Traffic accident information processing system based on vehicle and road cloud multi-dimensional fusion data

Abstract

The embodiment of the invention relates to a traffic accident information processing system based on vehicle and road cloud multi-dimensional fusion data. The system comprises vehicle end equipment, road side equipment and a cloud platform, the vehicle end equipment is used for continuously collecting and storing own vehicle data and periodically synchronizing with the cloud platform; the road side equipment is used for continuously collecting and storing road side data and periodically synchronizing with the cloud platform; the cloud platform is used for monitoring traffic accidents and generating a corresponding cloud data set, a vehicle-road data set and a vehicle-road supplementary data set for each piece of accident information; and carrying out accident data integration on the multi-dimensional data corresponding to the accident information. According to the invention, the traffic accident processing efficiency can be improved, and the processing difficulty is reduced.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a traffic accident information processing system based on multi-dimensional fusion data of vehicle, road and cloud. Background Technology

[0002] The standard procedure for handling traffic accidents is as follows: personnel arrive at the accident scene and determine liability based on information from the scene and vehicle-related data. Scene information includes photographs. Vehicle-related data includes chassis operating data (such as driving mode, vehicle identification, road / intersection markings, vehicle coordinates, speed, acceleration, heading angle, steering wheel angle, brake pedal engagement, accelerator pedal engagement, etc.), EDR (Electronic Event Data Recorder) data, and DVR (On-Board Driving Video Recording) video. This standard procedure becomes more challenging when the accident scene is disturbed or images are not promptly preserved.

[0003] With the continuous development and application of vehicle-to-everything (V2X) technology, intelligent connected vehicle (ICV) technology, and vehicle-road-cloud integrated collaborative technology, in principle, multi-dimensional data information of any event on the road network can be obtained from multiple dimensions of vehicle, road, and cloud. If a multi-dimensional data fusion processing mechanism specifically for traffic accident events can be customized based on this, then even if the accident scene is destroyed or the parties involved fail to preserve images in time, sufficient reference information can still be provided to accident handlers as a basis for accident handling. How to implement such a vehicle-road-cloud data processing mechanism for traffic accident handling is the technical problem that this invention aims to solve. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing a traffic accident information processing system based on multi-dimensional fusion data from vehicles, roads, and the cloud. This system includes: vehicle-mounted equipment, roadside equipment, and a cloud platform. The vehicle-mounted equipment continuously collects and stores chassis operation data, EDR data, DVR data, and fault data of the vehicle; synchronously receives and stores interaction messages between the roadside unit (RSU) equipment and the vehicle; periodically generates a vehicle-cloud synchronized dataset based on the stored chassis operation data and fault data and sends it to the cloud platform; and upon receiving a vehicle-mounted accident extraction request from the cloud platform, extracts the EDR, DVR, and vehicle-road message data matching the current request to form a vehicle-mounted accident dataset and sends it back to the cloud platform. Roadside equipment is used to continuously collect and store multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene (road segment or intersection) through roadside edge computing (MEC) devices; synchronously receive and store interaction messages between roadside unit devices and all passing vehicles; periodically generate road-cloud synchronized datasets based on the stored traffic participant, signal, and traffic event perception data and send them to the cloud platform; and when receiving a roadside accident extraction request from the cloud platform, extract the visual sensor data and vehicle-road message data that match the current request to form a roadside accident dataset and send it back to the cloud platform. The cloud platform is used to continuously collect and store meteorological and traffic data for the entire road network; it synchronizes data in the cloud based on vehicle-cloud / road-cloud synchronous datasets; it monitors traffic accident events in real time in the cloud, synthesizes cloud datasets of meteorological and traffic data for each monitored accident, synthesizes vehicle-road synchronous data of each monitored accident, and further obtains more detailed vehicle-road supplementary datasets by sending accident extraction requests to roadside devices on the vehicles related to each accident. Multi-dimensional accident data integration is then performed based on the vehicle-road datasets, vehicle-road supplementary datasets, and cloud datasets corresponding to each accident. This invention's system is used for real-time multi-dimensional vehicle-road-cloud data integration for any traffic accident occurring on the road network. This includes not only accident vehicle information collected from the vehicles involved, but also bystander information collected from other vehicles passing by the accident scene, global roadside information collected from roadside unit devices / roadside edge computing devices at the accident section / intersection, and meteorological and traffic information of the accident area obtained from the cloud. Based on this invention, the system can promptly output corresponding multidimensional datasets for any traffic accident. Even if the accident scene is damaged or the parties involved fail to preserve images in time, this dataset can still provide sufficient information for accident investigators. This invention can improve the efficiency and reduce the difficulty of handling traffic accidents.

[0005] To achieve the above objectives, embodiments of the present invention provide a traffic accident information processing system based on multi-dimensional fusion data of vehicle, road, and cloud, the system comprising: vehicle-side equipment, roadside equipment, and a cloud platform;

[0006] The vehicle-mounted equipment is connected to both the roadside equipment and the cloud platform, and the roadside equipment is connected to the cloud platform. The vehicle-mounted equipment includes a vehicle-mounted data acquisition and transmission module and a vehicle-mounted storage module, which are connected to each other. The roadside equipment includes a roadside data acquisition and transmission module and a roadside storage module, which are connected to each other. The cloud platform includes a cloud-based data acquisition and transmission module, a cloud-based storage module, and an accident handling module, which are connected to both the cloud-based data acquisition and transmission module and the accident handling module.

[0007] The vehicle-side data acquisition and transmission module interfaces with the vehicle's onboard system;

[0008] The vehicle-side data acquisition and transmission module is used to continuously collect chassis operation data, EDR data, DVR data, and fault data output by the vehicle system during vehicle operation and store the collected data in the vehicle-side storage module; it also synchronously receives roadside messages sent from roadside unit devices to the vehicle system and vehicle-side messages sent from the vehicle system to roadside unit devices and stores the messages in the vehicle-side storage module; it periodically generates corresponding vehicle-cloud synchronized datasets based on the stored chassis operation data and fault data and sends them to the cloud platform; and when it receives a vehicle-side accident extraction request from the cloud platform, it extracts the EDR, DVR, and vehicle-road message data matching the current request to form a corresponding vehicle-side accident dataset and sends it back to the cloud platform.

[0009] The vehicle-side storage module is used to store the first chassis dataset, the first EDR dataset, the first DVR dataset, the first fault dataset, the first roadside message set, and the first vehicle-side message set;

[0010] The roadside data acquisition and transmission module interfaces with the roadside edge computing device and the roadside unit device in the same traffic scenario as the roadside equipment; the scenario types of the traffic scenario include road segments and intersections;

[0011] The roadside data acquisition and transmission module is used to continuously acquire multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene through the roadside edge computing device and store the acquired data in the roadside storage module; it also synchronously receives roadside messages sent by the roadside unit device to the vehicle-mounted systems of all passing vehicles and vehicle-end messages sent by the vehicle-mounted systems of all passing vehicles to the roadside unit device and stores the received message data in the roadside storage module; it periodically generates corresponding road-cloud synchronous datasets based on the stored traffic participant, signal, and traffic event perception data and sends them to the cloud platform; and when it receives a roadside accident extraction request from the cloud platform, it extracts the visual sensor data and vehicle-road message data that match the current request to form a corresponding roadside accident dataset and sends it back to the cloud platform.

[0012] The roadside storage module is used to store a multimodal dataset, a first participant dataset, a first signal dataset, a first event dataset, a second roadside message set, and a second vehicle-side message set;

[0013] The cloud-based data acquisition and transmission module interfaces with a third-party data platform; the third-party data platform includes a meteorological data platform and a traffic management data platform.

[0014] The cloud-based data acquisition and transmission module is used to continuously collect meteorological and traffic data of the entire road network through the third-party data platform and store the collected data in the cloud storage module; and to synchronize vehicle data in the cloud according to the vehicle-to-cloud synchronization dataset periodically uploaded by each of the vehicle-side data acquisition and transmission modules; and to synchronize roadside data in the cloud according to the roadside synchronization dataset periodically uploaded by each of the roadside data acquisition and transmission modules.

[0015] The cloud storage module is used to store road network meteorological datasets, road network traffic datasets, vehicle synchronization datasets, roadside synchronization datasets, and road network accident datasets.

[0016] The accident handling module is used to monitor traffic accidents in the roadside synchronous dataset and generate corresponding monitoring accident information based on real-time monitoring results; and to collect meteorological and traffic data from the road network meteorological dataset and the road network traffic dataset based on each monitoring accident information to obtain corresponding cloud datasets; and to collect vehicle-road data from the vehicle synchronous dataset and the roadside synchronous dataset based on each monitoring accident information to obtain corresponding vehicle-road datasets; and to send corresponding vehicle-end and roadside accident extraction requests to vehicle-end and roadside devices related to the current accident based on each vehicle-road dataset, and generate corresponding vehicle-road supplementary datasets based on the request feedback data; and to perform multi-dimensional accident data integration on the vehicle-road datasets, vehicle-road supplementary datasets, and cloud datasets corresponding to each monitoring accident information, and refresh the road network accident dataset based on the integration results.

[0017] Preferably, the in-vehicle system includes an autonomous driving system, a vehicle event data recording system, and an in-vehicle video driving recorder system;

[0018] The first chassis dataset includes multiple first chassis data records; each first chassis data record includes a first data timestamp, a first vehicle identifier, a first driving mode, a first scene identifier, first vehicle coordinates, a first vehicle speed, a first acceleration, a first heading angle, a first steering wheel angle, a first brake pedal opening degree, and a first accelerator pedal opening degree; the first driving mode includes manual driving and automatic driving; the first scene identifier includes road segment lane identifiers and intersection lane identifiers.

[0019] The first EDR dataset includes multiple first EDR data records; each first EDR data record includes a second data timestamp, a first vehicle identifier, and multiple EDR sub-items; the multiple EDR sub-items consist of some or all of the following sub-items: vehicle motion data, driving operation data, engine operating status data, collision data, safety system data, tire pressure data, and vehicle warning data; each EDR sub-item consists of one or more corresponding secondary data sub-items.

[0020] The first DVR dataset includes multiple first DVR data records; each first DVR data record includes a third data timestamp, a first vehicle identifier, and multiple DVR sub-items; the multiple DVR sub-items consist of some or all of the following sub-items: video data sub-items, audio data sub-items, DVR working status data sub-items, and storage capacity sub-items; each DVR sub-item consists of one or more corresponding secondary data sub-items.

[0021] The first fault dataset includes multiple first fault data records; each first fault data record includes a fourth data timestamp, a first vehicle identifier, and one or more vehicle fault types.

