Abnormal slow vehicle detection method and abnormal slow vehicle detection system

Abstract

The present application provides an abnormal slow-moving vehicle detection method and an abnormal slow-moving vehicle detection system, belonging to the field of intelligent transportation technology. It includes: each roadside sensing device collects the sensing data of the sensed objects passing on each road section in real time and generates a sensed object recognition result, and the roadside sensing device sends the sensed object recognition result to the data collection unit corresponding to each road section; each task analysis unit receives the sensed object recognition results uploaded by at least one data collection unit subscribed in advance, and the spatial index unit receives the sensed object recognition results uploaded by each data collection unit and generates an abnormal slow-moving vehicle recognition result; the abnormal slow-moving vehicle warning module determines a first warning information and a second warning information according to the abnormal slow-moving vehicle recognition result and the sensing data of each vehicle, and sends them to the corresponding vehicle. The present application can achieve full-quantification collection of vehicle status information, improve the detection accuracy of abnormal slow-moving vehicles, and realize diversified warning of associated vehicles of abnormal slow-moving vehicles.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent transportation. Specifically, it relates to a method and system for detecting abnormally slow-moving vehicles. Background Art

[0002] With the rapid growth of the national economy, automobiles have become one of the essential means of transportation for people's daily travel. The increase in the number of automobiles will lead to a synchronous increase in potential safety hazards on road traffic. When the driving speeds of vehicles on a certain road are relatively fast, the driving speed of a vehicle in the front is slow, suddenly decelerates or stops, and the following distance of the vehicle behind is relatively close, a rear-end collision is very likely to occur. Therefore, there is an urgent need for a method to efficiently detect abnormally slow-moving vehicles on the road to reduce the occurrence of such traffic accidents.

[0003] In the related art, generally, an in-vehicle communication device is pre-installed on a vehicle, and the driving state information of the vehicle is actively uploaded through the in-vehicle communication device. According to the comparison result between the driving speed in the driving state information of the vehicle and a preset vehicle speed threshold, abnormally slow-moving vehicles are determined, and the vehicle identification code and geographical location are extracted from the driving state information of the abnormally slow-moving vehicles. Based on the vehicle identification code of the abnormally slow-moving vehicle, a safety warning instruction is sent to the abnormally slow-moving vehicle, and a safety prompt message is sent to the associated warning vehicle.

[0004] However, when detecting abnormally slow-moving vehicles based on the related art, it is necessary for the vehicle to pre-install an in-vehicle communication device and for the in-vehicle communication device to actively report the driving state information. However, in actual applications, not all vehicles are equipped with in-vehicle communication devices, which will result in missing the driving state information of some vehicles; in addition, the owners of in-vehicle communication devices may not be willing to share the driving state information with third parties, which will also lead to the lack of driving state information of some vehicles. Therefore, the solutions in the related art have problems such as difficulty in collecting vehicle state information and inaccurate detection results of abnormally slow-moving vehicles. Summary of the Invention

[0005] The purpose of the present application is to provide a method and system for detecting abnormally slow-moving vehicles, which can achieve full-quantification collection of vehicle state information, improve the detection accuracy of abnormally slow-moving vehicles, and realize diversified warnings for associated vehicles of abnormally slow-moving vehicles.

[0006] The embodiments of the present application are implemented as follows:

[0007] In the first aspect of the embodiments of the present application, an abnormal slow-moving vehicle detection method is provided. This method is applied to an abnormal slow-moving vehicle detection system, which includes: a perception data recognition and collection module, an abnormal slow-moving vehicle recognition module, and an abnormal slow-moving vehicle warning module. The perception data recognition and collection module includes: at least one roadside perception device arranged on the roadside and a data collection unit corresponding to each roadside perception device one by one. The abnormal slow-moving vehicle recognition module includes: a plurality of task analysis units and a spatial index unit. The method includes:

[0008] Each roadside perception device collects the perception data of the perceived objects passing on each road section in real time, generates a perceived object recognition result, and sends the perceived object recognition result to the data collection unit corresponding to each road section. The data collection unit stores the received perceived object recognition result in the data topic corresponding to the road section.

[0009] Each task analysis unit receives the perceived object recognition results uploaded by at least one pre-subscribed data collection unit. The spatial index unit receives the perceived object recognition results uploaded by each data collection unit. The spatial index unit and each task analysis unit generate an abnormal slow-moving vehicle recognition result based on the received perceived object recognition results. The abnormal slow-moving vehicle recognition result includes: at least one target abnormal slow-moving vehicle and associated warning vehicles related to the target abnormal slow-moving vehicle, and sends the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result to the abnormal slow-moving vehicle warning module.

[0010] The abnormal slow-moving vehicle warning module determines the first warning information of each target abnormal slow-moving vehicle and the second warning information corresponding to each associated warning vehicle based on the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result, and sends the first warning information to each target abnormal slow-moving vehicle and the second warning information to each associated warning vehicle.

[0011] As a possible implementation, the spatial index unit and each task analysis unit generate an abnormal slow-moving vehicle recognition result based on the received perceived object recognition results, including:

[0012] Each task analysis unit screens the vehicles to be recognized included in the perceived object recognition result according to the received perceived object recognition results to obtain an initial slow-moving vehicle sequence. The initial slow-moving vehicle sequence includes: at least one initial slow-moving vehicle.

[0013] Each task analysis unit determines the number of vehicle trips and the first average vehicle speed in the target area within a preset time period according to the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, the initialized lane information, and the perception data of the vehicles to be recognized passing through each lane within the preset time period.

[0014] The spatial index unit determines the number of associated vehicles corresponding to each initial slow-moving vehicle and the second average vehicle speed according to the received perception object recognition result, the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, and the initialized lane information.

[0015] The spatial index unit determines the total score of each initial slow-moving vehicle according to the train number, the first average vehicle speed, the number of associated vehicles, the second average vehicle speed, and the initialized lane information in the target area within a preset time period.

[0016] The spatial index unit generates an abnormal slow-moving vehicle recognition result according to the total score of each initial slow-moving vehicle and the received perception object recognition result.

[0017] As a possible implementation, each task analysis unit screens the vehicles to be recognized included in the received perception object recognition result for slow-moving vehicles to obtain an initial slow-moving vehicle sequence, including:

[0018] Each task analysis unit loads lane information from the lane database, and the lane information includes: lane range information, lane identifier, lane speed limit information, lane center line, lane starting point longitude and latitude information, and lane ending point longitude and latitude information.

[0019] Each task analysis unit performs initialization processing on the lane information through a preset segmentation algorithm to obtain initialized lane information, and the initialized lane information includes: the mapping relationship between each lane and multiple regions.

[0020] Each task analysis unit determines the longitude and latitude information and speed information of the vehicle to be recognized included in the received perception object recognition result.

[0021] Each task analysis unit determines the initialized lane information corresponding to the target lane where the vehicle to be recognized is currently located according to the longitude and latitude information of the vehicle to be recognized and the initialized lane information.

[0022] Each task analysis unit determines whether the vehicle to be recognized is an initial slow-moving vehicle according to the speed information of the vehicle to be recognized and the lane speed limit information in the initialized lane information corresponding to the target lane.

[0023] If so, the vehicle to be recognized is stored in the initial slow-moving vehicle sequence.

[0024] As a possible implementation, each task analysis unit determines whether the vehicle to be recognized is an initial slow-moving vehicle according to the speed information of the vehicle to be recognized and the lane speed limit information in the initialized lane information corresponding to the target lane, including:

[0025] Determine the minimum slow driving speed of the target lane according to the lane speed limit information in the initialized lane information corresponding to the target lane and a preset first threshold value;

[0026] Determine the maximum slow driving speed of the target lane according to the lane speed limit information in the initialized lane information corresponding to the target lane and a preset second threshold value;

[0027] If the speed information of the vehicle to be recognized is greater than or equal to the minimum slow driving speed of the target lane and less than or equal to the maximum slow driving speed of the target lane, then the vehicle to be recognized is an initial slow driving vehicle.

[0028] As a possible implementation manner, each task analysis unit determines the number of vehicle trips and the first average vehicle speed of the target area within a preset time period according to the perception data of each initial slow driving vehicle in the initial slow driving vehicle sequence, the initialized lane information, and the perception data of the vehicles to be recognized passing through each lane within the preset time period, including:

[0029] Each task analysis unit determines the target area where each initial slow driving vehicle is currently located according to the longitude and latitude information in the perception data of each initial slow driving vehicle and the initialized lane information;

[0030] Each task analysis unit obtains the perception data of each vehicle to be recognized passing through the target area within the preset time period;

[0031] Each task analysis unit determines the number of vehicle trips in the target area within the preset time period according to the number of perception data received within the preset time period;

[0032] Each task analysis unit determines the first average vehicle speed of the target area within the preset time period according to the vehicle speed information in each perception data within the preset time period and the number of vehicle trips in the target area.

[0033] As a possible implementation manner, the spatial index unit determines the number of associated vehicles and the second average vehicle speed corresponding to each initial slow driving vehicle according to the received perception object recognition result, the perception data of each initial slow driving vehicle in the initial slow driving vehicle sequence, and the initialized lane information, including:

[0034] The spatial index unit determines at least one target area corresponding to each initial slow driving vehicle according to the perception data of each initial slow driving vehicle and the initialized lane information;

[0035] The spatial index unit determines the associated vehicles and the number of associated vehicles corresponding to each initial slow driving vehicle according to at least one target area corresponding to each initial slow driving vehicle and the longitude and latitude information in each perception data in the received perception object recognition result;

[0036] The spatial index unit determines the second average speed of the associated vehicles corresponding to each initial slow-moving vehicle according to the perception data of each associated vehicle and the number of associated vehicles.

