dSoH can be interpreted as a rate of performance degradation of the battery, it can be utilized as an indicator that quantitatively measures the degree of battery degradation. A more specific embodiment in which the electronic device (100) checks information about dSoH for each of one or more factor combinations based on an artificial intelligence-based model will be described in detail with reference to FIGS. 7 to 9 below.

[0062] At step S280, the electronic device (100) may provide information on an optimal combination of factors for preventing battery degradation based on information on the impact on battery degradation. Here, the optimal combination of factors for preventing battery degradation may be, for example, a combination of factors that results in the smallest dSoH value.

[0063] Figure 3 illustrates a process for generating factor combination features according to one embodiment.

[0064] Referring to FIG. 3, the electronic device (100) may generate a factor combination feature for factor combination 1 consisting of a charging start SoC and a charging end SoC. Among a plurality of battery charging data for a plurality of vehicles, the battery charging data related to factor combination 1 may be in the form of a graph (300) with time as the x-axis and SoC as the y-axis. Specifically, the graph (300) may include SoC information of a specific vehicle battery for a preset period of time, and more specifically, may include information about a charging section confirmed each time the battery is charged. Accordingly, the graph (300) may indicate at what SoC level charging starts and at what SoC level charging ends each time the battery is charged. The electronic device (100) can analyze the frequency of a plurality of charging sections confirmed based on a time series graph (300) which is battery charging data, and estimate a probability distribution function regarding the frequency of the charging sections, and can generate a factor combination feature (310) regarding factor combination 1 based on the estimated probability distribution function.

[0065] At this time, the factor combination feature generated based on a single battery charging data, such as graph (300), may be a single factor combination feature, or, according to another embodiment, may be a plurality of factor combination features generated by dividing the battery charging data into certain time intervals and corresponding to each of the divided time intervals. In other words, the number of factor combination features generated corresponding to a single vehicle battery may be one or multiple.

[0066] The feature (310) may be a two-dimensional image with the charging start SoC as the x-axis and the charging end SoC as the y-axis, and the shading of the feature (310) may indicate the frequency of the charging section. That is, the darker the shading in the feature (310), the more frequently the battery is charged. Accordingly, the feature (310) may indicate a charging habit or tendency regarding the SoC at which charging mainly starts and ends in the vehicle, and as a more specific example, the feature (310) may indicate that the user of the first vehicle has a charging habit of mainly starting charging at a low SoC and ending charging at a low SoC.

[0067] Meanwhile, in an example of generating multiple factor combination features corresponding to factor combination 2, the electronic device (100) can check one or more battery charging data corresponding to the charging C-rate and one or more battery charging data corresponding to the charging temperature, and analyze the frequency of combinations of the charging C-rate and the charging temperature through a joint probability distribution estimation technique, which is a technique for estimating the probability that two or more random variables have specific values ​​simultaneously. Subsequently, a plurality of features representing the relationship between the analyzed charging C-rate and the charging temperature can be generated, and the generated features can be in the form of two-dimensional images with the charging C-rate and the charging temperature as the x-axis and the y-axis, respectively.

[0068] FIG. 4 illustrates a process of extracting one or more factor combinations based on a clustering algorithm according to one embodiment.

[0069] According to one embodiment, the electronic device (100) may obtain a plurality of battery charging data for a plurality of vehicles (step S410), identify a plurality of factors related to battery charging based on the plurality of battery charging data (step S420), generate a plurality of factor combinations based on the plurality of factors (step S430), and, for each of the plurality of factor combinations, generate a plurality of factor combination features indicating a charging pattern related to the factor combination based on one or more corresponding battery charging data (step S440).

[0070] Next, the electronic device (100) can perform clustering of a plurality of first factor combination features corresponding to a first factor combination among a plurality of factor combination features into a plurality of first clusters based on a clustering algorithm (step S450). As a result of performing clustering, the electronic device (100) can check whether the deviation between the distributions of the amount of variation in SoH (dSoH) corresponding to each of the plurality of first clusters satisfies a first condition (step S460), and if the deviation between the distributions of the dSoH corresponding to each of the plurality of first clusters satisfies the first condition, it can be confirmed that the first factor combination corresponds to a factor combination related to battery degradation (step S470).

