Method, device, storage medium and electronic device for determining target vehicle

Abstract

An embodiment of the present invention provides a method, device, storage medium, and electronic device for determining a target vehicle, wherein the method includes: obtaining a danger parameter for each vehicle among N vehicles, obtaining N danger parameters, wherein the danger parameter is calculated based on a set of vehicle attributes; selecting M vehicles with the top M danger parameters among the N vehicles, and obtaining an activity parameter for each of the M vehicles, obtaining M activity parameters; determining a vehicle feature vector for each of the M vehicles based on the M danger parameters and M activity parameters corresponding to the M vehicles, obtaining M vehicle feature vectors, wherein each vehicle feature vector includes the danger parameter and activity parameter of the corresponding vehicle; and determining a target vehicle as a controlled object among the M vehicles based on the M vehicle feature vectors. The embodiment of the present invention solves the problem of low efficiency in determining target vehicles in the related art.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of vehicle management technology, and in particular, to a method, device, storage medium, and electronic device for determining a target vehicle. Background Art

[0002] With the rapid development of urbanization, road traffic safety issues are becoming increasingly serious. Key vehicles, including those carrying passengers, dangerous goods, cargo, and vehicles with limited mobility, are committing serious traffic violations and presenting significant safety hazards. The management of key vehicles (or target vehicles) affects the safety of people's lives and property, making them a key focus for traffic management. However, due to the dramatic increase in the number of key vehicles, management departments' limited human resources have made it impossible to maintain comprehensive and strict control. This has led to a proliferation of traffic violations and irregularities, posing significant risks to public safety. While some methods currently exist for assessing vehicle risk based on vehicle characteristics to identify key vehicles for traffic management, the number of these vehicles remains high, making them ineffective in daily regulatory work. Consequently, prioritizing key vehicles based solely on vehicle characteristics in related technologies makes it difficult to effectively identify them from a large number of vehicles. In other words, the methods used in related technologies for identifying key vehicles are relatively simplistic, resulting in low efficiency.

[0003] With respect to the problem of low efficiency in determining target vehicles in related technologies, no effective solution has been proposed so far. Summary of the Invention

[0004] Embodiments of the present invention provide a method, device, storage medium, and electronic device for determining a target vehicle, so as to at least solve the problem of low efficiency in determining a target vehicle in the related art.

[0005] According to one embodiment of the present invention, a method for determining a target vehicle is provided, comprising: obtaining a danger parameter of each vehicle among N vehicles to obtain N danger parameters, wherein the danger parameter of each vehicle among the N vehicles is calculated based on a set of attributes of the each vehicle, and N is a positive integer greater than or equal to 2; selecting M vehicles from the N vehicles, and obtaining an activity parameter of each vehicle among the M vehicles to obtain M activity parameters, wherein the M vehicles are the vehicles ranked in descending order of danger parameters among the N vehicles, and M is a positive integer greater than or equal to 1 and less than N; determining a vehicle feature vector of each vehicle among the M vehicles based on the M danger parameters and the M activity parameters corresponding to the M vehicles among the N danger parameters to obtain M vehicle feature vectors, wherein each vehicle feature vector among the M vehicle feature vectors includes the danger parameter and the activity parameter of the corresponding vehicle among the M vehicles; and determining a target vehicle as a controlled object among the M vehicles based on the M vehicle feature vectors.

[0006] In an exemplary embodiment, the determining of a target vehicle as a controlled object among the M vehicles based on the M vehicle feature vectors includes: determining a dominant feature vector of each vehicle among the M vehicles based on the M vehicle feature vectors to obtain M dominant feature vectors, wherein the i-th dominant feature vector among the M dominant feature vectors includes an i-th dominant quantity and an i-th dominated quantity corresponding to the i-th vehicle feature vector among the M vehicle feature vectors, and the i-th dominant quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominance condition. , the dominating condition means that the values of each vector member in the vehicle feature vector are greater than the values of the vector members at the corresponding positions in the i-th vehicle feature vector; the i-th dominated number includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominated condition, and the dominated condition means that the values of each vector member in the vehicle feature vector are smaller than the values of the vector members at the corresponding positions in the i-th vehicle feature vector, and i is a positive integer less than or equal to M; based on the M dominating feature vectors, the target vehicle as the controlled object is determined among the M vehicles.

[0007] In an exemplary embodiment, determining the target vehicle as the controlled object among the M vehicles based on the M dominant feature vectors includes: sorting the M dominant feature vectors in ascending order of the dominance quantity included in each dominant feature vector in the M dominant feature vectors to obtain a sorting result; determining the top K dominant feature vectors in the sorting result, and determining K vehicles corresponding to the top K dominant feature vectors among the M vehicles as the target vehicles, wherein K is a positive integer greater than or equal to 1 and less than M.

[0008] In an exemplary embodiment, the M dominating feature vectors are sorted in ascending order of the dominating quantities included in each of the M dominating feature vectors to obtain a sorting result, comprising: when the i-th dominating quantity included in the i-th dominating feature vector in the M dominating feature vectors is equal to the j-th dominating quantity included in the j-th dominating feature vector in the M dominating feature vectors, and the i-th dominated quantity included in the i-th dominating feature vector is greater than the j-th dominated quantity included in the j-th dominating feature vector, the i-th dominating feature vector is sorted before the j-th dominating feature vector, where j is a positive integer greater than or equal to 1 and less than or equal to M; or when the i-th dominating quantity is equal to the j-th dominating quantity, When the i-th dominated number is equal to the j-th dominated number and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is greater than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is placed in front of the j-th dominating feature vector; when the i-th dominating number is equal to the j-th dominating number, the i-th dominated number is equal to the j-th dominated number, and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is less than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is placed behind the j-th dominating feature vector.

[0009] In an exemplary embodiment, the obtaining of the danger parameters of each of N vehicles to obtain N danger parameters includes: obtaining a set of attributes of each of the N vehicles to obtain N groups of attributes; determining the danger parameters of each of the N vehicles according to the danger sub-parameters corresponding to each attribute in each group of attributes in the N groups of attributes to obtain N danger parameters, wherein each different attribute in the N groups of attributes is pre-set with a corresponding danger sub-parameter.

[0010] In an exemplary embodiment, the method of determining the danger parameter of each of the N vehicles based on the danger sub-parameter corresponding to each attribute in each group of attributes in the N groups of attributes to obtain N danger parameters includes: determining the danger parameter of the i-th vehicle in the N vehicles in the following manner: danger parameter of the i-th vehicle = danger sub-parameter corresponding to the basic attribute of the i-th vehicle * (sum of danger sub-parameters corresponding to other attributes), wherein the group of attributes of the i-th vehicle includes the basic attribute and the other attributes, and the other attributes include attributes in the group of attributes of the i-th vehicle other than the basic attribute, and i is a positive integer greater than or equal to 1 and less than or equal to N; wherein, in the i-th vehicle When a group of attributes includes P attributes, the values of the P hazard sub-parameters corresponding to the group of attributes of the i-th vehicle are preset P values, where P is a positive integer greater than or equal to 2; or, when a group of attributes of the i-th vehicle includes P attributes, the values of the Q hazard sub-parameters corresponding to Q attributes in the group of attributes of the i-th vehicle are greater than the corresponding Q values in the preset P values, and the values of the hazard sub-parameters corresponding to the attributes other than the Q attributes in the group of attributes of the i-th vehicle are corresponding values in the P values, and the Q attributes are the top Q attributes with the greatest contribution to the hazard parameter among the P attributes determined according to a preset vehicle hazard prediction model, where Q is a positive integer greater than or equal to 1 and less than P.

[0011] In an exemplary embodiment, the acquiring of the activity parameters of each of the M vehicles to obtain the M activity parameters includes: determining the activity time data of each of the M vehicles and the activity space data of each of the M vehicles through the vehicle data in the target area over a predetermined time period, wherein the activity time data is used to represent the activity time of each of the M vehicles in the target area, and the active space data is used to represent the activity distance of each of the M vehicles in the target area; determining the activity parameters of each of the M vehicles based on the activity time data of each of the M vehicles and the activity space data of each of the M vehicles to obtain the M activity parameters.

