User portraying method and device, equipment and storage medium

Abstract

The invention provides a user portraying method, device and equipment and a storage medium, and relates to the technical field of computers, a charging behavior user portrait is constructed through clustering of two stages of charging behaviors and user types, and the method and the device have important significance for an electric power company to analyze the charging behaviors of electric vehicle users and formulate reasonable differentiated regulation and control measures. The method comprises the following steps: acquiring charging data of a charging station in a target area to obtain multiple pieces of charging data; according to the charging information, performing first clustering on the plurality of pieces of charging data to obtain a first clustering result; historical charging behavior information of the multiple target electric vehicle users is obtained, second clustering is conducted on the multiple target electric vehicle users according to the historical charging behavior information, a second clustering result is obtained, and portraying is conducted on the multiple target electric vehicle users according to the second clustering result; one cluster in the second clustering result corresponds to one user portrait.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a user portrait method and device, equipment and storage medium. BACKGROUND

[0002] The rapid development of electric vehicles powered by clean energy has great significance for promoting environmental sustainable development.

[0003] However, with the large-scale growth of electric vehicles, the disorderly charging of electric vehicles has brought many challenges to the power grid. For example, due to the disorderly charging of electric vehicles, it is difficult to regularly extract, leading to difficulties in power deployment of the power grid, and the power cannot be reasonably and economically distributed, which may further lead to the power grid unable to safely and stably operate.

[0004] In order to analyze the operation data of the charging station and analyze the charging behavior of the electric vehicle user, it is necessary to establish a user portrait method for the charging behavior of the electric vehicle user, so as to abstract each specific information of the user into a label, and to make the user image specific, and then to provide targeted services. SUMMARY

[0005] Based on the above technical problems, the present application provides a user portrait method, device, equipment and storage medium, which constructs a charging behavior user portrait through clustering of charging behavior and user type, and has important significance for the power company to analyze the charging behavior of the electric vehicle user and develop reasonable differentiated control measures.

[0006] In a first aspect, the present application provides a user portrait method, which comprises: obtaining charging data of charging stations in a target area to obtain a plurality of charging data; one charging data is used to reflect the charging information corresponding to one charging of an electric vehicle user at one charging station; according to the charging information, the first clustering of the plurality of charging data is performed to obtain a first clustering result; one cluster in the first clustering result is used to reflect a typical charging behavior; obtaining historical charging behavior information of a plurality of target electric vehicle users; one historical charging behavior information is used to reflect the number of times of corresponding each typical charging behavior of one target electric vehicle user in a historical time period; according to the historical charging behavior information, the second clustering of the plurality of target electric vehicle users is performed to obtain a second clustering result, and the plurality of target electric vehicle users are portraited according to the second clustering result; one cluster in the second clustering result corresponds to one user portrait.

[0007] In a possible implementation manner, the charging information includes a charging start time, a charging duration, a charging amount, a charging time interval, a charging location interval, and a charging station type; the first clustering is performed on the plurality of charging data according to the charging information, to obtain a first clustering result, including: taking the charging start time, the charging duration, the charging amount, the charging time interval, the charging location interval, and the charging station type as a feature vector of the charging data, and performing the first clustering on the plurality of charging data based on a K-means algorithm optimized by a particle swarm, to obtain the first clustering result; a distance between nodes in any one cluster in the first clustering result is less than or equal to a first threshold value, and the charging start time, the charging duration, the charging amount, the charging time interval, the charging location interval, and the charging station type corresponding to the same cluster in the first clustering result are the same.

[0008] In a possible implementation manner, the second clustering is performed on the plurality of target electric vehicle users according to the historical charging behavior information, to obtain a second clustering result, including: taking a number of times of each typical charging behavior of the target electric vehicle user in a historical time period as a feature vector of the target electric vehicle user, and performing the second clustering on the plurality of target electric vehicle users based on a K-means algorithm optimized by a particle swarm, to obtain the second clustering result; a distance between nodes in any one cluster in the second clustering result is less than or equal to a second threshold value.

[0009] In a possible implementation manner, the plurality of target electric vehicle users are profiled according to the second clustering result, including: for any one cluster in the second clustering result, determining a probability of each typical charging behavior corresponding to the cluster, and determining a user profile corresponding to the cluster according to a typical charging behavior with the highest probability.

[0010] In a possible implementation manner, the probability of each typical charging behavior corresponding to the cluster is determined, including: creating a probability density function by using a kernel density estimation algorithm, and calculating the probability of each typical charging behavior corresponding to the cluster by using the probability density function; the probability density function is

[0011] wherein, n is a sample size, h is a bandwidth, K (·) is a kernel function, and the kernel function is determined by an Epanechnikov kernel.

[0012] The technical scheme provided by the application brings at least the following beneficial effects: the application provides a portrait description method of electric vehicle users based on two-stage clustering of charging behavior-user type. First, parameters for describing the charging behavior of electric vehicle users are extracted from the charging session records recorded by the charging station. Second, a two-stage clustering framework based on charging behavior-user type is constructed. In the first stage, typical charging behavior is identified according to the distribution of charging behavior parameters of the user group. In the second stage, the historical typical charging behavior of the user is used as a label to cluster and divide the user type. This method does not require obtaining the privacy data of electric vehicle users, and is of great significance for power companies to analyze the charging behavior of electric vehicle users and develop reasonable differentiated control measures. The application avoids the use of complex prediction models by using simple and easily available but complex and incomprehensible data, and does not require additional management and metering equipment in the actual management process, and effectively expresses the charging station and charging load profile.

