Method for identifying multiple non-collision accident points of electric vehicle and people by considering spatial-temporal characteristics

Abstract

The invention provides a method for identifying multiple non-collision accident points of an electric vehicle and a person by considering spatial-temporal characteristics. The method comprises the following steps: acquiring non-collision accident data of the electric vehicle and the person and road network distribution data; preprocessing the accident data; the accident time distribution characteristics are quantitatively analyzed from the angles of monthly and hour distribution; introducing a category accident comprehensive influence index, determining an accident influence factor weight by adopting an analytic hierarchy process, and revealing the space aggregation of the non-collision accident between the electric vehicle and the human through a weighted network kernel density estimation method; aiming at space-time characteristics, a density peak value clustering algorithm is used as an accident space clustering model; time dimension features are introduced, and an ST-DBSCAN accident time-space multiple point segment identification model is constructed. According to the method, the problems that the accident gathering area range is relatively large, the dispersion degree is relatively low and a part of the area has a piece trend are avoided, and the accident space-time multi-point section with relatively high separation degree and the accident space-time multi-point section with relatively high separation degree near a certain accurate time point are accurately identified and obtained.

Description

Technical Field

[0001] The invention relates to the field of electric vehicle safety, and in particular to a method for identifying high-occurrence points of non-collision accidents between electric vehicles and people taking into account time and space characteristics. Background Art

[0002] Initially, the research on traffic accident distribution characteristics at home and abroad focused on the identification of frequent points in space, and most of the research ignored the temporal distribution of traffic accidents. As an important feature for identifying frequent points in traffic accidents, temporal characteristics have a significant impact on the location and severity of traffic accidents. The lack of temporal dimension information will lead to the inability to accurately identify frequent points in space and time, resulting in low efficiency in preventing and controlling collisions between electric vehicles and pedestrians / non-motor vehicles.

[0003] Density clustering methods such as network kernel density estimation can effectively reflect the spatial distribution characteristics of traffic accidents. Combining them with machine learning models, the characteristics of electric vehicle and pedestrian / non-motor vehicle collision accidents are considered to select and improve the existing machine learning models, thereby achieving accurate identification of high-incidence points and sections of electric vehicle and pedestrian / non-motor vehicle collision accidents.

[0004] For the research objects of traffic accident frequent point segment identification, vehicle accidents, fuel vehicle accidents, electric vehicle accidents, passenger and freight vehicle accidents, and autonomous vehicle accidents can be selected. Each object can be subdivided into collision accidents with different accident participants, and the consequences caused by different research objects are not the same. At present, there are rich research results on the identification of accident frequent point segments with vehicle accidents, fuel vehicle accidents and other research objects, and there are studies on the identification of electric vehicle accident frequent point segments. Exploring the spatiotemporal characteristics of collision accidents between electric vehicles and different accident participants, using model algorithms corresponding to the spatiotemporal characteristics, and identifying accurate multi-category electric vehicle accident frequent point segments are the key core issues in the research of electric vehicle accident prevention.

[0005] The research on identification of spatial high-incidence points of traffic accidents can be divided into three categories according to the research methods: from the perspective of traffic models, from statistical methods and from machine learning methods.

[0006] In early studies, the model assumed that the influencing factors of accidents are the same for collisions with different objects. In real accident scenarios, the impact of various influencing factors on collisions with different objects is different. This problem will lead to deviations in the calibration of model parameters, making it difficult to accurately determine the spatiotemporal frequent points of collisions between electric vehicles and different objects, and thus unable to prevent traffic accidents with large losses in a targeted manner. Therefore, in order to improve the accuracy of identifying spatiotemporal frequent points of accidents, it is a research focus to focus on the differential impact of influencing factors on collisions between vehicles and different objects.