[0012] In an exemplary embodiment, determining the activity parameters of each of the M vehicles based on the activity time data of each of the M vehicles and the activity space data of each of the M vehicles to obtain the M activity parameters includes: determining a first entropy weight of the activity time data of each of the M vehicles and a second entropy weight of the active space data of each of the M vehicles; determining the activity parameters of each of the M vehicles based on the activity time data of each of the M vehicles, the first entropy weight of the activity time data of each of the M vehicles, the active space data of each of the M vehicles, and the second entropy weight of the active space data of each of the M vehicles to obtain the M activity parameters.

[0013] In an exemplary embodiment, determining a first entropy weight of the activity time data of each of the M vehicles and a second entropy weight of the activity space data of each of the M vehicles includes: constructing an activity evaluation parameter matrix X of the M vehicles, wherein the element x in X is i,j Used to represent the jth evaluation parameter of the i-th vehicle among the M vehicles, when j is equal to 1, the j-th evaluation parameter of the i-th vehicle is the active time data of the i-th vehicle, when j is equal to 2, the j-th evaluation parameter of the i-th vehicle is the active space data of the i-th vehicle, wherein i is a positive integer greater than or equal to 1 and less than or equal to M, and j is equal to 1 or 2; the matrix X is normalized to obtain a normalized matrix Z, wherein the element z in the matrix Z is i,j for: According to the matrix Z, a probability matrix P is obtained, wherein the element p in the probability matrix P is i,j It is used to represent the proportion of the j-th evaluation parameter of the i-th vehicle among the M vehicles in the j-th evaluation parameter of the M vehicles; the information entropy of the j-th evaluation parameter is calculated according to the probability matrix P according to the following formula: Among them, e j The information entropy of the j-th evaluation parameter is represented; the entropy weight of the j-th evaluation parameter is calculated based on the information entropy of the j-th evaluation parameter according to the following formula: Among them, d j =1-e j , W1 is the first entropy weight, and W2 is the second entropy weight.

[0014] In an exemplary embodiment, the acquiring of a set of attributes for each of the N vehicles to obtain N sets of attributes includes: updating, based on received target configuration parameters, hazard sub-parameters corresponding to some of the attributes in the N sets of attributes to obtain updated hazard sub-parameters corresponding to each attribute in each set of the N sets of attributes; and determining, based on the hazard sub-parameters corresponding to each attribute in each set of the N sets of attributes to obtain N hazard parameters, includes: determining, based on the hazard sub-parameters corresponding to each attribute in each set of the N sets of attributes to obtain the N hazard parameters.

[0015] According to another embodiment of the present invention, a device for determining a target vehicle is provided, comprising: a first acquisition module for acquiring a danger parameter of each vehicle among N vehicles, thereby obtaining N danger parameters, wherein the danger parameter of each vehicle among the N vehicles is calculated based on a set of attributes of each vehicle, and N is a positive integer greater than or equal to 2; a second acquisition module for selecting M vehicles from the N vehicles and acquiring an activity parameter of each vehicle among the M vehicles, thereby obtaining M activity parameters, wherein the M vehicles are vehicles whose danger parameters are obtained from the N vehicles. The top M vehicles are ranked from largest to smallest, where M is a positive integer greater than or equal to 1 and less than N; a first determination module is used to determine the vehicle feature vector of each vehicle in the M vehicles based on the M danger parameters and the M activity parameters corresponding to the M vehicles among the N danger parameters, so as to obtain M vehicle feature vectors, wherein each vehicle feature vector in the M vehicle feature vectors includes the danger parameter and activity parameter of the corresponding vehicle in the M vehicles; a second determination module is used to determine the target vehicle as the controlled object among the M vehicles based on the M vehicle feature vectors.

[0016] According to yet another embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when run.

[0017] According to another embodiment of the present invention, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any one of the above method embodiments.

[0018] The present invention obtains N dangerousness parameters for each vehicle from N vehicles, then selects the top M vehicles from the N vehicles by dangerousness parameter, obtains activity parameters for each of the M vehicles, and then determines vehicle feature vectors for each of the M vehicles based on the M dangerousness parameters and M activity parameters corresponding to the M vehicles from the N dangerousness parameters, thereby obtaining M vehicle feature vectors. Each of the M vehicle feature vectors includes the dangerousness parameter and activity parameter of the corresponding vehicle from the M vehicles. The target vehicle to be controlled is then determined from the M vehicles based on the M vehicle feature vectors. This achieves the purpose of determining a target vehicle to be controlled from the M vehicles based on the vehicle's dangerousness parameters and activity parameters, avoiding the problem in the related art of determining target vehicles based solely on vehicle characteristics, which results in low efficiency. Therefore, the problem of low efficiency in determining target vehicles in the related art is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 1 is a block diagram of the hardware structure of a mobile terminal according to a method for determining a target vehicle according to an embodiment of the present invention;

[0020] Figure 2 is a flow chart of a method for determining a target vehicle according to an embodiment of the present invention;

[0021] Figure 3 is a flow chart of a method for analyzing and recommending key vehicles according to an embodiment of the present invention;

[0022] Figure 4 is an example diagram of an active space according to an embodiment of the present invention;

[0023] Figure 5 is an example diagram of a vehicle feature vector according to an embodiment of the present invention;

[0024] Figure 6 is an example diagram of dominant feature vectors according to an embodiment of the present invention;

[0025] Figure 7 4 is a structural block diagram of a device for determining a target vehicle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings and in combination with embodiments.

[0027] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0028] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a block diagram of the hardware structure of a mobile terminal according to a method for determining a target vehicle according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown) a processor 102 (the processor 102 may include but is not limited to a microprocessor MCU or a programmable logic device FPGA and other processing devices) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input and output device 108 for communication functions. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0029] The memory 104 can be used to store computer programs, such as software programs and modules of application software, such as the computer program corresponding to the target vehicle determination method in the embodiment of the present invention. The processor 102 executes the computer program stored in the memory 104 to execute various functional applications and data processing, that is, to implement the above-mentioned method. The memory 104 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely located relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0030] The transmission device 106 is used to receive or send data via a network. A specific example of the aforementioned network may include a wireless network provided by the mobile terminal's communications provider. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0031] In this embodiment, a method for determining a target vehicle is provided. Figure 2 : is a flow chart of a method for determining a target vehicle according to an embodiment of the present invention, such as Figure 2As shown, the process includes the following steps:

[0032] Step S202: Obtaining a danger parameter for each of N vehicles to obtain N danger parameters, wherein the danger parameter for each of the N vehicles is calculated based on a set of attributes of each vehicle, and N is a positive integer greater than or equal to 2;

[0033] Step S204: Select M vehicles from the N vehicles and obtain activity parameters for each of the M vehicles to obtain M activity parameters, where the M vehicles are the top M vehicles in descending order of risk parameters among the N vehicles, where M is a positive integer greater than or equal to 1 and less than N.

[0034] Step S206: Determine a vehicle feature vector for each of the M vehicles based on the M risk parameters and the M activity parameters corresponding to the M vehicles among the N risk parameters, thereby obtaining M vehicle feature vectors, wherein each of the M vehicle feature vectors includes the risk parameter and activity parameter of the corresponding vehicle among the M vehicles.

[0035] Step S208: determining a target vehicle to be controlled from among the M vehicles based on the M vehicle feature vectors.

[0036] Through the above steps, the dangerousness parameter of each of N vehicles is obtained to obtain N dangerousness parameters. Then, M vehicles ranked in the top M dangerousness parameters are selected from the N vehicles, and the activity parameters of each of the M vehicles are obtained to obtain M activity parameters. Then, based on the M dangerousness parameters and the M activity parameters corresponding to the M vehicles in the N dangerousness parameters, a vehicle feature vector is determined for each of the M vehicles to obtain M vehicle feature vectors, where each vehicle feature vector in the M vehicle feature vectors includes the dangerousness parameter and the activity parameter of the corresponding vehicle in the M vehicles. Then, based on the M vehicle feature vectors, a target vehicle to be controlled is determined from the M vehicles. This achieves the purpose of determining a target vehicle to be controlled from the M vehicles based on the vehicle's dangerousness parameter and activity parameter, avoiding the problem in the related art of determining target vehicles based solely on vehicle characteristics, which results in low efficiency. Therefore, the problem of low efficiency in determining target vehicles in the related art is solved.

[0037] Among them, the execution entity of the above steps can be a terminal, an application, a vehicle control system, a processor with human-computer interaction capabilities configured on a storage device, or a processing device or processing unit with similar processing capabilities, etc., but not limited to these.