[0071] Here, the distribution of dSoH corresponding to each of the plurality of first clusters may be, for example, a distribution formed based on the dSoH corresponding to each of one or more first factor combination features included in each of the plurality of first clusters. In other words, the distribution of dSoH corresponding to each of the plurality of first clusters may be a distribution of dSoH values ​​calculated corresponding to each of one or more factor combination features included in each cluster. A more specific example regarding the distribution of dSoH of each cluster identified as a clustering result related to a specific factor combination will be described in detail with reference to FIG. 5 below.

[0072] Meanwhile, according to one embodiment, the first condition that serves as the basis for determining whether a combination of factors related to battery degradation corresponds may include at least one of a condition regarding whether data distributions of each of the plurality of clusters overlap and a condition regarding a difference between data averages of each of the plurality of clusters. Specifically, the condition regarding whether data distributions of each of the plurality of clusters overlap may be a condition regarding whether a preset number or more of clusters whose data distributions of each cluster do not overlap each other are identified among the plurality of clusters, and the condition regarding the difference between data averages of each of the plurality of clusters may be a condition regarding whether a value that is greater than or equal to a preset value among the absolute values ​​of the differences between data averages of each of the plurality of clusters is identified, and the first condition may include at least one of these conditions. However, since the form of the data distribution of each of the plurality of clusters may vary greatly depending on the case, the first condition is not limited to the above examples, and depending on the embodiment, it may be set to a more appropriate condition that can be determined to correspond to a combination of factors related to battery degradation, such as a condition regarding whether a specific type of user input has been obtained.

[0073] Figure 5 shows the data distribution among multiple clusters related to factor combination 1 according to one embodiment.

[0074] Referring to FIG. 5, the data distribution among multiple clusters related to factor combination 1 may be in the form of a graph (500) with the x-axis representing the cluster index and the y-axis representing the dSoH. The cluster index may refer to an index arbitrarily labeled to distinguish multiple clusters generated by performing clustering on multiple factor combination features related to factor combination 1. Accordingly, as an example in which multiple factor combination features related to factor combination 1 are clustered into a total of 10 clusters, the graph (500) may represent the distribution of dSoH of each of the 11 clusters. More specifically, each distribution may be represented as a range from the minimum value to the maximum value of the dSoH values ​​of each factor combination feature within each cluster, and the illustrated point within each distribution may represent the average value of the dSoH values ​​of each factor combination feature within each cluster. Accordingly, cluster 0 may be a cluster composed of factor combination features having a mean dSoH value of about 0.35 and dSoH values ​​ranging from about 0.3 to about 0.42.

[0075] Meanwhile, in an example where the first condition is a condition regarding whether two or more clusters are confirmed among multiple clusters in which the data distributions of each cluster do not overlap with each other, the graph (500) includes clusters 0 and 9 in which the distributions of dSoH do not overlap with each other, so the electronic device (100) can confirm that factor combination 1 corresponds to a factor combination regarding battery degradation.

[0076] Figure 6 shows the data distribution among multiple clusters related to factor combination 2 according to one embodiment.

[0077] Referring to FIG. 6, the description regarding FIG. 5 may be equally applied, except that the graph (600) is a data distribution among multiple clusters related to factor combination 2. Furthermore, according to the same example related to the first condition in FIG. 5, the 11 clusters included in the graph (600) do not include any clusters whose distributions of dSoH do not overlap with each other, so the electronic device (100) can determine that factor combination 2 does not correspond to a factor combination related to battery degradation.

[0078] FIG. 7 illustrates a process for identifying information about the impact of each of one or more factor combinations on battery degradation based on an artificial intelligence-based model according to one embodiment.

