Server and operating method thereof

Abstract

According to an embodiment disclosed herein, a server includes: communication circuitry; a memory; and a processor operatively connected to the communication circuitry and the memory, where the processor may: obtain, for each user, time series data of a plurality of factors associated with the battery; generating feature data representing a relationship between at least two factors expected to have a cross-correlation among the plurality of factors based on each piece of time series data; and classifying the plurality of pieces of feature data into a plurality of clusters, and defining a plurality of modes based on features of each cluster.

Description

Technical Field

[0001] Cross-references to related applications

[0002] This application claims priority and benefit to Korean Patent Application No. 10-2023-0114896, filed with the Korean Intellectual Property Office on August 30, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The implementation methods disclosed herein relate to servers and their operation methods. Background Technology

[0004] Recently, research and development of rechargeable batteries have been actively pursued. Here, rechargeable batteries are batteries capable of charging and discharging, and include all recent lithium-ion batteries, such as conventional Ni / Cd and Ni / MH batteries. Among rechargeable batteries, lithium-ion batteries have the advantage of significantly higher energy density than conventional Ni / Cd and Ni / MH batteries. Furthermore, because lithium-ion batteries can be manufactured in small and lightweight forms, they are used as power sources for mobile devices. Recently, with the expanding use of electric vehicles as power sources, lithium-ion batteries have attracted attention as a next-generation energy storage medium.

[0005] When such batteries are installed in vehicles, battery data is typically collected and analyzed because battery performance, lifespan, and safety can vary depending on the vehicle driver's preferences. However, the problem lies in the large amount of data collected for each vehicle, and the difficulty in standardizing the collected data due to the different battery usage patterns of vehicle users. Summary of the Invention

[0006] Technical issues

[0007] The embodiments disclosed herein relate to providing a server and its operating methods that can standardize large amounts of data and take into account composite factors to define user types.

[0008] The embodiments disclosed herein also relate to providing a server and a method of operating the server, which can extract key factors related to the evaluation index of a battery.

[0009] The technical objectives of the embodiments disclosed herein are not limited to the above-described technical objectives, and other objectives not described will be clearly understood by those skilled in the art from the following description.

[0010] Technical solution

[0011] According to the embodiments disclosed herein, a server may include: a communication circuit; a memory; and a processor operatively coupled to the communication circuit and the memory, wherein the processor may be configured to: acquire multiple time-series data of multiple factors related to a battery for each user; generate feature data representing the relationship between at least two factors among the multiple factors that are expected to be cross-correlated based on each time-series data; and classify the multiple feature data into multiple clusters and define multiple patterns based on the characteristics of each cluster.

[0012] According to an implementation, the processor can generate the feature data based on the joint probability density function (joint PDF) of the at least two factors.

[0013] According to an implementation, the processor may be configured to: compute the joint PDF of the at least two factors based on the time series data; compute the joint PDF by applying a kernel density function (KDE) to the joint PDF; and generate the feature data representing the joint probability density of the at least two factors.

[0014] According to an implementation, the feature data may include two-dimensional (2D) contour image data representing the joint probability density of the at least two factors.

[0015] According to an implementation, the processor can classify the multiple feature data into multiple clusters based on the relationship between the at least two factors derived from the feature data and the distribution of each of the at least two factors.

[0016] According to an implementation, the processor may be configured to: fuse the plurality of feature data included in each cluster for each cluster to generate representative feature data; and define the plurality of patterns based on the relationship between the at least two factors derived from the representative feature data and the distribution of each of the at least two factors.

[0017] According to an implementation, the processor can determine multiple types for classifying users based on the defined multiple patterns.

[0018] According to an implementation, the processor can analyze the relationship between the plurality of types and the evaluation indicators of the battery, and extract the main factors related to the evaluation indicators from the plurality of factors.

[0019] According to the implementation method, the plurality of factors may include at least some of the following: State of Charge (SOC), Depth of Discharge (DOD), Charge Rate, Charging Start Temperature, Maximum Charging Temperature, Charging Start Time, Charging Start Date, Driving Start Time, Driving Distance, Driving Start Temperature, Maximum Driving Temperature, Driving DOD, Driving Speed ​​Distribution, Discharge Time, and Discharge End Date.

[0020] According to one implementation, the processor can acquire the time-series data from an on-board device (OBD) installed in each user's vehicle.