[0022] The first roadside message set includes multiple first roadside message records; the first roadside message record includes a first message timestamp, a first roadside identifier, a first message type, and first message content data; the first message type includes SPAT message, RSM message, RSI message, MAP message, SSM message, and SAM message;

[0023] The first vehicle-side message set includes multiple first vehicle-side message records; the first vehicle-side message record includes a second message timestamp, a first vehicle identifier, a second message type, and second message content data; the second message type includes BSM message, SSM message, ISM message, and SAM message.

[0024] Preferably, the multimodal dataset includes multiple sensor data records; the sensor data records include a fifth data timestamp, a second roadside identifier, a sensor identifier, a sensor type, and sensor data; the sensor types include cameras, lidar, millimeter-wave radar, ultrasonic radar, and traffic lights; when the sensor type is a camera, the sensor data is the corresponding image, image sequence, or video data; when the sensor type is lidar or millimeter-wave radar, the sensor data is the corresponding lidar dense point cloud or millimeter-wave radar sparse point cloud; when the sensor type is ultrasonic radar, the sensor data is the corresponding ranging dataset; when the sensor type is a traffic light, the sensor data is the corresponding traffic light status; the traffic light status includes red light, green light, and yellow light.

[0025] The first participant dataset includes multiple subsets of first participants; each subset of first participants corresponds one-to-one with a traffic participant who was or is currently in the current traffic scenario; each subset of first participants includes a second roadside identifier, a first participant identifier, a first participant type, and a first participant tracking trajectory; the first participant type includes pedestrians, motor vehicles, and non-motor vehicles; the first participant tracking trajectory includes multiple first tracking trajectory points; each first tracking trajectory point includes a first trajectory point timestamp, a first target image, a first target geometric dimension, a first target coordinates, a first target vehicle speed, and a first target heading angle; when the first participant type is a motor vehicle, the first participant identifier is the vehicle identifier of the current motor vehicle;

[0026] When the first signal dataset is not empty, it includes one or more first traffic light subsets; if there are traffic lights in the current traffic scene, the first traffic light subset corresponds one-to-one with the traffic lights in the current traffic scene; the first traffic light subset includes the second roadside sign, the first traffic light sign, the first traffic light coordinates, and the first signal tracking trajectory; the first signal tracking trajectory includes multiple second tracking trajectory points; the second tracking trajectory points include the second trajectory point timestamp, the first signal status, and the remaining time of the first signal; the first signal status includes red light, green light, and yellow light;

[0027] The first event dataset includes multiple first event data records; each first event data record includes a sixth data timestamp, a second roadside identifier, a first event type, and first event data; the first event type includes at least speeding violations, red light violations, lane departure violations, construction obstruction violations, and traffic accident events; when the first event type is a speeding violation, red light violation, or lane departure violation, the first event data consists of the participant identifier corresponding to the violating vehicle, the violation time period tracking trajectory, and the violation time period video; when the first event type is a construction obstruction violation, the first event data consists of the coordinates of the center point of the construction site, the construction... The first event data consists of the set of vertex coordinates of the accident site, the area of ​​the construction site, and the set of lane markers occupied by the construction site. When the first event type is a traffic accident, the first event data consists of the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident site, the area of ​​the accident site, the set of traffic participants within the accident site, and the set of lane markers occupied by the accident site. The set of traffic participants within the accident site consists of one or more traffic participant data, each of which consists of a corresponding traffic participant identifier and a traffic participant type, including pedestrians, motor vehicles, and non-motor vehicles. The set of lane markers occupied by the accident site consists of one or more lane markers.

[0028] The second pathside message set includes multiple second pathside message records; the second pathside message record includes a third message timestamp, a second pathside identifier, a third message type, and third message content data; the third message type includes SPAT message, RSM message, RSI message, MAP message, SSM message, and SAM message;

[0029] The second vehicle-side message set includes multiple second vehicle-side message records; the second vehicle-side message record includes a fourth message timestamp, a second vehicle-side identifier, a fourth message type, and fourth message content data; the fourth message type includes BSM message, SSM message, ISM message, and SAM message.

[0030] Preferably, the road network meteorological dataset includes multiple first regional subsets; each first regional subset corresponds one-to-one with a traffic scenario; each first regional subset includes multiple first meteorological data records; each first meteorological data record includes a seventh data timestamp, a first regional identifier, and a first meteorological dataset; the first meteorological dataset is composed of multiple types of meteorological data.

[0031] The road network traffic dataset includes multiple second region subsets; each second region subset corresponds one-to-one with a traffic scenario; each second region subset includes multiple first traffic data records; each first traffic data record includes an eighth data timestamp, a second region identifier, and a first traffic indicator set; the first traffic indicator set consists of multiple types of traffic indicator data.

[0032] The vehicle synchronization dataset includes multiple first vehicle subsets; each first vehicle subset corresponds one-to-one with the vehicle-end device; each first vehicle subset includes a chassis synchronization dataset and a fault synchronization dataset; the chassis synchronization dataset includes multiple first chassis data records; the fault synchronization dataset includes multiple first fault data records;

[0033] The roadside synchronization dataset includes multiple first roadside subsets; each first roadside subset corresponds one-to-one with a roadside device; each first roadside subset includes a participant synchronization dataset, a signal synchronization dataset, and an event synchronization dataset; the participant synchronization dataset includes multiple first participant subsets; the signal synchronization dataset includes multiple first traffic light subsets; the event synchronization dataset includes multiple first event data records;

[0034] The road network accident dataset includes multiple first accident subsets; the first accident subsets include basic accident information, meteorological data of the accident area, traffic data of the accident area, vehicle-side dataset, participant trajectory set, signal trajectory set, supplementary set of accident vehicles, supplementary set of surrounding vehicles, and supplementary roadside dataset.

[0035] The basic accident information includes the accident time, roadside signs, accident type, center point of the accident area, set of vertices of the accident area, area of ​​the accident area, set of accident participants, and set of accident lanes.

[0036] The vehicle-side dataset consists of vehicle-side data from all accident vehicles within the current accident range. The vehicle-side data includes accident vehicle identification, chassis data sequence, and fault data sequence.

[0037] The participant trajectory set consists of the participant trajectory data of all traffic participants within the current accident area; the participant trajectory data includes accident participant identifier, accident participant type, and accident participant trajectory;

[0038] When the signal trajectory set is not empty, it consists of the signal trajectory data of all traffic lights adjacent to the current accident area; the signal trajectory data includes the accident roadside marker, adjacent signal light marker, adjacent signal light coordinates, and adjacent signal light trajectory.

[0039] The accident vehicle supplementary set consists of a subset of all accident vehicles within the current accident scope; the accident vehicle subset includes EDR data sequences, DVR data sequences, roadside message sequences, and vehicle-side message sequences;

[0040] When the surrounding vehicle supplementary set is not empty, it consists of a subset of surrounding vehicles of all surrounding vehicles outside the current accident area; the surrounding vehicle subset includes the EDR data sequence, the DVR data sequence, the roadside message sequence, and the vehicle-side message sequence;

[0041] The roadside supplementary dataset consists of a roadside visual dataset, a roadside broadcast message set, and a roadside vehicle message set for the road segment or intersection where the current accident occurs.

[0042] Preferably, the vehicle-side data acquisition and transmission module is specifically used when the corresponding vehicle-to-cloud synchronized dataset is generated periodically based on the stored chassis operating data and fault data and sent to the cloud platform:

[0043] According to a preset vehicle-side synchronization frequency, the current time is periodically used as the first end time, and the time point obtained by subtracting a preset first duration from the first end time is used as the first start time. The first start time and the first end time form the corresponding current most recent time period. The first chassis data records in the first chassis dataset that are in the current most recent time period are extracted to form the corresponding first synchronization dataset. The first fault data records in the first fault dataset that are in the current most recent time period are extracted to form the corresponding second synchronization dataset. The obtained first and second synchronization datasets form the corresponding vehicle-cloud synchronization dataset and are sent to the cloud platform.

[0044] The vehicle-side data acquisition and transmission module is specifically used to extract the EDR, DVR, and vehicle-road message data that match the current request when receiving a vehicle-side accident extraction request from the cloud platform, form a corresponding vehicle-side accident dataset, and send it back to the cloud platform:

[0045] The current time period is extracted from the current vehicle-side accident extraction request as the current time period; the first EDR data records in the first EDR dataset that are in the current time period are extracted to form the corresponding EDR data sequence; the first DVR data records in the first DVR dataset that are in the current time period are extracted to form the corresponding DVR data sequence; the first roadside message records in the first roadside message set that are in the current time period are extracted to form the corresponding roadside message sequence; the first vehicle-side message records in the first vehicle-side message set that are in the current time period are extracted to form the corresponding vehicle-side message sequence; and the vehicle-side accident dataset obtained this time, composed of the EDR data sequence, the DVR data sequence, the roadside message sequence, and the vehicle-side message sequence, is sent back to the cloud platform.

[0046] Preferably, the roadside data acquisition and transmission module is specifically used when the corresponding road-cloud synchronized dataset is generated periodically based on the stored traffic participant, signal, and traffic event perception data and sent to the cloud platform:

[0047] According to a preset roadside synchronization frequency, the current time is periodically used as the second end time, and the time point obtained by subtracting a preset second duration from the second end time is used as the second start time. The second start time and the second end time constitute the corresponding current most recent time period. The subset of the first participants whose tracking trajectories have been updated in the first participant dataset during the current most recent time period is extracted to form the corresponding third synchronization dataset. The subset of the first traffic lights whose tracking trajectories have been updated in the first signal dataset during the current most recent time period is extracted to form the corresponding fourth synchronization dataset. The first event data records in the first event dataset that are in the current most recent time period are extracted to form the corresponding fifth synchronization dataset. The obtained third, fourth and fifth synchronization datasets are combined to form the corresponding road-cloud synchronization dataset and sent to the cloud platform.

[0048] The roadside data acquisition and transmission module is specifically used to extract visual sensor data and vehicle-road message data that match the current request when receiving a roadside accident extraction request from the cloud platform, form a corresponding roadside accident dataset, and send it back to the cloud platform:

[0049] The current accident collection time period is extracted from the current vehicle-side accident extraction request and used as the current time period; the sensor data records of the sensor type being camera and whose fifth data timestamp is within the current time period are extracted from the multimodal dataset to form the corresponding roadside visual dataset; the second roadside message records in the second roadside message set that are within the current time period are extracted to form the corresponding roadside broadcast message set; the second vehicle-side message records in the second vehicle-side message set that are within the current time period are extracted to form the corresponding roadside vehicle message set; and the roadside accident dataset, composed of the roadside visual dataset, the roadside broadcast message set, and the roadside vehicle message set obtained this time, is sent back to the cloud platform.

