Electric vehicle user charging demand activity feature analysis method and device and medium

Abstract

The invention relates to an electric vehicle user charging demand activity feature analysis method and device and a medium, and the method comprises the steps: carrying out the normalization processing of travel behavior data and charging behavior data of an electric vehicle user, and obtaining a plurality of feature quantities; establishing a feature matrix according to the plurality of feature quantities, and performing standardization processing on the feature matrix; and k-means-GSA clustering is carried out on the standardized feature matrix, and electric vehicle user charging demand activity feature analysis is realized. The method can accurately analyze the charging demand activity characteristics of the electric vehicle user.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicles, and in particular to a method, device and medium for analyzing characteristics of charging demand activities of electric vehicle users. Background Art

[0002] An electric vehicle is a vehicle that uses an onboard power source as a power source and uses an electric motor to drive the wheels. Its working principle is to provide current through a battery, which is regulated by a power regulator to drive the electric motor, and then the vehicle is driven by a power transmission system. At present, with the gradual development of new energy technologies, electric vehicles have occupied a certain proportion of people's travel methods. Since electric vehicles have a relatively small impact on the environment, their prospects are widely optimistic. During the travel process, electric vehicles often need to be charged to supplement the electric energy of electric vehicles. Therefore, in areas where electric vehicles are widely used, a large number of charging piles and other facilities need to be set up.

[0003] In order to meet the needs of users in various regions, electric vehicle charging companies need to master the charging data of users, so it is necessary to study the behavioral characteristics of charging users. At present, correctly identifying the behavioral characteristics of charging station users or electric vehicle users is of great significance for operating companies to improve operational efficiency, enhance the quality of operation and maintenance services, and guide user charging behavior. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method, device and medium for analyzing the characteristics of charging demand activities of electric vehicle users, which can accurately analyze the characteristics of charging demand activities of electric vehicle users.

[0005] The technical solution adopted by the present invention to solve the technical problem is: to provide a method for analyzing the characteristics of charging demand activities of electric vehicle users, comprising the following steps:

[0006] Normalizing the travel behavior data and charging behavior data of electric vehicle users to obtain multiple feature quantities;

[0007] Establishing a feature matrix according to the multiple feature quantities, and performing standardization processing on the feature matrix;

[0008] The standardized feature matrix is ​​clustered using k-means-GSA to analyze the characteristics of electric vehicle users’ charging demand activities.

[0009] The characteristic quantity is one or a combination of the following: the number of times an electric vehicle stops in a day, the average start time of an electric vehicle's stops, the number of times an electric vehicle is charged in a day, the average duration of an electric vehicle's stops, the average start time of an electric vehicle's charging, the average duration of an electric vehicle's charging, the average remaining power of an electric vehicle's charging, the average power required for an electric vehicle's charging, and the average charging power of an electric vehicle's charging.

[0010] The feature matrix is ​​expressed as: Among them, Z m×n is the feature matrix, m is the total number of electric vehicles involved in clustering, and n is the number of feature quantities.

[0011] The k-means-GSA clustering of the standardized feature matrix specifically includes:

[0012] The k-means clustering algorithm is used to cluster the standardized feature matrix with different cluster numbers to obtain clustering results with different cluster numbers;

[0013] According to the clustering results of different cluster numbers by GSA cluster analysis, the gap values ​​of different cluster numbers are obtained, and the optimal number of clusters is determined according to the gap values ​​of different cluster numbers;

[0014] The standardized feature matrix is ​​classified using the k-means clustering algorithm with the optimal number of clusters to obtain activity characteristics of each type of electric vehicle user regarding the stay time, charging time and charging demand.

[0015] The optimal number of clusters k is determined by analyzing the characteristic values ​​of different numbers of clusters according to the GSA clustering, specifically:

[0016] Calculating the clustering dispersion of different numbers of clusters and the mean of the reference distribution set, and obtaining the gap values ​​when the number of clusters is different based on the clustering dispersion and the mean of the reference distribution set;

[0017] The number of clusters corresponding to the maximum gap value is selected and this number of clusters is taken as the optimal number of clusters.