[0021] According to the embodiments disclosed herein, a method for operating a server may include: acquiring multiple time-series data of multiple factors related to a battery for each user; generating feature data based on each time-series data representing the relationship between at least two factors among the multiple factors that are expected to be mutually correlated; and classifying the multiple feature data into multiple clusters, and defining multiple patterns based on the characteristics of each cluster.

[0022] According to an implementation, generating the feature data may include: calculating the joint probability mass function (joint PMF) of the at least two factors based on the time series data; calculating the joint probability density function (PDF) of the at least two factors by applying a kernel density function (KDE) to the joint PMF; and generating the feature data representing the joint probability density of the at least two factors.

[0023] According to an implementation, defining the plurality of patterns may include: classifying the plurality of feature data into the plurality of clusters based on the relationship between the at least two factors derived from the feature data and the distribution of each of the at least two factors; fusing the plurality of feature data included in each cluster for each cluster to generate representative feature data; and defining the plurality of patterns based on the relationship between the at least two factors derived from the representative feature data and the distribution of each of the at least two factors.

[0024] According to an implementation, the method may further include determining multiple types for classifying users based on the defined multiple patterns.

[0025] Beneficial effects

[0026] According to the server and its operation method disclosed herein, unstructured data can be standardized and its patterns and types defined, taking into account composite factors.

[0027] Furthermore, based on the server and its operation method disclosed herein, key factors related to the battery's evaluation index can be extracted.

[0028] In addition, it can provide various effects that can be identified directly or indirectly through this article. Attached Figure Description

[0029] Figure 1 This is a view illustrating a data analysis system according to one embodiment disclosed herein.

[0030] Figure 2 This is a view showing an embodiment generated from feature data according to one embodiment disclosed herein.

[0031] Figure 3 This is a view illustrating multiple modal implementations of one embodiment disclosed herein.

[0032] Figure 4 This is a block diagram illustrating the configuration of a server according to one embodiment disclosed herein.

[0033] Figure 5 This is a flowchart describing a method of operating a server according to one embodiment disclosed herein.

[0034] Figure 6 This is a flowchart describing the process of generating feature data according to one embodiment disclosed herein.

[0035] Figure 7 It is a flowchart used to describe the process of defining multiple modes according to one implementation disclosed herein. Detailed Implementation

[0036] Various embodiments of the present disclosure will be described below with reference to the accompanying drawings. However, it should be understood that this is not intended to limit the present disclosure to a particular embodiment, and includes various variations, equivalents and / or alternatives to the embodiments of the present disclosure.

[0037] Unless the relevant context explicitly specifies otherwise, the singular form of the noun corresponding to an item in this document may include one or more items. In this document, 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 one of the items listed together in the corresponding phrases of these phrases or all possible combinations thereof. Terms such as “first,” “second,” “first,” or “second” may be used simply to distinguish a corresponding component from another component and do not limit the corresponding component in other respects (e.g., importance or order). When a particular (e.g., first) component is described as being “connected,” “linked,” or “joined,” “coupled,” or “connected” to another component, whether or not the terms “functionally” or “communically” are used, this means that the particular component can be connected to the other component directly (e.g., wired), wirelessly, or via a third component.

[0038] Each component described herein (e.g., a module or program) may include a single object or multiple objects. According to various embodiments, one or more of the corresponding components described above 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 perform one or more functions of each of the multiple components in the same or similar manner as the functions performed by the corresponding components of the multiple components prior to integration. According to various embodiments, the operations performed by modules, programs, or other components may be performed sequentially, in parallel, repeatedly, or heuristically, or may be performed in a different order, or one or more operations may be omitted, or one or more other operations may be added.

[0039] As used herein, the terms "module" or "part" can include units implemented in hardware, software, or firmware, and are used interchangeably with terms such as logic, logic block, component, or circuit. A module can be an integrated part or the smallest unit of parts that performs one or more functions, or a portion thereof. For example, according to one implementation, a module can be implemented as an application-specific integrated circuit (ASIC).

[0040] The various embodiments described herein can be implemented in the form of software (e.g., a program or application) comprising one or more commands stored in a machine-readable storage medium (e.g., memory). For example, a device's processor can retrieve at least one of the stored commands from the storage medium and execute those commands. This enables the device to operate according to the retrieved at least one command to perform at least one function. The one or more commands may include code generated by a compiler or code executable by a parser. The device-readable storage medium may be provided in the form of a non-temporary storage medium. Here, "non-temporary storage medium" is a tangible device that refers only to a storage medium that does not contain signals (such as electromagnetic waves), and this term does not distinguish between semi-permanent and temporary storage of data in the storage medium.