[0050] Preferably, the cloud acquisition and transmission module is specifically used when performing cloud vehicle data synchronization based on the vehicle-to-cloud synchronization dataset periodically uploaded by each of the vehicle-end acquisition and transmission modules:

[0051] The currently received vehicle-cloud synchronized dataset is taken as the current dataset; the first and second synchronized datasets of the current dataset are taken as the corresponding current chassis dataset and current fault dataset; the first vehicle subset in the vehicle synchronized dataset corresponding to the current dataset is taken as the current vehicle subset; all the first chassis data records of the current chassis dataset are added to the chassis synchronized dataset of the current vehicle subset; and all the first fault data records of the current fault dataset are added to the fault synchronized dataset of the current vehicle subset.

[0052] The cloud-based data acquisition and transmission module is specifically used when performing cloud-based roadside data synchronization based on the roadside cloud synchronization dataset periodically uploaded by each of the roadside data acquisition and transmission modules:

[0053] The currently received road-cloud synchronization dataset is taken as the current dataset; and the third, fourth, and fifth synchronization datasets of the current dataset are taken as the corresponding current participant dataset, current signal dataset, and current event dataset;

[0054] The first roadside subset corresponding to the current dataset in the roadside synchronization dataset is taken as the current roadside subset; and the participant synchronization dataset, the signal synchronization dataset, and the event synchronization dataset in the current roadside subset are taken as the corresponding current synchronization dataset A, current synchronization dataset B, and current synchronization dataset C.

[0055] The process iterates through all subsets of the first participants in the current participant dataset. During this iteration, the first participant subset currently being iterated through is taken as the corresponding current iteration subset. The subset of first participants in the current synchronization dataset A whose first participant identifier matches the first participant identifier of the current iteration subset is taken as the corresponding current matching subset. The process checks whether the current matching subset is empty. If it is, the current iteration subset is added to the current synchronization dataset A. If not, the current matching subset is replaced based on the current iteration subset.

[0056] The system iterates through all the first traffic light subsets in the current signal dataset. During this iteration, the currently iterated first traffic light subset is taken as the corresponding current iteration subset. The first traffic light subset in the current synchronization dataset B whose first traffic light identifier matches the first traffic light identifier of the currently iterated subset is taken as the corresponding current matching subset. The system checks whether the current matching subset is empty. If it is, the current iteration subset is added to the current synchronization dataset B. If not, the current matching subset is replaced based on the current iteration subset.

[0057] Then, all the first event data records in the current event dataset are added to the current synchronization dataset C.

[0058] Preferably, the accident processing module is specifically used when monitoring traffic accidents in the roadside synchronous dataset and generating corresponding monitoring accident information based on the real-time monitoring results:

[0059] The record addition operation of each event synchronization dataset in the roadside synchronization dataset is monitored in real time; and the latest added first event data record in any event synchronization dataset is taken as the current new record; and when the first event type of the current new record is a traffic accident event, a corresponding monitoring accident information is composed of the sixth data timestamp of the current new record, the second roadside identifier, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane identifiers occupied by the accident site.

[0060] Preferably, the accident handling module is specifically used when the road network meteorological dataset and the road network traffic dataset are collected based on the various monitored accident information to obtain the corresponding cloud dataset:

[0061] The sixth data timestamp and the second roadside identifier of the currently monitored accident information are used as the corresponding current time and current roadside identifier; the first region subset in the road network meteorological dataset corresponding to the current roadside identifier is used as the current region meteorological subset, and the second region subset in the road network traffic dataset corresponding to the current roadside identifier is used as the current region traffic subset; the first meteorological dataset of the first meteorological data record whose seventh data timestamp is closest to the current time in the current region meteorological subset is extracted to form the corresponding accident area meteorological data; the first traffic indicator set of the first traffic data record whose eighth data timestamp is closest to the current time in the current region traffic subset is extracted to form the corresponding accident area traffic data; and the accident area meteorological data and the accident area traffic data obtained this time are used to form the corresponding cloud dataset.

[0062] Preferably, the accident processing module is specifically used when the vehicle synchronous dataset and the roadside synchronous dataset are collected based on the various monitored accident information to obtain the corresponding vehicle-road dataset:

[0063] The sixth data timestamp, the second roadside identifier, the coordinates of the center point of the accident site, and the set of traffic participants within the accident site of the current monitored accident information are extracted as the corresponding current time, current roadside identifier, current center point coordinates, and current set of participants;

[0064] The current start time is obtained by subtracting the preset third duration from the current time, and the current end time is obtained by adding the third duration to the current time. The current start time and the current end time together form the corresponding current accident period.

[0065] The system iterates through all traffic participant data of type motor vehicle in the current participant set. During this iteration, the traffic participant identifier of the currently iterated traffic participant data is used as the current vehicle identifier. The first vehicle subset in the vehicle synchronization dataset that matches the current vehicle identifier is used as the current vehicle subset. All first chassis data records in the chassis synchronization dataset of the current vehicle subset that are in the current accident time period are extracted to form the corresponding chassis data sequence. All first fault data records in the fault synchronization dataset of the current vehicle subset that are in the current accident time period are extracted to form the corresponding fault data sequence. The current vehicle identifier is used as a corresponding accident vehicle identifier. The accident vehicle identifier, chassis data sequence, and fault data sequence obtained in this iteration form a corresponding vehicle-side data set. At the end of this iteration, all the obtained vehicle-side data sets form the corresponding vehicle-side dataset.

[0066] And the participant synchronization dataset and the signal synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset are taken as the corresponding current participant dataset and current signal dataset;

[0067] The system iterates through all traffic participant data in the current participant set. During this iteration, the traffic participant identifier of the currently iterated traffic participant data is used as the current participant identifier. The subset of first participants in the current participant dataset whose first participant identifier matches the current participant identifier is used as the current subset. The trajectory segments of the first participants in the current subset that are in the current accident time period are extracted as the corresponding accident participant trajectories. The traffic participant identifier and traffic participant type of the current participant data are used as a set of corresponding accident participant identifiers and accident participant types. The accident participant identifiers, accident participant types, and accident participant trajectories obtained in this iteration are combined to form a corresponding participant trajectory data set. At the end of this iteration, all the obtained participant trajectory data are combined to form the corresponding participant trajectory set.

[0068] The first traffic light subset in the current signal dataset whose Euclidean distance between the first traffic light coordinates and the current center point coordinates does not exceed a preset first distance threshold is taken as the corresponding neighboring traffic light subset; the total number of the neighboring traffic light subsets obtained this time is identified; if the total number of subsets obtained this time is 0, the corresponding signal trajectory set is set to empty; if the total number of subsets obtained this time is greater than 0, the trajectory segments on the first signal tracking trajectory of each neighboring traffic light subset that are in the current accident period are extracted as the corresponding neighboring traffic light trajectory, and the second roadside identifier, the first traffic light identifier, and the first traffic light coordinates of each neighboring traffic light subset are used as a set of corresponding accident roadside identifier, neighboring traffic light identifier, and neighboring traffic light coordinates, and the set of accident roadside identifier, neighboring traffic light identifier, neighboring traffic light coordinates, and neighboring traffic light trajectory corresponding to each neighboring traffic light subset is used to form a corresponding traffic light trajectory data, and all the obtained traffic light trajectory data are used to form the corresponding signal trajectory set;

[0069] The obtained vehicle-side dataset, participant trajectory set, and signal trajectory set constitute the corresponding vehicle-road dataset.

[0070] Preferably, the accident processing module is specifically used when, based on each of the vehicle-road datasets, it sends corresponding vehicle-side and roadside accident extraction requests to the vehicle-side and roadside devices related to the current accident and generates corresponding supplementary vehicle-road datasets based on the request-returned data:

[0071] Each of the vehicle-road datasets is taken as the current vehicle-road dataset; the vehicle-end dataset of the current vehicle-road dataset is taken as the corresponding current vehicle-end dataset; the sixth data timestamp and the second roadside identifier of the monitoring accident information corresponding to the current vehicle-road dataset are taken as the corresponding current time and current roadside identifier; and the participant synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset is taken as the corresponding current participant synchronization dataset.

[0072] The current start time is obtained by subtracting the preset third duration from the current time, and the current end time is obtained by adding the third duration to the current time. The current start time and the current end time together form the corresponding accident collection period.

[0073] The vehicle-end device corresponding to each vehicle-end data in the current vehicle-end dataset is taken as the accident vehicle device; the vehicle-end accident extraction request carrying the accident collection period is sent to each accident vehicle device; the vehicle-end accident dataset returned by each accident vehicle device is taken as the corresponding accident vehicle subset; and all the obtained accident vehicle subsets are used to form the corresponding accident vehicle supplementary set.

[0074] The current participant synchronization dataset includes the first participant subset whose type is motor vehicle, whose trajectory time period overlaps with the accident collection time period, and whose identifier does not overlap with any vehicle identifiers in the current vehicle dataset. The total number of the obtained surrounding participant subsets is then identified. If the total number of subsets is 0, the corresponding surrounding vehicle supplementary set is set to empty. If the total number of subsets is greater than 0, the vehicle-side devices corresponding to each surrounding participant subset are designated as surrounding vehicle devices, and the vehicle-side accident extraction request carrying the accident collection time period is sent to each surrounding vehicle device. The vehicle-side accident dataset returned by each surrounding vehicle device is then used as the corresponding surrounding vehicle subset, and all obtained surrounding vehicle subsets form the corresponding surrounding vehicle supplementary set.

[0075] The roadside device corresponding to the current roadside identifier is designated as the current device; the roadside accident extraction request carrying the accident collection period is sent to the current device; and the roadside accident dataset returned by the current device is designated as the corresponding roadside supplementary dataset.

[0076] The corresponding vehicle-road supplementary dataset is composed of the accident vehicle supplementary set, the surrounding vehicle supplementary set, and the roadside supplementary dataset obtained in this study.