[0038] In the above embodiment, the danger parameter of each vehicle in N vehicles is obtained to obtain N danger parameters, wherein the danger parameter of each vehicle in the N vehicles is calculated based on a set of attributes of each vehicle, and N is a positive integer greater than or equal to 2. For example, the above set of attributes may include some or all of the basic attributes, dangerous attributes, illegal behaviors, accident behaviors and active conditions of the vehicle. The basic attributes may be the type of vehicle, such as a large truck, a small truck, a hazardous chemical transport vehicle, a bus, a car, etc. The illegal behaviors may include speeding, overloading, or running a red light, etc. The danger parameter is calculated based on a set of attributes of each vehicle. For example, a different score is assigned to each attribute or behavior in a set of attributes, so as to construct a vehicle safety score system, that is, to calculate the safety score of each vehicle; M vehicles are selected from the N vehicles, and the activity parameters of each vehicle in the M vehicles are obtained to obtain M activity parameters. For example, from the N vehicles M vehicles with the top M hazard parameters are selected from the M vehicles. For example, 100 vehicles with the top 100 hazard parameters are selected from 10,000 vehicles in a certain area (such as a certain administrative district). Then, based on the M hazard parameters and M activity parameters corresponding to the M vehicles among the N hazard parameters, the vehicle feature vectors of each vehicle in the M vehicles are determined to obtain M vehicle feature vectors, that is, vehicle feature vectors of the M vehicles are established, wherein each vehicle feature vector includes the hazard parameter and activity parameter of the vehicle. Then, based on the M vehicle feature vectors, the target vehicle to be controlled is determined from the M vehicles. For example, the vehicle with the highest hazard parameter and activity parameter in the M vehicles is determined as the target vehicle to be controlled. In practical applications, the problem of determining the target vehicle can be abstracted into a minimization multi-objective problem. For example, a non-dominated sorting algorithm is established based on the two objectives of hazard parameter and activity parameter to obtain the target vehicle with the most reliable information and the one that needs to be controlled in a timely manner. This system achieves the goal of identifying a target vehicle from among M vehicles based on its danger and activity parameters, avoiding the problem of low efficiency in target vehicle identification, which occurs in related technologies due to the limited nature of the target vehicle identification method based solely on vehicle characteristics. Therefore, this solves the problem of low efficiency in target vehicle identification in related technologies.

[0039] In an optional embodiment, the determining of the target vehicle as a controlled object among the M vehicles based on the M vehicle feature vectors includes: determining the dominant feature vector of each vehicle among the M vehicles based on the M vehicle feature vectors to obtain M dominant feature vectors, wherein the i-th dominant feature vector among the M dominant feature vectors includes the i-th dominant quantity and the i-th dominated quantity corresponding to the i-th vehicle feature vector among the M vehicle feature vectors, and the i-th dominant quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominance condition. , the dominating condition means that the values of each vector member in the vehicle feature vector are greater than the values of the vector members at the corresponding positions in the i-th vehicle feature vector; the i-th dominated number includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominated condition, and the dominated condition means that the values of each vector member in the vehicle feature vector are smaller than the values of the vector members at the corresponding positions in the i-th vehicle feature vector, and i is a positive integer less than or equal to M; based on the M dominating feature vectors, the target vehicle as the controlled object is determined among the M vehicles. In this embodiment, the dominant feature vector and the vehicle feature vector are in one-to-one correspondence. Other vehicle feature vectors that satisfy the dominating condition (or dominating relationship) and the dominated condition (or dominated relationship) with each vehicle feature vector are determined among the M vehicle feature vectors. For example, the j-th vehicle feature vector dominates the ith vehicle feature vector, which means that the values of each vector member in the j-th vehicle feature vector are greater than the values of the corresponding vector members in the ith vehicle feature vector. For example, the vehicle feature vector includes the danger parameter and activity parameter of the vehicle. If the danger parameter and activity parameter in the j-th vehicle feature vector are greater than the values of the danger parameter and activity parameter in the ith vehicle feature vector, it means that the j-th vehicle feature vector dominates the ith vehicle feature vector. In actual application, In the application, as long as one vector member of the j-th vehicle feature vector (such as the danger parameter) is greater than the corresponding vector member of the ith vehicle feature vector (such as the danger parameter), and other vector members of the j-th vehicle feature vector (such as the activity parameter) are not less than other corresponding vector members of the ith vehicle feature vector (such as the activity parameter), it can be said that the j-th vehicle feature vector dominates the ith vehicle feature vector; for easier understanding, if the ith vehicle feature vector and the j-th vehicle feature vector are marked in a two-dimensional coordinate system (if the vehicle feature vector includes the danger parameter and the activity parameter, one parameter is used as the horizontal coordinate and the other parameter is used as the vertical coordinate), the coordinate point corresponding to the j-th vehicle feature vector is in the upper right corner, or to the right, or directly above the ith vehicle feature vector.The dominant feature vector includes the dominant number and the dominated number. For example, among the M vehicle feature vectors, the number of other vehicle feature vectors that satisfy the domination of the i-th vehicle feature vector is determined, and the number of other vehicle feature vectors that satisfy the domination of the i-th vehicle feature vector is determined. In this way, M dominant feature vectors corresponding to the M vehicle feature vectors can be obtained. Then, based on the M dominant feature vectors, the target vehicle to be controlled is determined among the M vehicles. For example, if the dominant number of the i-th vehicle feature vector is 0, it means that the danger level and activity level of the vehicle corresponding to the i-th vehicle feature vector are better than other vehicles. Therefore, the vehicle can be recommended as a target object to be focused on. Of course, in actual applications, multiple vehicles can be selected as target vehicles to be controlled based on the M dominant feature vectors. Through this embodiment, the purpose of establishing a dominant feature vector and determining a target vehicle among M vehicles based on the dominant feature vector is achieved.

[0040] In an optional embodiment, determining the target vehicle as the controlled object among the M vehicles based on the M dominant feature vectors includes: sorting the M dominant feature vectors in ascending order of the domination quantity included in each dominant feature vector in the M dominant feature vectors to obtain a sorting result; determining the top K dominant feature vectors in the sorting result, and determining K vehicles corresponding to the top K dominant feature vectors among the M vehicles as the target vehicles, wherein K is a positive integer greater than or equal to 1 and less than M. In this embodiment, M (e.g., M = 100, or other) dominant feature vectors are sorted in ascending order of the number of dominant features included in each of the M dominant feature vectors to obtain a sorting result. The top K (e.g., K = 10, or other) dominant feature vectors are then determined from the sorting result, and the K vehicles corresponding to the top K dominant feature vectors are determined as target vehicles. This indicates that the danger level parameters and activity level parameters corresponding to the vehicle feature vectors in the top K positions are greater than the danger level parameters and activity level parameters corresponding to the other vehicle feature vectors after the Kth position, and also indicates that the vehicles in the top K positions are relatively more dangerous. In this embodiment, the top K dominant feature vectors are determined by sorting the M dominant feature vectors, and the K vehicles corresponding to the top K dominant feature vectors are determined as target vehicles.

[0041] In an optional embodiment, the M dominating feature vectors are sorted in ascending order of the dominating quantities included in each dominating feature vector in the M dominating feature vectors to obtain a sorting result, comprising: when the i-th dominating quantity included in the i-th dominating feature vector in the M dominating feature vectors is equal to the j-th dominating quantity included in the j-th dominating feature vector in the M dominating feature vectors, and the i-th dominated quantity included in the i-th dominating feature vector is greater than the j-th dominated quantity included in the j-th dominating feature vector, the i-th dominating feature vector is sorted before the j-th dominating feature vector, where j is a positive integer greater than or equal to 1 and less than or equal to M; or when the i-th dominating quantity is equal to the j-th dominating quantity, When the i-th dominated number is equal to the j-th dominated number and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is greater than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is placed in front of the j-th dominating feature vector; when the i-th dominating number is equal to the j-th dominating number, the i-th dominated number is equal to the j-th dominated number, and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is less than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is placed behind the j-th dominating feature vector. In this embodiment, when the i-th dominating quantity included in the i-th dominating feature vector is equal to the j-th dominating quantity included in the j-th dominating feature vector, the i-th dominated quantity included in the i-th dominating feature vector is compared with the j-th dominated quantity included in the j-th dominating feature vector. If the i-th dominated quantity is greater than the j-th dominated quantity, the i-th dominating feature vector is placed before the j-th dominating feature vector. On the contrary, if the i-th dominated quantity is less than the j-th dominated quantity, the i-th dominating feature vector is placed after the j-th dominating feature vector. When the i-th dominating quantity is equal to the j-th dominating quantity and the i-th dominated quantity is also equal to the j-th dominated quantity, the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is compared with the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector. If the activity parameter in the i-th vehicle feature vector is greater than the activity parameter in the j-th vehicle feature vector, the i-th dominating feature vector is ranked before the j-th dominating feature vector; otherwise, the i-th dominating feature vector is ranked after the j-th dominating feature vector. This embodiment achieves the purpose of sorting M dominating feature vectors, and furthermore, the danger level of each of the M vehicles can be determined.

