Electronic device, recording medium, and method for analyzing vehicle driving pattern thereof

Abstract

Provided is an electronic device for clustering data pertaining to a driving pattern of a vehicle. The electronic device may be configured to: acquire a plurality of pieces of driving data for a plurality of vehicles; generate speed distribution data corresponding to each of the plurality of pieces of driving data; generate a plurality of features pertaining to a driving pattern on the basis of the speed distribution data of each of the plurality of pieces of driving data; cluster the plurality of features into a plurality of clusters on the basis of a clustering algorithm and an index for evaluating the clustered result; generate a driving pattern kernel indicating a driving pattern of each of the plurality of clusters; and determine, on the basis of similarities to the plurality of generated driving pattern kernels, a driving pattern of a feature generated after the plurality of driving pattern kernels are generated.

Description

Electronic device, recording medium and method for analyzing vehicle driving patterns thereof

[0001] The present disclosure relates to an electronic device, a recording medium, and a method for analyzing a vehicle driving pattern thereof, and more particularly, to a technology for analyzing a vehicle driving pattern of an electric vehicle user and providing a method for improving fuel efficiency for each driving pattern.

[0002] This application claims the benefit of priority to Republic of Korea Patent Application No. 2024-0121709, dated September 6, 2024, the entire contents of which are incorporated herein by reference.

[0003] A common method for providing fuel efficiency improvement measures involves quantifying the impact of electric vehicle users' driving habits on fuel economy. This is done by assessing driving behavior to determine which factors have the greatest impact. However, because multiple factors simultaneously impact fuel economy, simply analyzing the impact of individual factors may not be enough to provide optimal fuel efficiency improvement measures. In other words, simply improving specific factors is unlikely to yield substantial fuel efficiency improvements.

[0004] To address these issues, methods have been developed that individually analyze specific factors under identical conditions. However, even with this approach, the number of conditions and factors to consider can exponentially increase the number of cases to be analyzed, potentially leading to problems such as excessively high costs for providing fuel efficiency improvement plans. Therefore, there is a growing need for a method that can more efficiently and accurately analyze driving-related data from diverse electric vehicle users while also providing customized, optimal fuel efficiency improvement plans.

[0005] The present disclosure is proposed to solve the above-described problems, and provides an electronic device, a recording medium, and a vehicle driving pattern analysis method thereof.

[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 includes 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 driving data for a plurality of vehicles, generate speed distribution data corresponding to each of the plurality of driving data, generate a plurality of features regarding driving patterns based on the speed distribution data of each of the plurality of driving data, cluster the plurality of features into a plurality of clusters based on a clustering algorithm and an index for evaluating the clustering results, generate a driving pattern kernel representing a driving pattern of each of the plurality of clusters, and determine a driving pattern of a feature generated after the plurality of driving pattern kernels are generated based on a similarity with the plurality of generated driving pattern kernels.

[0008] According to one embodiment, each of the plurality of driving data may include information regarding the speed of each of the plurality of vehicles over time while driving.

[0009] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to identify one or more driving behavior types based on first driving data from among a plurality of driving data, identify a speed range of one or more driving behaviors corresponding to the first type from among the one or more driving behavior types, identify a driving behavior start speed and a driving behavior end speed for each of the one or more speed ranges of the driving behaviors, estimate a distribution of the driving behavior start speed corresponding to the first type and a distribution of the driving behavior end speed corresponding to the first type, respectively, and generate first speed distribution data including the distribution of the driving behavior start speed and the distribution of the driving behavior end speed.

[0010] According to one embodiment, the one or more instructions, when executed, are configured to cause the processor to further determine a travel distance of one or more driving actions corresponding to the first type, wherein a distribution of driving action start speeds may be a probability distribution estimated by weighting the travel distances of one or more driving actions, and a distribution of driving action end speeds may be a probability distribution estimated by weighting the travel distances of one or more driving actions.

[0011] According to one embodiment, the one or more instructions, when executed, may be configured to cause the processor to generate, based on the first speed distribution data, a first feature representing a frequency of a speed range corresponding to each of the one or more driving behavior types.

[0012] According to one embodiment, one or more instructions are configured to cause the processor, when executed, to sample a plurality of features, thereby classifying the plurality of features into a predetermined number of data sets, determine a number of clusters corresponding to each of the predetermined number of data sets based on a clustering algorithm and an index for evaluating a clustering result, perform a first clustering of the plurality of features included in each of the data sets into clusters of the number of clusters determined for each data set, generate a kernel for each of the one or more clusters generated by performing the first clustering, determine an optimal number of clusters corresponding to the one or more generated kernels based on the clustering algorithm and an index for evaluating a clustering result, and perform a second clustering of the one or more kernels into the clusters of the optimal number of clusters, wherein the driving pattern kernel may be a kernel representing a driving pattern of each of the clusters of the optimal number of clusters generated by performing the second clustering.

[0013] According to one embodiment, the similarity between the generated features and the generated plurality of driving pattern kernels after the plurality of driving pattern kernels are generated can be calculated based on the Jensen-Shannon Divergence (JSD) technique that calculates the similarity between probability distributions.

[0014] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to determine a driving pattern of a feature generated after the plurality of driving pattern kernels are generated by verifying that the feature generated after the plurality of driving pattern kernels are generated corresponds to a first driving pattern kernel among the plurality of driving pattern kernels having a similarity value greater than or equal to a threshold value.

[0015] According to one embodiment, one or more instructions may be configured to cause the processor, when executed, to perform clustering multiple times on a first data set among a preset number of data sets to have different numbers of clusters according to a clustering algorithm, and to determine the number of clusters corresponding to the first data set based on the results of the multiple clusterings performed and an index for evaluating the results of the clusterings.

[0016] According to one embodiment, the index includes a first index, a second index, and a third index, and the one or more instructions, when executed, cause the processor to calculate, based on clustering results performed multiple times, each of the first index according to the number of clusters, the second index according to the number of clusters, and the third index according to the number of clusters, and determine an optimal number of clusters for the first data set based on the first number of clusters corresponding to a maximum or minimum value among the calculated first indices according to the number of clusters, the second number of clusters corresponding to a maximum or minimum value among the calculated second indices according to the number of clusters, and the third number of clusters corresponding to a maximum or minimum value among the calculated third indices according to the number of clusters.