[0079] According to one embodiment, the electronic device (100) may extract a representative factor combination feature corresponding to a first type or a second type for each of one or more factor combinations (step S701). Here, the first type may be a factor combination feature corresponding to the smallest dSoH value among the plurality of factor combination features corresponding to each factor combination, which may be a factor combination feature that causes the least change in the health status of the battery, and the second type may be a factor combination feature corresponding to the largest dSoH value, which may be a factor combination feature that causes the greatest change in the health status of the battery. Accordingly, for convenience, the representative factor combination feature corresponding to the first type is referred to as a 'best feature', and the representative factor combination feature corresponding to the second type is referred to as a 'worst feature'. That is, the electronic device (100) may extract either the best feature or the worst feature for each extracted factor combination as a representative factor combination feature.

[0080] The electronic device (100) can generate a plurality of input data sets including a preset SoH value and representative factor combination features extracted as the first type or the second type as described above, wherein each of the plurality of input data sets can be generated such that only one of the representative factor combination features included corresponds to the second type (step S702). Specifically, the electronic device (100) can generate an input data set with a preset SoH value of an arbitrary value and a best feature or a worst feature corresponding to each extracted factor combination in order to check the dSoH according to the input data set. Here, the preset SoH value can be a reference SoH value set to determine the dSoH due to the factor combination features constituting the input data set. Meanwhile, each input data set can be generated such that only one representative factor combination feature among the representative factor combination features constituting each input data set corresponds to the worst feature, and all of the remaining representative factor combination features correspond to the best feature. That is, each of the generated multiple input data sets can be configured such that only the representative factor combination feature corresponding to a single factor combination corresponds to the worst feature. A more specific embodiment of each input data set will be described in detail with reference to FIG. 9 below.

[0081] The electronic device (100) inputs the plurality of input data sets generated in this manner into an artificial intelligence-based model (step S703), and can check the dSoH corresponding to each of the plurality of input data sets output from the artificial intelligence-based model as information on the influence of each of one or more factor combinations on battery degradation (step S704). For example, the electronic device (100) can check the input data set corresponding to the smallest value among the dSoH corresponding to each of the plurality of input data sets, and can check the factor combination corresponding to the worst feature among the representative factor combination features included in the corresponding input data set as the optimal factor combination and provide information thereon.

[0082] Figure 8 illustrates a process for training an artificial intelligence-based model according to one embodiment.

[0083] According to one embodiment, the artificial intelligence-based model utilized by the electronic device (100) to identify information on the influence of each of one or more factor combinations extracted as factor combinations corresponding to factor combinations regarding battery degradation on battery degradation may be a model constructed by modeling a correlation between a learning input data set and a learning output data set, using each of a plurality of factor combination features corresponding to each of one or more factor combinations and a SoH value of a vehicle corresponding to each of the plurality of factor combination features as a learning input data set, and a dSoH value of a vehicle corresponding to each of the plurality of factor combination features as a learning output data set.

[0084] In this disclosure, deep learning refers to the process of training a neural network model using experience processing data sets, and through deep learning, software can improve its data processing capabilities. An AI-based model is a model created by modeling correlations between data sets, and these correlations can be expressed by multiple parameters. An AI-based model can derive correlations between data sets by extracting and analyzing features from a given data set, and deep learning refers to the process of optimizing the parameters of an AI-based model by repeating this process. Specifically, an AI-based model can learn correlations between inputs and outputs for a data set given as input-output pairs according to a deep learning algorithm. Alternatively, an AI-based model can learn relationships by deriving regularities between data even when only input data is given. The deep learning algorithm here may be any of a deep neural network, a recurrent neural network, a convolutional neural network, a machine learning model for classification-regression analysis, or a reinforcement learning model.

[0085] The electronic device (100) can build an artificial intelligence model by modeling the correlation between the aforementioned training input data set and training output data set according to the deep learning algorithm. That is, the electronic device (100) can build an artificial intelligence-based model by modeling the correlation between the training input data set including each factor combination feature and the current SoH value of the vehicle corresponding to the factor combination feature, and each training output data set (label) of the dSoH of the vehicle corresponding to each training input data set. Accordingly, the artificial intelligence-based model can predict and output a dSoH value corresponding to an input data set including an input factor combination feature or multiple factor combination features and an arbitrary SoH value. The electronic device (100) can input a data set composed of factor combination features corresponding to each of various factor combinations into the artificial intelligence-based model and obtain the dSoH value output from the artificial intelligence-based model, thereby confirming quantitative information regarding the impact on the deterioration of the vehicle battery more quickly and efficiently.