[0018] The cluster dispersion is expressed by Calculated, where W k is the clustering dispersion when the number of clusters is k, |c i | is the number of data in the i-th cluster, D i is the distance between each data point in the i-th cluster.

[0019] The mean of the reference distribution set is given by E[lnW r (k)] = log (p0n / 12) - (2 / p0) log (k) + c, where E[lnWr (k)] is the mean of the reference distribution set when the number of clusters is k, p0 is the dimension of the space, n is the number of samples, and c is a constant.

[0020] The gap value is given by gap(k)=E[lnW r (k)]-ln(W k ), where gap(k) is the gap value when the number of clusters is k, and W k is the clustering dispersion when the number of clusters is k, E[lnW r (k)] is the mean of the reference distribution set when the number of clusters is k.

[0021] The technical solution adopted by the present invention to solve its technical problem is: to provide an electronic device, including a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned electric vehicle user charging demand activity characteristic analysis method when executing the computer program.

[0022] The technical solution adopted by the present invention to solve its technical problem is: providing a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned electric vehicle user charging demand activity characteristic analysis method are implemented.

[0023] Beneficial Effects

[0024] Due to the adoption of the above-mentioned technical solution, the present invention has the following advantages and positive effects compared with the prior art: the present invention determines the number of clusters through the GSA algorithm, and then performs clustering through the k-means clustering algorithm, which can accurately analyze the characteristics of electric vehicle users' charging demand activities, and provide a data basis for operating companies to improve operational efficiency, enhance the quality of operation and maintenance services, and guide user charging behavior. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flow chart of a method for analyzing characteristics of electric vehicle user charging demand activities according to a first embodiment of the present invention;

[0026] Figure 2 is a schematic diagram of the optimal number of clusters in an embodiment of the present invention;

[0027] Figure 3 is a schematic diagram of the number of electric vehicles of each type in an embodiment of the present invention;

[0028] Figure 4 Schematic diagram of the stop start time of each type of electric vehicle in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall within the scope limited by the appended claims of the application equally.

[0030] The first embodiment of the present invention relates to a method for analyzing the characteristics of electric vehicle user charging demand activities, such as Figure 1 As shown, the following steps are included:

[0031] Step 101 , normalizing the travel behavior data and charging behavior data of electric vehicle users to obtain a plurality of feature quantities.

[0032] This step first obtains the travel behavior data and charging behavior data of electric vehicle users, including: the number of stops of electric vehicles in a day, the start time of each stop, the duration of each stop, the number of charges of electric vehicles in a day, the start time of each charge, the duration of each charge, the remaining power of each charge, the required power of each charge, the charging power of each charge, the location of each stop, and the location of each charge. Since the number of stops and charges of each electric vehicle in a day is different, the problem of inconsistent data dimensions will be caused when performing cluster analysis. Therefore, this embodiment processes the relevant data of each stop or charge of the electric vehicle into the average data of one day, as follows:

[0033] Average start time of electric vehicle stop = ∑ start time of each stop / number of stops per day;

[0034] Average duration of electric vehicle stops = ∑ duration of each stop / number of stops per day;

[0035] Average start time of electric vehicle charging = ∑ start time of each charging / number of charging times per day;

[0036] Average duration of electric vehicle charging = ∑ duration of each charge / number of charges per day;

[0037] The average remaining power of electric vehicle charging = ∑ remaining power per charge / number of charges per day;

[0038] The average power demand for electric vehicle charging = ∑ power demand per charging / number of charging times per day;

[0039] Average charging power of electric vehicles = ∑ charging power per charge / number of charges per day;

[0040] Therefore, when selecting the clustering features of electric vehicle users in this step, the final selected feature quantities are one or more of the following: the number of times an electric vehicle stops in a day, the average start time of an electric vehicle stop, the number of times an electric vehicle is charged in a day, the average duration of an electric vehicle stop, the average start time of electric vehicle charging, the average duration of electric vehicle charging, the average remaining power of an electric vehicle charging, the average power required for electric vehicle charging, and the average charging power of an electric vehicle charging.

[0041] Step 102: establishing a feature matrix according to the plurality of feature quantities, and performing standardization processing on the feature matrix.