[0013] In a second aspect, the application provides a user portrait device, which comprises an acquisition unit and a processing unit; the acquisition unit is configured to acquire charging data of charging stations in a target area, and obtain a plurality of charging data; one charging data is used to reflect the charging information corresponding to one charging of an electric vehicle user at one charging station; the processing unit is configured to perform first clustering on the plurality of charging data according to the charging information, and obtain a first clustering result; one cluster in the first clustering result is used to reflect a typical charging behavior; the acquisition unit is further configured to acquire historical charging behavior information of a plurality of target electric vehicle users; one historical charging behavior information is used to reflect the number of times of corresponding typical charging behaviors of one target electric vehicle user in a historical time period; the processing unit is further configured to perform second clustering on the plurality of target electric vehicle users according to the historical charging behavior information, obtain a second clustering result, and portrait the plurality of target electric vehicle users according to the second clustering result; one cluster in the second clustering result corresponds to one user portrait.

[0014] In a possible implementation manner, the charging information comprises charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type; the processing unit is specifically configured to: take the charging start time, the charging duration, the charging power, the charging time interval, the charging location interval, and the charging station type as a feature vector of the charging data, and perform first clustering on the plurality of charging data based on a K-means algorithm optimized by a particle swarm, to obtain the first clustering result; the distance between any two nodes in any cluster in the first clustering result is less than or equal to a first threshold value, and the charging start time, the charging duration, the charging power, the charging time interval, the charging location interval, and the charging station type corresponding to the same cluster in the first clustering result are the same.

[0015] In a possible implementation, the processing unit is specifically configured to: take the number of times of each typical charging behavior of the target electric vehicle user in a historical time period as a feature vector of the target electric vehicle user, and perform second clustering on the plurality of target electric vehicle users based on a K-means algorithm optimized by a particle swarm to obtain a second clustering result; the distance between each node in any cluster in the second clustering result is less than or equal to a second threshold.

[0016] In a possible implementation, the processing unit is specifically configured to: for any cluster in the second clustering result, determine the probability of each typical charging behavior corresponding to the cluster, and determine a user portrait corresponding to the cluster according to the typical charging behavior with the highest probability.

[0017] In a possible implementation, the processing unit is specifically configured to: create a probability density function by using a kernel density estimation algorithm, and calculate the probability of each typical charging behavior corresponding to the cluster by using the probability density function; the probability density function is

[0018] wherein, n is a sample size, h is a bandwidth, K (·) is a kernel function, and the kernel function is determined by an Epanechnikov kernel.

[0019] In a third aspect, the present application provides an electronic device, including: a processor and a memory; the memory stores instructions executable by the processor; and the processor is configured to execute the instructions, so that the electronic device implements the method of the first aspect.

[0020] In a fourth aspect, the present application provides a computer program product, when the computer program product runs in an electronic device, makes the electronic device execute the method related to the first aspect, so as to implement the method of the first aspect.

[0021] In a fifth aspect, the present application provides a computer readable storage medium, including: software instructions; when the software instructions run in an electronic device, make the electronic device implement the method of the first aspect.

[0022] The beneficial effects of the second aspect to the fifth aspect can refer to the first aspect, which will not be described here. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0024] Figure 1 A structural schematic diagram of an electric vehicle charging system provided by an embodiment of the present application is shown in the figure. Figure 2 A component schematic diagram of an electronic device provided by an embodiment of the present application is shown in the figure. Figure 3 A flow schematic diagram of a user portrait method provided by an embodiment of the present application is shown in the figure. Figure 4 A test effect diagram of a K-means algorithm based on PSO improvement provided by an embodiment of the present application is shown in the figure. Figure 5 A typical charging behavior classification schematic diagram provided by an embodiment of the present application is shown in the figure. Figure 6 A probability schematic diagram of each typical charging behavior provided by an embodiment of the present application is shown in the figure. Figure 7 A component schematic diagram of a user portrait device provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0025] In order to make the ordinary person skilled in the art better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0027] In addition, in the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B. "And / or" herein is merely a description of the relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can mean that A exists alone, A and B exist together, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0028] Before the embodiments of the present application are explained in detail, some related terms and related technologies involved in the embodiments of the present application are introduced.

[0029] User portrait: a tagged user model constructed through multi-dimensional data such as user attributes, behavior records, and consumption habits, aiming to describe the key characteristics of the target group in a visual way.

[0030] Electric vehicles: mainly including pure electric vehicles (BEV), hybrid electric vehicles (HEV), plug-in hybrid electric vehicles (PHEV), extended-range electric vehicles (EREV), and fuel cell electric vehicles (FCEV). Among them, BEV relies entirely on battery power, while PHEV and EREV combine fuel and electric drive, and FCEV generates electricity through hydrogen-oxygen chemical reaction.

[0031] Charging stations: as a key link in the electric vehicle ecosystem, charging stations are showing a trend of technological upgrading, policy driving, and scene diversification. In the future, super-fast charging, intelligentization, and vehicle-network interaction will become the development direction, while solving the problem of uneven distribution and profitability is the key to the sustainable development of the industry. For users, factors such as power, location, and cost need to be considered when choosing a charging station; for the industry, technological innovation and policy support are still the core driving force to promote the perfection of charging infrastructure.

[0032] The rapid development of electric vehicles powered by clean energy has great significance in promoting environmental sustainability. However, the large-scale unordered access of electric vehicles to the power grid will cause problems such as load growth, decline in power quality, and increased difficulty in optimizing control of power grid operation, affecting the safe and stable operation of the power grid. Electric vehicle charging and discharging behavior characterization is of great significance to distribution network planning and operation as a basis for analyzing the impact of electric vehicles on the power grid.