[0007] The data of electric vehicle and pedestrian / non-motor vehicle collision accidents can provide the time and location of the accident, and the longitude and latitude information of the accident WGS-84 coordinate system can be picked up more accurately through coordinate matching. Summary of the invention

[0008] In view of this, the present invention starts from the characteristics of vulnerable groups on the road and electric vehicles, combines the temporal and spatial characteristics of electric vehicle and pedestrian / non-motor vehicle collision accidents, considers the influence of relevant factors on electric vehicle and pedestrian / non-motor vehicle collision accidents to weight them, uses continuous time variable input and output, and constructs a spatiotemporal multiple-point segment identification model based on ST-DBSCAN to provide targeted suggestions for preventing serious (fatal and injury) accidents of electric vehicles and improving the service level of road facilities. The non-people mentioned in the present invention refer to pedestrians or non-motor vehicles.

[0009] In a first aspect, the present invention provides a method for identifying high-occurrence points of non-collision accidents between electric vehicles and people taking into account spatiotemporal characteristics, the method comprising: S1: Obtain the specific data of non-collision accidents between electric vehicles and people and the distribution of road networks; S2: Preprocessing of non-collision accident data between electric vehicles and people; S3: Quantitatively analyze the time distribution characteristics of accidents from the perspective of monthly and hourly distribution; S4: Introducing comprehensive impact indicators for category accidents U i , the analytic hierarchy process was used to determine the weights of accident influencing factors, and the weighted network kernel density estimation method was used to reveal the spatial clustering of traffic accident locations; ; in, R , W , A , S They are road type, weather, age of accident participants, and accident severity; k 1 , k 2 , k 3 , k 4 The weight coefficients corresponding to road type, weather, age of accident participants and accident severity were determined using the analytic hierarchy process; S5: According to the spatiotemporal characteristics of non-collision accidents between electric vehicles and humans, the density peak clustering algorithm is used as the accident space clustering model, and the spatiotemporal DBSCAN algorithm with time dimension characteristics is introduced to construct a spatiotemporal frequent point segment identification model for accidents, which characterizes the spatiotemporal frequent locations of non-collision accidents between electric vehicles and humans.

[0010] Furthermore, the electric vehicle and human non-collision accident data is traffic accident attribute data, and the traffic accident attribute data specifically includes spatiotemporal location attribute data, road environment attribute data, personnel attribute data, and self attribute data; The spatiotemporal location attribute data includes accident date, accident time, longitude and latitude; The road environment attribute data includes the road type and weather conditions when the accident occurred; The personnel attribute data include the gender of the accident participants, the age of the accident participants, and the driving experience of the driver; The self-attribute data includes collision type and severity.

[0011] Furthermore, the preprocessing of the non-collision accident data between electric vehicles and people includes picking up the longitude and latitude of the accident location, cleaning up outliers, correcting erroneous data, and repairing missing data.

[0012] Furthermore, the weighted network kernel density estimation method is used to reveal the spatial clustering of traffic accident locations, including the road type R ,weather W , Age of accident participants A and the severity of the accident S The weight coefficients of the four indicators are determined by the hierarchical analysis method: ; in, is the network kernel density estimate; n is the total number of accident points in the study area; r is the window width; K(x) is a one-dimensional Gaussian kernel function; d i For the i The network distance from the accident to the corresponding nuclear center; U i It is the comprehensive impact indicator of the accident.

[0013] Furthermore, the density peak clustering algorithm is used as the accident space clustering model. Specifically include: ; in, ρ i For data objects x i The local density of , d ij For any two objects in the domain x i and x jThe distance between them is calculated using the Euclidean distance; d c Represents the cutoff distance.

[0014] Furthermore, the spatiotemporal DBSCAN algorithm with time dimension features specifically includes: introducing the time dimension into the DBSCAN algorithm, and expanding the circular search area into a spherical search area, through the spatial search distance, the temporal search distance and the minimum number of points contained in the spherical area centered on P.

[0015] Furthermore, the description of the spatiotemporal locations where non-collision accidents between electric vehicles and humans frequently occur is the spatiotemporal coordinates of the segments where non-collision accidents between electric vehicles and humans frequently occur, including time, longitude, and latitude.