[0086] Referring to FIG. 8, in the process (800) of training an artificial intelligence-based model, the electronic device (100) may set the selected feature (801) and the current SoH value of the vehicle corresponding thereto as a training input data set, and the dSoH value of the vehicle according to the selected feature (801) as a training output data set. Here, the selected feature refers to a factor combination extracted by being confirmed to correspond to a factor combination related to battery degradation in the present invention. The electronic device (100) may train the artificial intelligence-based model (803) based on the training input data set and the training output data set set in this manner.

[0087] Figure 9 illustrates an input data set and an output data set of an artificial intelligence-based model according to one embodiment.

[0088] Referring to FIG. 9, the electronic device (100) can input a total of k input data sets (901, 902, etc.) into the trained artificial intelligence-based model. The input of the k input data sets into the artificial intelligence-based model (803) can be performed simultaneously or in parallel, and, depending on the embodiment, can also be performed sequentially. In addition, each input data set can include a representative factor combination feature of each of the k factor combinations and an arbitrary SoH value (omitted in FIG. 9). Specifically, the representative factor combination feature of each of the k factor combinations can be extracted as a best feature or a worst feature. For example, the input data set (901) can be an input data set configured by extracting a best feature for each of factor combinations 1 to k-1, and a worst feature for factor combination k. For example, the input data set (902) may be an input data set configured by extracting the best feature for each of factor combinations 2 to k, and the worst feature for factor combination 1. In this way, each input data set may be configured such that only one of the k representative factor combination features is the worst feature.

[0089] The trained artificial intelligence-based model (803) can output a corresponding dSoH value for each of the k input data sets. Specifically, the dSoH corresponding to the input data set 901, in which only the representative factor combination feature corresponding to factor combination k is the worst feature, can be dSoH 1 (910), and the dSoH corresponding to the input data set 902, in which only the representative factor combination feature corresponding to factor combination 1 is the worst feature, can be dSoH k (920). In this way, the trained artificial intelligence-based model (803) can output k dSoH values.

[0090] The electronic device (100) can confirm the k dSoH values ​​output in this way as information regarding the influence of each of the k factor combinations on battery degradation. That is, each of the k dSoH values ​​can quantitatively represent the influence of the factor combination corresponding to the worst feature among the representative factor combination features of the corresponding input data set on the SoH of the vehicle battery. Accordingly, the electronic device (100) can compare the sizes of the k dSoH values ​​to confirm one worst feature of the input data set corresponding to the dSoH with the smallest value, and confirm the factor combination corresponding to the worst feature as the optimal factor combination that has the least influence on the SoH of the battery. Meanwhile, since the worst feature of the input data set corresponding to the largest dSoH value has the greatest influence on the SoH of the battery, the electronic device (100) can determine that this is a combination of factors that has the greatest adverse influence on battery degradation, and by sorting the input data sets in order of dSoH values, it is also possible to determine information about the relative order of influence on battery degradation among the combinations of factors corresponding to the worst features of each input data set.

[0091] Figure 10 illustrates a process for identifying an optimal set of factor combinations based on an artificial intelligence-based model according to one embodiment.

[0092] According to one embodiment, the electronic device (100) can extract a representative factor combination feature corresponding to a first type (best feature) or a second type (worst feature) for each of one or more factor combinations (step S1001), and can generate a plurality of input data sets including a preset SoH value and the extracted representative factor combination feature, wherein the plurality can correspond to all cases of extracting the representative factor combination feature for each of one or more factor combinations (step S1002). That is, unlike the above-described embodiment, each of the plurality of input data sets can be configured to include one of the best feature or the worst feature corresponding to each factor combination, without any condition on the number of best or worst features. Accordingly, when there are n number of one or more factor combinations extracted as factors related to battery degradation, the plurality of input data sets can be 2 n A dog can be created.