[0041] Figure 1 This is a view illustrating a data analysis system according to one embodiment disclosed herein.

[0042] refer to Figure 1 The data analysis system may include vehicle 10, server 20 and electronic device 30.

[0043] The data analytics system can standardize the large amounts of unstructured time-series data collected from each user's vehicles and define patterns that can represent the characteristics of the data from the standardized data. Furthermore, the data processing system can classify users by combining multiple patterns defined based on the standardized data to determine representative types.

[0044] Vehicle 10 may be an electric vehicle equipped with a battery or a hybrid electric vehicle. The battery may be a battery pack or a battery module, and may include at least one battery cell and a battery management system (BMS) for monitoring and managing the battery status.

[0045] According to one embodiment, vehicle 10 may be equipped with an on-board device (OBD) 11 for collecting battery data, and server 20 may obtain battery data from OBD 11. Battery data may broadly include data about battery state (e.g., state of charge (SOC), state of health (SOH), current or voltage), data about battery charging (e.g., charging time, temperature during charging, charging end time, or charging current distribution), data about the vehicle equipped with the battery (e.g., driving time, driving distance, speed, or acceleration), and is not limited to the embodiments described above. OBD 11 may receive battery data from BMS or directly obtain battery data.

[0046] Server 20 can acquire multiple time-series data points regarding various factors related to the battery and define multiple patterns. For example, server 20 can receive multiple time-series data points regarding multiple factors from the OBD 11 of vehicle 10. As an example, server 20 can acquire multiple time-series data points regarding the battery's initial state of charge (SOC) and depth of charge / discharge (DOD).

[0047] Server 20 can manage multiple time-series data entries obtained by users or vehicles 10. As an example, server 20 can store and manage multiple time-series data entries based on a unique ID assigned to each vehicle.

[0048] According to the implementation, the multiple factors may include at least some of the following: initial charge state of charge (SOC), charge time of discharge (DOD), charge rate (C-rate), initial charge temperature, maximum charge temperature, initial charge time, initial charge date, start of driving time, driving distance, start of driving temperature, maximum driving temperature, driving DOD, driving speed distribution, discharge time, and end of discharge date. The multiple factors are not limited to these and may include factors that may affect the battery state (such as the duration of the end of driving or the ambient air temperature), without any limitations. For example, the multiple factors may include factors that can separate the charging and discharging intervals from multiple time-series data and represent each interval.

[0049] Server 20 can generate feature data from multiple time series data for at least two factors. For example, at least two factors can be selected as those expected to be cross-correlated among multiple factors. As an example, at least two factors can be selected as the charging start-up state of charge (SOC) and charging destination of charge (DOD), which are expected to be cross-correlated. In one implementation, server 20 can select two factors from multiple factors to generate feature data in order to clearly distinguish the differences between the visualizations and patterns of the feature data.

[0050] According to the implementation, server 20 can generate feature data based on the joint probability density function (joint PDF) of at least two factors. For example, server 20 can generate feature data by representing the joint PDF as a curve. When server 20 generates feature data based on only one of the multiple factors, each factor is not independent, so the pattern definition result may be inaccurate and distorted. Therefore, server 20 can generate feature data based on the joint PDF, allowing for the complex consideration of at least two non-independent (i.e., expected to be cross-correlated) factors. Reference will be made below. Figure 2 This describes an implementation of a specific method for generating feature data by server 20.

[0051] According to an implementation, the feature data may include two-dimensional (2D) contour image data representing the joint probability density of at least two factors. The 2D contour image data may be image data in which the joint probability density is represented as contour data.

[0052] Server 20 can classify multiple feature data into multiple clusters. For example, server 20 can analyze the characteristics represented by multiple feature data and classify multiple feature data with similar characteristics into the same cluster. The characteristics of the multiple feature data can include, for example, the trend and distribution of the data values ​​included in the feature data.

[0053] According to the implementation, server 20 can classify multiple feature data into multiple clusters based on the relationship between at least two factors derived from the feature data and the distribution of each of the at least two factors. That is, the characteristics of multiple feature data can be analyzed using the relationship between at least two factors and the distribution of each of the two factors.

[0054] Server 20 can define multiple patterns based on the characteristics of clustering in classification. Multiple patterns can be patterns that represent the features of multiple data points with at least two factors.