[0017] According to one embodiment, for one or more kernels, clustering may be performed multiple times to have different numbers of clusters according to a clustering algorithm, and an optimal number of clusters corresponding to one or more kernels may be determined based on the results of the multiple times of clustering and an index for evaluating the results of the clustering.

[0018] According to one embodiment, the index includes a first index, a second index, and a third index, and the one or more instructions, when executed, cause the processor to calculate, based on clustering results performed multiple times, each of the first index according to the number of clusters, the second index according to the number of clusters, and the third index according to the number of clusters, and determine an optimal number of clusters corresponding to one or more kernels based on the first number of clusters corresponding to a maximum or minimum value among the first index according to the number of clusters calculated, the second number of clusters corresponding to a maximum or minimum value among the second index according to the number of clusters calculated, and the third number of clusters corresponding to a maximum or minimum value among the third index according to the number of clusters calculated.

[0019] According to one embodiment, the one or more instructions may be further configured to cause the processor, when executed, to: identify a driving road type corresponding to the determined driving pattern; identify a power consumption distribution of each of one or more driving pattern kernels corresponding to the driving road type among a plurality of driving pattern kernels; and provide a power consumption improvement method corresponding to a feature of the determined driving pattern based on an optimal power consumption distribution among the identified power consumption distributions of each of the one or more driving pattern kernels.

[0020] A method for analyzing a driving pattern of a vehicle performed by an electronic device according to one embodiment may include the steps of: acquiring a plurality of driving data for a plurality of vehicles; generating speed distribution data corresponding to each of the plurality of driving data; generating a plurality of features related to a driving pattern based on the speed distribution data of each of the plurality of driving data; clustering the plurality of features into a plurality of clusters based on a clustering algorithm and an index for evaluating a clustering result; generating a driving pattern kernel representing a driving pattern of each of the plurality of clusters; and determining a driving pattern of a feature generated after the plurality of driving pattern kernels are generated based on a similarity with the plurality of generated driving pattern kernels.

[0021] A computer-readable, non-transitory recording medium having recorded thereon a program for executing a method for analyzing a driving pattern of a vehicle according to one embodiment of the present invention on a computer, the method for analyzing a driving pattern of a vehicle comprising: acquiring a plurality of driving data for a plurality of vehicles; generating speed distribution data corresponding to each of the plurality of driving data; generating a plurality of features related to a driving pattern based on the speed distribution data of each of the plurality of driving data; clustering the plurality of features into a plurality of clusters based on a clustering algorithm and an index for evaluating the clustering result; generating a driving pattern kernel representing a driving pattern of each of the plurality of clusters; and determining a driving pattern of a feature generated after the plurality of driving pattern kernels are generated based on a similarity with the plurality of generated driving pattern kernels.

[0022] According to the present disclosure, a customized improvement plan can be provided by evaluating the fuel consumption according to the road environment and driving habits of the user.

[0023] Additionally, according to the present disclosure, any driving behavior can be classified into a predefined type, thereby inducing a change to a type having the optimal power efficiency within the same road type.

[0024] 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.

[0025] Figure 1 illustrates a block diagram of an electronic device according to one embodiment.

[0026] Figure 2 shows a flowchart of a method for analyzing a driving pattern of a vehicle according to one embodiment.

[0027] Figures 3a to 3e illustrate a feature generation process according to one embodiment.

[0028] Figure 4 illustrates a feature clustering process according to one embodiment.

[0029] Figure 5 illustrates a first clustering process according to one embodiment.

[0030] Figure 6 illustrates a secondary clustering process according to one embodiment.

[0031] Figure 7 illustrates a process for determining a driving pattern of a feature according to one embodiment.

[0032] Figure 8 shows the power distribution of each driving pattern kernel according to one embodiment.

[0033] Figures 9a and 9b illustrate a process for providing a power improvement method according to one embodiment.

[0034] 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.

[0035] 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.

[0036] 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'.

[0037] 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.

[0038]

[0039] 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.

[0040] Figure 1 illustrates a block diagram of an electronic device according to one embodiment.

[0041] 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.

[0042] 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 inputting 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.

[0043] 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 driving data for the multiple vehicles. The multiple driving 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 driving data is previously stored, via the transceiver (110) of the electronic device (100). The manner in which the electronic device (100) acquires the multiple driving 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 driving 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.

[0044] 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.

[0045] The processor (120) may be implemented as a computer or similar device according to 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 a control function, 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 vehicle driving pattern analysis system are executed, the modules may be interpreted as controlling the processor (120) to perform the operations of the electronic device (100) below.

[0046] 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.

[0047] 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.

[0048] 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.

[0049] 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.

[0050] Figure 2 shows a flowchart of a method for analyzing a driving pattern of a vehicle according to one embodiment.

[0051] At step S210, the electronic device (100) may acquire multiple driving data for multiple vehicles. Each of the plurality of driving data may include information regarding the speed of each of the multiple vehicles over time while driving. Specifically, each driving data may be time-series data representing a series of speed changes over time during a single driving period of the corresponding vehicle. A more specific embodiment of the acquired driving data will be described in detail below with reference to FIG. 3A.

[0052] In step S220, the electronic device (100) may generate speed distribution data corresponding to each of a plurality of driving data. Specifically, the electronic device (100) may identify one or more driving behavior types corresponding to each of the driving data, identify a speed range of one or more driving behaviors corresponding to each of the one or more driving behavior types, and identify a driving behavior start speed and a driving behavior end speed for each of the one or more driving behaviors.