[0016] Furthermore, the k 1 , k 2 , k 3 , k 4 They are 0.090, 0.042, 0.216 and 0.652 respectively.

[0017] In a second aspect, the present invention further provides an electronic device comprising a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement a method for identifying high-incidence points of non-collision accidents between electric vehicles and humans that takes into account temporal and spatial characteristics as described above.

[0018] In a third aspect, the present invention further provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for identifying high-incidence points of non-collision accidents between electric vehicles and humans taking into account spatiotemporal characteristics as described above.

[0019] Beneficial effects of the present invention: 1. The present invention effectively avoids the problems of large accident clustering areas, low degree of dispersion, and the tendency of some areas to form clusters, and accurately identifies accident spatial multiple-occurrence point segments with high separation and accident spatial multiple-occurrence point segments with high separation near a certain accurate time point.

[0020] 2. The calculation error of the present invention is small and highly targeted, which fits the actual distribution of traffic accidents and is beneficial to the management and prevention of non-collision accidents between electric vehicles and people.

[0021] 3. The present invention can accurately identify the spatiotemporal high-incidence points of collision accidents between electric vehicles and vulnerable groups on the road. The algorithms adopted are applicable to the spatiotemporal distribution characteristics of collision accidents between electric vehicles and pedestrians / non-motor vehicles. It can meet the needs of road traffic safety hazard identification, electric vehicle accident-prone points investigation, road traffic improvement design and other fields for the assessment of electric vehicle accident-prone points. At the same time, it can also provide technical support for preventing the occurrence of serious (fatal and injury) accidents involving electric vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The present invention has the following accompanying drawings: Figure 1 is a flow chart of the steps of the present invention; Figure 2 This is a schematic diagram of the time distribution characteristics of non-collision accidents between electric vehicles and people according to the method of the present invention; Figure 3 1. It is a schematic diagram comparing the spatial characteristics of the network kernel density before and after weighting of the non-collision accident between an electric vehicle and a person according to the present invention; Figure 4 It is a schematic diagram of the improvement idea of ​​ST-DBSCAN clustering in the method of the present invention; Figure 5 This is a schematic diagram of the spatial clustering effect of the method of the present invention on non-collision accidents between electric vehicles and people; Figure 6 It is a schematic diagram of the spatiotemporal clustering effect of non-collision accidents between electric vehicles and people according to the method of the present invention. DETAILED DESCRIPTION