[0033] With the large-scale growth of electric vehicles, the unordered charging of electric vehicles has brought many challenges to the power grid. Due to the unordered charging of electric vehicles, it is difficult to extract regular patterns, leading to difficulties in power allocation, and the power grid cannot be operated safely and stably. In order to analyze the charging behavior of electric vehicles by mining and analyzing the operation data of charging stations, it is urgent to establish a user portrait method for the charging behavior of electric vehicles in residential areas, abstract each specific information of the user into a tag, and provide targeted services.

[0034] However, there is less research on electric vehicle charging station user portrait in related technologies, and related research often uses clustering analysis methods, mainly including K-Means clustering, K-Medoids clustering, and hierarchical clustering methods. However, these clustering analysis methods have problems such as inflexible cluster shape, high sensitivity to outliers, and unsuitability for large-scale data sets, leading to inaccurate electric vehicle charging station user portrait and insufficient detailed description of user charging behavior.

[0035] In view of the above problems, the user portrait method provided in the embodiments of the present application is used to analyze the charging behavior of electric vehicle users by the clustering of the two stages of charging behavior and user type, and to construct a charging behavior user portrait, which is of great significance for electric power companies to formulate reasonable differentiated control measures.

[0036] The user portrait method provided in the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0037] The user portrait method provided in the embodiments of the present application can be applied to an electric vehicle charging system, Figure 1 A structural schematic diagram of the electric vehicle charging system is shown. As shown in the figure, Figure 1 The electric vehicle charging system 10 includes a user portrait device 11, a charging station management server 12, and a plurality of charging stations 13. The user portrait device 11 and the charging station management server 12 are connected in a wired or wireless manner, and the charging station management server 12 and the plurality of charging stations 13 are connected in a wired or wireless manner.

[0038] Each charging station 13 can be used to provide charging services for electric vehicle users, and the charging station management server 12 can be used to store the charging data of each charging station 13. In addition, the charging station management server 12 can also be used to provide business support for each charging station 13, for example, the charging station management server 12 can provide charging services, charging services, etc. for each charging station 13.

[0039] The user portrait device 11 can be used to obtain the charging data of each charging station 13 from the charging station management server 12, and to obtain the user portrait of the electric vehicle user according to the charging data. The specific user portrait process can refer to the user portrait method described in the following method embodiment, which will not be described here.

[0040] The user portrait device 11 can be any electronic device with data processing function, for example, the user portrait device 11 can be a server, a computer, or a server cluster composed of multiple servers. In some embodiments, the server cluster can also be a distributed cluster. Optionally, the server can be a central server, and the server can also be implemented on a cloud platform, for example, the cloud platform can include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an inter-cloud, and a multi-cloud, etc. or any combination thereof. The embodiments of the present application do not make any limitation.

[0041] The user terminal 12 can be a mobile phone, a vehicle-mounted terminal, a tablet computer, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (PDA), a desktop computer, an edge server, etc. The embodiments of the present application do not limit the specific type of the user terminal 12.

[0042] It should be noted that the above Figure 1 The information extraction device 11 and the user terminal 12 can also be combined into one device. For example, the information extraction device 11 or its corresponding function, and the user terminal 12 or its corresponding function can be integrated into one device. The embodiments of the present application do not limit this.

[0043] The execution subject of the user portrait method provided by the embodiments of the present application can be the user portrait device 11 described above. As described above, the user portrait device 11 can be a computer or a server, etc. electronic device with data processing function. Alternatively, the user portrait device 11 can also be a processor (for example, a central processing unit (CPU)) in the foregoing electronic device; or the user portrait device 11 can also be an application (application, APP) with model training function installed in the foregoing electronic device; or the user portrait device 11 can also be a functional module with model training function in the foregoing electronic device, etc. The embodiments of the present application do not limit this.

[0044] For simplicity of description, the following will be described by taking the user portrait device 11 as an electronic device.

[0045] Figure 2 The composition of the electronic device provided by the embodiments of the present application is shown in the following schematic diagram. Figure 2 As shown in the schematic diagram, the electronic device can include a processor 20, a memory 21, a communication line 22, a communication interface 23, and an input / output interface 24.

[0046] The processor 20, the memory 21, the communication interface 23, and the input / output interface 24 can be connected through the communication line 22.

[0047] The processor 20 is configured to execute instructions stored in the memory 21 to implement the fault analysis method provided in the embodiments of the present application. The processor 20 can be a CPU, a general-purpose processor, a network processor (NP), a digital signal processor (DSP), a microprocessor, a micro control unit (MCU), a programmable logic device (PLD) or any combination thereof. The processor 20 can also be any device or apparatus with processing capabilities, such as a circuit, a device or a software module, which is not limited in the embodiments of the present application. In an example, the processor 20 can include one or more CPUs, such as the CPU0 and the CPU1 in FIG. 1. Figure 2 As an optional implementation, the electronic device can include a plurality of processors, for example, in addition to the processor 20, the processor 25 (illustrated by a dashed line in FIG. 1) can also be included. Figure 2

[0048] The memory 21 is configured to store instructions. For example, the instructions can be a computer program. Alternatively, the memory 21 can be a read-only memory (ROM) or other type of static storage device that can store static information and / or instructions, or a random access memory (RAM) or other type of dynamic storage device that can store information and / or instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magneto-optical disk, a magnetic disk storage or other magnetic storage devices, etc., which is not limited in the embodiments of the present application.

[0049] It should be noted that the memory 21 can exist independently of the processor 20, or can be integrated with the processor 20. The memory 21 can be located in the electronic device, or can be located outside the electronic device, which is not limited in the embodiments of the present application.

[0050] The communication line 22 is configured to transmit information between components included in the electronic device.