[0042] In an optional embodiment, the obtaining of the danger parameters of each of the N vehicles to obtain N danger parameters includes: obtaining a set of attributes of each of the N vehicles to obtain N groups of attributes; determining the danger parameters of each of the N vehicles according to the danger sub-parameters corresponding to each attribute in each group of attributes in the N groups of attributes to obtain N danger parameters, wherein each different attribute in the N groups of attributes is pre-set with a corresponding danger sub-parameter. In this embodiment, a set of attributes for each of N vehicles can be obtained to obtain N sets of attributes. For example, the above set of attributes can include some or all of the basic attributes, dangerous attributes, violations, accident behaviors, and activity status of the vehicle. Each attribute can also include one or more sub-attributes. The basic attribute can be the type of vehicle, such as a large truck, a small truck, a hazardous chemical transport vehicle, a bus, a car, etc. The illegal behavior can include speeding, overloading, or running a red light, etc. The accident behavior can include whether the vehicle has been in an accident or the number of accidents, etc. The dangerous attribute can include whether the vehicle is scrapped or the age of the vehicle, etc. Different dangerousness sub-parameters (or scores) can be set in advance for each attribute or sub-attribute. In this way, the dangerousness parameters of each vehicle in the N vehicles can be determined to obtain N dangerousness parameters. Through this embodiment, the purpose of determining the dangerousness parameters of each vehicle based on the pre-set dangerousness sub-parameters of different attributes is achieved.

[0043] In an optional embodiment, the determining of the danger parameter of each vehicle in the N vehicles based on the danger sub-parameter corresponding to each attribute in each group of attributes in the N groups of attributes to obtain N danger parameters includes: determining the danger parameter of the ith vehicle in the N vehicles in the following manner: danger parameter of the ith vehicle = danger sub-parameter corresponding to the basic attribute of the ith vehicle * (sum of danger sub-parameters corresponding to other attributes), wherein the set of attributes of the ith vehicle includes the basic attributes and the other attributes, and the other attributes include the attributes of the set of attributes of the ith vehicle other than the basic attributes, and i is a positive integer greater than or equal to 1 and less than or equal to N; wherein, in the ith vehicle When a group of attributes includes P attributes, the values of the P hazard sub-parameters corresponding to the group of attributes of the i-th vehicle are preset P values, where P is a positive integer greater than or equal to 2; or, when a group of attributes of the i-th vehicle includes P attributes, the values of the Q hazard sub-parameters corresponding to Q attributes in the group of attributes of the i-th vehicle are greater than the corresponding Q values in the preset P values, and the values of the hazard sub-parameters corresponding to the attributes other than the Q attributes in the group of attributes of the i-th vehicle are corresponding values in the P values, and the Q attributes are the top Q attributes with the greatest contribution to the hazard parameter among the P attributes determined according to a preset vehicle hazard prediction model, where Q is a positive integer greater than or equal to 1 and less than P. In this embodiment, for any one of N vehicles (e.g., the i-th vehicle), the vehicle's danger parameter can be calculated according to the formula: Danger parameter of the i-th vehicle = Danger sub-parameter corresponding to the i-th vehicle's basic attribute * (sum of the danger sub-parameters corresponding to the other attributes). Optionally, the values of the P danger sub-parameters corresponding to the P attributes included in the set of attributes for the i-th vehicle are all preset; or, the values of the Q danger sub-parameters corresponding to some of the P attributes (e.g., Q attributes) are greater than the preset values. For example, if the Q attributes are attributes that contribute significantly to the vehicle's danger parameter as determined by the vehicle danger prediction model, the values of the Q danger sub-parameters corresponding to the Q attributes can be adjusted (or corrected) before the vehicle's danger parameter is calculated. This embodiment achieves the purpose of determining a vehicle's danger parameter based on the danger sub-parameters corresponding to each attribute in a set of attributes.

[0044] In an optional embodiment, the acquiring of the activity parameters of each of the M vehicles to obtain the M activity parameters includes: determining the activity time data of each of the M vehicles and the activity space data of each of the M vehicles through the vehicle data in the target area over a predetermined time period, wherein the activity time data is used to represent the activity time of each of the M vehicles in the target area, and the active space data is used to represent the activity distance of each of the M vehicles in the target area; determining the activity parameters of each of the M vehicles based on the activity time data of each of the M vehicles and the activity space data of each of the M vehicles to obtain the M activity parameters. In this embodiment, the active time data and active space data of each of M vehicles are determined based on vehicle data within a target area over a predetermined period of time. For example, statistical analysis can be performed on vehicle traffic data at checkpoints in the target area (e.g., an administrative district) over the past few weeks to determine the active time data and active space data of each vehicle. Then, based on the active time data and active space data of the M vehicles, the activity parameters of each of the M vehicles are determined, resulting in M activity parameters. In other words, a comprehensive evaluation of vehicle activity is performed based on the active time data and active space data of the vehicles.

[0045] In an optional embodiment, determining the activity parameters of each of the M vehicles based on the active time data of each of the M vehicles and the active space data of each of the M vehicles to obtain the M activity parameters includes: determining a first entropy weight of the active time data of each of the M vehicles and a second entropy weight of the active space data of each of the M vehicles; determining the activity parameters of each of the M vehicles based on the active time data of each of the M vehicles, the first entropy weight of the active time data of each of the M vehicles, the active space data of each of the M vehicles, and the second entropy weight of the active space data of each of the M vehicles to obtain the M activity parameters. In this embodiment, the activity level of vehicles is comprehensively evaluated through the two dimensions of active time data and active space data. The entropy weight method can be used to determine the weights of these two indicators (i.e., active time data and active space data), that is, to determine the first entropy weight of the active time data and the second entropy weight of the active space data of each vehicle, and then determine the activity parameter of each vehicle based on the active time data, the first entropy weight, the active space data, and the second entropy weight of each vehicle.

[0046] In an optional embodiment, determining the first entropy weight of the active time data of each of the M vehicles and the second entropy weight of the active space data of each of the M vehicles includes: constructing an activity evaluation parameter matrix X of the M vehicles, wherein the element x in X is i,j Used to represent the jth evaluation parameter of the i-th vehicle among the M vehicles, when j is equal to 1, the j-th evaluation parameter of the i-th vehicle is the active time data of the i-th vehicle, when j is equal to 2, the j-th evaluation parameter of the i-th vehicle is the active space data of the i-th vehicle, wherein i is a positive integer greater than or equal to 1 and less than or equal to M, and j is equal to 1 or 2; the matrix X is normalized to obtain a normalized matrix Z, wherein the element z in the matrix Z is i,j for: According to the matrix Z, a probability matrix P is obtained, wherein the element p in the probability matrix P is i,j It is used to represent the proportion of the j-th evaluation parameter of the i-th vehicle among the M vehicles in the j-th evaluation parameter of the M vehicles; the information entropy of the j-th evaluation parameter is calculated according to the probability matrix P according to the following formula: Among them, e j The information entropy of the j-th evaluation parameter is represented; the entropy weight of the j-th evaluation parameter is calculated based on the information entropy of the j-th evaluation parameter according to the following formula: Among them, d j =1-e j , W1 is the first entropy weight, and W2 is the second entropy weight. In this embodiment, an activity evaluation parameter matrix X of M vehicles is constructed, and the entropy weights of each evaluation parameter are determined according to the above method. For example, W1 is the entropy weight corresponding to the active time data (i.e., the first entropy weight), and W2 is the entropy weight corresponding to the active space data (i.e., the second entropy weight). After determining the first entropy weight and the second entropy weight, the activity parameter A of the i-th vehicle can be calculated according to the formula: i =(W1*x i,1 +W2*x i,2 ).