[0053] Here, one or more driving behavior types may be predefined driving habit types, and may be, for example, five subdivided types of starting acceleration, stopping deceleration, shift acceleration, shift deceleration, and constant speed. The starting acceleration type may be, for example, a type in which the slope of the speed is greater than a predetermined value and starts from a stop, the stop deceleration type may be, for example, a type in which the slope of the speed is less than a predetermined value and ends in a stop, the shift acceleration type may be, for example, a type in which the slope of the speed is greater than a predetermined value and starts from a non-stop state, the shift deceleration may be, for example, a type in which the slope of the speed is less than a predetermined value and ends in a non-stop state, and the constant speed may be, for example, the four types above and the remaining types excluding the stop. That is, the electronic device (100) can check multiple driving behavior types included in each driving data, and check the speed range of the driving section corresponding to each driving behavior type. Meanwhile, the driving action start speed may be the start speed of the driving section confirmed based on the speed range of the driving section corresponding to each driving action type, and the driving action end speed may be the end speed of the driving section.

[0054] Based on each speed range, the electronic device (100) that confirms the driving behavior start speed and the driving behavior end speed corresponding to each driving behavior as described above can estimate the distribution of the driving behavior start speed and the driving behavior end speed corresponding to each driving behavior type, respectively, and can generate speed distribution data for each driving data that includes both the distribution of the driving behavior start speed and the driving behavior end speed estimated for each driving behavior. Here, each of the distribution of the driving behavior start speed and the distribution of the driving behavior end speed may mean a probability distribution for the target speed, and accordingly, the speed distribution data may include multiple probability distributions. In addition, each of the probability distribution for the start speed and the probability distribution for the end speed may be a probability density function estimated using, for example, KDE (Kernel Density Estimation). KDE may be a technique that applies a kernel function, which is a type of probability distribution, to a target data point, and then approximates the entire data by adding kernel functions for all data points to generate a probability density function. Accordingly, a probability density function in the form of a continuous and smooth curve can be generated.

[0055] According to one embodiment, the electronic device (100) may further confirm the travel distance of one or more driving behaviors corresponding to each driving behavior type in order to estimate the distribution of the driving behavior start speed and the driving behavior end speed corresponding to each driving behavior type, respectively. Accordingly, the distribution of the driving behavior start speed may be a probability distribution estimated by assigning weights to the travel distances of one or more driving behaviors, and the distribution of the driving behavior end speed may be a probability distribution estimated by assigning weights to the travel distances of one or more driving behaviors. That is, when there are multiple driving behaviors corresponding to one driving behavior type, each speed distribution may be a probability distribution reflecting multiple driving behaviors such that data corresponding to driving behaviors with longer travel distances have a higher probability. Meanwhile, according to one embodiment, the electronic device (100) may further confirm energy consumption data of one or more driving behaviors corresponding to each driving behavior type. Information regarding the energy consumption data may be additionally acquired to distinguish and assign weights to multiple driving behaviors corresponding to the stop type when the predefined driving behavior type includes a stop behavior.

[0056] A more specific embodiment in which the electronic device (100) estimates the distribution of the starting speed and the distribution of the ending speed of the driving behavior corresponding to each driving behavior type will be described in detail with reference to FIGS. 3b and 3c below, and a more specific embodiment of the speed distribution data generated thereby will be described in detail with reference to FIG. 3d below.

[0057] At step S230, the electronic device (100) may generate multiple features related to a driving pattern based on the speed distribution data of each of the plurality of driving data. Specifically, the electronic device (100) may generate features indicating the frequency of a speed range corresponding to each of one or more driving behavior types based on the speed distribution data. In other words, features may be generated for each speed distribution data of each driving data and may represent a driving pattern identified based on the corresponding speed distribution data. A more specific embodiment of the generated features will be described in detail with reference to FIG. 3E below.

[0058] At step S240, the electronic device (100) can cluster multiple features into multiple clusters based on a clustering algorithm and an index used to evaluate the clustering results. That is, multiple features can be classified into the same cluster based on their degree of similarity. More specific examples of the clustering process for multiple features will be described in detail with reference to FIGS. 4 to 6 below.

[0059] At step S250, the electronic device (100) may generate a driving pattern kernel representing the driving pattern of each of the plurality of clusters. Here, the kernel is data representing the characteristics of data within the corresponding cluster and may have the same form as a feature. That is, the driving pattern kernel may be data representing a plurality of features within the corresponding cluster, and in one example, may be generated by overlapping features within the corresponding cluster. The driving pattern kernel generated in this manner may be the driving pattern itself defined by the present disclosure. A more specific embodiment of generating the driving pattern kernel will be described in detail with reference to FIG. 6 below.

[0060] In step S260, the electronic device (100) can determine the driving pattern of the feature generated after the plurality of driving pattern kernels are generated based on the similarity with the plurality of generated driving pattern kernels. In other words, the electronic device (100) can determine the driving pattern of the generated feature by determining which of the plurality of driving pattern kernels is most similar to the feature of the newly generated driving pattern after the plurality of driving pattern kernels are generated. The similarity between the generated feature and the plurality of driving pattern kernels can be calculated based on, for example, the Jensen-Shannon Divergence (JSD) technique that calculates the similarity between probability distributions. More specifically, the JSD technique can be a technique that quantifies and represents the difference between two probability distributions based on the distance between the two distributions, and the similarity calculated through the JSD technique can more intuitively represent the degree of similarity between the generated feature and the plurality of driving pattern kernels. The method for calculating similarity is not limited to the JSD technique, and it will be clear to those skilled in the art that the similarity between the newly generated feature and each of the multiple driving pattern kernels can be calculated through various methods that can quantify the degree of similarity between the two distributions.

[0061] Based on the similarity calculated in this way, the electronic device (100) can determine that among the plurality of driving pattern kernels, a driving pattern kernel having a similarity value greater than or equal to a threshold value corresponds to the driving pattern of the newly generated feature. That is, each driving pattern kernel is representative data representing the characteristics of each cluster in which the plurality of features are clustered as described above, and the electronic device (100) can practically determine which cluster the newly generated feature can be classified into through the aforementioned process, without the hassle of performing a new clustering on a large amount of data as driving data is additionally acquired and features are generated based on the acquired driving data. Accordingly, the electronic device (100) can more quickly and easily determine the driving pattern of the corresponding driving data based on the similarity between the features generated for the corresponding driving data and the plurality of previously generated driving pattern kernels, regardless of which driving data is acquired.

[0062] Figures 3a to 3e illustrate a feature generation process according to one embodiment.