[0051] ​The communication interface 23 is configured to communicate with other devices (e.g., the image acquisition apparatus 100) or other communication networks. The other communication networks can be an Ethernet, a radio access network (RAN), a wireless local area network (WLAN), or the like. The communication interface 23 can be a module, a circuit, a transceiver, or any device capable of communication.

[0052] The input / output interface 24 is configured to realize the human-computer interaction between the user and the electronic device. For example, the input / output interface 24 is configured to realize the action interaction or the information interaction between the user and the electronic device.

[0053] For example, the input / output interface 24 can be a mouse, a keyboard, a display screen, or a touch display screen, or the like. The action interaction or the information interaction between the user and the electronic device can be realized through the mouse, the keyboard, the display screen, or the touch display screen, or the like.

[0054] It should be noted that, Figure 2 The structure shown in FIG. 1 does not constitute a limitation on the electronic device, and the electronic device can include more or fewer components than those shown in the figure, or a combination of some components, or a different arrangement of components. Figure 2 The electronic device can include more or fewer components than those shown in the figure, or a combination of some components, or a different arrangement of components.

[0055] The user portrait method provided by the embodiment of the present application is described below.

[0056] Figure 3 The flowchart of the user portrait method provided by the embodiment of the present application is shown in FIG. 3. Optionally, the method can be executed by the electronic device with the hardware structure shown in FIG. 1, and the method includes S301 to S303. Figure 2 Figure 3 The method includes S301 to S303.

[0057] S301, obtaining charging data of charging stations in a target region, to obtain a plurality of charging data.

[0058] Each charging data is used to reflect the charging information corresponding to one charging of one electric vehicle user at one charging station. The target region can be pre-set by an operation and maintenance personnel. For example, the target region can be a city or a county, or the like.

[0059] In actual application, after the electric vehicle user charges at the charging station, the charging station generates the session record of the user charging, and uploads the session record to the background server, such as a charging station management server, and saves the session record by the charging station management server. The charging data can be extracted from the session record.

[0060] ​In some embodiments, electronic devices can collect user charging session records from charging stations. These charging session records, derived from charging station records and cost metering devices, typically include information such as the user's identity document (ID), charging start and end time, delivered energy, and cost. Furthermore, pre-processing is required during the collection process, including charging sessions that are abnormally interrupted due to charging device failures and charging sessions that are incorrectly recorded due to device failures, to avoid misidentification of charging behavior and user characteristics.

[0061] As you can understand, the raw data in this embodiment of the application is collected by recording devices installed by charging station operators. Each charging session record contains a unique index number corresponding to the charging station and charging port, as well as the user's charging behavior parameters. This eliminates the need to collect data that could potentially compromise user privacy, which is crucial for protecting user safety.

[0062] In other embodiments, the charging behavior description system must fully describe the load characteristics of electric vehicles, including spatial mobility, temporal flexibility, and power uncertainty, as well as user site selection preferences to assist in business decision-making. Therefore, in this embodiment, the following six parameters can be selected to describe charging behavior: charging start time, charging duration, charging power, charging interval, charging spatial distribution, and charging station type preference.

[0063] Specifically, the charging start time is used to describe the start time of a user's charging behavior. The format of this data can be set to time format (such as hh:mm). The charging duration is used to describe the duration of the energy delivered by the user's charging behavior. The format of this data can be set to time format (such as hh:mm). The charging power is used to describe the power required by the user's charging behavior. The format of this data can be set to floating point number. The charging interval is used to describe the length of the time interval between two charging behaviors of the user. The format of this data can be set to time format (such as hh:mm). The charging location interval is used to describe the spatial location interval between two charging behaviors of the user. It is calculated based on the longitude and latitude of the charging station. The format of this data can be set to floating point number. The charging station selection preference is used to describe the type of charging station selected by the user each time for charging, including fast charging station (marked as FC), slow charging station (marked as AC), highway charging pile (marked as HW), and intelligent charging station (marked as IC). The format of this data can be set to text type.

[0064] As described above, the six parameter systems describing the charging behavior are heterogeneous data, which need to be transformed into a unified format to quantify the charging behavior. For data in time format, a floating-point number in hours is used for transformation, and data in text format is processed using integer encoding. Therefore, the charging information extracted by the embodiments of the present application can include charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type.

[0065] In S302, first clustering is performed on the plurality of charging data according to the charging information, and a first clustering result is obtained.

[0066] In the first clustering result, one cluster is used to reflect a typical charging behavior.

[0067] In some embodiments, the charging information can include charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type. The electronic device can use the charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type as the feature vector of the charging data, and perform first clustering on the plurality of charging data based on the K-means algorithm of particle swarm optimization, to obtain the first clustering result. In the first clustering result, the distance between any two nodes in any cluster is less than or equal to a first threshold, and the charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type corresponding to the same cluster in the first clustering result are the same.

[0068] It should be noted that the particle swarm optimization (PSO) algorithm is a global random search evolutionary algorithm based on swarm intelligence, which is derived from the study of bird hunting behavior. The algorithm simulates a bird swarm by designing a particle swarm with no mass, no volume, and only two attributes of speed and position. The speed of the i-th particle is represented as and the position is represented as Each particle in the particle swarm represents a feasible solution in the optimization problem. In the improvement of the K-means algorithm, the clustering center to be found is mapped to the particle of the PSO algorithm, i.e., the position of each particle is the n clustering center.