[0055] According to the implementation method, server 20 can fuse multiple feature data included in each cluster to generate representative feature data for each cluster. Here, the fusion of multiple feature data can refer to the sum of the data values ​​of multiple feature data. That is, the multiple feature data can be data with the same dimensionality, and server 20 can add the data values ​​of multiple feature data. For example, server 20 can represent the outline of multiple data based on the summed data, thereby generating representative feature data.

[0056] According to the implementation, server 20 can define multiple patterns based on the relationship between at least two factors derived from representative feature data and the distribution of each of the at least two factors. Server 20 can generate representative feature data based on multiple feature data included in the cluster, and define multiple patterns accordingly to more clearly distinguish the trends and representative characteristics of each pattern.

[0057] In this way, server 20 can define multiple patterns for the selected factors. Multiple patterns can also be defined differently when the pairs of factors selected from the multiple factors are different. That is, server 20 can define different patterns based on the selected pairs of factors. For example, a pattern defined based on the charging start SoC-charging DOD can be different from a pattern defined based on the charging DOD-charging C-rate.

[0058] According to the implementation, server 20 can determine multiple types for classifying users based on multiple defined patterns. Server 20 can determine multiple types by combining patterns defined according to different pairwise factors. For example, server 20 can determine multiple types by combining a first set of patterns defined according to charging start SoC-charging DOD and a second set of patterns defined according to charging start time-charging start day.

[0059] Multiple types used to categorize users can be represented, for example, as combinations of characters that can identify a user's preferences. For instance, server 20 can use combinations of multiple characters to express user types to indicate a user's vehicle usage preferences, such as the Myers-Briggs Type Indicator (MBTI) representing an individual's personality type. Needless to say, server 20 can express user types in various ways, such as ranks, graphics, symbols, and shapes expressed as combinations of numbers and letters.

[0060] As an example, server 20 can classify users into "ready" or "balanced" based on a first set of patterns defined by charging start SOC-charging DOD. In another embodiment, server 20 can classify users into "regular" or "immediate" based on a second set of patterns defined by charging start time-charging start day. Server 20 can combine the above to classify users into four types. The above-described user types and number are merely implementation methods and are not limited thereto.

[0061] According to the implementation method, server 20 can analyze the relationship with battery evaluation metrics based on multiple types. For example, server 20 can analyze the behavioral correlation between battery evaluation metrics and specific types. Battery evaluation metrics may include battery life (e.g., SOH or Remaining Life (RUL)), frequency of abnormal behavior, average energy efficiency, etc.

[0062] According to the implementation method, server 20 can extract the main factors related to the evaluation index from multiple factors. Server 20 can analyze the evaluation index according to the user type and identify the types that show a high correlation with the evaluation index. For example, when a specific evaluation index is highly correlated with a type defined according to the pattern from charging start SOC to charging DOD, the main factors related to the corresponding evaluation index can be estimated as charging start SOC and charging DOD.

[0063] According to the implementation method, server 20 can classify users according to multiple determined types. For example, server 20 can provide the user with the results of user type classification using electronic device 30.

[0064] Electronic device 30 can be a user terminal carried by a user. Electronic device 30 can include various types of devices capable of data communication. For example, electronic device 30 can include portable devices such as smartphones or tablets, computer devices such as desktop or laptop computers, multimedia devices, cameras, wearable devices, virtual reality (VR) devices, etc., and is not limited to the above-mentioned devices. For example, electronic device 30 can include a server or gateway that can transmit data packets via an application. A user can use electronic device 30 to confirm their classification type.

[0065] Figure 2 This is a view showing an embodiment generated from feature data according to one embodiment disclosed herein.

[0066] refer to Figure 2 Server 20 can generate feature data based on multiple time series data.

[0067] Server 20 can acquire multiple time-series data points for multiple factors. For example, server 20 can acquire multiple time-series data points from the OBD 11 set in vehicle 10. As an example, the multiple time-series data points for multiple factors can be multiple data points acquired during multiple charge / discharge cycles. Figure 2 The multiple time series data 210 shown exemplarily illustrates multiple time series data of charge start SOC and charge DOD among multiple factors.

[0068] According to the implementation, server 20 can use multiple time series data of at least two factors to calculate the joint PMF 220 of at least two factors. For example, when server 20 generates feature data of two factors, the joint PMF 220 can have an x-axis representing the first factor and a y-axis representing the second factor, and the data of each coordinate can represent the data value of the first factor, the data value of the second factor, and the joint probability mass.