[0077] Preferably, the accident processing module is specifically used when performing multi-dimensional accident data integration on the vehicle-road dataset, the vehicle-road supplementary dataset, and the cloud dataset corresponding to each of the monitored accident information, and refreshing the road network accident dataset according to the integration result:

[0078] Each of the aforementioned monitored incident information is used as the current incident information;

[0079] The sixth data timestamp of the current accident information, the second roadside identifier, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane identifiers occupied by the accident site are used as a set of corresponding accident occurrence time, accident roadside identifier, accident type, center point of the accident area, set of vertex coordinates of the accident area, area of ​​the accident area, set of accident participants, and set of accident lanes to form a corresponding basic accident information;

[0080] And extract the corresponding vehicle-side dataset, participant trajectory set, signal trajectory set, accident vehicle supplementary set, surrounding vehicle supplementary set, roadside supplementary dataset, accident area meteorological data and accident area traffic data from the vehicle-road dataset, vehicle-road supplementary dataset and cloud dataset corresponding to the current accident information;

[0081] The first accident subset, composed of the basic accident information corresponding to the current accident information, the meteorological data of the accident area, the traffic data of the accident area, the vehicle-side dataset, the participant trajectory set, the signal trajectory set, the accident vehicle supplementary set, the surrounding vehicle supplementary set, and the roadside supplementary dataset, is added to the road network accident dataset.

[0082] This invention provides a traffic accident information processing system based on multi-dimensional fusion data from vehicles, roads, and the cloud. As described above, the system includes: vehicle-side equipment, roadside equipment, and a cloud platform. The vehicle-side equipment includes a vehicle-side data acquisition and transmission module and a vehicle-side storage module. The roadside equipment includes a roadside data acquisition and transmission module and a roadside storage module. The cloud platform includes a cloud-based data acquisition and transmission module, a cloud-based storage module, and an accident processing module. The vehicle-side data acquisition and transmission module continuously collects and stores chassis operation data, EDR data, DVR data, and fault data output during vehicle operation via the vehicle-mounted system. It also synchronously receives and stores roadside messages sent from the roadside unit equipment to the vehicle-mounted system and vehicle-side messages sent from the vehicle-mounted system to the roadside unit equipment. It periodically generates a vehicle-cloud synchronized dataset based on the stored chassis operation data and fault data and sends it to the cloud platform. Upon receiving a vehicle-side accident extraction request from the cloud platform, it extracts the EDR, DVR, and vehicle-road message data matching the current request to form a vehicle-side accident dataset and sends it back to the cloud platform. The roadside data acquisition and transmission module continuously collects and stores multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene (road segment or intersection) through roadside edge computing devices; it synchronously receives and stores roadside messages sent by roadside unit devices to the onboard systems of all passing vehicles, as well as vehicle-to-vehicle messages sent by the onboard systems of all passing vehicles to roadside unit devices; it periodically generates a road-cloud synchronized dataset based on the stored traffic participant, signal, and traffic event perception data and sends it to the cloud platform; and when it receives a roadside accident extraction request from the cloud platform, it extracts the visual sensor data and vehicle-road message data matching the current request to form a roadside accident dataset and sends it back to the cloud platform. The cloud-based data acquisition and transmission module continuously collects and stores meteorological and traffic data of the entire road network through a third-party data platform; and performs cloud-to-cloud vehicle data synchronization and cloud-to-roadside data synchronization based on the vehicle-to-cloud synchronized dataset and the road-to-cloud synchronized dataset. The accident handling module is used to monitor traffic accidents using the roadside synchronous dataset in the cloud; and to synthesize cloud datasets based on the road network meteorological dataset and road network traffic dataset based on the monitored accident information; and to synthesize vehicle-road datasets based on the vehicle synchronous dataset and roadside synchronous dataset in the cloud based on the monitored accident information; and to send corresponding vehicle-roadside accident extraction requests to the vehicle-side roadside devices related to the current accident according to each vehicle-road dataset and generate vehicle-road supplementary datasets based on the request feedback data; and to integrate multi-dimensional accident data of the vehicle-road dataset, vehicle-road supplementary dataset, and cloud dataset corresponding to each monitored accident information.This invention provides a system for real-time integration of vehicle-road-cloud multidimensional data for any traffic accident occurring on a road network. This data includes not only accident vehicle information collected from the vehicles involved, but also bystander information collected from other vehicles passing by the accident scene, global roadside information collected from roadside unit devices / roadside edge computing devices at the accident site, and meteorological and traffic information for the accident area obtained from the cloud. Based on embodiments of this invention, a corresponding multidimensional dataset can be output promptly for any traffic accident. Even if the accident scene is damaged or the parties involved fail to preserve images in time, this dataset can still provide sufficient information for accident handling personnel. This invention improves the efficiency of traffic accident handling while reducing its complexity. Attached Figure Description

[0083] Figure 1 A module structure diagram of a traffic accident information processing system based on multi-dimensional fusion data of vehicle, road, and cloud provided in an embodiment of the present invention;

[0084] Figure 2 A schematic diagram of six types of datasets provided in the vehicle-side storage module of this invention;

[0085] Figure 3 A schematic diagram of six types of datasets provided in the embodiments of the present invention for the roadside storage module;

[0086] Figure 4 This is a schematic diagram of five types of datasets provided in the cloud storage module of this invention. Detailed Implementation

[0087] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0088] This invention provides a traffic accident information processing system based on multi-dimensional fusion data from vehicle-road-cloud, such as... Figure 1 The module structure diagram of a traffic accident information processing system based on multi-dimensional fusion data of vehicle, road, and cloud provided in an embodiment of the present invention is shown. It mainly includes: vehicle-side equipment 1, roadside equipment 2, and cloud platform 3. Vehicle-side equipment 1 is connected to roadside equipment 2 and cloud platform 3 respectively, and roadside equipment 2 is connected to cloud platform 3.

[0089] Vehicle-side equipment 1 includes a vehicle-side data acquisition and transmission module 11 and a vehicle-side storage module 12, with the vehicle-side data acquisition and transmission module 11 and the vehicle-side storage module 12 connected together. Roadside equipment 2 includes a roadside data acquisition and transmission module 21 and a roadside storage module 22, with the roadside data acquisition and transmission module 21 and the roadside storage module 22 connected together. Cloud platform 3 includes a cloud-side data acquisition and transmission module 31, a cloud-side storage module 32, and an accident handling module 33, with the cloud-side storage module 32 connected to both the cloud-side data acquisition and transmission module 31 and the accident handling module 33.

[0090] (I) Vehicle-side data acquisition and transmission module 11:

[0091] In this embodiment of the invention, the vehicle-side data acquisition and transmission module 11 interfaces with the vehicle's onboard system. Here, the onboard system includes an autonomous driving system, a vehicle event data recording system, and an onboard video driving recorder system.

[0092] The vehicle-side data acquisition and transmission module 11 is used to continuously collect chassis operation data, EDR data, DVR data, and fault data output during the operation of the vehicle through the vehicle-mounted system and store the collected data into the vehicle-side storage module 12; it also synchronously receives roadside messages sent from roadside unit devices to the vehicle-mounted system and vehicle-side messages sent from the vehicle-mounted system to the roadside unit devices and stores the messages into the vehicle-side storage module 12; it periodically generates corresponding vehicle-cloud synchronous datasets based on the stored chassis operation data and fault data and sends them to the cloud platform 3; and when it receives a vehicle-side accident extraction request from the cloud platform 3, it extracts the EDR, DVR, and vehicle-road message data that match the current request to form the corresponding vehicle-side accident dataset and sends it back to the cloud platform 3.

[0093] In one specific implementation of this invention, the vehicle-side data acquisition and transmission module 11 is specifically used to periodically generate a corresponding vehicle-cloud synchronous dataset based on the stored chassis operating data and fault data and send it to the cloud platform 3:

[0094] According to the preset vehicle-side synchronization frequency, the current time is periodically used as the first end time, and the time point obtained by subtracting the preset first duration from the first end time is used as the first start time. The first start time and the first end time form the corresponding current most recent time period. The first chassis data records in the first chassis dataset that are in the current most recent time period are extracted to form the corresponding first synchronization dataset. The first fault data records in the first fault dataset that are in the current most recent time period are extracted to form the corresponding second synchronization dataset. The obtained first and second synchronization datasets form the corresponding vehicle-cloud synchronization dataset and are sent to the cloud platform 3.

[0095] Here, the vehicle-side synchronization frequency in this embodiment of the invention is a preset time frequency parameter; the first duration is a preset time length parameter.

[0096] In another specific implementation of this invention, the vehicle-side data acquisition and transmission module 11 is specifically used to extract the EDR, DVR, and vehicle-road message data matching the current request to form a corresponding vehicle-side accident dataset and send it back to the cloud platform 3 when it receives a vehicle-side accident extraction request from the cloud platform 3:

[0097] Extract the corresponding accident collection time period from the current vehicle-side accident extraction request as the current time period; extract the first EDR data record in the first EDR dataset that is in the current time period to form the corresponding EDR data sequence; extract the first DVR data record in the first DVR dataset that is in the current time period to form the corresponding DVR data sequence; extract the first roadside message record in the first roadside message set that is in the current time period to form the corresponding roadside message sequence; extract the first vehicle-side message record in the first vehicle-side message set that is in the current time period to form the corresponding vehicle-side message sequence; and send the corresponding vehicle-side accident dataset obtained this time, consisting of the EDR data sequence, DVR data sequence, roadside message sequence, and vehicle-side message sequence, back to the cloud platform 3.

[0098] (II) Vehicle-side storage module 12:

[0099] The vehicle-side storage module 12 in this embodiment of the invention is used to store a first chassis dataset, a first EDR dataset, a first DVR dataset, a first fault dataset, a first roadside message set, and a first vehicle-side message set.

[0100] like Figure 2 The diagram illustrates six types of datasets for the vehicle-side storage module provided in this embodiment of the invention. The first chassis dataset in this embodiment includes multiple first chassis data records. Each first chassis data record includes a first data timestamp, a first vehicle identifier, a first driving mode, a first scene identifier, first vehicle coordinates, a first vehicle speed, a first acceleration, a first heading angle, a first steering wheel angle, a first brake pedal opening degree, and a first accelerator pedal opening degree. The first driving mode includes manual driving and automatic driving; the first scene identifier includes road segment lane identifiers and intersection lane identifiers.

[0101] like Figure 2 As shown, the first EDR dataset in this embodiment of the invention includes multiple first EDR data records. Each first EDR data record includes a second data timestamp, a first vehicle identifier, and multiple EDR sub-items. These multiple EDR sub-items consist of some or all of the following: vehicle motion data, driving operation data, engine operating status data, collision data, safety system data, tire pressure data, and vehicle alarm data. Each EDR sub-item consists of one or more corresponding secondary data sub-items.

[0102] like Figure 2 As shown, the first DVR dataset in this embodiment of the invention includes multiple first DVR data records. Each first DVR data record includes a third data timestamp, a first vehicle identifier, and multiple DVR sub-items. These multiple DVR sub-items consist of some or all of the following: video data sub-items, audio data sub-items, DVR operating status data sub-items, and storage capacity sub-items; each DVR sub-item consists of one or more corresponding secondary data sub-items.