[0093] The electronic device (100) inputs a plurality of input data sets generated in this manner into a trained model (step S1003), confirms an optimal factor combination set corresponding to each of the plurality of input data sets output from the trained model (step S1004), and provides the optimal factor combination set corresponding to each of the plurality of input data sets as information on the optimal factor combination (step S1005).

[0094] Here, the trained model may refer to a model that is further trained to output an optimal factor combination set in which at least one factor combination is changed in response to the input data set based on the SoH change corresponding to the input data set by the AI-based model constructed and trained as described above. In other words, the trained model may be further trained to output an optimal factor combination set in which the factor combination features of the input data set are changed so that the dSoH value corresponding to each input data set can be improved (relatively reduced). For this purpose, in one example, the model may be trained to output a configuration of factor combination features corresponding to the smallest dSoH value as the optimal factor combination set. Meanwhile, in another example, the model may output a factor combination set in which one factor combination feature among the configurations of the factor combination features of the input current state is changed, wherein the one factor combination feature that is changed may be a factor combination feature selected so that the corresponding dSoH value is minimized most due to the change of such feature. That is, the trained model can provide a set of factor combinations that can reduce the dSoH value as much as possible by changing the configuration of the minimum number of factor combination features as the optimal battery degradation prevention solution.

[0095] FIG. 11 illustrates an input data set and an output data set of a trained artificial intelligence-based model according to one embodiment.

[0096] Referring to FIG. 11, the electronic device further trains an artificial intelligence model (1104) to output a set of optimal factor combinations, totaling 2 kA set of input data (1101, 1102, 1103, etc.) can be input. The input of the input data sets can be performed simultaneously or in parallel as in FIG. 9, and may also be performed sequentially, depending on the embodiment. Each input data set can include a best feature or a worst feature corresponding to each of k factor combinations and a preset SoH value for dSoH judgment, and the trained model (1104) corresponding to this can output a set of factor combinations corresponding to each input data set. For example, the trained model (1104) can output a set of factor combinations (1110) that maintains the current factor combination features corresponding to the input data set 1101. That is, since the dSoH value according to the input data set 1101 corresponds to the minimum value, it can be confirmed that the configuration of the factor combination set included in the input data set is the optimal factor combination set as it is, and accordingly, the electronic device (100) can provide this as information on the optimal factor combination. Meanwhile, the trained model (1104) can output a factor combination set (1120) in which the feature of factor combination 2 is changed to the best feature in response to the input data set 1102, and this output can be an output in which the dSoH value corresponding to 1102 is minimized the most when the feature of factor combination 2 is changed from the worst feature to the best feature among the representative factor combination features of each of the k factor combinations. Accordingly, the electronic device (100) can provide the factor combination set in which the feature of factor combination 2 is changed to the best feature as information on the optimal factor combination.

[0097] The above-described embodiments can be implemented as artificial intelligence (AI) through a processor and memory of an electronic device. The processor may be composed of one or more processors, and one or more processors may be a general-purpose processor such as a CPU, an AP, a digital signal processor (DSP), a graphics-only processor such as a GPU, a vision processing unit (VPU), or an AI-only processor such as an NPU. One or more processors may be controlled to process input data according to predefined operating rules or AI models stored in memory. Alternatively, if one or more processors are AI-only processors, the AI-only processor may be designed with a hardware structure specialized for processing a specific AI model.

[0098] The predefined operation rules or artificial intelligence models are characterized by being created through learning. Here, being created through learning means that the basic artificial intelligence model is learned by a learning algorithm using a plurality of learning data, thereby creating a predefined operation rules or artificial intelligence model set to perform a desired characteristic (or purpose). This learning may be performed in the electronic device itself on which the artificial intelligence according to the present disclosure is performed, or may be performed through a separate server and / or system. Examples of the learning algorithm include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0099] An artificial intelligence model may be composed of multiple neural network layers. Each of the multiple neural network layers has multiple weight values, and can perform neural network operations through operations between the operation results of the previous layer and the multiple weights. The multiple weights of the multiple neural network layers may be optimized based on the learning results of the artificial intelligence model. For example, the multiple weights may be updated so that the loss value or cost value obtained from the artificial intelligence model is reduced or minimized during the learning process. The artificial neural network may include, but is not limited to, a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), or a deep Q-network.