[0063] According to one embodiment, the electronic device (100) may segment each of the acquired plurality of driving data according to the driving behavior type. Referring to FIG. 3A, the driving data (300) may be in the form of a graph representing the speed over time during one driving of a corresponding specific vehicle, wherein the x-axis of the graph may represent time (sec) and the y-axis may represent speed (km / h). The electronic device (100) may identify five driving behavior types (start acceleration, stop deceleration, shift acceleration, shift deceleration, and constant speed) based on a series of speed changes in the driving data (300). That is, the electronic device (100) may identify each of the sections t1 to t2, t2 to t3, t3 to t4, t4 to t5, and t5 to t6 in the driving data (300) as corresponding to one of the driving behavior types. Meanwhile, checking the sections corresponding to the five types in this way is only an example, and depending on the embodiment, the number of driving behavior types confirmed for each driving data and the number of time sections corresponding to each driving behavior type may vary (i.e., there may be multiple driving behaviors corresponding to one driving behavior type).

[0064] Next, the electronic device (100) can segment the driving data to correspond to each driving behavior. The graph (310) of FIG. 3A is an example of segmented driving data, and as described above, may be a graph that simplifies and represents the change in speed over time for each driving behavior. By segmenting the driving data (300) in this way, the electronic device (100) can more easily confirm the speed range for each driving behavior.

[0065] Referring to FIG. 3B, the electronic device (100) can identify the driving behavior type (attribute), starting speed, ending speed, and moving distance corresponding to each driving behavior based on the speed range corresponding to each driving behavior. The information for each driving behavior can be organized in a listed form, for example, as in table (330). As k driving behaviors corresponding to the stop / deceleration type are identified in table (330), the electronic device (100) can estimate the start speed distribution and the stop speed distribution of the driving behavior of the stop / deceleration type so that the start speed and the end speed of "stop / deceleration 1", which has the longest moving distance, have a higher probability.

[0066] Next, referring to FIG. 3c, the electronic device (100) can estimate the time-speed distribution and the end speed distribution corresponding to each driving behavior type (340). In an example where all driving behaviors corresponding to each of the five types of starting acceleration, stopping deceleration, shifting acceleration, shifting deceleration, and constant speed are confirmed, five starting speed distributions and five ending speed distributions estimated for each of the five types can be estimated, as shown in 340. Meanwhile, each speed distribution graph may have the starting speed or the end speed (km / h) on the x-axis and the probability on the y-axis, as shown. Each speed distribution graph generated to represent the probability according to the speed in this way may have a different index. For example, as shown in 340, a total of ten speed distribution graphs may be expressed with indices from 0 to 9. At this time, the indexes may be assigned in the order of the starting speed distribution and the end speed distribution corresponding to the same driving behavior type. That is, in the example of 340, the probability distributions 0, 1, ..., 9 can be assigned in the order of the starting speed distribution and the ending speed distribution of the departure acceleration type, the starting speed distribution and the ending speed distribution of the variable acceleration type, the starting speed distribution and the ending speed distribution of the constant velocity type, the starting speed distribution and the ending speed distribution of the variable deceleration type, and the starting speed distribution and the ending speed distribution of the stop deceleration type.

[0067] According to one embodiment, the electronic device (100) may generate speed distribution data that lists all speed distributions so that all estimated start speed distributions and end speed distributions can be checked at a glance. Referring to FIG. 3d, the electronic device (100) may generate speed distribution data (350) in a form in which all estimated start speed distributions and end speed distributions for each driving behavior type are listed in the order of the index, as in 340 of FIG. 3c. Accordingly, speed distribution data (350) including a total of 10 speed distributions of probability distribution 0 to probability distribution 9 may be generated. Here, the speed distribution data may be in the form of a three-dimensional graph in which the x-axis represents the index of the probability distribution, the y-axis represents the speed, and the z-axis represents the probability of the speed.

[0068] Referring to FIG. 3E, the electronic device (100) may generate a feature (360) representing the frequency of a speed range corresponding to each of five driving behavior types based on the speed distribution data (350). The feature (360) may be, for example, a two-dimensional image generated by projecting the corresponding speed distribution data (350) with respect to the z-axis. More specifically, the feature (360) may be a form that visualizes data composed of an x-axis and a y-axis, looking at the speed distribution data (350) in the -z direction, and accordingly, the feature (360) may be a two-dimensional image form with the probability distribution index (n) as the x-axis and the speed (km / h) as the y-axis.

[0069] The feature (360) generated based on the speed distribution data (350) may be in the form of probability distributions listed in a form in which various intensities appear within each probability distribution, and here, the intensities of the feature (360) may mean the frequency of the corresponding speed. That is, the higher the intensities appear within each probability distribution included in the feature (360), the more frequently the user drives at the speed when driving according to the corresponding type of driving behavior. Accordingly, the feature (360) may indicate driving habits or tendencies regarding the speed at which driving mainly starts and ends when driving according to the corresponding type of driving behavior. In particular, since the feature (360) sequentially represents the starting speed distribution and the ending speed distribution of the starting acceleration type, the starting speed distribution and the ending speed distribution of the shift acceleration type, the starting speed distribution and the ending speed distribution of the constant velocity type, the starting speed distribution and the ending speed distribution of the shift deceleration type, and the starting speed distribution and the ending speed distribution of the stop deceleration type, it can provide intuitive information about the speed at which driving mainly starts and ends during starting acceleration driving, the speed at which driving mainly starts and ends during shift acceleration driving, etc. As a more specific example, the probability distributions 2 and 3 of the feature (360) can represent that the user of the corresponding vehicle has a driving habit of mainly starting driving in the range of 20 to 40 km / h and mainly ending driving in the range of 50 to 70 km / h when driving in the shift acceleration type.

[0070] The electronic device (100) can generate features corresponding to each driving data by performing the process described in the examples of FIGS. 3A to 3E on each of the plurality of driving data. Accordingly, each driving data can ultimately be formatted into a feature in the form of a two-dimensional image of the same size.

[0071] Figure 4 illustrates a feature clustering process according to one embodiment.