[0069] Specifically, a group of particles search for the target function in the solution space at a certain speed. Each particle has a position X i , a speed V i and a personal best Pb i , and the entire group has a global best position Gb:

[0070]

[0071]

[0072]

[0073] where i = 1, 2,..., N represents the number of particles, and d represents the dimension of the solution space. In finding the optimal solution of a problem, particles are iteratively updated in the solution space, whose position is Xi, based on the following motion equation:

[0074] where t and t+1 represent two consecutive iterations of the algorithm; V i is the vector of velocity components along d dimensions that collects the i-th particle. The velocity vector determines the way the particle moves in the search space, which consists of three components: the first is defined as inertia, which prevents the particle from changing direction sharply by tracking the previous direction of motion; the second is defined as the cognitive component, which describes the tendency of the particle to return to the best position it has ever reached; the last one, defined as the social component, describes the tendency of the particle to move towards the best position of the entire swarm. Based on these considerations, the velocity update of the i-th particle is defined as

[0075] where Pb i is the best position of the particle i, indicates the coordinates of the particle that obtain the best position, and Gb is the global best position, which indicates the coordinates of the swarm that obtain the best position. The inertia coefficient w i , the cognitive coefficient c1 and the social coefficient c2 are real values that regulate the variation of the particle velocity vector in the three components. R1 and R2 are two diagonal random number matrices generated by uniform distribution in [0, 1]. Therefore, the trajectory of the particle is essentially semi-random, which improves the searchability of the particle in the search space.

[0076] In the context of clustering, a particle represents a vector of K cluster centroids. That is, each particle X i is constructed as follows:

[0077] where c ij is the j-th (j = 1, 2,..., K, K is the number of clusters) cluster centroid vector of the i-th particle in the cluster C j Therefore, a cluster represents several candidate clusters for the current data vector. The fitness of the particle swarm can be easily measured as the quantization error. The measurement formula of the PSO-K-means optimization objective function is as follows:

[0078] where Z p is the pth data sample; d(x, y) is the distance between vectors x and y in Euclidean distance; |C j | is the number of sampling points in cluster Cj.

[0079] The test results of the K-means algorithm improved based on PSO are shown in Figure 4 It can be seen from Figure 4 that the positions of the centroid vector are always generated around the best solution of the previous iteration. As the number of iterations increases, these positions are closer and closer to the best solution. In particular, these positions do not overlap with the best solution, so it avoids the ability to stick to a local optimum. These centroid vector positions establish a new position update trend together with the optimal solution.

[0080] In practical applications, the electronic device can construct a charging behavior feature vector according to the above six parameters describing the charging behavior for each electric vehicle user i, and perform PSO-Kmeans clustering on all charging processes to complete the first stage clustering of the embodiments of the present application.

[0081] As shown in Figure 5 , after obtaining the charging data of each charging station in City A, the K-means algorithm based on particle swarm optimization is used to classify the charging features, and 9 clusters are obtained, which correspond to 9 typical charging behaviors in Figure 5 (in Figure 5 WD-1 to WD-9). Among them, Figure 5 Part a is used to reflect the characteristics corresponding to the 9 typical charging behaviors in the charging start time, WD-1 represents the charging start time as 8:00-10:00, WD-2 represents the charging start time as 00:00-6:00, …, and WD-9 represents the charging start time as 20:00-24:00. Figure 5 Part b is used to reflect the characteristics corresponding to the 9 typical charging behaviors in the charging duration, WD-1 represents the charging duration as half an hour (h), WD-2 represents the charging duration as 4 hours, …, and WD-9 represents the charging duration as 8 hours. Figure 5 Part c is used to reflect the characteristics corresponding to the 9 typical charging behaviors in the charging power, WD-1 represents the charging power as 40 kilowatts (kW), WD-2 represents the charging power as 60 kilowatts, …, and WD-9 represents the charging power as 100 kilowatts. Figure 5 Part d is used to reflect the characteristics corresponding to the 9 typical charging behaviors in the charging time interval, WD-1 represents the charging time interval as 1 day (day), WD-2 represents the charging time interval as 2.5 days, …, and WD-9 represents the charging time interval as 5 days. Figure 5The e part of the formula is used to reflect the corresponding characteristics of the nine typical charging behaviors on the charging location interval, WD-1 represents that the charging location interval is 100 kilometers, WD-2 represents that the charging location interval is 40 kilometers, and WD-9 represents that the charging location interval is 10 kilometers. Figure 5 The f part of the formula is used to reflect the corresponding characteristics of the nine typical charging behaviors on the charging station type, WD-1 represents that the charging station type is fast charging, WD-2 represents that the charging station type is intelligent charging, and WD-9 represents that the charging station type is slow charging.

[0082] S303, obtain historical charging behavior information of a plurality of target electric vehicle users.

[0083] Wherein, one historical charging behavior information of a plurality of target electric vehicle users is used to reflect the number of times of corresponding each typical charging behavior of one target electric vehicle user in a historical time period. The target electric vehicle user can be a user who has charging experience at the charging station in the target area. The number of target electric vehicle users to be obtained can be pre-set by the operation and maintenance personnel, and the present application does not limit this.

[0084] As a possible implementation manner, for any target electric vehicle user, the electronic device can determine the corresponding typical charging behavior of the charging experience according to the charging data corresponding to each charging, so as to obtain the number of times of corresponding each typical charging behavior of the target electric vehicle user in the historical time period.

[0085] For example, the parameters in the charging data corresponding to the completion of a charging of a certain target electric vehicle user all meet the first typical charging behavior (i.e. WD-1), and the number of times of the first typical charging behavior of the target electric vehicle user is recorded as one. Similarly, if the parameters in the charging data corresponding to the completion of a charging of the target electric vehicle user all meet the second typical charging behavior (i.e. WD-2), the number of times of the second typical charging behavior of the target electric vehicle user is recorded as one. After checking all the historical charging data of the target electric vehicle user, the number of times of corresponding each typical charging behavior is calculated, and then the historical charging behavior information of the target electric vehicle user is obtained.