[0103] like Figure 2 As shown, the first fault dataset in this embodiment of the invention includes multiple first fault data records. Each first fault data record includes a fourth data timestamp, a first vehicle identifier, and one or more vehicle fault types.

[0104] like Figure 2 As shown, the first roadside message set in this embodiment of the invention includes multiple first roadside message records. Each first roadside message record includes a first message timestamp, a first roadside identifier, a first message type, and first message content data. The first message types include Signal Phase and Timing Message (SPAT), Roadside Message (RSM), Roadside Information (RSI), Map Message (MAP), Sensor Sharing Message (SSM), and Service Announcement Message (SAM).

[0105] like Figure 2 As shown, the first vehicle-side message set in this embodiment of the invention includes multiple first vehicle-side message records. Each first vehicle-side message record includes a second message timestamp, a first vehicle identifier, a second message type, and second message content data. The second message type includes Basic Safety Message (BSM), SSM message, Intention Sharing Message (ISM), and SAM message.

[0106] (III) Roadside data acquisition and transmission module 21:

[0107] In this embodiment of the invention, the roadside data acquisition and transmission module 21 is connected to the roadside edge computing device and roadside unit device 2 in the same traffic scenario; the traffic scenario types in this embodiment of the invention include road segments and intersections.

[0108] The roadside data acquisition and transmission module 21 is used to continuously collect multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene through roadside edge computing devices and store the collected data into the roadside storage module 22; it also synchronously receives roadside messages sent by the roadside unit device to the vehicle-mounted systems of all passing vehicles and vehicle-end messages sent by the vehicle-mounted systems of all passing vehicles to the roadside unit device and stores the received message data into the roadside storage module 22; it periodically generates corresponding road-cloud synchronous datasets based on the stored traffic participant, signal, and traffic event perception data and sends them to the cloud platform 3; and when it receives a roadside accident extraction request issued by the cloud platform 3, it extracts the visual sensor data and vehicle-road message data that match the current request to form the corresponding roadside accident dataset and sends it back to the cloud platform 3.

[0109] In another specific implementation of this invention, the roadside data acquisition and transmission module 21 is specifically used to periodically generate a corresponding road-cloud synchronized dataset based on stored traffic participant, signal, and traffic event perception data and send it to the cloud platform 3:

[0110] According to the preset roadside synchronization frequency, the current time is periodically used as the second end time, and the time point obtained by subtracting the preset second duration from the second end time is used as the second start time. The second start time and the second end time constitute the corresponding current most recent time period. The subset of first participants whose tracking trajectories have been updated in the first participant dataset within the current most recent time period is extracted to form the corresponding third synchronization dataset. The subset of first traffic lights whose tracking trajectories have been updated in the first signal dataset within the current most recent time period is extracted to form the corresponding fourth synchronization dataset. The first event data records in the current most recent time period are extracted from the first event dataset to form the corresponding fifth synchronization dataset. The obtained third, fourth and fifth synchronization datasets are combined to form the corresponding road-cloud synchronization dataset and sent to the cloud platform 3.

[0111] Here, the roadside synchronization frequency in this embodiment of the invention is a preset time frequency parameter; the second duration is a preset time length parameter.

[0112] In another specific implementation of this invention, the roadside data acquisition and transmission module 21 is specifically used to extract the visual sensor data and vehicle-road message data that match the current request to form a corresponding roadside accident dataset and send it back to the cloud platform 3 when it receives a roadside accident extraction request from the cloud platform 3:

[0113] Extract the corresponding accident collection time period from the current vehicle-side accident extraction request as the current time period; extract sensor data records with camera sensor type and fifth data timestamp within the current time period from the multimodal dataset to form the corresponding roadside visual dataset; extract the second roadside message records within the current time period from the second roadside message set to form the corresponding roadside broadcast message set; extract the second vehicle-side message records within the current time period from the second vehicle-side message set to form the corresponding roadside vehicle message set; and send the corresponding roadside accident dataset, composed of the roadside visual dataset, roadside broadcast message set, and roadside vehicle message set obtained this time, back to the cloud platform 3.

[0114] (iv) Roadside storage module 22:

[0115] The roadside storage module 22 in this embodiment of the invention is used to store a multimodal dataset, a first participant dataset, a first signal dataset, a first event dataset, a second roadside message set, and a second vehicle-side message set.

[0116] like Figure 3 The diagram illustrates six types of datasets for the roadside storage module provided in this embodiment of the invention. The multimodal dataset in this embodiment includes multiple sensor data records. Each sensor data record includes a fifth data timestamp, a second roadside identifier, a sensor identifier, a sensor type, and sensor data. Sensor types include cameras, lidar, millimeter-wave radar, ultrasonic radar, and traffic lights. When the sensor type is a camera, the sensor data is the corresponding image, image sequence, or video data. When the sensor type is lidar or millimeter-wave radar, the sensor data is the corresponding lidar dense point cloud or millimeter-wave radar sparse point cloud. When the sensor type is ultrasonic radar, the sensor data is the corresponding ranging dataset. When the sensor type is a traffic light, the sensor data is the corresponding traffic light status; traffic light status includes red, green, and yellow lights.

[0117] like Figure 3 As shown, the first participant dataset in this embodiment of the invention includes multiple subsets of first participants. Each subset of first participants corresponds one-to-one with a traffic participant who has been or is currently in the current traffic scenario. The first participant subset includes a second roadside identifier, a first participant identifier, a first participant type, and a first participant tracking trajectory. The first participant type includes pedestrians, motor vehicles, and non-motor vehicles; the first participant tracking trajectory includes multiple first tracking trajectory points; each first tracking trajectory point includes a first trajectory point timestamp, a first target image, a first target geometric dimension, first target coordinates, a first target vehicle speed, and a first target heading angle. It should be noted that when the first participant type is a motor vehicle, the first participant identifier is also the vehicle identifier of the current motor vehicle.

[0118] like Figure 3 As shown, when the first signal dataset of this embodiment is not empty, it includes one or more first traffic light subsets. If traffic lights are present in the current traffic scene, the first traffic light subset corresponds one-to-one with the traffic lights in the current traffic scene. The first traffic light subset includes a second roadside marker, a first traffic light marker, first traffic light coordinates, and a first signal tracking trajectory. The first signal tracking trajectory includes multiple second tracking trajectory points; each second tracking trajectory point includes a second trajectory point timestamp, a first signal status, and the remaining time of the first signal; the first signal status includes red, green, and yellow lights.

[0119] like Figure 3 As shown, the first event dataset in this embodiment of the invention includes multiple first event data records. Each first event data record includes a sixth data timestamp, a second roadside identifier, a first event type, and first event data. The first event type includes at least speeding violations, running red lights, driving outside designated lanes, road construction obstruction, and traffic accidents.

[0120] It should be noted that: When the first event type is a speeding violation, running a red light violation, or driving outside of designated lanes, the first event data consists of the participant identifier corresponding to the violating vehicle, the tracking trajectory during the violation period, and the video of the violation period. When the first event type is a road construction violation, the first event data consists of the coordinates of the center point of the construction site, the coordinate set of the vertices of the construction area, the area of ​​the construction area, and the set of lane markers occupied by the construction site. When the first event type is a traffic accident, the first event data consists of the traffic accident type, the coordinates of the center point of the accident site, the coordinate set of the vertices of the accident area, the area of ​​the accident area, the set of traffic participants within the accident site, and the set of lane markers occupied by the accident site; wherein, the set of traffic participants within the accident site consists of one or more traffic participant data, each of which consists of a corresponding traffic participant identifier and traffic participant type, including pedestrians, motor vehicles, and non-motor vehicles; the set of lane markers occupied by the accident site consists of one or more lane markers.

[0121] like Figure 3 As shown, the second pathside message set in this embodiment of the invention includes multiple second pathside message records; the second pathside message record includes a third message timestamp, a second pathside identifier, a third message type, and third message content data; the third message type includes SPAT message, RSM message, RSI message, MAP message, SSM message, and SAM message.

[0122] like Figure 3As shown, the second vehicle-side message set in this embodiment of the invention includes multiple second vehicle-side message records; the second vehicle-side message record includes a fourth message timestamp, a second vehicle-side identifier, a fourth message type, and fourth message content data; the fourth message type includes BSM message, SSM message, ISM message, and SAM message.

[0123] (V) Cloud Acquisition and Transmission Module 31:

[0124] The cloud-based data acquisition and transmission module 31 of this embodiment of the invention interfaces with a third-party data platform; the third-party data platform of this embodiment of the invention includes at least a meteorological data platform and a traffic management data platform.

[0125] The cloud-based data acquisition and transmission module 31 is used to continuously collect meteorological and traffic data of the entire road network through a third-party data platform and store the collected data in the cloud storage module 32; and to synchronize vehicle data in the cloud according to the vehicle-cloud synchronization dataset uploaded periodically by each vehicle-side data acquisition and transmission module 11; and to synchronize roadside data in the cloud according to the road-cloud synchronization dataset uploaded periodically by each roadside data acquisition and transmission module 21.

[0126] In another specific implementation of this invention, the cloud acquisition and transmission module 31 is specifically used to synchronize vehicle data in the cloud based on the vehicle-cloud synchronization dataset periodically uploaded by each vehicle-end acquisition and transmission module 11:

[0127] The currently received vehicle-cloud synchronization dataset is used as the current dataset; the first and second synchronization datasets of the current dataset are used as the corresponding current chassis dataset and current fault dataset; the first vehicle subset in the vehicle synchronization dataset corresponding to the current dataset is used as the current vehicle subset; all first chassis data records of the current chassis dataset are added to the chassis synchronization dataset of the current vehicle subset; and all first fault data records of the current fault dataset are added to the fault synchronization dataset of the current vehicle subset.

[0128] In another specific implementation of this invention, the cloud acquisition and transmission module 31 is specifically used for synchronizing cloud-cloud roadside data based on the roadside acquisition and transmission modules 21 periodically uploaded by each roadside acquisition and transmission module 21:

[0129] Step A1: Take the currently received road-cloud synchronization dataset as the current dataset; and take the third, fourth and fifth synchronization datasets of the current dataset as the corresponding current participant dataset, current signal dataset and current event dataset;

[0130] Step A2, and take the first roadside subset in the roadside synchronization dataset that corresponds to the current dataset as the current roadside subset; and take the participant synchronization dataset, signal synchronization dataset, and event synchronization dataset of the current roadside subset as the corresponding current synchronization dataset A, current synchronization dataset B, and current synchronization dataset C;

[0131] Step A3 involves iterating through all first participant subsets of the current participant dataset; during this iteration, the first participant subset being iterated through is taken as the corresponding current iteration subset; and the first participant subset in the current synchronization dataset A whose first participant identifier matches the first participant identifier of the current iteration subset is taken as the corresponding current matching subset; and it checks whether the current matching subset is empty; if so, the current iteration subset is added to the current synchronization dataset A; otherwise, the current matching subset is replaced based on the current iteration subset.