[0072] According to one embodiment, the electronic device (100) can sample a plurality of features generated as described above, thereby classifying the plurality of features into a preset number of data sets, determine the number of clusters corresponding to each of the preset number of data sets based on a clustering algorithm and an index for evaluating the clustering results, and perform primary clustering of the plurality of features included in each of the data sets into clusters of the number of clusters determined for each data set.

[0073] Next, the electronic device (100) may generate a kernel for each of one or more clusters generated by performing the primary clustering, determine an optimal number of clusters corresponding to the generated one or more kernels based on a clustering algorithm and an index for evaluating the clustering result, and perform secondary clustering with the one or more kernels as clusters with the optimal number of clusters. The electronic device (100) may generate a driving pattern kernel representing the driving pattern of each cluster with the optimal number of clusters generated by performing the secondary clustering in this manner.

[0074] Referring to FIG. 4, a preset number (k) of data sets including data set 1 (422), data set 2 (424), ..., data set k (426) can be generated based on features regarding a plurality of driving patterns. The electronic device (100) can sample features regarding a plurality of driving patterns into k data sets to generate the data sets, and the sampling of the features can be performed randomly, for example. Each of the preset number of data sets can include, for example, at least some features, and as a specific example, when a total of 840,000 driving data are acquired and a total of 840,000 features are generated, and the k value is set to 100, the number of at least some features included in each data set can be 8400.

[0075] The electronic device (100) can perform primary clustering on each of the k data sets. As the primary clustering is performed, multiple clusters corresponding to each of the k data sets can be generated. Specifically, data set 1 (422) can be clustered into A1 clusters as a result of the primary clustering, and data set 2 (424) can be clustered into A2 clusters as a result of the primary clustering. The electronic device (100) that performs primary clustering on each data set can generate a kernel for each generated cluster. That is, as a result, A1 clusters and A1 kernels (432) corresponding thereto, A2 clusters and A2 kernels (434) corresponding thereto, etc. can be generated. The kernel of each cluster is data representing the characteristics of multiple features within each cluster, as described above with respect to the driving pattern kernel, and can be a type of representative feature generated by overlapping multiple features within the cluster. By performing primary clustering in this manner, the amount of data can be significantly reduced. In an example with hundreds of thousands of features, the number of kernels (A1, A2, etc.) generated through primary clustering can each be in the hundreds, significantly reducing the amount of data.

[0076] Next, the electronic device (100) can perform secondary clustering on all of the generated kernels, such as A1 kernels, A2 kernels, etc. That is, if the primary clustering was performed for each data set, the secondary clustering can be performed on all of the multiple kernels representing the multiple clusters generated as a result of the primary clustering. In addition, since the number of multiple kernels is significantly reduced compared to the number of multiple features, the process of sampling data performed prior to the primary clustering can also be omitted considering the large number of data. The electronic device (100) can perform secondary clustering on the multiple kernels to cluster them into clusters with an optimal number of clusters (m). The set of driving pattern kernels (440) can include a driving pattern kernel generated for each of the m clusters generated by performing the secondary clustering. By performing the secondary clustering, the amount of data reduced through the primary clustering can be reduced once again. Accordingly, hundreds of thousands of features can be reduced to tens of data through the first clustering, and the final data can be reduced to tens through the second clustering, and the number of driving pattern kernels m obtained by performing clustering based on 840,000 data can be, for example, 25.

[0077] Figure 5 illustrates a first clustering process according to one embodiment.

[0078] According to one embodiment, the electronic device (100) may perform clustering multiple times for each of a preset number of data sets to have different numbers of clusters according to a clustering algorithm, and may determine the number of clusters corresponding to each data set based on the results of the multiple clustering performed and an index for evaluating the results of the clustering.

[0079] Referring to FIG. 5, the electronic device (100) can perform clustering multiple times on data set 1 (422) so that the number of clusters is p (i∈{2, 3, ..., p}) according to a clustering algorithm. The electronic device (100) can cluster data set 1 (422) multiple times so that the number of clusters changes each time clustering is performed.

[0080] In one example, the clustering algorithm performed on the data set can be any type of clustering algorithm that allows the number of clusters to be set as a parameter (i.e., hyperparameterized). For example, it can be a hierarchical clustering algorithm that hierarchizes at least some features contained in each data set based on similarity. Hierarchical clustering algorithms can have the advantage of very low computational overhead, as they calculate an initial distance matrix and then repeatedly merge or subdivide clusters with close distances based on this matrix.

[0081] In another example, the clustering algorithm may be the K-means clustering algorithm, which clusters at least some features within each data set based on their centroids. The K-means clustering algorithm performs clustering by initially assigning data points to the closest cluster based on the selected centroids, calculating the cluster centroids, and repeating the process until the clusters have the same centroids. This iterative operation can be parallelized, allowing for faster computation on large data sets.

[0082] Next, the electronic device (100) can determine the optimal number of clusters for the data set 1 (422) based on the index for each of the clustering results performed multiple times. The indexes may include, for example, a first index (522), a second index (524), and a third index (526), ​​and the electronic device (100) can calculate the first index according to the number of clusters, the second index according to the number of clusters, and the third index according to the number of clusters, based on the clustering results performed multiple times. The index according to the number of clusters may refer to a performance index that evaluates the results of performing clustering with different p-values, and may be organized in the form of a graph, for example. Meanwhile, the types of indices that the electronic device (100) calculates based on the clustering results performed multiple times are not limited to the three types of indices, and it will be clearly understood by those skilled in the art that more types of indices may be calculated depending on the embodiment.

[0083] In one example, each of the first index, the second index, and the third index may include at least one of the Silhouette Index technique, the Calinski-Harabasz Index technique, and the Davies-Bouldin Index technique, which evaluate the clustering results based on the cohesion of each cluster and the separation between clusters, for each clustering result performed according to the clustering algorithm. The techniques have differences in the process of calculating the cohesion and separation. For example, according to the Silhouette Index technique, the cohesion may be calculated based on the average of the distances between each data point and all other data points in the same cluster, and the separation may be calculated based on the average of the distances between each cluster and all other data points in the closest other cluster. In contrast, according to the Kalinsky-Harabats index technique, cohesion is calculated based on the average square of the distance between each data point and its cluster centroid, while separability can be calculated based on the average square of the distance between each cluster's centroid and the entire data set's centroid. Furthermore, according to the Davis-Boldin index technique, cohesion is calculated based on the average of the distance between each data point and its cluster centroid, while separability is calculated based on the average of the distance between each cluster's centroid.