[0100] The electronic device according to the above-described embodiments may include a processor, a memory for storing and executing program data, permanent storage such as a disk drive, a communication port for communicating with an external device, a user interface device such as a touch panel, a key, a button, etc. The methods implemented as software modules or algorithms may be stored on a computer-readable recording medium as computer-readable codes or program instructions executable on the processor. Here, the computer-readable recording medium includes a magnetic storage medium (e.g., read-only memory (ROM), random-access memory (RAM), floppy disk, hard disk, etc.) and an optical reading medium (e.g., CD-ROM, DVD: Digital Versatile Disc)). The computer-readable recording medium may be distributed to computer systems connected to a network, so that the computer-readable code may be stored and executed in a distributed manner. The medium may be readable by a computer, stored in a memory, and executed by a processor.

[0101] The present embodiment may be represented by functional block configurations and various processing steps. These functional blocks may be implemented by various hardware and / or software configurations that perform specific functions. For example, the embodiment may employ direct circuit configurations such as memory, processing, logic, look-up tables, etc., which may perform various functions under the control of one or more microprocessors or other control devices. Similarly, the present embodiment may be implemented in a programming or scripting language such as C, C++, Java, assembler, etc., including various algorithms implemented as a combination of data structures, processes, routines, or other programming configurations. Functional aspects may be implemented as algorithms that execute on one or more processors. Furthermore, the present embodiment may employ conventional techniques for electronic configuration, signal processing, and / or data processing. Terms like "mechanism," "element," "means," and "composition" can be used broadly and are not limited to mechanical or physical components. These terms can also encompass a series of software routines, such as those associated with a processor.

[0102] The above-described embodiments are merely examples, and other embodiments may be implemented within the scope of the claims set forth below.

Claims

1. In electronic devices, transceiver; processor; and Contains one or more memories that store one or more instructions, The one or more instructions, when executed, cause the processor to: Obtain multiple battery charging data for multiple vehicles, Based on the above multiple battery charging data, multiple factors related to battery charging are identified, Based on the above multiple factors, multiple factor combinations are generated, For each of the plurality of factor combinations, generate a plurality of factor combination features representing a charging pattern related to the factor combination based on one or more corresponding battery charging data, Based on a clustering algorithm, clustering is performed targeting multiple factor combination features that exhibit charging patterns related to the same factor combination, Based on the deviation of the data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations, one or more factor combinations corresponding to the factor combination related to battery degradation are extracted from the plurality of factor combinations, Based on an artificial intelligence-based model, information on the impact of each combination of one or more factors on battery degradation is identified, An electronic device configured to provide information on an optimal combination of factors for preventing battery degradation based on information on the impact on battery degradation.

2. In paragraph 1, The above multiple battery charging data is, For each of the plurality of batteries included in the above plurality of vehicles, data collected over a preset period of time, An electronic device comprising at least one of information about a state of charge (SoC) of a battery over time, information about a charging current rate (C-rate) of the battery over time, information about a charging temperature of the battery over time, and information about a charging cycle of the battery over time.

3. In paragraph 1, The above multiple arguments are, An electronic device comprising at least one of a charge start SoC, a charge end SoC, a C-rate, a charge temperature, a charge cycle, and a charge time zone.

4. In paragraph 1, The one or more instructions, when executed, cause the processor to: Check at least one battery charging data corresponding to a first factor included in a first factor combination among the plurality of factor combinations and at least one battery charging data corresponding to a second factor included in the first factor combination, An electronic device configured to generate a plurality of first factor combination features representing a charging pattern related to the first factor combination based on one or more battery charging data corresponding to the first factor and one or more battery charging data corresponding to the second factor.