[0047] In an optional embodiment, obtaining a set of attributes for each of the N vehicles to obtain N sets of attributes includes: updating, based on the received target configuration parameters, risk sub-parameters corresponding to some of the attributes in the N sets of attributes to obtain updated risk sub-parameters corresponding to each attribute in each of the N sets of attributes; and determining, based on the risk sub-parameters corresponding to each attribute in each of the N sets of attributes to obtain N risk parameters, includes: determining, based on the risk sub-parameters corresponding to each attribute in each of the N sets of attributes to obtain the N risk parameters. In this embodiment, the risk sub-parameters corresponding to some of the attributes in the N sets of attributes may be updated based on the target configuration parameters, and then the risk parameter for each of the N vehicles may be determined based on the risk sub-parameters corresponding to each of the updated attributes to obtain the N risk parameters. For example, the target configuration parameters can be parameters configured by the vehicle management department, such as the area under the vehicle management department's jurisdiction or the number of times a vehicle of interest appears. Alternatively, the target configuration parameters can be modifications or configurations made by the vehicle management department to the dangerous attributes. This embodiment achieves a comprehensive determination of the target vehicle's purpose by combining user-configured parameters, thereby enhancing the user experience.

[0048] Obviously, the above-described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The present invention will now be described in detail with reference to the embodiments.

[0049] The embodiment of the present invention provides a method for identifying and recommending key vehicles based on vehicle risk characteristics and active behaviors. Figure 3 This is a flow chart of a method for identifying and recommending key vehicles according to an embodiment of the present invention, comprising the following steps:

[0050] Step 1: High Danger Score Vehicles

[0051] (1) Total vehicle safety score

[0052] A vehicle safety evaluation labeling system is constructed based on basic attributes, hazardous attributes, illegal behaviors, accident behaviors, and active status. Different attributes and behaviors are assigned different scores to create a vehicle safety score. The vehicle safety score formula is Z = basic attributes * (hazardous attributes + illegal behaviors + accident behaviors + active status).

[0053] (2) Vehicles with high risk scores

[0054] Traffic police need to distinguish between vehicles that require special control, vehicles that require special attention, and vehicles that require regular attention. Vehicles with higher scores represent a higher risk level and can be included in the high-risk score category.

[0055] (3) Hazard characteristics by vehicle type

[0056] Ranking the danger levels of vehicles is essentially a multi-attribute decision-making process. Based on the multiple tags of a vehicle, the decision maker subjectively determines the importance (integral value) of each tag based on his or her own experience, and determines the degree of danger and danger ranking of the vehicle through integration.

[0057] The scoring process relies on the subjective experience of decision-makers, which can lead to unreasonable results for dangerous vehicles. Therefore, an objective weighting method is introduced into the research and recommendation process. For each vehicle type, the data shows that the significant characteristics of the labels shared by high-risk and low-risk vehicles are further distinguished to further enhance the differentiation of vehicle dangerousness and better identify vehicles with dangerous characteristics.

[0058] Using labels to distinguish safe from dangerous vehicles is essentially a feature-based classification process. Classification trees are a common approach to this problem. Starting from the root node, a recursive method is used to split the training set data (feature vectors and their corresponding classification results) based on the different values of the features at the leaf nodes. Decision tree algorithms are prone to overfitting, resulting in insufficient generalization and finding a local optimum rather than a global one. Therefore, the concept of random forest bagging is introduced. The classification results of multiple weak classifiers are voted together to form a strong classifier. This approach can address data overfitting and improve accuracy.

[0059] For vehicles with high risk scores, the labels used for integration and the attributes used to determine whether a vehicle is considered high risk serve as the independent variables, while the scores for the labels serve as the feature values. A random forest model is constructed based on the correlation between the individual data points within the independent variable data and their importance to the high-risk vehicle data. A high-risk vehicle prediction model is trained using data from both high-risk and non-high-risk vehicles. When the random forest model's prediction accuracy exceeds 0.8, the vehicle type is considered high risk, and n features with the highest contribution are selected as representative features.

[0060] (4) Correction of vehicles with high risk scores

[0061] After obtaining the high-risk characteristics for each type of vehicle, the risk score for that type of vehicle is adjusted. The adjustment methods include, but are not limited to: 1) adding x points for each risk characteristic; 2) accumulating the scores based on the risk characteristics and their contributions to form the total risk score; 3) ranking vehicles based on the risk characteristics and their scores, comparing them with the original risk ranking. For each vehicle, the vehicle with the lower ranking is selected as the final risk ranking.

[0062] Step 2: Vehicle Active Behavior

[0063] (1) Active period

[0064] Based on the statistical analysis of the vehicle's checkpoint passing data in the past n weeks, we can get the total number of hours the vehicle appeared, and use this as the number of active periods for the vehicle (corresponding to the aforementioned active time data).

[0065] (2) Active spatial range

[0066] Since vehicles may wander within a small space, resulting in a long active path, but they are not active vehicles that traffic police pay attention to, the space is divided into a grid network, and the active distance of vehicles between grids is considered (corresponding to the aforementioned active space data).

[0067] 1) Establish spatial grid: Divide the urban area into one or more grids according to longitude and latitude;

[0068] 2) Grid distance: Based on the complete trajectory of the vehicle within n weeks, each trajectory point is placed in a spatial grid and connected into a trajectory. The total distance between grids is recorded as the spatial activity level. Figure 4 is an example diagram of active space according to an embodiment of the present invention. The active space range of a vehicle can be determined based on the grid distance.

[0069] (3) Active in time and space

[0070] In order to objectively use the two dimensions of active time period and active spatiotemporal range to comprehensively evaluate the activity level of vehicles, the entropy weight method is used to determine the weights of these two indicators.

[0071] 1) Data standardization

[0072] There are n vehicle objects (corresponding to the aforementioned M vehicles), and m activity evaluation indicators, where m=2, representing the active time period and active spatiotemporal range respectively.

[0073] The matrix before normalization is:

[0074]

[0075] The normalized matrix is:

[0076]

[0077] 2) Calculate the proportion of the i-th sample under the j-th indicator and regard it as the probability used in the relative entropy calculation. Based on the previous step, calculate the probability matrix P. Each element in P is as follows:

[0078]

[0079] 3) Calculate the information entropy of each indicator, calculate the information utility value, and normalize it to obtain the entropy weight of each indicator. For the jth indicator, the calculation formula for its information entropy is:

[0080]

[0081] e j The larger it is, the greater the information entropy of the j-th indicator is, and the smaller the corresponding amount of information is.

[0082] Define the information utility value d j , the formula is as follows:

[0083] d j =1-e j

[0084] Normalize the information utility value to obtain the entropy weight of each indicator:

[0085]

[0086] 4) Spatial and temporal activity

[0087] The spatiotemporal activity of the vehicle is: A i =W1*x i1 +W2*x i2 W1 corresponds to the aforementioned first entropy weight, and W2 corresponds to the aforementioned second entropy weight.

[0088] Step 3: Potential Recommended Vehicles

[0089] When traffic police users' management preferences are unknown, vehicle recommendations are designed to more quickly recommend vehicles that meet both dangerousness and activity criteria. The vehicle recommendation problem can be abstracted into a multi-objective minimization problem. By setting dangerousness and activity as two objectives and establishing a non-dominated sorting algorithm, we can identify the most reliable vehicles (corresponding to the aforementioned target vehicles) that require timely supervision.

[0090] (1) Construct all vehicle feature vectors and sort vehicle feature vectors

[0091] Let all t vehicles V={v1,v1,…,v i}(i=1,…,t), the characteristics of the vehicle include the danger level a and the activity level p, then the characteristic vector of the vehicle is v i (p i ,a i )(i=1,…,t). Since the data corresponding to different vehicles may have the same characteristics, there are different vehicle feature vectors X(p i ,a i ,s i )(i=1,…,x),s i Indicates the number of vehicles with the same vehicle characteristics.

[0092] (2) Calculate the dominant solution number m and the dominated solution number n of the vehicle characteristics

[0093] For the vehicle feature vector X i (p i ,a i ,s i ), if there exists an eigenvector X j (p j ,a j ,s j ), if:

[0094] p i ≥p j

[0095] a i >a j

[0096] or

[0097] p i >p j

[0098] a i ≥a j

[0099] like Figure 5 As shown, the vehicle feature vector X i Dominating vehicle feature vector X j , vehicle feature vector X j The vehicle feature vector X i Domination; that is, X i It's X j The dominant solution, total s i The number of dominant solutions, X j It's X i The dominated solution, total s j The number of dominated solutions.