[0086] As another possible implementation manner, the charging station management server can also pre-statistically obtain the number of times of corresponding each typical charging behavior of each target electric vehicle user in the historical time period, and store it. Further, the electronic device can directly obtain the historical charging behavior information of a plurality of target electric vehicle users from the charging station management server.

[0087] S304, according to the historical charging behavior information, a second clustering is performed on the plurality of target electric vehicle users, a second clustering result is obtained, and a plurality of target electric vehicle users are profiled according to the second clustering result.

[0088] Among them, one cluster in the second clustering result corresponds to a user portrait.

[0089] As a possible implementation method, the electronic device can use the number of times the target electric vehicle user corresponds to each typical charging behavior in a historical time period as the feature vector of the target electric vehicle user, and perform a second clustering on multiple target electric vehicle users based on the K-means algorithm of particle swarm optimization to obtain a second clustering result; wherein the distance between each node in any cluster in the second clustering result is less than or equal to the second threshold.

[0090] In some embodiments, for any cluster in the second clustering results, the electronic device may determine the probability of each typical charging behavior corresponding to the cluster, and determine the user portrait corresponding to the cluster based on the typical charging behavior with the highest probability.

[0091] Optionally, determining the probability of each typical charging behavior corresponding to the cluster includes: creating a probability density function through a kernel density estimation algorithm, and calculating the probability of each typical charging behavior corresponding to the cluster using the probability density function; the probability density function is

[0092] in, is the probability density function, is any sample data, x is any point data, is the probability density corresponding to a bandwidth, n is the sample size, h is a smoothing parameter called bandwidth, and h>0, K (⋅) is the kernel function, which is determined by the Epanechnikov kernel.

[0093] In practical applications, electronic devices can be used to distinguish the types of electric vehicle users based on their historical charging behavior clusters. The feature vector of the user cluster is constructed as the number of times the user used each charging behavior during the study period, which can be expressed as:

[0094] Among them, n i WdBj (j = 1, 2, ...) and n i WeBj (k = 1, 2, ..) represents the number of typical charging behaviors of user i on working days and non-working days in the historical time period.

[0095] After constructing the user behavior feature label of electric vehicle, the electronic device can perform the K-means algorithm based on particle swarm optimization clustering to complete the second stage clustering of the behavior-user framework according to the feature. Further, the electronic device can create a probability model according to the number of times of using the charging behavior of each type of user. In order to deal with the unknown nature of the user's use of the charging behavior, kernel density estimation is used, which avoids any prior assumptions about the distribution of data. The basic idea of kernel density estimation (KDE) is to regard each observed sample point as a "kernel" (usually a probability distribution function such as a Gaussian distribution), and then superimpose these kernels to form a smooth probability density estimator. The specific probability model can be the above probability density function. Through a large number of experiments, the embodiment of the present application determines the kernel function as the Epanechnikov kernel, so that the error is significantly reduced.

[0096] As shown in Figure 6 , it is the probability of each typical charging behavior corresponding to each cluster of the second clustering result of the embodiment of the present application. Among them, the electronic device performs second clustering on a plurality of target electric vehicle users, and obtains 12 clusters in the second clustering result, which respectively correspond to Figure 6For the user type 1-12 in the user profile database, each user type corresponds to a user portrait. For the user type 1, the method provided in the above embodiment is used to sample the probability of each typical charging behavior, and it is found that the probability of using the first typical charging behavior (i.e., WD-1) is the largest. Therefore, the electronic device can use the first typical charging behavior (i.e., WD-1) to describe the user portrait of the user of this type, for example, the user portrait of this type can be described as “charging start time is between 8:00-10:00”, “charging duration is half an hour”, “charging power is 40 kilowatts”, “charging time interval is 1 day”, “charging location interval is 100 kilometers”, “charging station type is fast charging”, or the user portrait of this type can also be described as “emergency charging user”. For the user type 2, the method provided in the above embodiment is used to sample the probability of each typical charging behavior, and it is found that the probability of using the second typical charging behavior (i.e., WD-2) is the largest. Therefore, the electronic device can use the second typical charging behavior (i.e., WD-2) to describe the user portrait of the user of this type, for example, the user portrait of this type can be described as “charging start time is between 00:00-6:00”, “charging duration is 4 hours”, “charging power is 60 kilowatts”, “charging time interval is 2.5 days”, “charging location interval is 40 kilometers”, “charging station type is smart charging”, or the user portrait of this type can also be described as “non-emergency charging user”. For the user type 3, the method provided in the above embodiment is used to sample the probability of each typical charging behavior, and it is found that the probability of using the ninth typical charging behavior (i.e., WD-9) is the largest. Therefore, the electronic device can use the ninth typical charging behavior (i.e., WD-9) to describe the user portrait of the user of this type, for example, the user portrait of this type can be described as “charging start time is between 20:00-24:00”, “charging duration is 8 hours”, “charging power is 100 kilowatts”, “charging time interval is 5 days”, “charging location interval is 100 kilometers”, “charging station type is slow charging”, or the user portrait of this type can also be described as “stable charging user”.