5. In paragraph 1, The one or more instructions, when executed, cause the processor to: Based on the clustering algorithm, clustering is performed on a plurality of first factor combination features corresponding to a first factor combination among the plurality of factor combination features into a plurality of first clusters, It is checked whether the deviation between the distributions of the state of health (SoH) change amount of the battery corresponding to each of the plurality of first clusters satisfies the first condition, An electronic device configured to determine that the first factor combination corresponds to the factor combination related to battery degradation when the deviation between the distributions of the SoH variation corresponding to each of the plurality of first clusters satisfies the first condition.

6. In paragraph 5, The distribution of SoH change corresponding to each of the above plurality of first clusters is, An electronic device, wherein the distribution is formed based on the SoH variation corresponding to each of one or more first factor combination features included in each of the plurality of first clusters.

7. In paragraph 5, The first condition above is, An electronic device comprising at least one of a condition regarding whether there is overlap between data distributions of each of a plurality of clusters and a condition regarding a difference between data mean values ​​of each of a plurality of clusters.

8. In paragraph 1, The above artificial intelligence-based model is, An electronic device, wherein a model is constructed by modeling the correlation between the learning input data set and the learning output data set, using each of a plurality of factor combination features corresponding to each of the one or more factor combinations and the SoH value of the vehicle corresponding to each of the plurality of factor combination features as a learning input data set, and using the SoH change amount of the vehicle corresponding to each of the plurality of factor combination features as a learning output data set.

9. In paragraph 8, The one or more instructions, when executed, cause the processor to: For each of the above one or more factor combinations, a representative factor combination feature corresponding to the first type or the second type is extracted, Generate a plurality of input data sets including a preset SoH value and the extracted representative factor combination feature, wherein each of the plurality of input data sets is generated such that only one of the representative factor combination features it includes corresponds to the second type, Inputting the above multiple input data sets into the artificial intelligence-based model, An electronic device configured to check the SoH change amount corresponding to each of the plurality of input data sets output from the artificial intelligence-based model as information regarding the influence of each of the one or more factor combinations on battery degradation.

10. In paragraph 9, The one or more instructions, when executed, cause the processor to: Identify the input data set corresponding to the smallest value among the SoH changes corresponding to each of the above multiple input data sets, An electronic device configured to identify a factor combination corresponding to the second type among the representative factor combination features included in the above-mentioned confirmed input data set as the optimal factor combination.

11. In paragraph 8, The above artificial intelligence-based model is an electronic device that is further trained to output an optimal factor combination set in which at least one factor combination is changed in response to the input data set based on the SoH change amount corresponding to the input data set.

12. In paragraph 11, The one or more instructions, when executed, cause the processor to: For each of the above one or more factor combinations, a representative factor combination feature corresponding to the first type or the second type is extracted, Generate a plurality of input data sets including preset SoH values ​​and the extracted representative factor combination features, wherein the plurality of sets correspond to all cases of extracting representative factor combination features for each of the one or more factor combinations, Inputting the above multiple input data sets into the trained model, Identifying an optimal set of factor combinations corresponding to each of the plurality of input data sets output from the trained model, An electronic device configured to provide a set of optimal factor combinations corresponding to each of the plurality of input data sets as information regarding the optimal factor combinations.

13. In a method for analyzing battery degradation factors performed by an electronic device, A step of acquiring multiple battery charging data for multiple vehicles; A step of checking a plurality of factors related to battery charging based on the plurality of battery charging data; A step of generating a plurality of factor combinations based on the plurality of factors; For each of the plurality of factor combinations, a step of generating a plurality of factor combination features representing a charging pattern related to the factor combination based on one or more corresponding battery charging data; A step of performing clustering targeting multiple factor combination features that exhibit a charging pattern related to the same factor combination based on a clustering algorithm; A step of extracting one or more factor combinations corresponding to a factor combination related to battery degradation among the plurality of factor combinations based on a deviation in data distribution between the plurality of clusters generated corresponding to each of the plurality of factor combinations; A step of identifying information on the impact of each of the one or more factor combinations on battery degradation based on an artificial intelligence-based model; and A method for analyzing battery degradation factors, comprising a step of providing information on an optimal combination of factors for preventing battery degradation based on information on the influence on the above battery degradation.

14. A non-transitory computer-readable recording medium recording a program for executing the method of Article 13 on a computer.

Images