[0100] Based on this, the vehicle feature vector X can be calculated i The number of dominant solutions and the number of dominated solutions are recorded as the dominant eigenvector Xi (m i ,n i )(i=1,…,x).

[0101] (3) Sort the dominant eigenvectors and calculate the ranking r

[0102] One-time sorting:

[0103] For the dominant vector X of the vehicle feature vector i (m i ,n i )(i=1,…,x), first sort in ascending order by the dominating solution number. In particular, if the dominating solution number of a vehicle feature is 0, it indicates that the vehicle is more dangerous and active than other vehicles, that is, the feature vector of this vehicle dominates all other vehicle feature vectors (frequently active and relatively reliable), and its ranking level r should be 1.

[0104] Secondary sorting:

[0105] Assume that vehicle feature X i With vehicle feature X j The number of dominant solutions m i =m j , indicating that vehicle characteristics do not dominate each other, e.g. Figure 6 As shown, X i The risk level a i Higher than X j The risk level a j , but the activity p i <p j ,At this time, it is necessary to consider the number of dominated solutions n of the two vehicle characteristics and perform a secondary sorting in descending order;

[0106] When n i >n j When , it indicates that the vehicle feature dataset is more inclined to be distributed on the activity index. Vehicle feature X j The move towards more centralized data needs to be prioritized. i The sorting level X i (r i ) is higher than X j (r j ).

[0107] Three-time sorting:

[0108] Assume that vehicle feature X i With vehicle feature X j The number of dominated solutions n i =n j , then it is necessary to sort the vehicles three times again according to their activity. Here, the vehicles with higher activity p are assumed to have more reliable information and need to be recommended first.

[0109] Finally, the ranking level r of vehicle feature X is obtained i (i=1,…,x), integrated into the sorted vehicle feature vector X i (p i ,a i ,r i )(i=1,…,x).

[0110] (4) Calculate the recommended ranking of all vehicles

[0111] According to the danger and activity, if the vehicle v i (p i ,a i )(i=1,…,t) and the vehicle x in the sorted vehicle dataset satisfies the sorted vehicle feature vector X j (p j ,a j ,r j )(j=1,…,x)

[0112] If satisfied:

[0113]

[0114]

[0115] but

[0116]

[0117] Among them, if there are multiple vehicles with the same danger level and activity level, they are marked with the same serial number, and finally the recommended ranking data set T(p i ,a i ,r i )(i=1,…,t).

[0118] Step 4: Recommended vehicles based on the daily work of police officers

[0119] The previous step recommends vehicles based on the danger dimension and the activity dimension. In order to better meet the daily work needs of traffic police, this step takes into account the parameters configured by traffic police.

[0120] (1) Vehicles of potential concern to traffic police in the jurisdiction

[0121] Traffic police support configuration parameters include: the scope of the traffic police's jurisdiction and the number of times the vehicle of interest appears.

[0122] Based on these two configurations, vehicles that meet the conditions are filtered out from all vehicles as a set of vehicles of potential interest for work on that day.

[0123] (2) Corrected risk score

[0124] Traffic police support configuring parameters such as danger level labels. The initial correction of high-risk vehicles uses a random forest algorithm to mine important labels for each vehicle type. Once police identify key features of a particular vehicle type, they use these as revised danger level labels to adjust the danger level of potential vehicles of concern.

[0125] (3) Recommended vehicles

[0126] Referring to the method for recommending potential vehicles in step 3, the vehicles with corrected dangerousness levels are ranked to obtain a recommended order for potential vehicles of interest.

[0127] Traffic police support configuration parameters: daily recommended vehicle count. Based on the daily recommended vehicle count, the police recommend vehicles that ultimately meet the police's requirements.

[0128] In the above embodiment, a comprehensive judgment is made from the two perspectives of activity dimension and danger dimension, and a dominance ranking method is used to recommend the control order of key vehicles; the danger level of the vehicle is determined based on the total score obtained by accumulating various danger sub-features of the vehicle, and the random forest algorithm is used to extract important features with higher contribution, and based on these features, the danger level of the vehicle is corrected; considering the spatial activity level of the vehicle, the grid distance theory is used to eliminate the phenomenon of vehicles wandering in a shorter spatial range, so as to better discover spatially active vehicles; the activity level of the vehicle is determined by using the entropy weight method for temporal activity and spatial activity to obtain the comprehensive spatiotemporal activity level; based on the user-configured danger tag, a comprehensive comparison is made with the original total danger score to obtain a list of vehicles that incorporate user concerns and are more dangerous in themselves.

[0129] Compared with related technologies, the embodiments of the present invention have the following advantages: 1) The embodiments of the present invention are based on the actual key vehicle supervision scenarios of traffic police, fully consider the dangerous characteristics and active conditions of vehicles, and recommend daily controlled vehicles for them. This is an important means to truly solve the vehicle management problems of traffic police; 2) When considering the dangerous characteristics of vehicles, the embodiments of the present invention not only consider the total score obtained by accumulating various dangerous sub-features of the vehicle, but also fully consider which sub-features contribute more to the degree of danger for this type of key vehicle model, thereby correcting the vehicle in the dangerous dimension; 3) When considering the active behavior of vehicles, the embodiments of the present invention simultaneously consider the active status of vehicles in both time and space dimensions; 4) Based on the integration of dangerous characteristics that users are concerned about, the number of vehicles they want to pay attention to, the active status, etc., the embodiments of the present invention comprehensively obtain key vehicles that meet the daily management work of traffic police.

[0130] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0131] In this embodiment, a device for determining a target vehicle is also provided. Figure 7 is a structural block diagram of a device for determining a target vehicle according to an embodiment of the present invention, such as Figure 7 As shown, the device includes:

[0132] A first acquisition module 702 is configured to acquire a danger parameter of each of N vehicles, where the danger parameter of each of the N vehicles is calculated based on a set of attributes of each vehicle, and N is a positive integer greater than or equal to 2.

[0133] A second acquisition module 704 is configured to select M vehicles from the N vehicles and acquire an activity parameter of each of the M vehicles to obtain M activity parameters, wherein the M vehicles are the top M vehicles in descending order of riskiness parameters among the N vehicles, where M is a positive integer greater than or equal to 1 and less than N.

[0134] A first determining module 706 is configured to determine a vehicle feature vector for each of the M vehicles based on the M risk parameters among the N risk parameters corresponding to the M vehicles and the M activity parameters, thereby obtaining M vehicle feature vectors, wherein each of the M vehicle feature vectors includes the risk parameter and activity parameter of the corresponding vehicle among the M vehicles;

[0135] The second determining module 708 is configured to determine a target vehicle as a controlled object among the M vehicles based on the M vehicle feature vectors.

[0136] In an optional embodiment, the second determination module 708 includes: a first determination submodule, configured to determine a dominant feature vector of each of the M vehicles based on the M vehicle feature vectors to obtain M dominant feature vectors, wherein the i-th dominant feature vector in the M dominant feature vectors includes the i-th dominant quantity and the i-th dominated quantity corresponding to the i-th vehicle feature vector in the M vehicle feature vectors, and the i-th dominant quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet a dominant condition, where the dominant condition refers to the vehicle feature vector. The value of each vector member in the eigenvector is greater than the value of the vector member at the corresponding position in the i-th vehicle feature vector; the i-th dominated number includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominated condition, and the dominated condition means that the value of each vector member in the vehicle feature vector is smaller than the value of the vector member at the corresponding position in the i-th vehicle feature vector, and i is a positive integer less than or equal to M; the second determination submodule is used to determine the target vehicle as the controlled object among the M vehicles based on the M dominating feature vectors.

[0137] In an optional embodiment, the above-mentioned second determination submodule includes: a sorting unit, used to sort the M dominant feature vectors in order from small to large according to the dominance quantity included in each dominant feature vector in the M dominant feature vectors, to obtain a sorting result; a first determination unit, used to determine the dominant feature vectors ranked in the top K positions in the sorting result, and determine the K vehicles corresponding to the dominant feature vectors ranked in the top K positions among the M vehicles as the target vehicles, wherein K is a positive integer greater than or equal to 1 and less than M.