[0084] Next, the electronic device (100) may determine an optimal number of clusters for the first data set based on a first cluster number corresponding to a maximum or minimum value among the first indexes according to the calculated number of clusters, a second cluster number corresponding to a maximum or minimum value among the second indexes according to the calculated number of clusters, and a third cluster number corresponding to a maximum or minimum value among the third indexes according to the calculated number of clusters. Here, each of the maximum or minimum value among the first indexes, the maximum or minimum value among the second indexes, and the maximum or minimum value among the third indexes may correspond to an optimal value among the maximum and minimum values ​​set based on the characteristics of the indexes. For example, when the index is a silhouette index technique or a Kalinsky-Harabats index technique, the maximum value may correspond to the optimal value, and when the index is a Davis-Boldin index technique, the minimum value may correspond to the optimal value. Meanwhile, the optimal value for each index is not limited to the maximum or minimum value, and can be set to an appropriate value depending on the characteristics of the index.

[0085] The electronic device (100) may set the optimal number of clusters (A1) for data set 1 (422) to an overlapping value among the first number of clusters, the second number of clusters, and the third number of clusters. That is, the number of clusters where the maximum or minimum values ​​intersect as evaluated by each of the three indices may be set as the optimal number of clusters. For example, if the first number of clusters and the second number of clusters are 184 and the third number of clusters is 180, the optimal number of clusters may be set to 184. In this way, by setting the number of clusters evaluated as having the optimal value according to various index techniques as the optimal number of clusters for data set 1 (422), clustering with guaranteed high performance may be performed. According to another embodiment, if the first number of clusters, the second number of clusters, and the third number of clusters all have different values, their average value may be set as the optimal number of clusters for data set 1 (422).

[0086] According to the above-described process, the first clustering can be performed with A1 clusters, which is the optimal number of clusters for data set 1 (422), and then representative kernels for each of the A1 clusters can be generated. That is, A1 clusters and A1 kernels (432) can be generated corresponding to data set 1 (422). Meanwhile, the same process can be performed for each data set, and thus, clusters and kernels corresponding to each of the k data sets in total can be generated.

[0087] Figure 6 illustrates a secondary clustering process according to one embodiment.

[0088] According to one embodiment, the electronic device (100) performs clustering multiple times with different numbers of clusters according to a clustering algorithm for multiple kernels generated as a result of the primary clustering, and determines an optimal number of clusters corresponding to the multiple kernels based on the results of the multiple clustering and an index for evaluating the clustering results. The process of performing the secondary clustering is the same as the process (process of determining the optimal number of clusters for each data set) described in detail with respect to FIG. 5, and thus, a description thereof is omitted.

[0089] Referring to Fig. 6, the optimal number of clusters determined for multiple kernels may be m. Accordingly, multiple kernels (kernels as numerous as the sum of A1, A2, etc.) may be clustered into cluster 1 (610), cluster 2 (612), ..., cluster m (614). Each cluster includes at least some kernels, and since multiple kernels are clustered based on similarity, they may include kernels having similar patterns. Meanwhile, as in Fig. 6, the more similar kernels or features are to each other, the more similar the brightness patterns within each kernel or feature image may appear.

[0090] The electronic device (100) can generate a driving pattern kernel representing the characteristics of each cluster by overlapping (∑) kernels within each cluster. For example, driving pattern kernel 1 (442) corresponding to cluster 1 (610) can be a kernel derived by overlapping all kernels included in cluster 1 (610), and the m driving pattern kernels generated in this way can be in a form that more clearly reflects the kernel shape of the corresponding cluster, and at the same time, the differences between the m clusters can be more clearly revealed through the driving pattern kernels.

[0091] Figure 7 illustrates a process for determining a driving pattern of a feature according to one embodiment.

[0092] Referring to FIG. 7, the graph (700) may be a graph regarding the similarity between the newly generated feature and m driving pattern kernels, which is confirmed by the electronic device (100), in order to determine which of the m kernels the newly generated feature corresponds to after all m driving pattern kernels are generated. Accordingly, the x-axis of the similarity graph (700) may be a driving pattern kernel index (m), and the y-axis may be a similarity (%). Here, the driving pattern kernel index may be an index assigned to each of the generated multiple driving pattern kernels, and according to an example in which a total of 25 driving pattern kernels are generated, the driving pattern kernel index may have a value from 0 to 24. Referring to the graph (700), the threshold value of the similarity, which is the criterion for determining similarity, may be 95, and accordingly, the driving pattern kernel 15 having a similarity of 95% or more may be confirmed to be the driving pattern kernel corresponding to the feature. Therefore, the feature can be determined to have a driving pattern of driving pattern kernel 15.

[0093] Figure 8 shows the power distribution of each driving pattern kernel according to one embodiment.

[0094] According to one embodiment, the electronic device (100) can generate m driving pattern kernels and check the power consumption distribution corresponding to each driving pattern kernel. Referring to FIG. 8, the graph (800) may be a graph representing the power consumption for each driving pattern kernel, with the x-axis representing the driving pattern kernel index (m) and the y-axis representing the power consumption (km / kWh).