[0042] In this embodiment, the feature matrix established based on multiple feature quantities is expressed as:

[0043]

[0044] Among them, Z m×n is the feature matrix, m is the total number of electric vehicles involved in clustering, and n is the number of feature quantities.

[0045] The characteristic quantities that characterize the electric vehicle travel behavior data and charging behavior data have different dimensions, and the dispersion of the values ​​of each variable varies greatly. In order to avoid affecting the subsequent dimensionality reduction processing, this step also requires standardization of the characteristic matrix to obtain the characteristic matrix X m×n , feature matrix X m×n The element x in ij It is obtained by the following formula:

[0046]

[0047] Among them, z min,j and z max,j They represent the minimum and maximum values ​​in the j-th feature value respectively.

[0048] Step 103: normalize the feature matrix X m×n Perform k-means-GSA clustering to analyze the characteristics of electric vehicle users’ charging demand activities.

[0049] Clustering technology can automatically divide data into different categories according to the characteristic performance of each sample. At present, K-means clustering is the most widely used, convenient and fast unsupervised clustering method, but the disadvantage of unsupervised clustering is that it is necessary to formulate its classification categories. In order not to affect the classification effect due to artificial settings, this step first uses the k-means clustering algorithm to cluster the standardized feature matrix with different numbers of clusters to obtain clustering results with different numbers of clusters; then, according to the GSA clustering analysis of the clustering results with different numbers of clusters, the gap values ​​with different numbers of clusters are obtained, and the optimal number of clusters is determined according to the gap values ​​with different numbers of clusters; finally, the standardized feature matrix is ​​classified using the k-means clustering algorithm with the optimal number of clusters to obtain the activity characteristics of each type of electric vehicle user regarding the stay time, charging time and charging demand.

[0050] The principle of the k-means algorithm used in this step is to minimize the similarity of elements in different categories and maximize the similarity of elements in the same category. The similarity is measured by the average value of the data in a class.

[0051] If there are n data now, set the number of clusters to k, and this k needs to meet three conditions: first, k < n, that is, the number of clusters is less than the total number of data; second, each class after clustering cannot be an empty set, and there must be at least one number; third, in the clustering algorithm with multiple iterations, each data can only belong to one class in each iteration.

[0052] When performing clustering, first determine the initial cluster center. Usually, the k-means algorithm randomly selects k numbers as the initial cluster centers, and then uses these k data as the centers to calculate the Euclidean distance of other data to these k centers according to the following formula.

[0053]

[0054] Where: x j is the sample data, y j is the cluster center; n is the dimension of the data; d(x,y) is the sample data x j and cluster center y j The Euclidean distance.

[0055] Finally, the data with the smallest similarity are clustered into one category to obtain the final clustering result. The same steps are repeated multiple times, and the termination of the iteration is determined by checking whether the standard measurement function converges.

[0056]

[0057] Where E is the sum of the mean square error from each data to the cluster center of the belonging class; c i For each cluster; m iis the cluster center; k is the number of clusters; p refers to a data in a class.

[0058] The k-means algorithm divides the data according to the mean value of each class after classification. It needs to input the final number of clusters k and the n data to be clustered. The output is: the class of each data, and the Euclidean distance of each data in each class to the cluster center. The specific steps are as follows:

[0059] ① Randomly select k data as the initial clustering center;

[0060] ② Calculate the distance from each data to the cluster center and add it to the cluster with the smallest distance from the cluster center;

[0061] ③ Calculate the average value of each cluster as the cluster center of the new cluster, calculate the distance from each data to the new cluster center again, and then put it into the cluster with the smallest distance;

[0062] ④ Check whether the cluster centers of the two clusterings have changed. If so, repeat steps ② and ③. If the cluster centers of all classes no longer change, stop the iteration.

[0063] The k-means algorithm clusters the data in different numbers, and uses GSA to analyze the eigenvalues ​​of each number of clusters to determine the optimal number of clusters. The fundamental method of the GSA algorithm is to compare the mean value E[lnW r (k)] and the natural logarithm of cluster dispersion to determine the optimal number of clusters using k-means clustering. The variable that represents their relationship is called the gap value gap(k). The larger the gap value, the better the clustering result, which can be expressed as:

[0064] gap(k)=E[lnW r (k)]-ln(W k );

[0065] W k It is the cluster dispersion when the number of clusters is k, which refers to the sum of the distances of all data in each cluster from the cluster center, and can reflect the degree of concentration between data in each class. The calculation formula is:

[0066]

[0067] Among them, |c i | is the number of data in the i-th cluster, D i is the distance between each data point in the i-th cluster, The distance metric used is the Euclidean square distance, and the distance calculation formula between points j and j' is:

[0068] djj′ =∑(x j -x j′ ) 2 .