[0069] According to the implementation method, server 20 can apply the kernel density function (KDE) to the generated joint PDF 220 and calculate joint PDF 230. Since the statistics of joint PDF 220 may be distorted by discrete values, and the analysis may change according to data density, server 20 can employ KDE and generate joint PDF 230 for continuous domain.

[0070] According to an implementation, server 20 can generate feature data 240 representing the generated joint probability density. In this implementation, the feature data may include 2D contour image data representing the joint probability density of at least two factors. For example, when server 20 generates feature data 240 for two factors, the x-axis of feature data 240 may represent the first factor, its y-axis may represent the second factor, and the data at each coordinate may represent the data value of the first factor, the data value of the second factor, and the joint probability density. As an example, in... Figure 2 In the feature data 240, the joint probability density at specific coordinates can be represented by color. That is, the darker the color of the coordinates in the feature data 240, the higher the probability density is likely to be.

[0071] According to the implementation method, server 20 can analyze feature data. For example, by analyzing... Figure 2 The feature data 240 shown can be used to identify trends in low initial charging SOC, widely distributed charging DOD, and low buffer ratio for corresponding users based on the relationship between charging start SOC and charging DOD. Server 20 can analyze the characteristics of the feature data and classify multiple feature data into multiple clusters.

[0072] Figure 3 This is a view illustrating an embodiment of several modes according to one implementation disclosed herein.

[0073] refer to Figure 3 Server 20 can define multiple patterns based on multiple representative feature data of clustering.

[0074] Server 20 can merge multiple feature data belonging to each cluster and generate representative feature data. Figure 3 In this context, each curve within each pattern region represents representative feature data.

[0075] Server 20 can classify multiple representative feature data based on the relationship between at least two factors and the distribution of each of the at least two factors, and define multiple patterns. For example, such as Figure 3 As shown, server 20 can define multiple patterns P1 to P13 based on the distribution of each factor (charging start area, distribution expansion degree) and the relationship between factors derived from representative feature data (whether buffering is present). For example, pattern P13 can indicate that the user's charging start SOC is high, the distribution expansion degree of charging DOD is low, and there is a buffering trend.

[0076] Figure 4 This is a block diagram illustrating the configuration of a server according to one embodiment disclosed herein.

[0077] refer to Figure 4According to one implementation, server 400 (e.g., Figure 1 The server 20 may include a first communication circuit 410, a first processor 420, and a first memory 430.

[0078] The first communication circuit 410 can support establishing wired or wireless communication connections between the server 400 and external electronic devices (e.g., vehicles or electronic devices) and communicating via the established connections. According to one embodiment, the first communication circuit 410 may include a wireless communication circuit (e.g., a cellular communication circuit, a short-range wireless communication circuit, or a Global Navigation Satellite System (GNSS) communication circuit) or a wired communication circuit (e.g., a local area network (LAN) communication circuit or a power line communication circuit), and communicate with external electronic devices via short-range communication networks (such as Bluetooth, WiFi Direct, or Infrared Data Association (IrDA)) or long-range communication networks (such as cellular networks, the Internet, or computer networks) using the corresponding communication circuits described above. Various types of first communication circuits 410 can be implemented as a single chip, or each can be implemented as a separate chip. According to one embodiment, the first communication circuit 410 of the server 400 can be connected to... Figure 1 The electronic device 30 communicates with the vehicle 10.

[0079] The first processor 420 can control the overall operation of the server 400. In various embodiments, the first processor 420 may include one processor core (single core) or multiple processor cores. For example, the processor 420 may include multi-core processors, such as dual-core, quad-core, or hexa-core processors. According to embodiments, the processor 420 may also include internal or external cache memory. According to embodiments, the first processor 420 may be configured as one or more processors. For example, the first processor 420 may include at least one of an application processor, a communication processor, or a graphics processing unit (GPU).

[0080] All or some of the first processors 420 may be electrically or operationally connected to or linked to other components of the server 400 (e.g., the first communication circuit 410 or the first memory 430). The first processor 420 may receive commands from other components, parse the received commands, and perform calculations or data processing based on the parsed commands. The first processor 420 may parse and process messages, data, commands, or signals received from the first communication circuit 410 and the first memory 430. The first processor 420 may generate new messages, data, commands, or signals based on the received messages, data, commands, or signals. The first processor 420 may provide the processed or generated messages, data, commands, or signals to the first communication circuit 410 or the first memory 430.