[0095] Meanwhile, although not illustrated in FIG. 8, the electronic device (100) can additionally classify each driving pattern kernel by road type and also display the road type of each driving pattern kernel together on the power distribution graph (800). For example, based on 25 driving pattern kernels, the road type corresponding to each driving pattern kernel can be identified in response to the input of a worker performing the data analysis task, and a driving pattern kernel index can be assigned so that adjacent indices are provided for each of the same road types. As a more specific example, if, in response to the operator input, 9 of the 25 driving pattern kernels are identified as being of the urban driving type, 10 are identified as being of the main road driving type, 4 are identified as being of the highway driving type, and the remaining 2 are identified as being of the mixed driving type, then the 9 driving pattern kernels of the urban driving can be sequentially assigned indices from 0 to 8, the 10 driving pattern kernels of the main road driving can be sequentially assigned indices from 9 to 18, the 4 driving pattern kernels of the highway driving can be sequentially assigned indices from 19 to 22, and the 2 driving pattern kernels of the mixed driving can be sequentially assigned indices from 23 and 24. By confirming the fuel efficiency distribution in the order of the assigned indices and graphing it, the electronic device (100) can generate and provide a fuel efficiency distribution graph that allows for easier confirmation of the fuel efficiency distribution for each road type.

[0096] Figures 9a and 9b illustrate a process for providing a power improvement method according to one embodiment.

[0097] According to one embodiment, an electronic device (100) that has generated a plurality of driving pattern kernels can then acquire a plurality of driving data. Based on the acquired driving data, the electronic device (100) can determine the similarity between the plurality of features generated in the same manner as described above and each of the plurality of driving pattern kernels, and determine a driving pattern corresponding to each feature based on the similarity.

[0098] According to one embodiment, the electronic device (100) that determines a driving pattern corresponding to each feature may identify a driving road type corresponding to the determined driving pattern, and may identify the fuel efficiency distribution of each of one or more driving pattern kernels corresponding to the identified driving road type among a plurality of driving pattern kernels, and may provide a fuel efficiency improvement method corresponding to each feature for which the driving pattern is determined based on the optimal fuel efficiency distribution among them. For example, the electronic device (100) that identifies that the driving pattern kernel corresponding to feature 1 is driving pattern kernel 15 may then identify, in response to an operator input, a main road type, which is a road type of the previously identified driving pattern kernel 15, as the road type of feature 1. Subsequently, the electronic device (100) may identify the fuel efficiency distribution of driving pattern kernels of the main road type through a fuel efficiency distribution graph or the like, and identify a driving pattern kernel (for example, driving pattern kernel 11 on the graph (800)) having the optimal fuel efficiency distribution among them. Accordingly, the electronic device (100) can provide information on a fuel efficiency improvement method required to improve the fuel efficiency of feature 1 based on the difference between the driving pattern kernel 15, which is the driving pattern kernel of feature 1, and the driving pattern kernel having the optimal fuel efficiency distribution. In other words, information inducing a change to the driving pattern of the driving pattern kernel having the optimal fuel efficiency distribution can be provided. This information on the fuel efficiency improvement method can be provided, for example, to the OBD of the vehicle corresponding to feature 1.

[0099] Referring to Fig. 9a, as in the schematic diagram (900), the electronic device (100) can classify each driving pattern kernel into a road type and identify the main driving habits of each driving pattern kernel. Meanwhile, the class in Fig. 9a may be the driving pattern kernel of the present invention. For the convenience of vehicle users who are provided with a fuel efficiency improvement plan, the state type (920) of each road type (910) may be subdivided into at least one of congested driving and smooth driving so that the classes can be classified more finely. Accordingly, for example, if the classes corresponding to the urban driving types are 0, 2, 4 to 6, 8, and 10, the electronic device (100) can further classify and assign the congested driving type as classes 0, 4, and 6, and the smooth driving type as classes 5, 8, 10, and 2. Next, the electronic device (100) can check the main driving habits (930) for each class assigned to correspond to the detailed state type (920). The main driving habits (930) of the class may be information on the main habits extracted for speed and gear shifting based on the kernel of each class (e.g., one of slow / normal / fast and the habits of gear shifting, constant speed, and wide gear shifting, etc.).

[0100] Referring to FIG. 9B, the analysis result (940) may be information regarding the result of analyzing the difference from the class confirmed to have the optimal fuel efficiency distribution for each road type. Specifically, the electronic device (100) may analyze the difference from the main driving habits of the class having the optimal fuel efficiency distribution among the classes classified by the road type corresponding to each class based on the main driving habits of each class confirmed as in FIG. 9A. Here, the class having the optimal fuel efficiency distribution among the classes classified by the same road type may be the class having the optimal fuel efficiency distribution among the classes belonging to the broadest classification, that is, the road type, or, depending on the embodiment, may be the class having the optimal fuel efficiency distribution among the classes belonging to the condition type (congested driving / smooth driving type) of each road type, which is a narrower classification.

[0101] For example, since Class 4 belongs to traffic congestion driving among urban road driving types, if Class 6 has the optimal fuel efficiency distribution among Classes 0, 4, and 6, which belong to traffic congestion driving types of urban road types, which are narrow categories, the electronic device (100) can analyze the difference between Class 6 and Class 4. The difference between each class can be analyzed based on major driving habits (930). As a result of the analysis, if the difference between Class 4 and Class 6 is speed, and Class 4 drives more slowly than Class 6, the electronic device (100) can determine that Class 4 corresponds to the 'low-speed driving' category among the analysis results (940). In addition, accordingly, information of 'Increase speed -> Present target speed' among the fuel efficiency improvement plans (950) can be provided to a vehicle corresponding to Class 4. Here, the target speed can be determined as an appropriate value based on the speed distribution of Class 6.

[0102] As described above, the electronic device (100) can determine the driving pattern of driving data based on the driving pattern kernels, and thus, even with a limited number of driving pattern kernels, it can more quickly provide the most appropriate fuel efficiency improvement method that can achieve the optimal fuel efficiency distribution depending on the driving situation and on which road. This fuel efficiency improvement method is the most appropriate user-customized fuel efficiency improvement method for countless different cases. Therefore, it can significantly contribute to the direct cost reduction of the user by enabling the user to use the battery more efficiently.

[0103]

[0104] 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.

[0105] 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.

[0106] 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: Acquire multiple driving data for multiple vehicles, Generate speed distribution data corresponding to each of the above plurality of driving data, Based on the speed distribution data of each of the plurality of driving data, multiple features related to the driving pattern are generated, Based on the clustering algorithm and the index for evaluating the clustering results, the plurality of features are clustered into multiple clusters, Generate a driving pattern kernel representing the driving pattern of each of the above multiple clusters, An electronic device configured to determine a driving pattern of a feature generated after the plurality of driving pattern kernels are generated based on the similarity with the plurality of driving pattern kernels generated above.