[0037] As a possible implementation, the spatial index unit determines the total score of each initial slow-moving vehicle according to the train number, the first average speed, the number of associated vehicles, the second average speed, and the initialized lane information in the target area within a preset time period, including:

[0038] The spatial index unit determines the target area score of each initial slow-moving vehicle according to the train number and the first average speed in the target area within a preset time period;

[0039] The spatial index unit determines the surrounding score of each initial slow-moving vehicle according to the number of associated vehicles corresponding to each initial slow-moving vehicle and the second average speed;

[0040] The spatial index unit determines the speed limit score of each initial slow-moving vehicle according to the speed information in the perception data of each initial slow-moving vehicle and the initialized lane information;

[0041] The spatial index unit determines the total score of each initial slow-moving vehicle according to the target area score, the surrounding score, and the speed limit score.

[0042] As a possible implementation, the spatial index unit generates an abnormal slow-moving vehicle recognition result according to the total score of each initial slow-moving vehicle and the received perception object recognition result, including:

[0043] The spatial index unit determines at least one target abnormal slow-moving vehicle according to the total score of each initial slow-moving vehicle;

[0044] The spatial index unit determines the associated vehicles corresponding to each target abnormal slow-moving vehicle according to each target abnormal slow-moving vehicle and the received perception object recognition result;

[0045] The spatial index unit determines the associated warning vehicles related to each target abnormal slow-moving vehicle according to the perception data of the associated vehicles corresponding to each target abnormal slow-moving vehicle.

[0046] As a possible implementation, the abnormal slow-moving vehicle warning module determines the first warning information of each target abnormal slow-moving vehicle and the second warning information corresponding to each associated warning vehicle according to the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result, including:

[0047] The abnormal slow vehicle warning module determines the front-to-back driving order of each target abnormal slow vehicle and each associated warning vehicle based on the perception data of each target abnormal slow vehicle in the abnormal slow vehicle recognition result and the perception data of each associated warning vehicle, and determines the associated warning vehicles in the same lane as each target abnormal slow vehicle;

[0048] The abnormal slow vehicle warning module generates a first warning message for each target abnormal slow vehicle and a second warning message corresponding to each associated warning vehicle based on the front-to-back driving order of each target abnormal slow vehicle and each associated warning vehicle and the perception data of the associated warning vehicles in the same lane as each target abnormal slow vehicle.

[0049] In a second aspect of the embodiments of the present application, an abnormal slow vehicle detection system is provided. The abnormal slow vehicle detection system includes: a perception data recognition and acquisition module, an abnormal slow vehicle recognition module, and an abnormal slow vehicle warning module. The perception data recognition and acquisition module includes: at least one roadside perception device arranged on the roadside and a data acquisition unit corresponding to the roadside perception device one by one. The abnormal slow vehicle recognition module is communicatively connected to each data acquisition unit in the perception data recognition and acquisition module and the abnormal slow vehicle warning module respectively;

[0050] The abnormal slow vehicle detection system is used to execute the steps of the abnormal slow vehicle detection method described in the first aspect above.

[0051] In a third aspect of the embodiments of the present application, an electronic device is provided, and the abnormal slow vehicle detection system described in the second aspect above is deployed in the electronic device.

[0052] The beneficial effects of the embodiments of the present application include:

[0053] An abnormal slow-moving vehicle detection method provided by an embodiment of the present application collects perception data of perceptible objects passing on each road section in a traffic network through various roadside perception devices in real time, generates a perception object recognition result based on the collected perception data of the perceptible objects, and sends the perception object recognition result to the data collection unit corresponding to each road section. The data collection unit stores the received perception object recognition result in the data topic corresponding to the road section; the task analysis unit in the abnormal slow-moving vehicle recognition module obtains the perception object recognition result in the data topic through a subscription method, and the location indexing unit receives the perception object recognition results sent by all data collection units. The abnormal slow-moving vehicle recognition module determines an abnormal slow-moving vehicle recognition result according to the received perception object recognition result and the perception data of the vehicle to be recognized included in the perception object recognition result, and transmits the abnormal slow-moving vehicle recognition result, the perception data of the target abnormal slow-moving vehicle in the abnormal slow-moving vehicle recognition result, and the perception data of the associated warning vehicle related to the target abnormal slow-moving vehicle to the abnormal slow-moving vehicle warning module; the abnormal slow-moving vehicle warning module determines a first warning message corresponding to the target abnormal slow-moving vehicle and a second warning message corresponding to each associated warning vehicle based on the received abnormal slow-moving vehicle recognition result, the perception data of the target abnormal slow-moving vehicle, and the perception data of the associated warning vehicle related to the target abnormal slow-moving vehicle, and sends the first warning message to the corresponding target abnormal slow-moving vehicle and the second warning message to the corresponding associated warning vehicle. Among them, by deploying multiple roadside perception devices on each road in the traffic network, full-quantitative collection of the perception data of each vehicle passing on each road is realized, and the perception object recognition results of each roadside perception device are transmitted to the abnormal slow-moving vehicle detection module through the data collection unit. The abnormal slow-moving vehicle detection module comprehensively determines abnormal slow-moving vehicles based on the perception data of each vehicle to be recognized, the perception data of each vehicle passing around each vehicle to be recognized, and the road conditions of each road, and sends the abnormal slow-moving vehicle recognition result to the abnormal slow-moving vehicle warning module. The abnormal slow-moving vehicle warning module generates diversified second warning messages according to the threat degree of the target abnormal slow-moving vehicle to the vehicles passing around it. In this way, full-quantitative collection of vehicle status information can be achieved, the detection accuracy of abnormal slow-moving vehicles can be improved, and diversified warning of associated vehicles of abnormal slow-moving vehicles can be realized. In this way, full-quantitative collection of vehicle status information can be achieved, the detection accuracy of abnormal slow-moving vehicles can be improved, and diversified warning of associated vehicles of abnormal slow-moving vehicles can be realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0055] Figure 1 Schematic diagram of a structure of an abnormal slow - moving vehicle detection system provided by an embodiment of the present application;

[0056] Figure 2 Schematic diagram of a structure of an abnormal slow - moving vehicle recognition module provided by an embodiment of the present application;

[0057] Figure 3 Flowchart of the first abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0058] Figure 4 Schematic diagram of a structure of a perception data recognition and acquisition module provided by an embodiment of the present application;

[0059] Figure 5 Flowchart of the second abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0060] Figure 6 Flowchart of the third abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0061] Figure 7 Flowchart of the fourth abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0062] Figure 8 Flowchart of the fifth abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0063] Figure 9 Flowchart of the sixth abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0064] Figure 10 Flowchart of the seventh abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0065] Figure 11 Flowchart of the eighth abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0066] Figure 12 Flowchart of the ninth abnormal slow - moving vehicle detection method provided by an embodiment of the present application;

[0067] Figure 13 Schematic diagram of a structure of an electronic device provided by an embodiment of the present application.

[0068] Description of the Drawings: 10: Abnormal Slow-moving Vehicle Detection System; 101: Perception Data Identification and Acquisition Module; 1011: Roadside Perception Device; 111: Millimeter-wave Radar; 1012: Data Acquisition Unit; 102: Abnormal Slow-moving Vehicle Identification Module; 1021: Task Analysis Unit; 1022: Spatial Index Unit; 103: Abnormal Slow-moving Vehicle Warning Module; 20: Electronic Device. Detailed Implementation Manner

[0069] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. Usually, the components of the embodiments of the present application described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0070] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application claimed, but merely represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.

[0071] It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0072] Currently, vehicle-mounted communication devices are often pre-installed on vehicles. The vehicle-mounted communication devices actively upload the driving state information of the vehicles to the cloud server. The cloud server determines abnormal slow-moving vehicles based on the comparison result between the driving speed in the driving state information of the vehicles and a preset vehicle speed threshold, extracts the vehicle identification code and geographical location from the driving state information of the abnormal slow-moving vehicles, and sends a safety warning instruction to the abnormal slow-moving vehicles and a safety prompt message to associated warning vehicles based on the extracted vehicle identification code and geographical location. However, this solution requires vehicles to be pre-installed with vehicle-mounted communication devices and for the vehicle-mounted communication devices to actively report the driving state information. In actual applications, not all vehicles are equipped with vehicle-mounted communication devices, which may lead to missing the driving state information of some vehicles; in addition, the owners of vehicle-mounted communication devices may not be willing to share the driving state information with third parties, which may also lead to the missing of the driving state information of some vehicles.

[0073] To this end, the embodiments of the present application provide an abnormal slow-moving vehicle detection method. The roadside perception device collects the perception data of the perceived objects passing through each section in real time, and sends the perceived object recognition result to the data collection unit corresponding to each section. The data collection unit stores the received perceived object recognition result in the data topic corresponding to the section. The task analysis unit obtains the perceived object recognition results in at least one data collection unit by subscribing to the data topic. The spatial index unit receives the perceived object recognition results sent by all data collection units. The task analysis unit and the spatial index unit generate an abnormal slow-moving vehicle recognition result according to the received perceived object recognition results, and send the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result to the abnormal slow-moving vehicle warning module. The abnormal slow-moving vehicle warning module generates a first warning message corresponding to each target abnormal slow-moving vehicle and a second warning message of the associated warning vehicle related to each target abnormal slow-moving vehicle according to the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result, and sends them to the corresponding vehicle. In this way, it is possible to achieve full-quantitative collection of vehicle status information, improve the detection accuracy of abnormal slow-moving vehicles, and realize diversified warning of associated vehicles of abnormal slow-moving vehicles.