[0023] In order to make the objects, advantages and features of the present invention more obvious, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] Example 1 like Figure 1 As shown, the present invention provides a method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering spatiotemporal characteristics, which mainly includes the following steps: Step 1: Obtain electric vehicle accident data from the traffic management department for the area where the road network is studied. According to the accident records such as the road traffic accident identification after the accident, extract the accident severity, accident type, time of occurrence, location of occurrence, road type, weather conditions, vehicles involved in the accident, and driver information; Step 2: Obtain the longitude and latitude of the WGS-84 coordinate system through coordinate matching, and perform data preprocessing such as outlier cleaning, error data correction, and missing data repair; Step 3: Classify electric vehicle accidents according to accident forms and build a database of electric vehicle and pedestrian / non-motor vehicle collision accidents; Step 4: Quantitatively analyze the accident time distribution characteristics from the perspective of monthly and hourly distribution; Step 5: Introduce comprehensive impact indicators of category accidents U i , the weights of accident influencing factors are determined by using the analytic hierarchy process; Step 6: Reveal the spatial clustering of traffic accident locations through weighted network kernel density estimation method; Step 7, using the obtained spatial density of collision accidents between electric vehicles and pedestrians / non-motor vehicles as the model dependent variable; Step 8: Using the acquired traffic accident attribute data such as spatial location attribute information, road environment attribute information, personnel attribute information and accident attribute information as model independent variables; Step 9: Establish a spatial clustering model for collision accidents between electric vehicles and pedestrians / non-motor vehicles based on the DPC algorithm (clustering by fast search and find of density peaks, DPC); Step 10: Introduce time dimension information; Step 11: Establish a spatiotemporal clustering model for collision accidents between electric vehicles and pedestrians / non-motor vehicles based on the ST-DBSCAN (Spatial Temporal-DBSCAN, ST-DBSCAN) algorithm; Step 12: In terms of spatial multiple-shot point segment identification, the DBSCAN, OPTICS, DPC and Mean Shift algorithms are used to train data, and the clustering evaluation indicators of the four algorithms, namely, silhouette coefficient, DBI and CHI, are output, and a comparative verification is performed based on the training results; Step 13: Use the DPC algorithm to identify the spatial frequent point segments of the electric vehicle and pedestrian / non-motor vehicle collision accident data, select different category parameters, and cluster the coordinates of the frequent accident spatial points; Step 14: In the identification of spatiotemporal multiple-point segments, the ST-DBSCAN, ST-OPTICS, ST-DPC and ST-Mean Shift algorithms are used to train data, and the clustering evaluation indicators of the four algorithms, namely, the silhouette coefficient, DBI and CHI, are output, and a comparative verification is performed based on the training effect; Step 15: Use the ST-Mean Shift algorithm to identify the spatiotemporal frequent point segments of the electric vehicle and pedestrian / non-motor vehicle collision accident data, select different bandwidth parameters, and cluster the spatiotemporal frequent point coordinates of the accidents; In the method of the present invention, the electric vehicle accidents are classified into three categories according to the accident form classification described in step 3: collision between electric vehicles and pedestrians / non-motor vehicles, collision between electric vehicles and stationary obstacles, and collision between electric vehicles and motor vehicles. The data of collision accidents between electric vehicles and pedestrians / non-motor vehicles are retained, and a database of collision accidents between electric vehicles and pedestrians / non-motor vehicles is constructed by quantizing fields and associating longitude and latitude; Figure 2 This is a schematic diagram of the time distribution characteristics of a collision accident between an electric vehicle and a pedestrian / non-motor vehicle according to the method of the present invention; In the method of the present invention, the comprehensive impact index of the introduced category accident in step 5 is U i , the analytic hierarchy process was used to determine the weights of accident influencing factors, and the weighted network kernel density estimation method was used to reveal the spatial clustering of traffic accident locations; ; in, R , W , A , S They are road type, weather, age of accident participants, and accident severity; k 1 , k 2 , k 3 , k 4 The weight coefficients corresponding to road type, weather, age of accident participants and accident severity were determined using the analytic hierarchy process; Construct a hierarchical structure of weight coefficients of factors affecting collision accidents between electric vehicles and pedestrians / non-motor vehicles, generate an accident judgment matrix, and obtain the accident weight coefficients through calculation and normalization. k 1 ~k 4 They are 0.090, 0.042, 0.216 and 0.652 respectively; In the method of the present invention, the accident spatial characteristics described in step 6 are obtained using an improved weighted network kernel density estimation method, wherein the road type R ,weather W , Age of accident participants A and the severity of the accident S The weight coefficients of the four indicators are determined by the hierarchical analysis method: ; in, is the network kernel density estimate; n is the total number of accident points in the study area; r is the window width; K(x)is a one-dimensional Gaussian kernel function; d i For the i The network distance from the accident to the corresponding nuclear center; U i It is the comprehensive impact index of the accident; Figure 3 1. It is a schematic diagram of the comparison of the network kernel density spatial characteristics before and after weighting of the collision accident between electric vehicles and pedestrians / non-motor vehicles of the present invention; In the method of the present invention, the input vector described in step 8 is traffic accident attribute data, including spatial location attributes such as accident date, accident longitude, latitude, etc.; road environment attributes such as road type and weather conditions at the time of the accident; personnel attributes such as gender of accident participants, age of accident participants, and driving experience of the driver; and accident attributes such as collision type and severity; In the method of the present invention, the DPC algorithm described in step 9 takes into account the characteristics of small amount of collision data between electric vehicles and pedestrians / non-motor vehicles, large area of ​​multiple accident points, low degree of dispersion between multiple accident points, and concentration on the main roads with large traffic volume. The core of the algorithm lies in local density and relative distance: ; in, ρ i For data objects x i The local density of , d ij For any two objects in the domain x i and x j The relative distance between them is calculated using the Euclidean distance; d c Represents the cutoff distance.