[0069] E[ln W r (k)] is the mean of the reference distribution set when the number of clusters is k. When n samples are clustered into k classes in the po-dimensional space, the mathematical expectation of the discreteness of the reference distribution set can be approximated as:

[0070] E[ln W r (k)]=log(p0n / 12)-(2 / p0)log(k)+c;

[0071] Among them, c is a constant. If the optimal number of clusters K is reached, that is, k = K, the gap value of the algorithm reaches the maximum. In order to make the data clustering adaptability range more flexible, it is necessary to find a better model with logarithmic concave function properties, that is, a density function with normal distribution to model, such as uniform distribution and normal distribution. In fact, the essence of the GSA algorithm is to compare the eigenvalues ​​extracted from the known distribution with the reference distribution, so as to determine the type of discrete points to be clustered and whether they can be classified into one category.

[0072] According to the above steps, the characteristic quantities can be used to cluster electric vehicles, and the behavioral characteristics of each type of electric vehicle user such as residence time, charging time and charging demand can be obtained, thereby realizing the electric vehicle user portrait.

[0073] The present invention is further illustrated by a specific embodiment below.

[0074] Since the travel and charging characteristics of electric taxis and electric private cars are relatively obvious, the travel of these two types of electric vehicles in one day is simulated. The total number of simulated electric vehicles is 10,000, and the ratio of taxis is 0.022, that is, 220 taxis and 9,780 private cars. The traffic model uses a regional road network with a total of 245 traffic nodes, 424 road sections, a total length of 355 kilometers, and an average of 0.84 kilometers per road section. The road network nodes are divided into 110 residential areas, 85 industrial areas and 50 commercial areas. The charging method of residential areas and is slow charging, with a charging power of 7kW. 80% of the charging methods in industrial areas are slow charging, and 20% are fast charging, with a fast charging power of 60kW. The charging method in commercial areas is fast charging, with a charging power of 60kW. 30 public fast charging stations are set up in the road network, and each charging station is equipped with 50 charging piles, with a charging pile power of 60kW.

[0075] This embodiment adopts uniform distribution. The reference distribution is generated within the range of the data to be tested. F groups of reference distribution sets need to be generated to detect a set of data. The F value is 10. The calculation formula of the expected value of the mean of the reference distribution set is:

[0076]

[0077] ln(W k ) standard deviation sd k and E[lnW r (k)] k The definition is as follows:

[0078]

[0079] The optimal number of clusters k is determined by the following formula:

[0080] gap(k)≥gap(k+1)-s k+1 .

[0081] The optimal number of clusters is obtained through the GSA algorithm. The results are as follows Figure 2 As shown in the figure, when the number of clusters is 5, the gap value is the largest, that is, the clustering effect is the best.

[0082] Considering the number of stops of electric vehicles in a day, the average start time of electric vehicle stops, the number of electric vehicle charging times in a day, the average duration of electric vehicle stops, the average start time of electric vehicle charging, the average duration of electric vehicle charging, the average remaining power of electric vehicle charging, the average power required for electric vehicle charging and the average charging power of electric vehicle charging, the electric vehicle users are clustered based on the k-means clustering algorithm. The number of each of the five types of electric vehicle users is as follows: Figure 3 Among the 10,000 electric vehicles, there are 2,750 electric vehicles in the first category, 1,197 electric vehicles in the second category, 2,658 electric vehicles in the third category, 811 electric vehicles in the fourth category, and 2,584 electric vehicles in the fifth category.

[0083] The dwell start time of each type of electric vehicle is as follows: Figure 4 As shown by Figure 4 It can be seen that the fourth type of electric vehicles are parked all day long, and the time distribution is scattered. The other four types of electric vehicles are concentrated in the morning 6:00-10:00 and afternoon 15:00-22:00 to start parking.