[0074] In addition, the traditional method for identifying abnormal slow-moving vehicles mainly determines whether a vehicle is abnormally slow-moving by comparing the vehicle's own driving speed with the speed experience values of each road in the traffic network. For example, the normal speed of vehicles passing through Road A is usually 60 km / h, but the driving speed of vehicle a on Road A is 20 km / h, then vehicle a is determined to be an abnormal slow-moving vehicle. However, when there are sudden traffic accidents, peak travel during holidays, road damage and repair, etc., this determination scheme will seriously affect the driving speed of the vehicle, and the speed experience value of the road will not be of reference significance. At this time, misjudgment results will occur if the speed experience value of the road is still relied on.

[0075] The abnormal slow-moving vehicle detection method and the abnormal slow-moving vehicle detection system provided by the embodiments of the present application are explained in detail below with reference to the accompanying drawings.

[0076] Figure 1 The structural schematic diagram of an abnormal slow-moving vehicle detection system provided by the present application is shown in Figure 1, the abnormal slow vehicle detection system 10 provided by the embodiments of the present application includes: a perception data identification and collection module 101, an abnormal slow vehicle identification module 102, and an abnormal slow vehicle warning module 103. Among them, the perception data identification and collection module 101 includes: at least one roadside perception device 1011 arranged on the roadside and a data collection unit 1012 corresponding to the roadside perception device 1011 one by one. The roadside perception device 1011 is communicatively connected to its corresponding data collection unit 1012, and the data collection unit 1012 is also communicatively connected to the abnormal slow vehicle identification module 102, and the abnormal slow vehicle identification module 102 is also communicatively connected to the abnormal slow vehicle warning module 103.

[0077] Optionally, the perception data identification and collection module 101 is mainly used to collect the perception data of the perception objects passing on each road in the traffic network to achieve the full-volume collection of the perception data. Among them, the perception objects can be pedestrians, bicycles, non-motor vehicles, cars, trucks, trains, etc., and the present application does not make specific limitations on this.

[0078] Optionally, the perception data identification and collection module 101 is composed of multiple roadside perception devices 101 and multiple data collection units 1012. The roadside perception devices 1011 are deployed on the support rods of the basic traffic facilities installed on the roadside. The basic traffic facilities installed on the roadside can be street lights, traffic lights, etc., and the present application does not make specific limitations on this.

[0079] It should be noted that one roadside perception device 1011 can be deployed on each road in the traffic network, or multiple roadside perception devices 1011 can be deployed. The specific number of roadside perception devices 1011 deployed on each road is determined by the actual road conditions of each road, and the present application does not make specific limitations on this.

[0080] Optionally, the roadside perception device 1011 in the perception data identification and collection module 101 corresponds to the data collection unit 1012 one by one, that is, the perception data set collected by one roadside perception device 1011 is transmitted to its corresponding data collection unit 1012. Among them, one roadside perception device 1011 is responsible for a section of the road. The perception data of the perception objects passing on a section of the road can be collected by multiple perception data collection devices. The roadside perception device 1011 generates a perception object identification result based on the perception data of the perception objects collected by the multiple perception data collection devices integrated inside, and transmits the perception object identification result to the data collection unit 1012 corresponding to the roadside perception device 1011.

[0081] In an optional implementation manner, refer to Figure 2, in the abnormal slow vehicle detection system provided by this application, the abnormal slow vehicle recognition module 102 includes: a plurality of task analysis units 1021 and a spatial index unit 1022. Among them, the task analysis units 1021 in the abnormal slow vehicle recognition module 102 obtain the perception object recognition results stored in each data acquisition unit 1012 by subscribing. The spatial index unit 1022 in the abnormal slow vehicle recognition module 102 receives the perception data sent by all data acquisition units 1012. The abnormal slow vehicle recognition module 102 determines whether there are abnormal slow vehicles on each road in the transportation network based on the received perception object recognition results and perception data, and transmits the judgment results and the perception data of each vehicle included in the judgment results to the abnormal slow vehicle warning module 103.

[0082] Optionally, under the premise of a large amount of perception data, the computing power of a single task analysis unit 1021 is limited. A single task analysis unit 1021 can only process part of the perception data in a perception object recognition result for data analysis. Thus, the abnormal slow vehicle recognition module 102 requires a combination of multiple task analysis units 1021 to jointly complete the data analysis task of a perception object recognition result. Collecting and distributing the perception data of the same road section or the same road intersection uniformly is beneficial to the orderly input of data for the task analysis unit 1021, and can also avoid the problem of data timing disorder caused by different network transmission rates of different data acquisition units 1012.

[0083] Optionally, the abnormal slow vehicle warning module 103 further determines the warning information of the abnormal slow vehicle and the warning information of the associated warning vehicles related to the abnormal slow vehicle according to the received judgment results and the received perception data of each vehicle.

[0084] Figure 3 is a flowchart of an abnormal slow vehicle detection method provided by this application. This method is applied to the above abnormal slow vehicle detection system 10. See Figure 3 , an embodiment of this application provides an abnormal slow vehicle detection method, including:

[0085] S301. Each roadside perception device real-time collects the perception data of the perception objects passing on each road section, and generates a perception object recognition result. The roadside perception device sends the perception object recognition result to the data acquisition unit corresponding to each road section, and the data acquisition unit stores the received perception object recognition result into the data topic corresponding to the road section.

[0086] Optionally, the perception data refers to various environmental signals and information obtained by roadside perception devices through internal algorithms such as sensing technology and camera technology. The perception data includes: the identifier of the perceived object, the type of the perceived object, the latitude and longitude information of the perceived object, the speed information of the perceived object, the heading angle information of the perceived object, etc. The perceived object refers to pedestrians, vehicles, etc. passing on each section of the transportation network. The perception data of the perceived object identification result refers to the set of perception data of the vehicles to be identified passing on each section collected by each roadside perception device at a certain moment, that is, the perception data of multiple perceived objects is included in the perception data of the perceived object identification result transmitted by a roadside perception device to the corresponding data collection unit. Among them, the vehicle to be identified refers to the objects belonging to vehicles among the perceived objects passing on each road in the transportation network. The vehicles to be identified may include abnormally slow vehicles, normally traveling vehicles, etc. The vehicles to be identified may include: passenger cars, trucks, freight cars, etc. The present application does not make specific limitations on this.

[0087] It should be noted that the roadside perception device can collect the perception data of all passing objects passing on each road in the transportation network, but the roadside perception device can filter out the perception data of the perceived objects that do not belong to vehicles among the perceived objects, so that the perception data of only vehicles is included in the perception data of the perceived object identification result.

[0088] Optionally, each roadside perception device perceives passing objects such as vehicles and pedestrians passing on each section of the transportation network to identify information such as the type, latitude and longitude information, and speed of the perceived object. Among them, each roadside perception device performs global perception on a section of the road to achieve full-scale collection of vehicle status information. In this way, the driving data of a certain vehicle will not be missing because the vehicle does not pre-install an in-vehicle communication device.

[0089] In addition, it is necessary to deploy roadside perception devices on various traffic arteries in the city to obtain the full-scale vehicle driving data of each traffic artery in the city, so as to achieve more accurate positioning of abnormally slow vehicles, and through the abnormally slow vehicle warning module, warn the drivers of other vehicles around the abnormally slow vehicle to pay attention to avoiding the abnormally slow vehicle, so as to reduce the occurrence of traffic accidents.

[0090] Optionally, the roadside perception device sends the perception data of the perceived object identification result to the corresponding data collection unit, and the data collection unit stores the received perception data of the perceived object identification result in the data topic corresponding to each section. Among them, the name of the data topic can be named by the unique identifier of the section or by the identifier of the data collection unit. The present application does not make specific limitations on this. It should be noted that each data collection unit has its corresponding data topic. The data topic is used as the middleware for data transmission. The data collection unit is the message publisher, and the task analysis unit in the abnormally slow vehicle identification module is the message subscriber.

[0091] Optionally, the data acquisition unit obtains the perception object recognition results generated by the corresponding roadside perception device through the UDP network communication protocol and conducts unified integration, so that the perception data of the vehicles to be recognized passing on each section of the traffic network are grouped by section, that is, the perception object recognition results stored in a single data acquisition unit are the perception data of the vehicles passing on one section, facilitating the fragmentation and diversion of the perception data.

[0092] It should be noted that the perception object data recognition and acquisition module can not only send the perception object recognition results to the abnormal slow-moving vehicle recognition module, but also store the perception object recognition results in the database for subsequent traceability.

[0093] S302. Each task analysis unit receives the perception object recognition results uploaded by at least one pre-subscribed data acquisition unit. The spatial index unit receives the perception object recognition results uploaded by each data acquisition unit. The spatial index unit and each task analysis unit generate the abnormal slow-moving vehicle recognition results based on the received perception object recognition results. The abnormal slow-moving vehicle recognition results include: at least one target abnormal slow-moving vehicle and the associated warning vehicles related to the target abnormal slow-moving vehicle, and send the abnormal slow-moving vehicle recognition results and the perception data of each vehicle in the abnormal slow-moving vehicle recognition results to the abnormal slow-moving vehicle warning module.