[0025] The DPC algorithm directly identifies the cluster center and cluster structure by drawing the density and relative distance diagram of each point, thereby effectively identifying accident-prone points.

[0026] In the method of the present invention, the ST-DBSCAN algorithm described in step 11 aims to identify independent accident-prone segments as much as possible, narrow the scope of accident clustering, and improve the accident spatiotemporal clustering model based on the characteristics of large spatial clustering areas of electric vehicle and pedestrian / non-motor vehicle collision accidents, low degree of dispersion, and some areas connected into pieces.

[0027] By introducing the time dimension into the DBSCAN algorithm, the ST-DBSCAN algorithm is obtained. A time search parameter is introduced on the original basis, and the circular search area is expanded to a spherical search area. Its three parameters are: spatial search distance (eps1), time search distance (eps2) and the minimum number of points contained in the spherical area centered on P (MinPts).

[0028] ST-DBSCAN retains DBSCAN's ability to accurately identify spatial density features and adds time constraints (Eps1 and Eps2), giving full play to the advantages of both. It is suitable for data with uneven density distribution and large discreteness of electric vehicle / pedestrian / non-motor vehicle collision accidents.

[0029] Figure 4 It is a schematic diagram of the improvement idea of ​​ST-DBSCAN clustering in the method of the present invention; Figure 5 This is a schematic diagram of the spatial clustering effect of the method of the present invention on collision accidents between electric vehicles and pedestrians / non-motor vehicles; In the method of the present invention, the values ​​of the four algorithm parameters described in step 14 are eps=1.8, min_samples=11; min_samples=15; classnum=10; bandwidth=1.8.

[0030] Figure 6 This is a schematic diagram of the spatiotemporal clustering effect of the method of the present invention on collision accidents between electric vehicles and pedestrians / non-motor vehicles; The comparison results show that the method for identifying high-incidence points of non-collision accidents between electric vehicles and pedestrians taking into account the spatiotemporal characteristics can take into account the spatiotemporal clustering of traffic accidents, deeply explore the spatiotemporal characteristics of collision accidents between electric vehicles and pedestrians / non-motor vehicles, and has a small calculation error, strong pertinence, and is in line with the actual distribution of traffic accidents.

[0031] The present invention effectively avoids the problems of large accident clustering areas, low discreteness, and clustering trends in some areas, and accurately identifies accident spatial multiple-occurrence point segments with a high degree of separation and accident temporal multiple-occurrence point segments with a high degree of separation near a certain accurate time point. The identification effect is good, which is beneficial to the management and prevention of collision accidents between electric vehicles and pedestrians / non-motor vehicles.

[0032] Embodiment 2: An electronic device includes a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the above-mentioned method for identifying high-incidence points of non-collision accidents between electric vehicles and people taking into account time and space characteristics.

[0033] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0034] Embodiment 3: A computer-readable storage medium stores computer instructions, which implement the steps of the method in Example 1 when executed by a processor.

[0035] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0036] The present invention is described with reference to the flowcharts and / or block diagrams of the method, terminal device (system), and computer program product according to the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks 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 terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal 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.

[0037] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0038] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0039] The above embodiments describe the technical solutions of the present invention in detail. Obviously, the present invention is not limited to the described embodiments. Based on the embodiments of the present invention, people familiar with the technical field can also make various changes accordingly, but any changes that are equivalent or similar to the present invention belong to the scope of protection of the present invention.