[0132] Step A4 involves iterating through all first traffic light subsets in the current signal dataset. During this iteration, the first traffic light subset being iterated through is taken as the corresponding current iteration subset. The first traffic light subset in the current synchronization dataset B whose first traffic light identifier matches the first traffic light identifier of the current iteration subset is taken as the corresponding current matching subset. The system then checks whether the current matching subset is empty. If it is, the current iteration subset is added to the current synchronization dataset B. If not, the current matching subset is replaced based on the current iteration subset.

[0133] Step A5, and add all first event data records of the current event dataset to the current synchronization dataset C.

[0134] (vi) Cloud storage module 32:

[0135] The cloud storage module 32 in this embodiment of the invention is used to store road network meteorological datasets, road network traffic datasets, vehicle synchronization datasets, roadside synchronization datasets, and road network accident datasets.

[0136] like Figure 4 The diagram illustrates five types of datasets for the cloud storage module provided in this embodiment of the invention. The road network meteorological dataset in this embodiment includes multiple first regional subsets. Each first regional subset corresponds one-to-one with a traffic scenario. Each first regional subset includes multiple first meteorological data records. Each first meteorological data record includes a seventh data timestamp, a first regional identifier, and a first meteorological dataset; wherein, the first meteorological dataset consists of multiple types of meteorological data.

[0137] like Figure 4As shown, the road network traffic dataset in this embodiment of the invention includes multiple second region subsets. Each second region subset corresponds one-to-one with a traffic scenario. Each second region subset includes multiple first traffic data records. Each first traffic data record includes an eighth data timestamp, a second region identifier, and a first traffic indicator set; wherein, the first traffic indicator set consists of multiple types of traffic indicator data.

[0138] like Figure 4 As shown, the vehicle synchronization dataset in this embodiment of the invention includes multiple first vehicle subsets. Each first vehicle subset corresponds one-to-one with the vehicle-side device 1. The first vehicle subset includes a chassis synchronization dataset and a fault synchronization dataset; wherein, the chassis synchronization dataset includes multiple first chassis data records; and the fault synchronization dataset includes multiple first fault data records.

[0139] like Figure 4 As shown, the roadside synchronization dataset in this embodiment of the invention includes multiple first roadside subsets. Each first roadside subset corresponds one-to-one with a roadside device 2. The first roadside subsets include a participant synchronization dataset, a signal synchronization dataset, and an event synchronization dataset; wherein, the participant synchronization dataset includes multiple first participant subsets; the signal synchronization dataset includes multiple first traffic light subsets; and the event synchronization dataset includes multiple first event data records.

[0140] like Figure 4 As shown, the road network accident dataset in this embodiment of the invention includes multiple first accident subsets. The first accident subsets include basic accident information, meteorological data of the accident area, traffic data of the accident area, vehicle-side datasets, participant trajectory sets, signal trajectory sets, supplementary accident vehicle datasets, supplementary surrounding vehicle datasets, and supplementary roadside datasets.

[0141] Basic accident information includes the time of the accident, roadside signs, accident type, center point of the accident zone, set of vertices of the accident zone, area of ​​the accident zone, set of accident participants, and set of accident lanes.

[0142] The vehicle-side dataset consists of vehicle-side data from all accident vehicles within the current accident range. The vehicle-side data includes accident vehicle identification, chassis data sequence, and fault data sequence.

[0143] The participant trajectory set consists of the participant trajectory data of all traffic participants within the current accident area; the participant trajectory data includes the accident participant identifier, the accident participant type, and the accident participant trajectory.

[0144] When the signal trajectory set is not empty, it consists of the signal trajectory data of all traffic lights adjacent to the current accident area; the signal trajectory data includes the accident roadside markers, adjacent signal light markers, adjacent signal light coordinates, and adjacent signal light trajectories.

[0145] The accident vehicle supplement set consists of a subset of all accident vehicles within the current accident scope; the accident vehicle subset includes EDR data sequences, DVR data sequences, roadside message sequences, and vehicle-side message sequences.

[0146] When the surrounding vehicle supplement set is not empty, it consists of a subset of surrounding vehicles of all surrounding vehicles outside the current accident area; the surrounding vehicle subset includes EDR data sequence, DVR data sequence, roadside message sequence, and vehicle-side message sequence.

[0147] The roadside supplementary dataset consists of the roadside visual dataset, the roadside broadcast message set, and the roadside vehicle message set of the road segment or intersection where the current accident occurs.

[0148] (vii) Accident Handling Module 33:

[0149] The accident processing module 33 of this embodiment of the invention is used to monitor traffic accidents in the roadside synchronous dataset and generate corresponding monitoring accident information based on the real-time monitoring results; and to collect meteorological and traffic data from the road network meteorological dataset and the road network traffic dataset based on each monitoring accident information to obtain the corresponding cloud dataset; and to collect vehicle-road data from the vehicle synchronous dataset and the roadside synchronous dataset based on each monitoring accident information to obtain the corresponding vehicle-road dataset; and to send corresponding vehicle-end and roadside accident extraction requests to the vehicle-end device 1 and roadside device 2 related to the current accident according to each vehicle-road dataset, and to generate the corresponding vehicle-road supplementary dataset based on the request feedback data; and to perform multi-dimensional accident data integration on the vehicle-road dataset, vehicle-road supplementary dataset and cloud dataset corresponding to each monitoring accident information, and to refresh the road network accident dataset according to the integration result.

[0150] In another specific implementation of this invention, the accident handling module 33 is specifically used to monitor traffic accidents on the roadside synchronous dataset and generate corresponding monitoring accident information based on the real-time monitoring results:

[0151] The system monitors the record addition operations of each event synchronization dataset in the roadside synchronization dataset in real time; and takes the latest added first event data record of any event synchronization dataset as the current newly added record; and when the first event type of the current newly added record is a traffic accident event, a corresponding monitoring accident information is composed of the sixth data timestamp of the current newly added record, the second roadside identifier, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane identifiers occupied by the accident site.

[0152] In another specific implementation of this invention, the accident handling module 33 is specifically used to obtain the corresponding cloud dataset by collecting meteorological and traffic data from the road network meteorological dataset and the road network traffic dataset based on various monitored accident information:

[0153] The sixth data timestamp and the second roadside identifier of the current monitored accident information are used as the corresponding current time and current roadside identifier; the first regional subset corresponding to the current roadside identifier in the road network meteorological dataset is used as the current regional meteorological subset, and the second regional subset corresponding to the current roadside identifier in the road network traffic dataset is used as the current regional traffic subset; the first meteorological dataset of the first meteorological data record whose seventh data timestamp is closest to the current time in the current regional meteorological subset is extracted to form the corresponding accident area meteorological data; the first traffic indicator set of the first traffic data record whose eighth data timestamp is closest to the current time in the current regional traffic subset is extracted to form the corresponding accident area traffic data; and the accident area meteorological data and accident area traffic data obtained in this study form the corresponding cloud dataset.

[0154] In another specific implementation of this invention, the accident processing module 33 is specifically used to obtain the corresponding vehicle-road dataset by collecting vehicle-road data from the vehicle synchronization dataset and the roadside synchronization dataset based on various monitored accident information:

[0155] Step B1: Extract the sixth data timestamp, the second roadside sign, the coordinates of the center point of the accident site, and the set of traffic participants within the accident site from the current monitored accident information as the corresponding current time, current roadside sign, current center point coordinates, and current set of participants.

[0156] Step B2, and take the current time minus the preset third duration as the current start time, and take the current time plus the third duration as the current end time, and make up the current accident period by the current start time and the current end time;

[0157] Here, the third duration in this embodiment of the invention is a pre-set time length parameter;

[0158] Step B3 involves iterating through all traffic participant data of motor vehicle type in the current participant set. During this iteration, the traffic participant identifier of the currently iterated traffic participant data is used as the current vehicle identifier. The first vehicle subset in the vehicle synchronization dataset that matches the current vehicle identifier is used as the current vehicle subset. All first chassis data records in the chassis synchronization dataset of the current vehicle subset that are in the current accident period are extracted to form the corresponding chassis data sequence. All first fault data records in the fault synchronization dataset of the current vehicle subset that are in the current accident period are extracted to form the corresponding fault data sequence. The current vehicle identifier is used as a corresponding accident vehicle identifier. The accident vehicle identifier, chassis data sequence, and fault data sequence obtained in this iteration are combined to form a corresponding vehicle-side data set. At the end of this iteration, all the obtained vehicle-side data are combined to form the corresponding vehicle-side dataset.

[0159] Step B4, and take the participant synchronization dataset and signal synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset as the corresponding current participant dataset and current signal dataset;

[0160] Step B5 involves iterating through all traffic participant data in the current participant set. During this iteration, the traffic participant identifier of the currently iterated traffic participant data is used as the current participant identifier. The subset of first participants whose first participant identifier matches the current participant identifier in the current participant dataset is taken as the current subset. The trajectory segments of the first participant's tracking trajectory in the current subset that are in the current accident time period are extracted as the corresponding accident participant trajectories. The traffic participant identifier and traffic participant type of the current participant data are used as a set of corresponding accident participant identifiers and accident participant types. The accident participant identifiers, accident participant types, and accident participant trajectories obtained in this iteration are combined to form a corresponding participant trajectory data set. At the end of this iteration, all the obtained participant trajectory data are combined to form the corresponding participant trajectory set.

[0161] Step B6: The first subset of traffic lights whose Euclidean distance between the first traffic light coordinates and the current center point coordinates in the current signal dataset does not exceed a preset first distance threshold is taken as the corresponding neighboring traffic light subset; the total number of neighboring traffic light subsets obtained this time is identified; if the total number of subsets obtained this time is 0, the corresponding signal trajectory set is set to empty; if the total number of subsets obtained this time is greater than 0, the trajectory segments in the current accident period on the first signal tracking trajectory of each neighboring traffic light subset are extracted as the corresponding neighboring traffic light trajectories, and the second roadside identifier, the first traffic light identifier, and the first traffic light coordinates of each neighboring traffic light subset are used as a set of corresponding accident roadside identifier, neighboring traffic light identifier, and neighboring traffic light coordinates, and the set of accident roadside identifier, neighboring traffic light identifier, neighboring traffic light coordinates, and neighboring traffic light trajectory corresponding to each neighboring traffic light subset is used to form a corresponding traffic light trajectory data, and all the obtained traffic light trajectory data are used to form the corresponding signal trajectory set;

[0162] Here, the first distance threshold in this embodiment of the invention is a pre-set distance parameter;

[0163] Step B7, and the obtained vehicle-side dataset, participant trajectory set, and signal trajectory set are combined to form the corresponding vehicle-road dataset.