[0094] Optionally, the data acquisition unit transmits the perception object recognition results to the task analysis unit through communication protocols such as MQTT. The MQTT communication protocol can support the message publishing and message subscription modes. The task analysis unit in the abnormal slow-moving vehicle recognition module can subscribe to the data topic through the MQTT communication protocol. The perception data recognition and acquisition unit can send the perception data of the vehicles to be recognized in the data topic pre-subscribed by the abnormal slow-moving vehicle recognition module to the corresponding task analysis unit. Among them, the task analysis unit can obtain the perception object recognition results of different data acquisition units by subscribing to different data topics.

[0095] Optionally, the spatial index unit receives the perception object recognition results sent by all data acquisition units. The abnormal slow-moving vehicle recognition module determines the abnormal slow-moving vehicles based on the perception object recognition results corresponding to the data topics pre-subscribed by each task analysis unit and the comprehensive perception object recognition results received by the spatial index unit to obtain the abnormal slow-moving vehicle recognition results. Among them, the abnormal slow-moving vehicle recognition results include: at least one target abnormal slow-moving vehicle and the associated warning vehicles related to the target abnormal slow-moving vehicle. It should be noted that the perception object recognition results contain the perception data of each perception object.

[0096] Optionally, a single task analysis unit can process the recognition results of sensed objects transmitted by one data acquisition unit, or can process the recognition results of sensed objects transmitted by multiple data acquisition units, and output the recognition results of abnormally slow-moving vehicles; the spatial index unit receives the recognition results of sensed objects transmitted by all data acquisition units, and updates the positioning information of all sensed objects in real time. The spatial index unit participates in the algorithm for identifying abnormally slow-moving vehicles in each task analysis unit.

[0097] Optionally, the target abnormally slow-moving vehicle refers to a vehicle that actually moves abnormally slowly among all vehicles passing on each road in the transportation network. The associated warning vehicle refers to a vehicle that has a risk of traffic accident among the vehicles passing around the target abnormally slow-moving vehicle. Specifically, the associated warning vehicle is determined by the threat level of the target abnormally slow-moving vehicle to the vehicles passing around the target abnormally slow-moving vehicle. If the target abnormally slow-moving vehicle poses no threat to the vehicles passing around it, then the vehicle passing around the target abnormally slow-moving vehicle cannot be used as an associated warning vehicle.

[0098] It should be noted that when the abnormally slow-moving vehicle recognition module conducts the recognition of abnormally slow-moving vehicles, it will make a comprehensive judgment in combination with the surrounding environment of the abnormally slow-moving vehicle, such as aggregating reference information such as the vehicle speeds and vehicle distances of the vehicles passing around the abnormally slow-moving vehicle, so as to reduce the influence of the driving environment on the recognition results of abnormally slow-moving vehicles.

[0099] Optionally, the abnormally slow-moving vehicle recognition module determines abnormally slow-moving vehicles based on the received recognition results of sensed objects to obtain the recognition results of abnormally slow-moving vehicles, and sends the recognition results of abnormally slow-moving vehicles and the sensed data of each vehicle included in the recognition results of abnormally slow-moving vehicles to the abnormally slow-moving vehicle warning module.

[0100] S303. The abnormally slow-moving vehicle warning module determines the first warning information for each target abnormally slow-moving vehicle and the second warning information corresponding to each associated warning vehicle according to the recognition results of abnormally slow-moving vehicles and the sensed data of each vehicle in the recognition results of abnormally slow-moving vehicles, and sends the first warning information to each target abnormally slow-moving vehicle and sends the second warning information to each associated warning vehicle.

[0101] Optionally, the first warning message refers to the warning message sent to the mobile application of the driver of a vehicle that is actually abnormally slow. The first warning message is mainly used to warn the driver of the target abnormally slow vehicle to increase the speed. Among them, the first warning messages received by the target abnormally slow vehicles passing on each road in the transportation network may be the same or different, which is specifically determined by the driving data and driving environment of the target abnormally slow vehicle, and the present application does not make specific limitations in this regard; the second warning message refers to the warning message sent to the mobile application of the driver of the surrounding vehicles that may be affected by the target abnormally slow vehicle. The second warning message is mainly used to warn the drivers of the vehicles passing around the target abnormally slow vehicle to avoid the target abnormally slow vehicle.

[0102] It should be noted that when there are multiple associated warning vehicles related to the target abnormally slow vehicle, the second warning messages for each associated warning vehicle are generated based on the threat level of the target abnormally slow vehicle to each associated warning vehicle. That is, the second warning messages for the associated warning vehicles related to the same target abnormally slow vehicle can be warned in grades. The second warning messages for the associated warning vehicles related to the same target abnormally slow vehicle may be the same or different, and the present application does not make specific limitations in this regard.

[0103] Optionally, the abnormally slow vehicle warning module determines the first warning message corresponding to the target abnormally slow vehicle and the second warning messages corresponding to each associated warning vehicle based on the received abnormally slow vehicle recognition result, the perception data of the target abnormally slow vehicle in the abnormally slow vehicle recognition result, and the perception data of the associated warning vehicles related to the target abnormally slow vehicle.

[0104] In the embodiment of the present application, the perception data of the perceived objects passing on each road section in the traffic network is collected in real time by various roadside perception devices, and the perception object recognition result is generated based on the collected perception data of the perceived objects, and the perception object recognition result is sent to the data collection unit corresponding to each road section. The data collection unit stores the received perception object recognition result into the data topic corresponding to the road section; the task analysis unit in the abnormal slow-moving vehicle recognition module obtains the perception object recognition result in the data topic through subscription, and the positioning index unit receives the perception object recognition results sent by all data collection units. The abnormal slow-moving vehicle recognition module determines the abnormal slow-moving vehicle recognition result according to the received perception object recognition result and the perception data of the vehicle to be recognized included in the perception object recognition result, and transmits the abnormal slow-moving vehicle recognition result, the perception data of the target abnormal slow-moving vehicle in the abnormal slow-moving vehicle recognition result, and the perception data of the associated warning vehicle related to the target abnormal slow-moving vehicle to the abnormal slow-moving vehicle warning module; the abnormal slow-moving vehicle warning module determines the first warning information corresponding to the target abnormal slow-moving vehicle and the second warning information corresponding to each associated warning vehicle based on the received abnormal slow-moving vehicle recognition result, the perception data of the target abnormal slow-moving vehicle, and the perception data of the associated warning vehicle related to the target abnormal slow-moving vehicle, and sends the first warning information to the corresponding target abnormal slow-moving vehicle and sends the second warning information to the corresponding associated warning vehicle. Among them, by deploying a plurality of roadside perception devices on each road in the traffic network, the perception data of each vehicle passing on each road is collected in a full-quantitative manner, and the perception object recognition result of each roadside perception device is transmitted to the abnormal slow-moving vehicle detection module through the data collection unit. The abnormal slow-moving vehicle detection module comprehensively determines the abnormal slow-moving vehicle based on the perception data of each vehicle to be recognized, the perception data of each vehicle passing around each vehicle to be recognized, and the road conditions of each road, and sends the abnormal slow-moving vehicle recognition result to the abnormal slow-moving vehicle warning module. The abnormal slow-moving vehicle warning module generates diversified second warning information according to the threat degree of the target abnormal slow-moving vehicle to the surrounding passing vehicles. In this way, it is possible to achieve full-quantitative collection of vehicle status information, improve the detection accuracy of abnormal slow-moving vehicles, and realize diversified warning of associated vehicles of abnormal slow-moving vehicles.

[0105] Figure 4 It is a schematic structural diagram of a perception data recognition and collection module provided by the present application. Refer to Figure 4, in the abnormal slow vehicle detection system 10 provided by the embodiments of the present application, the roadside sensing device 1011 in the sensing data recognition and acquisition module 101 includes at least one millimeter-wave radar 111. The sensing data acquisition device in the roadside sensing device 1011 can be implemented by the millimeter-wave radar 111, and the roadside sensing device 1011 includes a group of millimeter-wave radars 111. Among them, the millimeter-wave radars 111 belonging to the same intersection are a group of millimeter-wave radars, and the millimeter-wave radars 111 belonging to the same road in the non-intersection area are a group of millimeter-wave radars. The present application does not make specific limitations on this.

[0106] It should be noted that the number of millimeter-wave radars 111 included in each roadside sensing device 1011 is not necessarily equal, which is specifically determined by the road conditions of each road, such as road width, etc. The present application does not make specific limitations on this.

[0107] Optionally, a millimeter-wave radar 111 can detect and sense a road intersection direction or a section of road. Through multiple millimeter-wave radars 111, the whole area of multiple traffic roads can be sensed, and through multiple roadside sensing devices 1011, a whole area sensing network at the road network level can be jointly realized.

[0108] Optionally, from a planar view, the scanning range of the millimeter-wave radar 111 is a combination of a short-distance fan-shaped area with an angle of about 30° and a long-distance fan-shaped area with an angle of about 15°. The scanning range of the millimeter-wave radar 111 is about 200 meters. One millimeter-wave radar 111 in the roadside sensing device 1011 can perform whole-area sensing for an intersection direction, and each road and each road intersection in the traffic network can be covered by the roadside sensing device 1011. If it is necessary to collect the full amount of vehicle passing data of an intersection, a millimeter-wave radar needs to be deployed in each intersection direction to ensure the integrity of the vehicle passing data of the intersection.