[0081] The first processor 420 can process data or signals generated or produced by a program. For example, the first processor 420 can request commands, data, or signals from the first memory 430 to execute or control a program. The first processor 420 can record (or store) or update commands, data, or signals in the first memory 430 to execute or control a program. According to one embodiment, the first processor 420 can define multiple patterns based on multiple time-series data stored in the first memory 430.

[0082] The first memory 430 may store commands, control command codes, control data, or user data used to control the server. For example, the first memory 430 may include at least one of an application program, an operating system (OS), middleware, and a device driver. The first memory 430 may include one or more of volatile and non-volatile memory. Volatile memory may include dynamic random access memory (DRAM), static RAM (SRAM), synchronous DRAM (SDRAM), phase-change RAM (PRAM), magnetic RAM (MRAM), resistive RAM (RRAM), ferroelectric RAM (FeRAM), etc. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, etc. Memory 430 may also include non-volatile media such as hard disk drives (HDDs), solid-state drives (SSDs), embedded multimedia cards (eMMC), or universal flash memory (UFS). According to one embodiment, the first memory 430 may store multiple time-series data acquired from a user's vehicle.

[0083] Figure 5 This is a flowchart describing a method of operating a server according to one embodiment disclosed herein. Figure 5 The embodiment shown is only one embodiment, and the order of operation according to the various embodiments of this disclosure may differ. Figure 5 The order shown can be omitted. Figure 5 Some of the operations shown can change the order of operations or can be combined.

[0084] refer to Figure 5 The server's operation method may include: acquiring multiple time series data of multiple factors related to the battery for each user (S100); generating feature data representing the relationship between at least two factors that are expected to be mutually correlated among the multiple factors based on each time series data (S200); classifying the multiple feature data into multiple clusters and defining multiple patterns based on the characteristics of each cluster (S300); and determining multiple types for classifying users based on the defined multiple patterns (S400).

[0085] In operation S100, server 20 can obtain multiple time-series data points related to multiple factors associated with the battery for each user. For example, server 20 can obtain multiple time-series data points for each user from the OBD 11 set in vehicle 10.

[0086] In operation S200, server 20 can generate feature data representing the relationship between at least two factors that are expected to be cross-correlated among multiple factors, based on each time series data. For example, server 20 can select at least two factors from multiple factors and generate feature data based on the joint PDF of the at least two factors.

[0087] In operation S300, server 20 can classify multiple feature data into multiple clusters and define multiple patterns based on the characteristics of each cluster. For example, server 20 can classify multiple feature data into multiple clusters based on the relationship between at least two factors derived from the feature data and the distribution of each of the at least two factors.

[0088] In operation S400, server 20 can determine multiple types for classifying users based on multiple defined patterns. For example, server 20 can determine multiple types by combining multiple patterns defined based on a combination of factors selected from multiple factors.

[0089] Figure 6 This is a flowchart describing the process of generating feature data according to one embodiment disclosed herein.

[0090] refer to Figure 6 Server 20 can generate feature data representing the relationship between at least two factors.

[0091] In operation S210, server 20 can calculate the joint probability mass function (PMF) of at least two factors based on time series data. For example, time series data can be acquired over multiple charge / discharge cycles.

[0092] In operation S220, server 20 can derive a joint PDF of at least two factors by applying KDE to the joint PMF.

[0093] In operation S230, server 20 can generate feature data representing the joint probability density of at least two factors. For example, the feature data may include 2D contour image data.

[0094] Figure 7 It is a flowchart used to describe the process of defining multiple modes according to one implementation disclosed herein.

[0095] refer to Figure 7Server 20 can define multiple patterns based on multiple feature data.

[0096] In operation S310, server 20 can classify multiple feature data into multiple clusters based on the relationship between at least two factors derived from the feature data and the distribution of each of the at least two factors.

[0097] In operation S320, server 20 can fuse multiple feature data included in each cluster to generate representative feature data.

[0098] In operation S330, server 20 can define multiple patterns based on the relationship between at least two factors derived from representative feature data and the distribution of each of the at least two factors.

[0099] As stated above, although all components constituting the embodiments disclosed herein are described as operating by connection or by connection, the embodiments disclosed herein are not necessarily limited to these embodiments. In other words, one or more of all components may operate by selective connection without departing from the purpose and scope of the embodiments disclosed herein.