[0097] The technical scheme provided by the application has at least the following beneficial effects: the application provides a portrait description method of electric vehicle users based on two-stage clustering of charging behavior-user type. First, parameters for describing the charging behavior of electric vehicle users are extracted from the charging session records recorded by the charging station. Second, a two-stage clustering framework based on charging behavior-user type is constructed, and the typical charging behavior is identified in the first stage according to the charging behavior parameter distribution of the user group. In the second stage, the historical typical charging behavior of the user is used as a label to cluster and divide the user type. The method does not need to obtain the private data of the electric vehicle user, and is of great significance for the power company to analyze the charging behavior of the electric vehicle user and develop reasonable differentiated control measures. The application avoids the use of complex prediction models through simple and easily available but complex and incomprehensible data, and does not need to increase additional management and metering equipment in the actual management process, and effectively expresses the charging station and charging load profile.

[0098] The above mainly introduces the scheme provided by the embodiments of the application from the perspective of the method. In order to realize the above functions, it contains the hardware structure and / or software module corresponding to the execution of each function. The skilled in the art should easily realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present application can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the form of hardware or computer software driven hardware depends on the specific application and design constraints of the technical scheme. The skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the application.

[0099] In an exemplary embodiment, the embodiments of the application also provide a user portrait device. Figure 7 The composition schematic diagram of the user portrait device provided by the embodiments of the application is shown in FIG. 1. As shown in the figure, the user portrait device includes an acquisition unit 401 and a processing unit 402. Figure 7

[0100] ​The acquisition unit 401 is configured to acquire charging data of charging stations in a target region, to obtain a plurality of charging data; one charging data is used to reflect charging information corresponding to one charging of one electric vehicle user at one charging station; the processing unit 402 is configured to perform first clustering on the plurality of charging data according to the charging information, to obtain a first clustering result; one cluster in the first clustering result is used to reflect one typical charging behavior; the acquisition unit 401 is further configured to acquire historical charging behavior information of a plurality of target electric vehicle users; one historical charging behavior information is used to reflect a number of times of corresponding each typical charging behavior of one target electric vehicle user in a historical time period; the processing unit 402 is further configured to perform second clustering on the plurality of target electric vehicle users according to the historical charging behavior information, to obtain a second clustering result, and to profile the plurality of target electric vehicle users according to the second clustering result; one cluster in the second clustering result corresponds to one user portrait.

[0101] In a possible implementation, the charging information includes a charging start time, a charging duration, a charging electric quantity, a charging time interval, a charging location interval, and a charging station type; the processing unit 402 is specifically configured to: take the charging start time, the charging duration, the charging electric quantity, the charging time interval, the charging location interval, and the charging station type as a feature vector of the charging data, and perform first clustering on the plurality of charging data based on a K-means algorithm optimized by a particle swarm, to obtain the first clustering result; a distance between each node in any one cluster in the first clustering result is less than or equal to a first threshold value, and the charging start time, the charging duration, the charging electric quantity, the charging time interval, the charging location interval, and the charging station type corresponding to the same cluster in the first clustering result are the same.

[0102] In a possible implementation, the processing unit 402 is specifically configured to: take the number of times of corresponding each typical charging behavior of the target electric vehicle user in the historical time period as a feature vector of the target electric vehicle user, and perform second clustering on the plurality of target electric vehicle users based on a K-means algorithm optimized by a particle swarm, to obtain the second clustering result; a distance between each node in any one cluster in the second clustering result is less than or equal to a second threshold value.

[0103] In a possible implementation, the processing unit 402 is specifically configured to: for any one cluster in the second clustering result, determine a probability of each typical charging behavior corresponding to the cluster, and determine a user portrait corresponding to the cluster according to a typical charging behavior with the highest probability.

[0104] In a possible implementation, the processing unit 402 is specifically configured to: create a probability density function by using a kernel density estimation algorithm, and calculate the probability of each typical charging behavior corresponding to the cluster by using the probability density function; the probability density function is

[0105] wherein, n is the sample size, h is the bandwidth, K (·) is a kernel function, the kernel function being determined by an Epanechnikov kernel.

[0106] It should be noted that, Figure 7 The division of the modules in the above description is illustrative only, and is merely a logical function division. In actual implementation, another division manner can be used. For example, two or more functions can be integrated into one processing module. The integrated module can be implemented in the form of hardware or in the form of a software function unit.

[0107] In the example embodiments, the embodiments of the present application further provide a computer readable storage medium including software instructions, which, when executed on an electronic device, cause the electronic device to perform any of the methods provided by the above-described embodiments.

[0108] In the example embodiments, the embodiments of the present application further provide a computer program product including computer execution instructions, which, when executed on an electronic device, cause the electronic device to perform any of the methods provided by the above-described embodiments.

[0109] In the above-described embodiments, all or part of the embodiments can be implemented by software, hardware, firmware, or any combination thereof. When implemented by using a software program, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer execution instructions. When the computer execution instructions are loaded and executed on a computer, all or part of the processes or functions according to the embodiments of the present application are produced. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable apparatus. The computer execution instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer execution instructions can be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center through a wired (for example, coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium that can be accessed by a computer or include one or more data storage devices that can be integrated with a server, data center, etc. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a solid state disk (SSD), etc.

[0110] Although the application has been described in connection with the embodiments thereof with reference to the various drawings, it will be apparent to those of ordinary skill in the art that variations in the embodiments can be used and that it is not intended to limit the application to the particular form described. Within the scope of the appended claims, the application can be practiced otherwise than is specifically described. For example, the order of elements in the description of the embodiments can be changed. The application can be practiced in alternating or interleave the elements of the embodiments. The elements of the application can be practiced in an order different than the specific circulation of elements. Various changes can be made without departing from the spirit or scope of the application. One of ordinary skill in the art will further appreciate that the application can be implemented anywhere a processor, controller, or computer is used to process information. Accordingly, it is intended that all such alterations and modifications be considered as equivalents thereof. The application is to be limited only by the claims appended hereto and equivalents thereof.