[0109] Optionally, the scanning range of the millimeter-wave radar is determined by the millimeter-wave emission range of the millimeter-wave radar. The millimeter-wave radar performs real-time scanning within the scanning range by emitting millimeter-waves to obtain data such as the size and distance of each sensing object within the scanning range, and then combines internal algorithms to calculate information such as the speed, direction, and type of each sensing object to output structured sensing data.

[0110] Optionally, each roadside sensing device has its corresponding target section. The target section is used to indicate the area range for each roadside sensing device to perform whole-area sensing. The target section can be any road or any intersection in the traffic network. The area range of the target section is determined by the scanning ranges of the millimeter-wave radars in the roadside sensing device. The roadside sensing device collects the sensing data of the passing objects on the target section through each millimeter-wave radar and generates the recognition result of the sensing objects on the target section.

[0111] In a possible implementation, refer to Figure 5 , the operation of step S302 can specifically be:

[0112] S501. Each task analysis unit screens slow-moving vehicles from the to-be-identified vehicles included in the received perception object recognition result to obtain an initial slow-moving vehicle sequence, where the initial slow-moving vehicle sequence includes: at least one initial slow-moving vehicle.

[0113] Optionally, after receiving the perception object recognition result, the task analysis unit performs pre-data analysis on the perception data of the to-be-identified vehicles included in the perception object recognition result to implement slow-moving vehicle screening, so as to obtain an initial slow-moving vehicle sequence. An initial slow-moving vehicle refers to a vehicle that may have abnormal slow movement, and an initial slow-moving vehicle sequence refers to a set including at least one vehicle that may have abnormal slow movement.

[0114] S502. Each task analysis unit determines the number of vehicle trips and the first average vehicle speed in the target area within a preset time period according to the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, the initialized lane information, and the perception data of the to-be-identified vehicles passing through each lane within the preset time period.

[0115] Optionally, the preset time period is a time range preset by the user. The preset time period can be 5 minutes, 10 minutes, etc., and the present application does not make specific limitations thereto.

[0116] Optionally, the task analysis unit determines the number of vehicle trips and the first average speed passing through the target area within the preset time period according to the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, the initialized lane information, and the perception data of the to-be-identified vehicles passing through each lane in the traffic network within the preset time period. Wherein, the number of vehicle trips refers to the number of vehicles passing through the target area within the preset time period, the first average speed refers to the average value of the vehicle speeds of all vehicles passing through the target area within the preset time period, and the target area refers to the area grid where the initial slow-moving vehicle is located.

[0117] S503. The spatial index unit determines the number of associated vehicles and the second average vehicle speed corresponding to each initial slow-moving vehicle according to the received perception object recognition result, the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, and the initialized lane information.

[0118] Optionally, based on all the received perception object recognition results, the perception data of each initial slow vehicle in the initial slow vehicle sequence, and the initialized lane information, the spatial indexing unit can determine the number of associated vehicles corresponding to each initial slow vehicle and the second average speed. Here, the associated vehicle refers to the vehicle traveling around the initial slow vehicle during the driving process of the initial slow vehicle. The associated vehicle is the surrounding vehicle traveling within a preset range around the initial slow vehicle. The preset range can be 50 meters, 60 meters, etc., and the present application does not make specific limitations in this regard; the second average speed refers to the average speed of the associated vehicles of the initial slow vehicle.

[0119] S504. The spatial indexing unit determines the total score of each initial slow vehicle according to the train number, the first average speed, the number of associated vehicles, the second average speed, and the initialized lane information in the target area within a preset time period.

[0120] Optionally, the total score of the initial slow vehicle is used to evaluate the likelihood of the initial slow vehicle actually experiencing abnormal slowdown. The task analysis unit and the spatial indexing unit calculate the total score of each initial slow vehicle based on the received perception object recognition results and each initial slow vehicle in the initial slow vehicle sequence.

[0121] S505. The spatial indexing unit generates an abnormal slow vehicle recognition result according to the total score of each initial slow vehicle and the received perception object recognition results.

[0122] Optionally, the spatial indexing unit generates an abnormal slow vehicle recognition result according to the total score of each initial slow vehicle and all the received perception object recognition results.

[0123] In a possible implementation manner, referring to Figure 6 , the operation of step S501 can specifically be:

[0124] S601. Each task analysis unit loads lane information from the lane database. The lane information includes: lane range information, lane identifier, lane speed limit information, lane centerline, lane starting point longitude and latitude information, and lane ending point longitude and latitude information.

[0125] Optionally, the lane database refers to a database used to store the lane information of all roads in the traffic network, and the lane information refers to the basic information of each road in the traffic network. Among them, the lane range information refers to the lane range enclosed by multiple longitude and latitude points of each road, the lane identifier refers to the name or number of each road in the traffic network, the lane speed limit information refers to the restricted speed of each road in the traffic network, the lane centerline indicates the centerline of each road in the traffic network, the lane starting point longitude and latitude information is used to indicate the starting point of each road in the traffic network, and the lane ending point longitude and latitude information is used to indicate the ending point of each road in the traffic network.

[0126] Optionally, after the abnormal slow vehicle recognition module starts running, it loads the lane information of each road in the lane database so that a road network structure is formed in the abnormal slow vehicle recognition module.

[0127] S602. Each task analysis unit performs initialization processing on the lane information through a preset segmentation algorithm to obtain initialized lane information. The initialized lane information includes: the mapping relationship between each lane and multiple regions.

[0128] Optionally, the preset segmentation algorithm can be implemented by Google's S2 algorithm. After each task analysis unit loads the lane information of each road in the traffic network, it extracts the lane range information of each road in the traffic network and divides the lane range of each road into multiple equal-area regional grids through the preset cutting algorithm. Each regional grid can be 300 square meters, and the present application does not make specific limitations on this.

[0129] It should be noted that the preset segmentation algorithm can cut each road into regional grids with similar sizes, and the size of the cut regional grids can be adjusted, and the regional grids in this dimension can be quickly located through longitude and latitude.

[0130] Optionally, the task analysis unit can implement the mapping relationship between the cut regional grids and the lane information through the HashMap data structure in the Java language to implement the initialization processing of the lane information. After the input cut regional grids are input, the lane information corresponding to the regional grids can be quickly located, and the regional grid where the vehicle to be recognized is located can be quickly located according to the perception data of the vehicle to be recognized in the perception object recognition result.

[0131] It should be noted that all the content of the lane information is included in the initialized lane information, and the initialized lane information only cuts the lane range information in the lane information into equal areas.

[0132] S603. Each task analysis unit determines the longitude and latitude information and speed information of the vehicle to be recognized included in the perception object recognition result according to the received perception object recognition result.

[0133] Optionally, the task analysis unit can determine the longitude and latitude information and speed information of each vehicle to be recognized according to the perception data of each vehicle to be recognized in the received perception object recognition result. Among them, the longitude and latitude information is used to indicate the position point where the vehicle to be recognized is located, and the speed information is used to indicate the driving speed of the vehicle to be recognized.

[0134] S604. Each task analysis unit determines the initialized lane information corresponding to the target lane where the vehicle to be recognized is currently located according to the longitude and latitude information of the vehicle to be recognized and the initialized lane information.

[0135] Optionally, the task analysis unit can quickly locate the target lane to which the vehicle to be recognized belongs and the initialization lane information of the target lane according to the longitude and latitude information in the perception data of the vehicle to be recognized. Among them, the target lane is used to indicate the lane in which the vehicle to be recognized travels.

[0136] S605. Each task analysis unit determines whether the vehicle to be recognized is an initially slow-moving vehicle according to the speed information of the vehicle to be recognized and the lane speed limit information in the initialization lane information corresponding to the target lane.

[0137] Optionally, the task analysis unit determines whether there may be abnormal slowdown of the vehicle to be recognized according to the speed information in the perception data of the vehicle to be recognized and the lane speed limit information in the initialization lane information corresponding to the target lane.

[0138] It should be noted that if the vehicle to be recognized is not in any road in the traffic network but at the center of a certain intersection, the lane speed limit information given to the vehicle to be recognized can be a preset speed limit threshold, and the preset speed limit threshold can be 60 km / h. This application does not make specific limitations on this.

[0139] S606. If so, store the vehicle to be recognized in the initially slow-moving vehicle sequence.

[0140] Optionally, when the vehicle to be recognized belongs to an initially abnormally slow-moving vehicle, store the vehicle to be recognized in the initially slow-moving vehicle sequence.

[0141] In a possible implementation manner, referring to Figure 7 , the operation of step S605 can specifically be:

[0142] S701. Determine the minimum slow-moving speed of the target lane according to the lane speed limit information in the initialization lane information corresponding to the target lane and a preset first threshold.

[0143] Optionally, the minimum slow-moving speed of the target lane can be calculated according to the following formula (1). The minimum slow-moving speed refers to the lowest slow-moving speed of the target lane. Formula (1) is as follows:

[0144] Vmin = Vm × a (1)

[0145] It should be noted that Vmin is used to indicate the minimum slow-moving speed of the target lane, a is used to indicate the preset first threshold, and Vm is used to indicate the lane speed limit information of the target lane. Among them, the preset first threshold is a constant threshold preset by the user. The preset first threshold can be 3%, 5%, etc. This application does not make specific limitations on this.

[0146] S702. Determine the maximum slow-down speed of the target lane according to the lane speed limit information in the initialized lane information corresponding to the target lane and a preset second threshold.