2. In paragraph 1, Each of the above plurality of driving data is: An electronic device comprising information regarding the speed of each of the plurality of vehicles over time while they are driving.

3. In paragraph 1, The one or more instructions, when executed, cause the processor to: Based on the first driving data among the plurality of driving data, one or more driving behavior types are identified, Check the speed range of one or more driving behaviors corresponding to the first type among the one or more driving behavior types above, For each speed range of one or more of the driving actions, check the driving action start speed and the driving action end speed, Estimate the distribution of the starting speed of the driving behavior corresponding to the first type and the distribution of the ending speed of the driving behavior corresponding to the first type, respectively. An electronic device configured to generate first speed distribution data including a distribution of the driving action start speed and a distribution of the driving action end speed.

4. In paragraph 3, The one or more instructions, when executed, cause the processor to: It is configured to further check the movement distance of one or more driving actions corresponding to the first type above, The distribution of the start speed of the above driving behavior is a probability distribution estimated by weighting the movement distance of one or more driving behaviors, An electronic device wherein the distribution of the above driving behavior termination speed is a probability distribution estimated by weighting the movement distance of one or more driving behaviors.

5. In paragraph 3, The one or more instructions, when executed, cause the processor to: An electronic device configured to generate a first feature representing a frequency of a speed range corresponding to each of the one or more driving behavior types based on the first speed distribution data.

6. In paragraph 1, The one or more instructions, when executed, cause the processor to: By sampling the above plurality of features, the plurality of features are classified into a preset number of data sets, Based on the above clustering algorithm and the index for evaluating the clustering result, the number of clusters corresponding to each of the preset number of data sets is determined, Performing a first clustering of multiple features included in each of the above data sets into clusters of a number of clusters determined for each of the above data sets, Generating a kernel for each of one or more clusters generated by performing the above first clustering, Based on the above clustering algorithm and the index for evaluating the clustering result, the optimal number of clusters corresponding to one or more generated kernels is determined, The above one or more kernels are configured to perform a second clustering into clusters of the optimal number of clusters, An electronic device wherein the driving pattern kernel is a kernel representing the driving pattern of each cluster of the optimal number of clusters generated by performing the second clustering.

7. In paragraph 1, The similarity between the generated features and the generated plurality of driving pattern kernels after the plurality of driving pattern kernels are generated is An electronic device that is calculated based on the Jensen-Shannon Divergence (JSD) technique, which calculates the similarity between probability distributions.

8. In paragraph 1, The one or more instructions, when executed, cause the processor to: An electronic device configured to determine a driving pattern of a feature generated after the plurality of driving pattern kernels are generated by confirming that the feature generated after the plurality of driving pattern kernels are generated corresponds to a first driving pattern kernel among the plurality of driving pattern kernels whose similarity value is greater than or equal to a threshold value.

9. In paragraph 6, The one or more instructions, when executed, cause the processor to: For the first data set among the above-described number of data sets, clustering is performed multiple times to have different numbers of clusters according to the clustering algorithm, An electronic device configured to determine the number of clusters corresponding to the first data set based on the results of the clustering performed multiple times and an index for evaluating the results of the clustering.

10. In paragraph 9, The above index includes a first index, a second index, and a third index, The one or more instructions, when executed, cause the processor to: Based on the results of the clustering performed multiple times, the first index according to the number of clusters, the second index according to the number of clusters, and the third index according to the number of clusters are each calculated, An electronic device configured to determine an optimal number of clusters for the first data set based on a first number of clusters corresponding to a maximum or minimum value among the first indexes according to the calculated number of clusters, a second number of clusters corresponding to a maximum or minimum value among the second indexes according to the calculated number of clusters, and a third number of clusters corresponding to a maximum or minimum value among the third indexes according to the calculated number of clusters.

11. In paragraph 6, For the above one or more kernels, clustering is performed multiple times to have different numbers of clusters according to the clustering algorithm, An electronic device configured to determine an optimal number of clusters corresponding to one or more kernels based on the results of the clustering performed multiple times and an index for evaluating the results of the clustering.

12. In paragraph 11, The above index includes a first index, a second index, and a third index, The one or more instructions, when executed, cause the processor to: Based on the results of the clustering performed multiple times, the first index according to the number of clusters, the second index according to the number of clusters, and the third index according to the number of clusters are each calculated, An electronic device configured to determine an optimal number of clusters corresponding to the one or more kernels based on a first number of clusters corresponding to a maximum or minimum value among the first indexes according to the calculated number of clusters, a second number of clusters corresponding to a maximum or minimum value among the second indexes according to the calculated number of clusters, and a third number of clusters corresponding to a maximum or minimum value among the third indexes according to the calculated number of clusters.

13. In paragraph 1, The one or more instructions, when executed, cause the processor to: Check the driving road type corresponding to the driving pattern determined above, Check the power distribution of each of one or more driving pattern kernels corresponding to the driving road type among the plurality of driving pattern kernels, An electronic device further configured to provide a power consumption improvement method corresponding to a feature for which the driving pattern is determined based on an optimal power consumption distribution among the power consumption distributions of each of the one or more driving pattern kernels identified above.

14. In a method for analyzing a vehicle's driving pattern, performed by an electronic device, A step of acquiring multiple driving data for multiple vehicles; A step of generating speed distribution data corresponding to each of the plurality of driving data; A step of generating a plurality of features related to a driving pattern based on speed distribution data of each of the plurality of driving data; A step of clustering the plurality of features into a plurality of clusters based on a clustering algorithm and an index for evaluating the clustering results; A step of generating a driving pattern kernel representing the driving pattern of each of the plurality of clusters; and A method for analyzing a driving pattern of a vehicle, comprising a step of determining a driving pattern of a feature generated after the plurality of driving pattern kernels are generated based on the similarity with the plurality of generated driving pattern kernels.

15. A non-transitory computer-readable recording medium having recorded thereon a program for executing the method of Article 14 on a computer.

Images