[0111] Although the application has been described in connection with specific embodiments thereof, it will be evident that many alternatives, modifications, and variations can be made to the application set forth herein without departing from the spirit and scope thereof. Accordingly, it is intended to embrace all such alternatives, modifications, and variations as fall within the scope of the appended claims. Obviously, many modifications and variations of the present application are possible in light of the above teachings. It is, therefore, to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

[0112] The specific embodiments of the application have been disclosed herein. One of ordinary skill in the art will immediately appreciate that myriad modifications can be made to the application without departing from the scope and spirit of the application. Therefore, it is not intended that the application be limited to the specific embodiments disclosed herein, but rather that the scope of the application be defined by the following claims and their equivalents.

Claims

1. A user profiling method, characterized in that: The method comprises: Acquire charging data of charging stations in the target area to obtain multiple charging data; one charging data is used to reflect the charging information corresponding to an electric vehicle user completing a charge at a charging station; performing a first clustering on the plurality of charging data according to the charging information to obtain a first clustering result; wherein a cluster in the first clustering result is used to reflect a typical charging behavior; Acquire historical charging behavior information of multiple target electric vehicle users; each historical charging behavior information is used to reflect the number of times a target electric vehicle user has performed each typical charging behavior within a historical time period; Based on the historical charging behavior information, the multiple target electric vehicle users are subjected to a second clustering to obtain a second clustering result, and the multiple target electric vehicle users are profiled based on the second clustering result; one cluster in the second clustering result corresponds to a user profile.

2. The method according to claim 1, characterized in that The charging information includes charging start time, charging duration, charging power, charging time interval, charging location interval and charging station type; The performing a first clustering on the plurality of charging data according to the charging information to obtain a first clustering result includes: The charging start time, the charging duration, the charging power, the charging time interval, the charging location interval, and the charging station type are used as feature vectors of the charging data, and a first clustering is performed on the multiple charging data based on a K-means algorithm of particle swarm optimization to obtain a first clustering result; the distance between each node in any cluster in the first clustering result is less than or equal to a first threshold, and the charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type corresponding to the same cluster in the first clustering result are the same.

3. The method according to claim 1, characterized in that The second clustering of the plurality of target electric vehicle users according to the historical charging behavior information to obtain a second clustering result includes: The number of times the target electric vehicle user performs each typical charging behavior within a historical time period is used as a feature vector of the target electric vehicle user, and a second clustering is performed on the multiple target electric vehicle users based on a K-means algorithm using particle swarm optimization to obtain a second clustering result; the distance between each node in any cluster in the second clustering result is less than or equal to a second threshold.

4. The method according to claim 1 or 3, characterized in that The generating user portraits of the plurality of target electric vehicles according to the second clustering result includes: For any cluster in the second clustering results, the probability of each typical charging behavior corresponding to the cluster is determined, and the user portrait corresponding to the cluster is determined according to the typical charging behavior with the highest probability.

5. The method according to claim 4, characterized in that Determining the probability of each typical charging behavior corresponding to the cluster includes: A probability density function is created by a kernel density estimation algorithm, and the probability of each typical charging behavior corresponding to the cluster is calculated using the probability density function; the probability density function is in, is the probability density function, is any sample data, x is any point data, is the probability density corresponding to a bandwidth, n is the sample size, h is the bandwidth, K (⋅) is a kernel function, which is determined by the Epanechnikov kernel.

6. A user portrait device, characterized in that: The device includes an acquisition unit and a processing unit; The acquisition unit is used to acquire charging data of charging stations in the target area to obtain multiple charging data; one charging data is used to reflect the charging information corresponding to a charging completed by an electric vehicle user at a charging station; The processing unit is configured to perform a first clustering on the plurality of charging data according to the charging information to obtain a first clustering result; wherein a cluster in the first clustering result is used to reflect a typical charging behavior; The acquisition unit is further configured to acquire historical charging behavior information of a plurality of target electric vehicle users; one piece of historical charging behavior information is configured to reflect the number of times a target electric vehicle user has performed each typical charging behavior within a historical time period; The processing unit is further used to perform a second clustering of the multiple target electric vehicle users based on the historical charging behavior information to obtain a second clustering result, and to create portraits of the multiple target electric vehicle users based on the second clustering result; one cluster in the second clustering result corresponds to a user portrait.

7. The user portrait device according to claim 6, characterized in that: The charging information includes charging start time, charging duration, charging power, charging time interval, charging location interval and charging station type; the processing unit is specifically configured to: The charging start time, the charging duration, the charging power, the charging time interval, the charging location interval, and the charging station type are used as feature vectors of the charging data, and a first clustering is performed on the multiple charging data based on a K-means algorithm of particle swarm optimization to obtain a first clustering result; the distance between each node in any cluster in the first clustering result is less than or equal to a first threshold, and the charging start time, charging duration, charging power, charging time interval, charging location interval, and charging station type corresponding to the same cluster in the first clustering result are the same.

8. The user portrait device according to claim 6, characterized in that: The processing unit is specifically configured to: The number of times the target electric vehicle user performs each typical charging behavior within a historical time period is used as a feature vector of the target electric vehicle user, and a second clustering is performed on the multiple target electric vehicle users based on a K-means algorithm using particle swarm optimization to obtain a second clustering result; the distance between each node in any cluster in the second clustering result is less than or equal to a second threshold.

9. An electronic device, characterized in that: include: processor and memory; The memory stores instructions executable by the processor; When the processor is configured to execute the instructions, the electronic device implements the method according to any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that The readable storage medium includes: software instructions; When the software instructions are executed in an electronic device, the electronic device is enabled to implement the method according to any one of claims 1 to 5.

Images