[0147] Optionally, the maximum slow-down speed of the target lane can be calculated according to the following formula (2). The maximum slow-down speed refers to the highest slow-down speed of the target lane. Formula (2) is as follows:

[0148] Vmax = Vm × b (2)

[0149] It should be noted that Vmax is used to indicate the maximum slow-down speed of the target lane, b is used to indicate the preset second threshold, and Vm is used to indicate the lane speed limit information of the target lane. Among them, the preset second threshold is a constant threshold preset by the user. The preset first threshold can be 40%, 50%, etc. This application does not make specific limitations on this.

[0150] Among them, the preset second threshold is much larger than the preset first threshold.

[0151] S703. If the speed information of the vehicle to be recognized is greater than or equal to the minimum slow-down speed of the target lane and less than or equal to the maximum slow-down speed of the target lane, then the vehicle to be recognized is an initial slow-down vehicle.

[0152] Optionally, when the speed of the vehicle to be recognized is between the minimum slow-down speed Vmin and the maximum slow-down speed Vmax of the target lane, it is determined that the vehicle to be recognized may have abnormal slow-down, and the vehicle to be recognized is the initial slow-down vehicle.

[0153] In a possible implementation manner, referring to Figure 8 , the operation of step S502 can specifically be:

[0154] S801. Each task analysis unit determines the target area where each initial slow-down vehicle is currently located according to the latitude and longitude information in the perception data of each initial slow-down vehicle and the initialized lane information.

[0155] Optionally, the task analysis unit determines the area grid where the initial slow-down vehicle is located according to the latitude and longitude information in the perception data of the initial slow-down vehicle and the initialized lane information, that is, determines the target area where the initial slow-down vehicle is currently located.

[0156] S802. Each task analysis unit obtains the perception data of each vehicle to be recognized passing through the target area within a preset time period.

[0157] Optionally, the task analysis unit takes the target area where the initial slow vehicle is located as a queue data structure alone, caches the perception data of each vehicle passing through the target area within a preset time period into the queue data structure, and removes the historical perception data before the preset time period at the same time.

[0158] S803. Each task analysis unit determines the number of vehicle trips in the target area within a preset time period according to the number of perception data received within the preset time period.

[0159] Optionally, the task analysis unit can determine the number of vehicle trips in the target area within a preset time period by counting the formats of the perception data of the vehicles passing through the target area within the preset time period.

[0160] S804. Each task analysis unit determines the first average vehicle speed in the target area within a preset time period according to the vehicle speed information in each perception data and the number of vehicle trips in the target area within the preset time period.

[0161] Optionally, the first average vehicle speed in the target area within a preset time period can be calculated according to the following formula (3), and the formula (3) is as follows:

[0162] V1 = Vf ÷ N (3)

[0163] It should be noted that V1 is used to indicate the first average vehicle speed in the target area within a preset time period, Vf is used to indicate the sum of the vehicle speeds of each vehicle passing through the target area within the preset time period, and N is used to indicate the number of vehicle trips in the target area within the preset time period.

[0164] In a possible implementation manner, refer to Figure 9 , the operation of step S503 can specifically be:

[0165] S901. The spatial index unit determines at least one target area corresponding to each initial slow vehicle according to the perception data of each initial slow vehicle and the initialized lane information.

[0166] Optionally, the spatial index unit determines at least one area grid passed by each initial slow vehicle during the driving process according to the perception data of each initial slow vehicle and the initialized lane information, that is, determines at least one target area corresponding to each initial slow vehicle, and creates a mapping relationship between each area grid passed by each initial slow vehicle and the perception data of each initial slow vehicle.

[0167] Optionally, the spatial index unit determines the longitude and latitude information of the first appearance and the longitude and latitude information of the last appearance of each initial slow vehicle according to the perception data of each initial slow vehicle, and determines at least one target area passed by each initial slow vehicle according to the first longitude and latitude information, the last longitude and latitude information and the intermediate longitude and latitude information of the initial slow vehicle.

[0168] S902. The spatial indexing unit determines the associated vehicles corresponding to each initial slow-moving vehicle and the number of associated vehicles according to at least one target area corresponding to each initial slow-moving vehicle and the latitude and longitude information in each perception data in the received perception object recognition result.

[0169] Optionally, the spatial indexing unit determines the associated vehicles related to each initial slow-moving vehicle and the number of associated quantities according to at least one target area corresponding to each initial slow-moving vehicle and the latitude and longitude information in each perception data in all the received perception object recognition results.

[0170] S903. The spatial indexing unit determines the second average speed of the associated vehicles corresponding to each initial slow-moving vehicle according to the perception data of each associated vehicle and the number of associated vehicles.

[0171] Optionally, the second average speed of the associated vehicles corresponding to each initial slow-moving vehicle can be calculated according to the following formula (4), and the formula (4) is as follows:

[0172] V2 = Ve÷M (4)

[0173] It should be noted that V2 is used to indicate the second average speed of the associated vehicles corresponding to each initial slow-moving vehicle, Ve is used to indicate the sum of the speeds of the associated vehicles corresponding to the initial slow-moving vehicles in the target area, and M is used to indicate the number of associated vehicles corresponding to the initial slow-moving vehicles in the target area.

[0174] In a possible implementation, referring to Figure 10 , the operation of step S504 can specifically be:

[0175] S1001. The spatial indexing unit determines the target area score of each initial slow-moving vehicle according to the train number and the first average speed in the target area within a preset time period.

[0176] Optionally, the target area score is used to indicate the driving environment score of the target area passed by the initial slow-moving vehicle. When the train number in the target area where the initial slow-moving vehicle belongs is greater than the preset train number threshold, the target area score is the preset weight value. The preset train number threshold can be 50, 60, etc., and the preset area weight value can be 60%, 65%, etc. The present application does not make specific limitations on this.

[0177] Optionally, when the train number in the target area where the initial slow-moving vehicle belongs is less than the preset threshold, the weight value of the target area score can be calculated by the following formula (5), and the formula (5) is as follows:

[0178] Q1 = (P ÷x)×L (5)

[0179] It should be noted that Q1 is used to indicate the weight value of the target area score, P is used to indicate the number of vehicle trips in the target area within a preset time period, x is used to indicate the preset vehicle trip threshold, and L is used to indicate the preset area weight value.

[0180] Optionally, when the traveling speed of the initial slow-moving vehicle is less than the first average speed of the target area within a preset time period, the target area score S1 of the initial slow-moving vehicle is 0.

[0181] Optionally, when the traveling speed of the initial slow-moving vehicle is greater than or equal to the first average speed of the target area within a preset time period, the target area score of the initial slow-moving vehicle can be calculated by the following formula (6), and formula (6) is as follows:

[0182] S1 = [(V1 - V) ÷ V1] × Q1 (6)

[0183] It should be noted that S1 is used to indicate the target area score of the initial slow-moving vehicle, V1 is used to indicate the first average vehicle speed of the target area within a preset time period, V is used to indicate the traveling speed of the initial slow-moving vehicle, and Q1 is used to indicate the weight value of the target area score.

[0184] S1002. The spatial index unit determines the surrounding scores of each initial slow-moving vehicle according to the number of associated vehicles corresponding to each initial slow-moving vehicle and the second average speed.

[0185] Optionally, when the number of associated vehicles passing around the initial slow-moving vehicle is greater than the preset associated vehicle threshold, the weight value of the surrounding score of the initial slow-moving vehicle can be calculated by the following formula (7), and the preset associated vehicle threshold can be 8, 10, etc., and formula (7) is as follows:

[0186] Q2 = (1 - Q1) × L (7)

[0187] It should be noted that Q2 is used to indicate the weight value of the surrounding score of the initial slow-moving vehicle, Q1 is used to indicate the weight value of the target area score, and L is used to indicate the preset area weight value.

[0188] Optionally, when the number of associated vehicles passing around the initial slow-moving vehicle is less than the preset associated vehicle threshold, the weight value of the surrounding score of the initial slow-moving vehicle can be calculated by the following formula (8), and formula (8) is as follows:

[0189] Q2 = (M ÷ m) × (1 - Q1) × L (8)

[0190] It should be noted that Q2 is used to indicate the weight value of the surrounding score of the initial slow-moving vehicle, M is used to indicate the number of associated vehicles corresponding to the initial slow-moving vehicle in the target area, m is used to indicate the preset associated vehicle threshold, Q1 is used to indicate the weight value of the target area score, and L is used to indicate the preset area weight value.

[0191] Optionally, when the traveling speed of the initial slow vehicle is greater than or equal to a preset ratio of the second average speed of the associated vehicle related to the initial slow vehicle in the target area, the surrounding score S2 of the initial slow vehicle is 0. The preset ratio can be 60%, 65%, etc., and the present application does not make specific limitations thereto.

[0192] Optionally, when the traveling speed of the initial slow vehicle is less than the preset ratio of the second average speed of the associated vehicle related to the initial slow vehicle in the target area, the traveling speed of the initial slow vehicle can be calculated by the following formula (9), and the formula (9) is as follows:

[0193] S2 = [(V2 × y - V) ÷ V2] × Q2 (9)

[0194] It should be noted that S2 is used to indicate the surrounding score of the initial slow vehicle, V2 is used to indicate the second average speed of the associated vehicle corresponding to each initial slow vehicle, y is used to indicate the preset ratio, V is used to indicate the traveling speed of the initial slow vehicle, and Q2 is used to indicate the weight value of the surrounding score.