[0100] Furthermore, unless otherwise stated, terms such as “comprising,” “constituting,” or “having” above mean that the corresponding component may be inherent and should therefore be interpreted as further including rather than excluding another component. 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 consistent with their meaning in the context of the relevant art and not as having an ideal or overly formal meaning unless explicitly defined herein.

[0101] The above description is merely an exemplary description of the technical ideas disclosed herein, and those skilled in the art will be able to modify and alter this document in various ways without departing from its essential characteristics. Therefore, the embodiments disclosed herein are not intended to limit the technical ideas disclosed herein, but are for illustrative purposes, and the scope of the technical ideas disclosed herein is not limited by these embodiments. The scope of the technical ideas disclosed herein should be interpreted by the appended claims, and all technical ideas within the equivalent scope should be interpreted as included within the scope of this document.

Claims

1. A server, the server comprising: Communication circuits; Memory; as well as A processor, operatively coupled to the communication circuitry and the memory, The processor is configured as follows: For each user, obtain multiple time-series data points related to various factors associated with the battery; Based on each time series data, feature data representing the relationship between at least two factors among the plurality of factors that are expected to be cross-correlated is generated; as well as Multiple feature data are classified into multiple clusters, and multiple patterns are defined based on the characteristics of each cluster.

2. The server according to claim 1, wherein, The processor generates the feature data based on the joint probability density function (joint PDF) of the at least two factors.

3. The server according to claim 2, wherein, The processor is configured as follows: The joint PDF of the at least two factors is calculated based on the time series data; The joint PDF is computed by applying the kernel density function (KDE) to the joint PDF; and Generate the feature data representing the joint probability density of the at least two factors.

4. The server according to claim 3, wherein, The feature data includes two-dimensional (2D) contour image data representing the joint probability density of the at least two factors.

5. The server according to claim 1, wherein, The processor classifies the multiple feature data into multiple clusters based on the relationship between the at least two factors derived from the feature data and the distribution of each of the at least two factors.

6. The server according to claim 1, wherein, The processor is configured as follows: For each cluster, the multiple feature data included in each cluster are merged to generate representative feature data; as well as The plurality of patterns are defined based on the relationship between the at least two factors derived from the representative feature data and the distribution of each of the at least two factors.

7. The server according to claim 1, wherein, The processor determines multiple types for classifying users based on the defined multiple patterns.

8. The server according to claim 7, wherein, The processor analyzes the relationship between the multiple types and the evaluation indicators of the battery, and extracts the main factors related to the evaluation indicators from the multiple factors.

9. The server according to claim 1, wherein, The multiple factors include at least some of the following: State of Charge (SOC) at the start of charging, Depth of Discharge (DOD), Charge Rate (C-rate), Starting Temperature of Charging, Maximum Charging Temperature, Starting Time of Charging, Starting Date of Charging, Start Time of Driving, Driving Distance, Starting Temperature of Driving, Maximum Driving Temperature, Driving DOD, Driving Speed ​​Distribution, Discharge Time, and End Date of Discharge.

10. The server according to claim 1, wherein, The processor acquires the time-series data from the on-board device OBD installed in each user's vehicle.

11. A method for operating a server, the method comprising the following steps: For each user, obtain multiple time-series data points related to various factors associated with the battery; Based on each time series data, feature data representing the relationship between at least two factors among the plurality of factors that are expected to be cross-correlated is generated; as well as Multiple feature data are classified into multiple clusters, and multiple patterns are defined based on the characteristics of each cluster.

12. The method according to claim 11, wherein, The steps for generating the feature data include: Calculate the joint probability mass function (joint PMF) of the at least two factors based on the time series data. The joint probability density function (joint PDF) of the at least two factors is calculated by applying the kernel density function (KDE) to the joint PMF; and Generate the feature data representing the joint probability density of the at least two factors.

13. The method according to claim 11, wherein, The steps for defining the multiple patterns include: Based on the relationship between the at least two factors derived from the feature data and the distribution of each of the at least two factors, the multiple feature data are classified into the multiple clusters; For each cluster, the multiple feature data included in each cluster are merged to generate representative feature data; and The plurality of patterns are defined based on the relationship between the at least two factors derived from the representative feature data and the distribution of each of the at least two factors.

14. The method according to claim 11, further comprising the step of: Based on the defined multiple patterns, multiple types are determined for classifying users.

Images