[0084] It is not difficult to find that the present invention determines the number of clusters through the GSA algorithm and then performs clustering through the k-means clustering algorithm, which can accurately analyze the characteristics of electric vehicle users' charging demand activities, and provide a data basis for operating companies to improve operational efficiency, enhance the quality of operation and maintenance services, and guide user charging behavior.

[0085] A second embodiment of the present invention relates to an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method for analyzing characteristics of charging demand activities of electric vehicle users of the first embodiment when executing the computer program.

[0086] The third embodiment of the present invention relates to a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for analyzing characteristics of charging demand activities of electric vehicle users of the first embodiment are implemented.

[0087] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.

[0088] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0089] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction method, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0090] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1A step that specifies a function in one or more boxes.

[0091] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for analyzing the characteristics of electric vehicle user charging demand activities, characterized in that: The following steps are involved: Normalizing the travel behavior data and charging behavior data of electric vehicle users to obtain multiple feature quantities; Establishing a feature matrix according to the multiple feature quantities, and performing standardization processing on the feature matrix; The standardized feature matrix is ​​clustered using k-means-GSA to analyze the characteristics of electric vehicle users’ charging demand activities.

2. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 1, characterized in that: The characteristic quantity is one or a combination of the following: the number of times an electric vehicle stops in a day, the average start time of an electric vehicle's stops, the number of times an electric vehicle is charged in a day, the average duration of an electric vehicle's stops, the average start time of an electric vehicle's charging, the average duration of an electric vehicle's charging, the average remaining power of an electric vehicle's charging, the average power required for an electric vehicle's charging, and the average charging power of an electric vehicle's charging.

3. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 1, characterized in that: The feature matrix is ​​expressed as: Among them, Z m×n is the feature matrix, m is the total number of electric vehicles involved in clustering, and n is the number of feature quantities.

4. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 1, characterized in that: The k-means-GSA clustering of the standardized feature matrix specifically includes: The k-means clustering algorithm is used to cluster the standardized feature matrix with different cluster numbers to obtain clustering results with different cluster numbers; According to the clustering results of different cluster numbers by GSA cluster analysis, the gap values ​​of different cluster numbers are obtained, and the optimal number of clusters is determined according to the gap values ​​of different cluster numbers; The standardized feature matrix is ​​classified using the k-means clustering algorithm with the optimal number of clusters to obtain activity characteristics of each type of electric vehicle user regarding the stay time, charging time and charging demand.

5. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 4 is characterized in that: The clustering results of different cluster numbers are analyzed according to the GSA clustering to obtain the gap values ​​of different cluster numbers, and the optimal number of clusters is determined according to the gap values ​​of different cluster numbers, specifically: Calculating the clustering dispersion of different numbers of clusters and the mean of the reference distribution set, and obtaining the gap values ​​when the number of clusters is different based on the clustering dispersion and the mean of the reference distribution set; The number of clusters corresponding to the maximum gap value is selected and this number of clusters is taken as the optimal number of clusters.

6. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 5, characterized in that: The cluster dispersion is expressed by Calculated, where W k is the clustering dispersion when the number of clusters is k, |c i | is the number of data in the i-th cluster, D i is the distance between each data point in the i-th cluster.

7. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 5, characterized in that: The mean of the reference distribution set is given by E[lnW r (k)] = log (p0n / 12) - (2 / p0) log (k) + c, where E[lnW r (k)] is the mean of the reference distribution set when the number of clusters is k, p0 is the dimension of the space, n is the number of samples, and c is a constant.

8. The method for analyzing the characteristics of electric vehicle user charging demand activities according to claim 5, characterized in that: The gap value is given by gap(k)=E[lnW r (k)]-ln(W k ), where gap(k) is the gap value when the number of clusters is k, and W k is the clustering dispersion when the number of clusters is k, E[lnW r (k)] is the mean of the reference distribution set when the number of clusters is k.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for analyzing the characteristics of electric vehicle user charging demand activities as described in any one of claims 1-8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for analyzing the characteristics of electric vehicle user charging demand activities as described in any one of claims 1 to 8 are implemented.

Images