[0195] S1003. The spatial index unit determines the speed limit score of each initial slow vehicle according to the vehicle speed information and the initialized lane information in the perception data of each initial slow vehicle.

[0196] Optionally, the weight value of the speed limit score of the initial slow vehicle is calculated according to the following formula (10), and the formula (10) is as follows:

[0197] Q3 = (1 - Q1 - Q2) × 100% (10)

[0198] It should be noted that Q3 is used to indicate the weight value of the speed limit score of the initial slow vehicle, Q1 is used to indicate the weight value of the target area score of the initial slow vehicle, and Q2 is used to indicate the weight value of the surrounding score of the initial slow vehicle.

[0199] Optionally, when the initial slow vehicle is greater than or equal to the lane speed limit of its target lane, the speed limit score S3 of the initial slow vehicle is 0.

[0200] Optionally, when the initial slow vehicle is less than the lane speed limit of its target lane, the speed limit score of the initial slow vehicle is calculated by the following formula (11), and the formula (11) is as follows:

[0201] S3 = (V ÷ Vm) × Q3 (11)

[0202] It should be noted that S3 is used to indicate the speed limit score of the initial slow-moving vehicle, V is used to indicate the driving speed of the initial slow-moving vehicle, Vm is used to indicate the lane speed limit information of the target lane, and Q3 is used to indicate the weight value of the speed limit score of the initial slow-moving vehicle.

[0203] S1004. The spatial indexing unit calculates the total score of each initial slow-moving vehicle based on the target area score, the surrounding score, and the speed limit score.

[0204] Optionally, the total score of the initial slow-moving vehicle is calculated according to the following formula (12), and formula (12) is as follows:

[0205] S4 = S1 + S2 + S3 (12)

[0206] It should be noted that S4 is used to indicate the total score of the initial slow-moving vehicle, S1 is used to indicate the target area score of the initial slow-moving vehicle, S2 is used to indicate the surrounding score of the initial slow-moving vehicle, and S3 is used to indicate the speed limit score of the initial slow-moving vehicle.

[0207] Optionally, when the total score of the initial slow-moving vehicle is less than the preset value, it is preliminarily determined that the initial slow-moving vehicle is an abnormal slow-moving vehicle, and the confidence level of the initial slow-moving vehicle is incremented by 1. The initial value of the confidence level is 0. When the cumulative confidence level of the total scores of the regional grids to which the initial slow-moving vehicle belongs is greater than or equal to 3, it is determined that the initial slow-moving vehicle actually has abnormal slow movement.

[0208] It should be noted that the weight values of the target area score, the surrounding score, and the speed limit score of the initial slow-moving vehicle are different, that is, the degrees of influence on the total score of the initial slow-moving vehicle are also different. This application does not make specific limitations on this.

[0209] In a possible implementation manner, referring to Figure 11 , the operation of step S505 can specifically be:

[0210] S1101. The spatial indexing unit determines at least one target abnormal slow-moving vehicle according to the total score of each initial slow-moving vehicle.

[0211] Optionally, the spatial indexing unit obtains the confidence level corresponding to each initial slow-moving vehicle according to the scores of the regional grids to which each initial slow-moving vehicle belongs, and determines the target abnormal slow-moving vehicle that actually has abnormal slow movement according to the confidence level of each initial slow-moving vehicle.

[0212] S1102. The spatial indexing unit determines the associated vehicle corresponding to each target abnormal slow-moving vehicle according to each target abnormal slow-moving vehicle and the received perception object recognition result.

[0213] Optionally, the spatial index unit determines associated vehicles related to each target abnormally slow-moving vehicle according to the sensed data of the determined target abnormally slow-moving vehicle and multiple sensed data included in all received sensed object recognition results.

[0214] S1103. The spatial index unit determines associated warning vehicles related to each target abnormally slow-moving vehicle according to the sensed data of the associated vehicles corresponding to each target abnormally slow-moving vehicle.

[0215] As an optional implementation manner, the abnormally slow-moving vehicle detection method provided in this application can, in the form of selling warning services, sell in-vehicle communication devices to drivers who want to subscribe to warning services or let them install corresponding application programs on their mobile phones to establish a communication connection. If a driver purchases a warning service and establishes a communication connection with the abnormally slow-moving vehicle detection system through a mobile phone application program, after the driver starts the application program, the application program will send positioning information to the continuously running abnormally slow-moving vehicle detection system program. The positioning information at least includes vehicle identification, vehicle longitude and latitude information, and vehicle heading angle information, which is received by the spatial index unit. After receiving the positioning information of the application program, the spatial index unit updates the positioning information to its own lane information data structure and can quickly retrieve the lane information in the corresponding range according to the grid. According to the longitude and latitude information of the abnormally slow-moving vehicle, calculate the influence range of the abnormally slow-moving vehicle and the area grid corresponding to this range, and determine the positioning information of this influence range. After finding the positioning information, transfer the positioning information and the sensed data of the abnormally slow-moving vehicle to the abnormally slow-moving vehicle warning module for presenting a warning effect.

[0216] In a possible implementation manner, referring to Figure 12 , the operation of step S303 can specifically be:

[0217] S1201. The abnormally slow-moving vehicle warning module determines the forward and backward driving order of each target abnormally slow-moving vehicle and each associated warning vehicle according to the sensed data of each target abnormally slow-moving vehicle and the sensed data of each associated warning vehicle in the abnormally slow-moving vehicle recognition result, and determines the associated warning vehicles in the same lane as each target abnormally slow-moving vehicle. ]>

[0218] Optionally, the abnormally slow vehicle warning module extracts the perception data of the target abnormally slow vehicle and the perception data of each associated warning vehicle from the abnormally slow vehicle recognition result, and extracts the longitude and latitude information of the associated warning vehicles from the perception data of each associated warning vehicle. Taking the associated warning vehicle as the starting point and the target abnormally slow vehicle as the ending point, draw a directed straight line, and calculate the azimuth angle of the directed straight line. The azimuth angle is the included angle between the directed straight line and the due north direction, which is 0° when the directed straight line coincides with the due north direction and 90° when it coincides with the due east direction. Calculate the angle difference between these two directions based on the azimuth angle and the heading angle information of the associated warning vehicle. When the difference is less than or equal to 45°, it means that these two directions are the same, that is, the target abnormally slow vehicle is considered to be in front of the associated warning vehicle's driving direction.

[0219] Optionally, calculate in which lane each vehicle is located according to the longitude and latitude information in the perception data of the target abnormally slow vehicle and the longitude and latitude information in the perception data of the associated vehicle to be warned. If the calculated lanes are the same, it is considered that these two vehicles are in the same lane. If it is calculated that neither of the two vehicles is in a lane, it is also considered that these two vehicles are in the same lane.

[0220] S1202. The abnormally slow vehicle warning module generates the first warning information for each target abnormally slow vehicle and the second warning information corresponding to each associated warning vehicle according to the front and rear driving order of each target abnormally slow vehicle and each associated warning vehicle and the perception data of the associated warning vehicle in the same lane as each target abnormally slow vehicle.

[0221] Optionally, after the application establishes a communication connection with the abnormally slow vehicle detection system provided in this embodiment of the present application, the abnormally slow vehicle detection system will record the vehicle identifier uploaded by the application. When the warning effect of this vehicle needs to be presented, the effect information to be presented will be sent to the application through the communication protocol according to the identifier of this vehicle for presentation.

[0222] Optionally, if the target abnormally slow vehicle is not in front of the associated warning vehicle's driving direction, it is considered that the threat of this target slow vehicle to the associated warning vehicle is low, and no warning effect is presented; if the target abnormally slow vehicle is in front of the associated warning vehicle's driving direction, further observe whether the target abnormally slow vehicle is in the same lane as the associated warning vehicle. If not in the same lane, it is considered that the threat of this target slow vehicle to the associated warning vehicle is average, and the longitude and latitude information of this target slow vehicle is displayed on the application. If the two vehicles are in the same lane, it is considered that the threat of this target slow vehicle to the associated warning vehicle is high. In addition to displaying the position of this target abnormally slow vehicle on the application, a flashing effect of the target abnormally slow vehicle is added to improve the warning effect.

[0223] Figure 13A schematic structural diagram of an electronic device provided for this application, see Figure 13 In the electronic device 20 provided in the embodiment of the present application, the above-mentioned abnormal slow vehicle detection system 10 is deployed. The electronic device 20 implements each embodiment of the above-mentioned abnormal slow vehicle detection method via the abnormal slow vehicle detection system 10, which will not be elaborated herein.