[0138] In an optional embodiment, the above-mentioned sorting unit includes: a first sorting subunit, which is used to sort the i-th dominating feature vector in front of the j-th dominating feature vector when the i-th dominating quantity included in the i-th dominating feature vector in the M dominating feature vectors is equal to the j-th dominating quantity included in the j-th dominating feature vector in the M dominating feature vectors, and the i-th dominated quantity included in the i-th dominating feature vector is greater than the j-th dominated quantity included in the j-th dominating feature vector, where j is a positive integer greater than or equal to 1 and less than or equal to M; or a second sorting subunit, which is used to sort the i-th dominating feature vector in front of the j-th dominating feature vector when the i-th dominating quantity is equal to the j-th dominating quantity, the i-th dominated quantity is equal to the j-th dominated quantity, and when the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominant feature vector is greater than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominant feature vector, the i-th dominant feature vector is arranged in front of the j-th dominant feature vector; a third sorting subunit is used to arrange the i-th dominant feature vector behind the j-th dominant feature vector when the i-th dominating number is equal to the j-th dominating number, the i-th dominated number is equal to the j-th dominated number, and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominant feature vector is less than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominant feature vector.

[0139] In an optional embodiment, the above-mentioned first acquisition module 702 includes: a first acquisition sub-module, used to obtain a set of attributes of each vehicle in the N vehicles, and obtain N groups of attributes; a third determination sub-module, used to determine the danger parameter of each vehicle in the N vehicles according to the danger sub-parameter corresponding to each attribute in each group of attributes in the N groups of attributes, and obtain N danger parameters, wherein each different attribute in the N groups of attributes is pre-set with a corresponding danger sub-parameter.

[0140] In an optional embodiment, the third determination submodule includes: a second determination unit, for determining the danger parameter of the i-th vehicle among the N vehicles in the following manner: the danger parameter of the i-th vehicle = the danger sub-parameter corresponding to the basic attribute of the i-th vehicle * (the sum of the danger sub-parameters corresponding to other attributes), wherein the set of attributes of the i-th vehicle includes the basic attributes and the other attributes, and the other attributes include the attributes of the set of attributes of the i-th vehicle other than the basic attributes, and i is a positive integer greater than or equal to 1 and less than or equal to N; wherein, in the case where the set of attributes of the i-th vehicle includes P attributes, the i-th vehicle The values of P hazard sub-parameters corresponding to a set of attributes of the i-th vehicle are preset P values, where P is a positive integer greater than or equal to 2; or, when the set of attributes of the i-th vehicle includes P attributes, the values of Q hazard sub-parameters corresponding to Q attributes in the set of attributes of the i-th vehicle are greater than the corresponding Q values in the preset P values, and the values of the hazard sub-parameters corresponding to the attributes other than the Q attributes in the set of attributes of the i-th vehicle are corresponding values in the P values, and the Q attributes are the top Q attributes with the largest contribution to the hazard parameter in the P attributes determined according to a preset vehicle hazard prediction model, where Q is a positive integer greater than or equal to 1 and less than P.

[0141] In an optional embodiment, the above-mentioned second acquisition module 704 includes: a fourth determination submodule, used to determine the active time data of each of the M vehicles and the active space data of each of the M vehicles through the vehicle data in the target area over a predetermined time period, wherein the active time data is used to represent the activity time of each of the M vehicles in the target area, and the active space data is used to represent the activity distance of each of the M vehicles in the target area; a fifth determination submodule, used to determine the activity parameters of each of the M vehicles based on the active time data of each of the M vehicles and the active space data of each of the M vehicles, and obtain the M activity parameters.

[0142] In an optional embodiment, the above-mentioned fifth determination submodule includes: a third determination unit, used to determine the first entropy weight of the active time data of each vehicle in the M vehicles, and the second entropy weight of the active space data of each vehicle in the M vehicles; a fourth determination unit, used to determine the activity parameter of each vehicle in the M vehicles based on the active time data of each vehicle in the M vehicles, the first entropy weight of the active time data of each vehicle in the M vehicles, the active space data of each vehicle in the M vehicles, and the second entropy weight of the active space data of each vehicle in the M vehicles, to obtain the M activity parameters.

[0143] In an optional embodiment, the third determining unit includes: a constructing subunit for constructing an activity evaluation parameter matrix X of the M vehicles, wherein the element x in X i,j Used to represent the jth evaluation parameter of the i-th vehicle among the M vehicles, when j is equal to 1, the j-th evaluation parameter of the i-th vehicle is the active time data of the i-th vehicle, when j is equal to 2, the j-th evaluation parameter of the i-th vehicle is the active space data of the i-th vehicle, wherein i is a positive integer greater than or equal to 1 and less than or equal to M, and j is equal to 1 or 2; a normalization subunit, used to normalize the matrix X to obtain a normalized matrix Z, wherein the element z in the matrix Z is i,j for: The obtaining subunit is used to obtain a probability matrix P according to the matrix Z, wherein the element p in the probability matrix P is i,j is used to represent the proportion of the j-th evaluation parameter of the i-th vehicle among the M vehicles in the j-th evaluation parameter of the M vehicles; a first calculation subunit is used to calculate the information entropy of the j-th evaluation parameter according to the probability matrix P according to the following formula: Among them, e j represents the information entropy of the j-th evaluation parameter; and a second calculation subunit, configured to calculate the entropy weight of the j-th evaluation parameter based on the information entropy of the j-th evaluation parameter according to the following formula: Among them, d j =1-e j , W1 is the first entropy weight, and W2 is the second entropy weight.

[0144] In an optional embodiment, the above-mentioned first acquisition sub-module includes: an updating unit, which is used to update the danger sub-parameters corresponding to some attributes in the N groups of attributes according to the received target configuration parameters, and obtain the danger sub-parameters corresponding to each attribute in each group of attributes in the N groups of attributes after the update; the above-mentioned third determination sub-module includes: a fifth determination unit, which is used to determine the danger parameter of each vehicle in the N vehicles according to the danger sub-parameters corresponding to each attribute in each group of attributes in the N groups of attributes after the update, and obtain the N danger parameters.

[0145] It should be noted that the above modules can be implemented through software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0146] An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored. The computer program is configured to execute the steps of any one of the above method embodiments when running.

[0147] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0148] An embodiment of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0149] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0150] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail here.

[0151] Obviously, those skilled in the art will appreciate that the various modules or steps of the present invention described above can be implemented using a general-purpose computing device, can be centralized on a single computing device, or can be distributed across a network of multiple computing devices. They can be implemented using program code executable by the computing device, and thus, can be stored in a storage device and executed by the computing device. In some cases, the steps shown or described herein can be performed in a different order than that shown, or can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0152] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for determining a target vehicle, characterized in that: include: Obtaining a danger parameter for each of N vehicles to obtain N danger parameters, wherein the danger parameter for each of the N vehicles is calculated based on a set of attributes of each of the vehicles, and N is a positive integer greater than or equal to 2; Selecting M vehicles from the N vehicles and obtaining an activity parameter for each of the M vehicles to obtain M activity parameters, wherein the M vehicles are the top M vehicles in the N vehicles ranked in descending order of riskiness parameters, where M is a positive integer greater than or equal to 1 and less than N; Determining a vehicle feature vector for each of the M vehicles based on M danger parameters and the M activity parameters corresponding to the M vehicles among the N danger parameters, to obtain M vehicle feature vectors, wherein each of the M vehicle feature vectors includes the danger parameter and the activity parameter of the corresponding vehicle among the M vehicles; Determining a dominant feature vector for each of the M vehicles based on the M vehicle feature vectors to obtain M dominant feature vectors, wherein an i-th dominant feature vector in the M dominant feature vectors includes an i-th domination quantity corresponding to the i-th vehicle feature vector in the M vehicle feature vectors, and the i-th domination quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet a domination condition, where the domination condition is that a value of each vector member in the vehicle feature vector is greater than a value of a vector member at a corresponding position in the i-th vehicle feature vector; A target vehicle as a controlled object is determined among the M vehicles based on the M dominant feature vectors.

2. The method according to claim 1, characterized in that The i-th dominant feature vector among the M dominant feature vectors also includes the i-th dominated quantity corresponding to the i-th vehicle feature vector among the M vehicle feature vectors, and the i-th dominated quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet the dominated condition, and the dominated condition means that the values of each vector member in the vehicle feature vector are smaller than the values of the vector members at the corresponding positions in the i-th vehicle feature vector, and i is a positive integer less than or equal to M.