[0040] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

Claims

1. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering spatiotemporal characteristics, characterized in that: The method comprises: S1: Obtain the specific data of non-collision accidents between electric vehicles and people and the distribution of road networks; S2: Preprocessing of non-collision accident data between electric vehicles and people; S3: Quantitatively analyze the time distribution characteristics of accidents from the perspective of monthly and hourly distribution; S4: Introducing comprehensive impact indicators for category accidents U i , the analytic hierarchy process was used to determine the weights of accident influencing factors, and the weighted network kernel density estimation method was used to reveal the spatial clustering of traffic accident locations; ； in, R , W , A , S They are road type, weather, age of accident participants, and accident severity; k 1 , k 2 , k 3 , k 4 The weight coefficients corresponding to road type, weather, age of accident participants and accident severity were determined using the analytic hierarchy process; S5: According to the spatiotemporal characteristics of non-collision accidents between electric vehicles and humans, the density peak clustering algorithm is used as the accident space clustering model, and the spatiotemporal DBSCAN algorithm with time dimension characteristics is introduced to construct a spatiotemporal frequent point segment identification model for accidents, which characterizes the spatiotemporal frequent locations of non-collision accidents between electric vehicles and humans.

2. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering spatiotemporal characteristics as claimed in claim 1, characterized in that: The electric vehicle and human non-collision accident data is traffic accident attribute data, and the traffic accident attribute data specifically includes spatiotemporal location attribute data, road environment attribute data, personnel attribute data, and self attribute data; The spatiotemporal location attribute data includes accident date, accident time, longitude and latitude; The road environment attribute data includes the road type and weather conditions when the accident occurred; The personnel attribute data include the gender of the accident participants, the age of the accident participants, and the driving experience of the driver; The self-attribute data includes collision type and severity.

3. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering time and space characteristics as claimed in claim 2, characterized in that: The preprocessing of the electric vehicle and human non-collision accident data includes picking up the longitude and latitude of the accident location, cleaning up abnormal values, correcting erroneous data, and repairing missing data.

4. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering time and space characteristics as claimed in claim 3, characterized in that: The weighted network kernel density estimation method is used to reveal the spatial clustering of traffic accident locations, including road types. R ,weather W , Age of accident participants A and the severity of the accident S The weight coefficients of the four indicators are determined by the hierarchical analysis method: ； in, is the network kernel density estimate; n is the total number of accident points in the study area; r is the window width; K(x) is a one-dimensional Gaussian kernel function; d i For the i The network distance from the accident to the corresponding nuclear center; U i It is the comprehensive impact indicator of the accident.

5. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering time and space characteristics as claimed in claim 4, characterized in that: The density peak clustering algorithm is used as the accident space clustering model, specifically including: ； in, ρ i For data objects x i The local density of , d ij For any two objects in the domain x i and x j The distance between them is calculated using the Euclidean distance; d c Represents the cutoff distance.

6. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering spatiotemporal characteristics as claimed in claim 5, characterized in that: The spatiotemporal DBSCAN algorithm with time dimension features specifically includes: introducing the time dimension into the DBSCAN algorithm, and expanding the circular search area into a spherical search area, through the spatial search distance, the temporal search distance and the minimum number of points contained in the spherical area centered on P.

7. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering time and space characteristics as claimed in claim 6, characterized in that: The spatial and temporal coordinates of the frequent non-collision accidents between electric vehicles and humans are the spatial and temporal coordinates of the frequent non-collision accidents between electric vehicles and humans, including time, longitude, and latitude.

8. A method for identifying high-occurrence points of non-collision accidents between electric vehicles and people considering time and space characteristics as claimed in claim 7, characterized in that: Said k 1 , k 2 , k 3 , k 4 They are 0.090, 0.042, 0.216 and 0.652 respectively.

9. An electronic device, characterized in that: It includes a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement a method for identifying high-incidence points of non-collision accidents between electric vehicles and people taking into account time and space characteristics as described in any one of claims 1-8.

10. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of a method for identifying high-occurrence points of non-collision accidents between electric vehicles and humans taking into account time and space characteristics as described in any one of claims 1 to 8 are implemented.

Images