[0224] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

[0225] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An abnormal slow vehicle detection method, characterized in that The method is applied to an abnormal slow-moving vehicle detection system, which includes: a perception data recognition and acquisition module, an abnormal slow-moving vehicle recognition module, and an abnormal slow-moving vehicle warning module. The perception data recognition and acquisition module includes: at least one roadside perception device arranged on the roadside and a data acquisition unit corresponding to the roadside perception device one by one. The abnormal slow-moving vehicle recognition module includes: a plurality of task analysis units and a spatial index unit. The method includes: Each roadside perception device collects the perception data of the perceived objects passing on each road section in real time, generates a perceived object recognition result, and the roadside perception device sends the perceived object recognition result to the data acquisition unit corresponding to each road section. The data acquisition unit stores the received perceived object recognition result into the data topic corresponding to the road section. Each task analysis unit receives the perceived object recognition results uploaded by at least one pre-subscribed data acquisition unit. The spatial index unit receives the perceived object recognition results uploaded by each data acquisition unit. The spatial index unit and each task analysis unit generate an abnormal slow-moving vehicle recognition result according to the received perceived object recognition results. The abnormal slow-moving vehicle recognition result includes: at least one target abnormal slow-moving vehicle and associated warning vehicles related to the target abnormal slow-moving vehicle, and sends the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result to the abnormal slow-moving vehicle warning module. The abnormal slow-moving vehicle warning module determines the first warning information of each target abnormal slow-moving vehicle and the second warning information corresponding to each associated warning vehicle according to the abnormal slow-moving vehicle recognition result and the perception data of each vehicle in the abnormal slow-moving vehicle recognition result, sends the first warning information to each target abnormal slow-moving vehicle, and sends the second warning information to each associated warning vehicle. The associated warning vehicle refers to a vehicle with a traffic accident risk among the vehicles passing around the target abnormal slow-moving vehicle. The spatial index unit and each task analysis unit generate an abnormal slow-moving vehicle recognition result according to the received perceived object recognition results, including: Each task analysis unit filters the slow-moving vehicles in the perceived object recognition result according to the received perceived object recognition results to obtain an initial slow-moving vehicle sequence. The initial slow-moving vehicle sequence includes: at least one initial slow-moving vehicle. Each task analysis unit determines the number of vehicle trips and the first average vehicle speed of the target area within a preset time period according to the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, the initialized lane information, and the perception data of the perceived objects passing through each lane within the preset time period. The number of vehicle trips refers to the number of vehicles passing through the target area within the preset time period. The first average vehicle speed refers to the average vehicle speed of all vehicles passing through the target area within the preset time period. The spatial index unit determines the number of associated vehicles corresponding to each of the initial slow-moving vehicles and the second average speed according to the received perception object recognition result, the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, and the initialized lane information. The associated vehicles are the surrounding vehicles traveling within a preset range around the initial slow-moving vehicle, and the second average speed refers to the average speed of the associated vehicles of the initial slow-moving vehicle. The spatial index unit determines the total score of each of the initial slow-moving vehicles according to the number of vehicle trips in the target area within a preset period, the first average speed, the number of associated vehicles, the second average speed, and the initialized lane information. The spatial index unit generates an abnormal slow-moving vehicle recognition result according to the total score of each of the initial slow-moving vehicles and the received perception object recognition result.

2. The abnormal slow vehicle detection method according to claim 1, wherein, Each task analysis unit screens the to-be-identified vehicles included in the perception object recognition result according to the received perception object recognition result to obtain an initial slow-moving vehicle sequence, including: Each task analysis unit loads lane information from the lane database. The lane information includes: lane range information, lane identification, lane speed limit information, lane center line, lane starting point longitude and latitude information, and lane ending point longitude and latitude information. Each task analysis unit performs initialization processing on the lane information through a preset segmentation algorithm to obtain initialized lane information. The initialized lane information includes: the mapping relationship between each lane and multiple regions. Each task analysis unit determines the longitude and latitude information and speed information of the to-be-identified vehicles included in the perception object recognition result according to the received perception object recognition result. Each task analysis unit determines the initialized lane information corresponding to the target lane where the to-be-identified vehicle is currently located according to the longitude and latitude information of the to-be-identified vehicle and the initialized lane information. Each task analysis unit determines whether the to-be-identified vehicle is an initial slow-moving vehicle according to the speed information of the to-be-identified vehicle and the lane speed limit information in the initialized lane information corresponding to the target lane. If so, the to-be-identified vehicle is stored in the initial slow-moving vehicle sequence.

3. The abnormal slow vehicle detection method according to claim 2, wherein Each task analysis unit determines whether the to-be-identified vehicle is an initial slow-moving vehicle according to the speed information of the to-be-identified vehicle and the lane speed limit information in the initialized lane information corresponding to the target lane, including: Determining the minimum slow-moving speed of the target lane according to the lane speed limit information in the initialized lane information corresponding to the target lane and a preset first threshold. Determining the maximum slow-moving speed of the target lane according to the lane speed limit information in the initialized lane information corresponding to the target lane and a preset second threshold. If the speed information of the to-be-identified vehicle is greater than or equal to the minimum slow-moving speed of the target lane and the speed information of the to-be-identified vehicle is less than or equal to the maximum slow-moving speed of the target lane, then the to-be-identified vehicle is an initial slow-moving vehicle.

4. The abnormal slow vehicle detection method according to claim 1, characterized in that, Each of the task analysis units determines the train number and the first average vehicle speed of the target area within a preset time period according to the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, the initialized lane information, and the perception data of the vehicles to be recognized passing through each lane within the preset time period, including: Each of the task analysis units determines the target area where each initial slow-moving vehicle is currently located according to the longitude and latitude information in the perception data of each initial slow-moving vehicle and the initialized lane information; Each of the task analysis units acquires the perception data of each vehicle to be recognized passing through the target area within a preset time period; Each of the task analysis units determines the train number of the target area within a preset time period according to the number of perception data received within the preset time period; Each of the task analysis units determines the first average vehicle speed of the target area within a preset time period according to the vehicle speed information in each of the perception data within the preset time period and the train number of the target area.

5. The abnormal slow vehicle detection method according to claim 1, characterized in that The spatial indexing unit determines the number of associated vehicles and the second average vehicle speed corresponding to each initial slow-moving vehicle according to the received perception object recognition result, the perception data of each initial slow-moving vehicle in the initial slow-moving vehicle sequence, and the initialized lane information, including: The spatial indexing unit determines at least one target area corresponding to each initial slow-moving vehicle according to the perception data of each initial slow-moving vehicle and the initialized lane information; The spatial indexing unit determines the associated vehicles corresponding to each initial slow-moving vehicle and the number of the associated vehicles according to at least one target area corresponding to each initial slow-moving vehicle and the longitude and latitude information in each perception data in the received perception object recognition result; The spatial indexing unit determines the second average vehicle speed of the associated vehicles corresponding to each initial slow-moving vehicle according to the perception data of each associated vehicle and the number of the associated vehicles.

6. The abnormal slow vehicle detection method according to claim 1, characterized in that The spatial indexing unit determines the total score of each initial slow-moving vehicle according to the train number of the target area within a preset time period, the first average vehicle speed, the number of associated vehicles, the second average vehicle speed, and the initialized lane information, including: The spatial indexing unit determines the target area score of each initial slow-moving vehicle according to the train number of the target area within a preset time period and the first average vehicle speed; The spatial indexing unit determines the surrounding score of each initial slow-moving vehicle according to the number of associated vehicles corresponding to each initial slow-moving vehicle and the second average vehicle speed; The spatial indexing unit determines the speed limit score of each initial slow-moving vehicle according to the vehicle speed information in the perception data of each initial slow-moving vehicle and the initialized lane information; The spatial indexing unit determines the total score of each initial slow-moving vehicle according to the target area score, the surrounding score, and the speed limit score.

7. The abnormal slow vehicle detection method according to claim 1, wherein, The spatial indexing unit generates an abnormal slow-moving vehicle recognition result according to the total score of each initial slow-moving vehicle and the received perception object recognition result, including: The spatial indexing unit determines at least one target abnormal slow-moving vehicle according to the total score of each initial slow-moving vehicle; The spatial index unit determines the associated vehicles corresponding to each of the target abnormally slow-moving vehicles according to each of the target abnormally slow-moving vehicles and the received perception object recognition results; The spatial index unit determines the associated warning vehicles related to each of the target abnormally slow-moving vehicles according to the perception data of the associated vehicles corresponding to each of the target abnormally slow-moving vehicles.

8. The abnormal slow vehicle detection method according to claim 1, characterized in that The abnormally slow-moving vehicle warning module determines the first warning information for each of the target abnormally slow-moving vehicles and the second warning information corresponding to each of the associated warning vehicles according to the abnormally slow-moving vehicle recognition results and the perception data of each vehicle in the abnormally slow-moving vehicle recognition results, including: The abnormally slow-moving vehicle warning module determines the front-back driving order of each of the target abnormally slow-moving vehicles and each of the associated warning vehicles, and determines the associated warning vehicles in the same lane as each of the target abnormally slow-moving vehicles according to the perception data of each of the target abnormally slow-moving vehicles and the perception data of each of the associated warning vehicles in the abnormally slow-moving vehicle recognition results; The abnormally slow-moving vehicle warning module generates the first warning information for each of the target abnormally slow-moving vehicles and the second warning information corresponding to each of the associated warning vehicles according to the front-back driving order of each of the target abnormally slow-moving vehicles and each of the associated warning vehicles and the perception data of the associated warning vehicles in the same lane as each of the target abnormally slow-moving vehicles.

9. An abnormal slow-moving vehicle detection system, characterized in that, The abnormally slow-moving vehicle detection system includes: a perception data recognition and acquisition module, an abnormally slow-moving vehicle recognition module, and an abnormally slow-moving vehicle warning module. The perception data recognition and acquisition module includes: at least one roadside perception device arranged on the roadside and a data acquisition unit corresponding to the roadside perception device one by one. The abnormally slow-moving vehicle recognition module is communicatively connected to each of the data acquisition units in the perception data recognition and acquisition module and the abnormally slow-moving vehicle warning module; The abnormally slow-moving vehicle detection system is used to execute the steps of the abnormally slow-moving vehicle detection method according to any one of claims 1-8.

Images