3. The method according to claim 2, characterized in that Determining the target vehicle as a controlled object among the M vehicles based on the M dominant feature vectors includes: Sorting the M dominant eigenvectors in ascending order of the dominance quantity included in each dominant eigenvector in the M dominant eigenvectors to obtain a sorting result; Determine the top K dominant feature vectors in the sorting result, and determine K vehicles corresponding to the top K dominant feature vectors among the M vehicles as the target vehicles, where K is a positive integer greater than or equal to 1 and less than M.

4. The method according to claim 3, characterized in that The M dominant eigenvectors are sorted in ascending order of the dominance quantity included in each dominant eigenvector in the M dominant eigenvectors to obtain a sorting result, including: When the i-th dominating quantity included in the i-th dominating feature vector among the M dominating feature vectors is equal to the j-th dominating quantity included in the j-th dominating feature vector among the M dominating feature vectors, and the i-th dominated quantity included in the i-th dominating feature vector is greater than the j-th dominated quantity included in the j-th dominating feature vector, the i-th dominating feature vector is arranged before the j-th dominating feature vector, where j is a positive integer greater than or equal to 1 and less than or equal to M; or If the i-th dominating number is equal to the j-th dominating number, the i-th dominated number is equal to the j-th dominated number, and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is greater than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is arranged before the j-th dominating feature vector; When the i-th dominating number is equal to the j-th dominating number, the i-th dominated number is equal to the j-th dominated number, and the activity parameter in the i-th vehicle feature vector corresponding to the i-th dominating feature vector is less than the activity parameter in the j-th vehicle feature vector corresponding to the j-th dominating feature vector, the i-th dominating feature vector is arranged behind the j-th dominating feature vector.

5. The method according to claim 1, wherein The step of obtaining the risk parameter of each of the N vehicles to obtain the N risk parameters includes: Obtaining a set of attributes for each of the N vehicles to obtain N sets of attributes; The hazard parameter of each of the N vehicles is determined based on the hazard sub-parameter corresponding to each attribute in each of the N groups of attributes to obtain N hazard parameters, wherein each different attribute in the N groups of attributes is pre-set with a corresponding hazard sub-parameter.

6. The method according to claim 5, characterized in that The determining of the hazard parameter of each of the N vehicles based on the hazard sub-parameter corresponding to each attribute in each of the N groups of attributes to obtain the N hazard parameters includes: The risk parameter of the i-th vehicle among the N vehicles is determined as follows: Danger parameter of the i-th vehicle = Danger sub-parameter corresponding to the basic attribute of the i-th vehicle * (sum of the dangerousness sub-parameters corresponding to other attributes), where the set of attributes of the i-th vehicle includes the basic attribute and the other attributes, and the other attributes include the attributes of the set of attributes of the i-th vehicle other than the basic attribute, and i is a positive integer greater than or equal to 1 and less than or equal to N; Among them, when the group of attributes of the i-th vehicle includes P attributes, the values of the P hazard sub-parameters corresponding to the group of attributes of the i-th vehicle are preset P values, and P is a positive integer greater than or equal to 2; or, when the group of attributes of the i-th vehicle includes P attributes, the values of the Q hazard sub-parameters corresponding to Q attributes in the group of attributes of the i-th vehicle are greater than the corresponding Q values in the preset P values, and the values of the hazard sub-parameters corresponding to the attributes other than the Q attributes in the group of attributes of the i-th vehicle are corresponding values in the P values, and the Q attributes are the top Q attributes of the P attributes that contribute the most to the hazard parameter determined according to a preset vehicle hazard prediction model, and Q is a positive integer greater than or equal to 1 and less than P.

7. The method according to claim 1, characterized in that The acquiring the activity parameter of each of the M vehicles to obtain the M activity parameters includes: Determining, based on vehicle data within a target area over a predetermined period of time, active time data for each of the M vehicles and active spatial data for each of the M vehicles, wherein the active time data is used to indicate the time during which each of the M vehicles is active within the target area, and the active spatial data is used to indicate the distance over which each of the M vehicles is active within the target area; An activity parameter of each of the M vehicles is determined according to the activity time data of each of the M vehicles and the activity space data of each of the M vehicles to obtain the M activity parameters.

8. The method according to claim 7, characterized in that The determining, based on the activity time data of each of the M vehicles and the activity space data of each of the M vehicles, an activity parameter of each of the M vehicles to obtain the M activity parameters includes: Determining a first entropy weight for active time data of each of the M vehicles and a second entropy weight for active spatial data of each of the M vehicles; Based on the active time data of each vehicle in the M vehicles, the first entropy weight of the active time data of each vehicle in the M vehicles, the active space data of each vehicle in the M vehicles, and the second entropy weight of the active space data of each vehicle in the M vehicles, the activity parameter of each vehicle in the M vehicles is determined to obtain the M activity parameters.

9. The method according to claim 8, characterized in that Determining a first entropy weight of the active time data of each of the M vehicles and a second entropy weight of the active space data of each of the M vehicles includes: Construct the activity evaluation parameter matrix X of the M vehicles, where the element x in X i,j used to represent the jth evaluation parameter of the i-th vehicle among the M vehicles, when j is equal to 1, the j-th evaluation parameter of the i-th vehicle is the active time data of the i-th vehicle, and when j is equal to 2, the j-th evaluation parameter of the i-th vehicle is the active space data of the i-th vehicle, wherein i is a positive integer greater than or equal to 1 and less than or equal to M, and j is equal to 1 or 2; The matrix X is normalized to obtain a normalized matrix Z, where the element z in the matrix Z is i,j for: According to the matrix Z, a probability matrix P is obtained, wherein the element p in the probability matrix P is i,j used to represent the proportion of the j-th evaluation parameter of the i-th vehicle among the M vehicles in the j-th evaluation parameter of the M vehicles; The information entropy of the j-th evaluation parameter is calculated according to the probability matrix P according to the following formula: Among them, e j represents the information entropy of the j-th evaluation parameter; The entropy weight of the j-th evaluation parameter is calculated based on the information entropy of the j-th evaluation parameter according to the following formula: Among them, d j =1-e j , W1 is the first entropy weight, and W2 is the second entropy weight.

10. The method according to claim 5, characterized in that Acquiring a set of attributes for each of the N vehicles to obtain N groups of attributes includes: updating, based on the received target configuration parameters, risk sub-parameters corresponding to some of the attributes in the N groups of attributes to obtain updated risk sub-parameters corresponding to each attribute in each of the N groups of attributes; The method of determining the danger parameter of each of the N vehicles based on the danger sub-parameter corresponding to each attribute in each group of the N groups of attributes to obtain N danger parameters includes: determining the danger parameter of each of the N vehicles based on the danger sub-parameter corresponding to each attribute in each group of the N groups of attributes after update to obtain the N danger parameters.

11. A device for determining a target vehicle, characterized in that: include: a first acquisition module configured to acquire a danger parameter of each of N vehicles, thereby obtaining N danger parameters, wherein the danger parameter of each of the N vehicles is calculated based on a set of attributes of each of the vehicles, and N is a positive integer greater than or equal to 2; a second acquisition module configured to select M vehicles from the N vehicles and acquire an activity parameter of each of the M vehicles to obtain M activity parameters, wherein the M vehicles are the top M vehicles in descending order of riskiness parameters among the N vehicles, where M is a positive integer greater than or equal to 1 and less than N; a first determining module, configured to determine a vehicle feature vector for each of the M vehicles based on M danger parameters among the N danger parameters corresponding to the M vehicles and the M activity parameters, thereby obtaining M vehicle feature vectors, wherein each of the M vehicle feature vectors includes a danger parameter and an activity parameter of a corresponding vehicle among the M vehicles; A second determination module is configured to determine a dominant feature vector of each of the M vehicles based on the M vehicle feature vectors to obtain M dominant feature vectors, wherein the i-th dominant feature vector among the M dominant feature vectors includes an i-th domination quantity corresponding to the i-th vehicle feature vector among the M vehicle feature vectors, and the i-th domination quantity includes the number of vehicle feature vectors in the M vehicle feature vectors that meet a domination condition, where the domination condition refers to that the values of each vector member in the vehicle feature vector are greater than the values of the vector members at corresponding positions in the i-th vehicle feature vector; and based on the M dominant feature vectors, determine a target vehicle as a controlled object among the M vehicles.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the method described in any one of claims 1 to 10 when executed by a processor.

13. 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 according to any one of claims 1 to 10 are implemented.

Images