[0164] In another specific implementation of this invention, the accident processing module 33 is specifically used to send corresponding vehicle-end devices 1 and roadside devices 2 related to the current accident according to each vehicle-road dataset and generate corresponding vehicle-road supplementary datasets based on the request feedback data:

[0165] Step C1: Take each vehicle-road dataset as the current vehicle-road dataset; take the vehicle-side dataset of the current vehicle-road dataset as the corresponding current vehicle-side dataset; take the sixth data timestamp and the second roadside identifier of the monitoring accident information corresponding to the current vehicle-road dataset as the corresponding current time and current roadside identifier; take the participant synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset as the corresponding current participant synchronization dataset.

[0166] Step C2, and take the current time minus the third duration as the current start time, and the current time plus the third duration as the current end time, and the current start time and the current end time form the corresponding accident collection period.

[0167] Step C3: Take the vehicle-end device 1 corresponding to each vehicle-end data in the current vehicle-end dataset as the accident vehicle device; send the vehicle-end accident extraction request carrying the accident collection period to each accident vehicle device; take the vehicle-end accident dataset returned by each accident vehicle device as the corresponding accident vehicle subset; and form the corresponding accident vehicle supplementary set by all the obtained accident vehicle subsets.

[0168] Step C4: In the current participant synchronization dataset, the first participant type is a motor vehicle, the trajectory time period of the first participant's tracking trajectory intersects with the accident collection time period, and the first participant's identifier does not intersect with any vehicle identifiers in the current vehicle-end dataset. This first participant subset is then identified as the corresponding surrounding participant subset. The total number of surrounding participant subsets obtained this time is then identified. If the total number of subsets obtained this time is 0, the corresponding surrounding vehicle supplementary set is set to empty. If the total number of subsets obtained this time is greater than 0, the vehicle-end device 1 corresponding to each surrounding participant subset is then identified as the surrounding vehicle device. The vehicle-end accident extraction request carrying the accident collection time period is sent to each surrounding vehicle device, and the vehicle-end accident dataset returned by each surrounding vehicle device is taken as the corresponding surrounding vehicle subset. All the obtained surrounding vehicle subsets are then combined to form the corresponding surrounding vehicle supplementary set.

[0169] Step C5, and designate the roadside device 2 corresponding to the current roadside identifier as the current device; send the roadside accident extraction request carrying the accident collection period to the current device; and use the roadside accident dataset sent back by the current device as the corresponding roadside supplementary dataset;

[0170] Step C6, and the corresponding vehicle-road supplementary dataset is composed of the accident vehicle supplementary set, the surrounding vehicle supplementary set, and the roadside supplementary dataset obtained in this step.

[0171] In another specific implementation of this invention, the accident processing module 33 is specifically used to integrate multi-dimensional accident data from the vehicle-road dataset, vehicle-road supplementary dataset, and cloud dataset corresponding to each monitored accident information, and to refresh the road network accident dataset based on the integration result:

[0172] Step D1: Use each monitored incident information as the current incident information;

[0173] Step D2, and the sixth data timestamp of the current accident information, the second roadside sign, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane signs occupied by the accident site are used as a set of corresponding accident occurrence time, accident roadside sign, accident type, accident area center point, accident area vertex set, accident area area, set of accident participants, and set of accident lanes to form a corresponding basic accident information;

[0174] Step D3: Extract the corresponding vehicle-side dataset, participant trajectory set, signal trajectory set, accident vehicle supplementary set, surrounding vehicle supplementary set, roadside supplementary dataset, accident area meteorological data, and accident area traffic data from the vehicle-road dataset, vehicle-road supplementary dataset, and cloud dataset corresponding to the current accident information.

[0175] Step D4 involves adding a corresponding first accident subset to the road network accident dataset, which consists of the basic accident information corresponding to the current accident information, meteorological data of the accident area, traffic data of the accident area, vehicle-side dataset, participant trajectory set, signal trajectory set, accident vehicle supplementary set, surrounding vehicle supplementary set, and roadside supplementary dataset.

[0176] This invention provides a traffic accident information processing system based on multi-dimensional fusion data from vehicles, roads, and the cloud. As described above, the system includes: vehicle-side equipment, roadside equipment, and a cloud platform. The vehicle-side equipment includes a vehicle-side data acquisition and transmission module and a vehicle-side storage module. The roadside equipment includes a roadside data acquisition and transmission module and a roadside storage module. The cloud platform includes a cloud-based data acquisition and transmission module, a cloud-based storage module, and an accident processing module. The vehicle-side data acquisition and transmission module continuously collects and stores chassis operation data, EDR data, DVR data, and fault data output during vehicle operation via the vehicle-mounted system. It also synchronously receives and stores roadside messages sent from the roadside unit equipment to the vehicle-mounted system and vehicle-side messages sent from the vehicle-mounted system to the roadside unit equipment. It periodically generates a vehicle-cloud synchronized dataset based on the stored chassis operation data and fault data and sends it to the cloud platform. Upon receiving a vehicle-side accident extraction request from the cloud platform, it extracts the EDR, DVR, and vehicle-road message data matching the current request to form a vehicle-side accident dataset and sends it back to the cloud platform. The roadside data acquisition and transmission module continuously collects and stores multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene (road segment or intersection) through roadside edge computing devices; it synchronously receives and stores roadside messages sent by roadside unit devices to the onboard systems of all passing vehicles, as well as vehicle-to-vehicle messages sent by the onboard systems of all passing vehicles to roadside unit devices; it periodically generates a road-cloud synchronized dataset based on the stored traffic participant, signal, and traffic event perception data and sends it to the cloud platform; and when it receives a roadside accident extraction request from the cloud platform, it extracts the visual sensor data and vehicle-road message data matching the current request to form a roadside accident dataset and sends it back to the cloud platform. The cloud-based data acquisition and transmission module continuously collects and stores meteorological and traffic data of the entire road network through a third-party data platform; and performs cloud-to-cloud vehicle data synchronization and cloud-to-roadside data synchronization based on the vehicle-to-cloud synchronized dataset and the road-to-cloud synchronized dataset. The accident handling module is used to monitor traffic accidents using the roadside synchronous dataset in the cloud; and to synthesize cloud datasets based on the road network meteorological dataset and road network traffic dataset based on the monitored accident information; and to synthesize vehicle-road datasets based on the vehicle synchronous dataset and roadside synchronous dataset in the cloud based on the monitored accident information; and to send corresponding vehicle-roadside accident extraction requests to the vehicle-side roadside devices related to the current accident according to each vehicle-road dataset and generate vehicle-road supplementary datasets based on the request feedback data; and to integrate multi-dimensional accident data of the vehicle-road dataset, vehicle-road supplementary dataset, and cloud dataset corresponding to each monitored accident information.This invention provides a system for real-time integration of vehicle-road-cloud multidimensional data for any traffic accident occurring on a road network. This data includes not only accident vehicle information collected from the vehicles involved, but also bystander information collected from other vehicles passing by the accident scene, global roadside information collected from roadside unit devices / roadside edge computing devices at the accident site, and meteorological and traffic information for the accident area obtained from the cloud. Based on embodiments of this invention, a corresponding multidimensional dataset can be output promptly for any traffic accident. Even if the accident scene is damaged or the parties involved fail to preserve images in time, this dataset can still provide sufficient information for accident handling personnel. This invention improves the efficiency of traffic accident handling while reducing its complexity.

[0177] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0178] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0179] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A traffic accident information processing system based on multi-dimensional fusion data of vehicle, road, and cloud, characterized in that, The system includes: vehicle-mounted equipment, roadside equipment, and a cloud platform; The vehicle-mounted equipment is connected to both the roadside equipment and the cloud platform, and the roadside equipment is connected to the cloud platform. The vehicle-mounted equipment includes a vehicle-mounted data acquisition and transmission module and a vehicle-mounted storage module, which are connected to each other. The roadside equipment includes a roadside data acquisition and transmission module and a roadside storage module, which are connected to each other. The cloud platform includes a cloud-based data acquisition and transmission module, a cloud-based storage module, and an accident handling module, which are connected to both the cloud-based data acquisition and transmission module and the accident handling module. The vehicle-side data acquisition and transmission module interfaces with the vehicle's onboard system; The vehicle-side data acquisition and transmission module is used to continuously collect chassis operation data, EDR data, DVR data, and fault data output by the vehicle system during vehicle operation and store the collected data in the vehicle-side storage module; it also synchronously receives roadside messages sent from roadside unit devices to the vehicle system and vehicle-side messages sent from the vehicle system to roadside unit devices and stores the messages in the vehicle-side storage module; it periodically generates corresponding vehicle-cloud synchronized datasets based on the stored chassis operation data and fault data and sends them to the cloud platform; and when it receives a vehicle-side accident extraction request from the cloud platform, it extracts the EDR, DVR, and vehicle-road message data matching the current request to form a corresponding vehicle-side accident dataset and sends it back to the cloud platform. The vehicle-side storage module is used to store the first chassis dataset, the first EDR dataset, the first DVR dataset, the first fault dataset, the first roadside message set, and the first vehicle-side message set; The roadside data acquisition and transmission module interfaces with the roadside edge computing device and the roadside unit device in the same traffic scenario as the roadside equipment; the scenario types of the traffic scenario include road segments and intersections; The roadside data acquisition and transmission module is used to continuously acquire multimodal sensor data, traffic participant perception data, signal perception data, and traffic event perception data of the current traffic scene through the roadside edge computing device and store the acquired data in the roadside storage module; it also synchronously receives roadside messages sent by the roadside unit device to the vehicle-mounted systems of all passing vehicles and vehicle-end messages sent by the vehicle-mounted systems of all passing vehicles to the roadside unit device and stores the received message data in the roadside storage module; it periodically generates corresponding road-cloud synchronous datasets based on the stored traffic participant, signal, and traffic event perception data and sends them to the cloud platform; and when it receives a roadside accident extraction request from the cloud platform, it extracts the visual sensor data and vehicle-road message data that match the current request to form a corresponding roadside accident dataset and sends it back to the cloud platform. The roadside storage module is used to store a multimodal dataset, a first participant dataset, a first signal dataset, a first event dataset, a second roadside message set, and a second vehicle-side message set; The cloud-based data acquisition and transmission module interfaces with a third-party data platform; the third-party data platform includes a meteorological data platform and a traffic management data platform. The cloud-based data acquisition and transmission module is used to continuously collect meteorological and traffic data of the entire road network through the third-party data platform and store the collected data in the cloud storage module; and to synchronize vehicle data in the cloud according to the vehicle-to-cloud synchronization dataset periodically uploaded by each of the vehicle-side data acquisition and transmission modules; and to synchronize roadside data in the cloud according to the roadside synchronization dataset periodically uploaded by each of the roadside data acquisition and transmission modules. The cloud storage module is used to store road network meteorological datasets, road network traffic datasets, vehicle synchronization datasets, roadside synchronization datasets, and road network accident datasets. The accident handling module is used to monitor traffic accidents in the roadside synchronous dataset and generate corresponding monitoring accident information based on real-time monitoring results; and to collect meteorological and traffic data from the road network meteorological dataset and the road network traffic dataset based on each monitoring accident information to obtain corresponding cloud datasets; and to collect vehicle-road data from the vehicle synchronous dataset and the roadside synchronous dataset based on each monitoring accident information to obtain corresponding vehicle-road datasets; and to send corresponding vehicle-end and roadside accident extraction requests to vehicle-end and roadside devices related to the current accident based on each vehicle-road dataset, and generate corresponding vehicle-road supplementary datasets based on the request feedback data; and to perform multi-dimensional accident data integration on the vehicle-road datasets, vehicle-road supplementary datasets, and cloud datasets corresponding to each monitoring accident information, and refresh the road network accident dataset based on the integration results.

2. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data as described in claim 1, characterized in that, The in-vehicle system includes an autonomous driving system, a vehicle event data recording system, and an in-vehicle video driving recorder system; The first chassis dataset includes multiple first chassis data records; each first chassis data record includes a first data timestamp, a first vehicle identifier, a first driving mode, a first scene identifier, first vehicle coordinates, a first vehicle speed, a first acceleration, a first heading angle, a first steering wheel angle, a first brake pedal opening degree, and a first accelerator pedal opening degree; the first driving mode includes manual driving and automatic driving; the first scene identifier includes road segment lane identifiers and intersection lane identifiers. The first EDR dataset includes multiple first EDR data records; each first EDR data record includes a second data timestamp, a first vehicle identifier, and multiple EDR sub-items; the multiple EDR sub-items consist of some or all of the following sub-items: vehicle motion data, driving operation data, engine operating status data, collision data, safety system data, tire pressure data, and vehicle warning data; each EDR sub-item consists of one or more corresponding secondary data sub-items. The first DVR dataset includes multiple first DVR data records; The first DVR data record includes a third data timestamp, the first vehicle identifier, and multiple DVR sub-items; the multiple DVR sub-items consist of some or all of the video data sub-items, audio data sub-items, DVR working status data sub-items, and storage capacity sub-items; each DVR sub-item consists of one or more corresponding secondary data sub-items; The first fault dataset includes multiple first fault data records; The first fault data record includes a fourth data timestamp, the first vehicle identifier, and one or more vehicle fault types; The first roadside message set includes multiple first roadside message records; the first roadside message record includes a first message timestamp, a first roadside identifier, a first message type, and first message content data; the first message type includes SPAT message, RSM message, RSI message, MAP message, SSM message, and SAM message; The first vehicle-side message set includes multiple first vehicle-side message records; the first vehicle-side message record includes a second message timestamp, a first vehicle identifier, a second message type, and second message content data; the second message type includes BSM message, SSM message, ISM message, and SAM message.

3. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 2, characterized in that, The multimodal dataset includes multiple sensor data records; each sensor data record includes a fifth data timestamp, a second roadside identifier, a sensor identifier, a sensor type, and sensor data; the sensor types include cameras, lidar, millimeter-wave radar, ultrasonic radar, and traffic lights; when the sensor type is a camera, the sensor data is the corresponding image, image sequence, or video data; when the sensor type is lidar or millimeter-wave radar, the sensor data is the corresponding lidar dense point cloud or millimeter-wave radar sparse point cloud; when the sensor type is ultrasonic radar, the sensor data is the corresponding ranging dataset; when the sensor type is a traffic light, the sensor data is the corresponding traffic light status; the traffic light status includes red, green, and yellow lights. The first participant dataset includes multiple subsets of first participants; each subset of first participants corresponds one-to-one with a traffic participant who was or is currently in the current traffic scenario; each subset of first participants includes a second roadside identifier, a first participant identifier, a first participant type, and a first participant tracking trajectory; the first participant type includes pedestrians, motor vehicles, and non-motor vehicles; the first participant tracking trajectory includes multiple first tracking trajectory points; each first tracking trajectory point includes a first trajectory point timestamp, a first target image, a first target geometric dimension, a first target coordinates, a first target vehicle speed, and a first target heading angle; when the first participant type is a motor vehicle, the first participant identifier is the vehicle identifier of the current motor vehicle; When the first signal dataset is not empty, it includes one or more first traffic light subsets; if there are traffic lights in the current traffic scene, the first traffic light subset corresponds one-to-one with the traffic lights in the current traffic scene; the first traffic light subset includes the second roadside sign, the first traffic light sign, the first traffic light coordinates, and the first signal tracking trajectory; the first signal tracking trajectory includes multiple second tracking trajectory points; the second tracking trajectory points include the second trajectory point timestamp, the first signal status, and the remaining time of the first signal; the first signal status includes red light, green light, and yellow light; The first event dataset includes multiple first event data records; each first event data record includes a sixth data timestamp, a second roadside identifier, a first event type, and first event data; the first event type includes at least speeding violations, red light violations, lane departure violations, construction obstruction violations, and traffic accident events; when the first event type is a speeding violation, red light violation, or lane departure violation, the first event data consists of the participant identifier corresponding to the violating vehicle, the violation time period tracking trajectory, and the violation time period video; when the first event type is a construction obstruction violation, the first event data consists of the coordinates of the center point of the construction site, the construction... The first event data consists of the set of vertex coordinates of the accident site, the area of ​​the construction site, and the set of lane markers occupied by the construction site. When the first event type is a traffic accident, the first event data consists of the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident site, the area of ​​the accident site, the set of traffic participants within the accident site, and the set of lane markers occupied by the accident site. The set of traffic participants within the accident site consists of one or more traffic participant data, each of which consists of a corresponding traffic participant identifier and a traffic participant type, including pedestrians, motor vehicles, and non-motor vehicles. The set of lane markers occupied by the accident site consists of one or more lane markers. The second pathside message set includes multiple second pathside message records; the second pathside message record includes a third message timestamp, a second pathside identifier, a third message type, and third message content data; the third message type includes SPAT message, RSM message, RSI message, MAP message, SSM message, and SAM message; The second vehicle-side message set includes multiple second vehicle-side message records; the second vehicle-side message record includes a fourth message timestamp, a second vehicle-side identifier, a fourth message type, and fourth message content data; the fourth message type includes BSM message, SSM message, ISM message, and SAM message.

4. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 3, characterized in that, The road network meteorological dataset includes multiple first regional subsets; each first regional subset corresponds one-to-one with a traffic scenario; each first regional subset includes multiple first meteorological data records. The first meteorological data record includes a seventh data timestamp, a first regional identifier, and a first meteorological dataset; the first meteorological dataset consists of multiple types of meteorological data. The road network traffic dataset includes multiple second region subsets; each second region subset corresponds one-to-one with a traffic scenario; each second region subset includes multiple first traffic data records; each first traffic data record includes an eighth data timestamp, a second region identifier, and a first traffic indicator set. The first traffic indicator set consists of multiple types of traffic indicator data; The vehicle synchronization dataset includes multiple first vehicle subsets; each first vehicle subset corresponds one-to-one with the vehicle-end device; each first vehicle subset includes a chassis synchronization dataset and a fault synchronization dataset; the chassis synchronization dataset includes multiple first chassis data records. The fault synchronization dataset includes multiple records of the first fault data; The roadside synchronization dataset includes multiple first roadside subsets; each first roadside subset corresponds one-to-one with a roadside device; each first roadside subset includes a participant synchronization dataset, a signal synchronization dataset, and an event synchronization dataset; the participant synchronization dataset includes multiple first participant subsets; the signal synchronization dataset includes multiple first traffic light subsets; the event synchronization dataset includes multiple first event data records; The road network accident dataset includes multiple first accident subsets; the first accident subsets include basic accident information, meteorological data of the accident area, traffic data of the accident area, vehicle-side dataset, participant trajectory set, signal trajectory set, supplementary set of accident vehicles, supplementary set of surrounding vehicles, and supplementary roadside dataset. The basic accident information includes the accident time, roadside signs, accident type, center point of the accident area, vertex set of the accident area, area of ​​the accident area, set of accident participants, and set of accident lanes; The vehicle-side dataset consists of vehicle-side data of all accident vehicles within the current accident range, and the vehicle-side data includes accident vehicle identification, chassis data sequence, and fault data sequence. The participant trajectory set consists of the participant trajectory data of all traffic participants within the current accident area; the participant trajectory data includes accident participant identifier, accident participant type, and accident participant trajectory; When the signal trajectory set is not empty, it consists of the signal trajectory data of all traffic lights adjacent to the current accident area; the signal trajectory data includes the accident roadside marker, adjacent signal light marker, adjacent signal light coordinates, and adjacent signal light trajectory. The accident vehicle supplementary set consists of a subset of all accident vehicles within the current accident scope; the accident vehicle subset includes EDR data sequences, DVR data sequences, roadside message sequences, and vehicle-side message sequences; When the surrounding vehicle supplementary set is not empty, it consists of a subset of surrounding vehicles of all surrounding vehicles outside the current accident area; the surrounding vehicle subset includes the EDR data sequence, the DVR data sequence, the roadside message sequence, and the vehicle-side message sequence; The roadside supplementary dataset consists of a roadside visual dataset, a roadside broadcast message set, and a roadside vehicle message set for the road segment or intersection where the current accident occurs.

5. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The vehicle-side data acquisition and transmission module is specifically used when the corresponding vehicle-cloud synchronized dataset is generated periodically based on the stored chassis operating data and fault data and sent to the cloud platform: According to the preset vehicle-side synchronization frequency, the current time is periodically used as the first end time, and the time point obtained by subtracting the preset first duration from the first end time is used as the first start time. The first start time and the first end time form the corresponding current most recent time period. The first chassis data records in the first chassis dataset that are in the current most recent time period are extracted to form the corresponding first synchronization dataset. The first fault data records in the most recent time period are extracted from the first fault dataset to form a corresponding second synchronization dataset; and the vehicle-cloud synchronization dataset is formed by the obtained first and second synchronization datasets and sent to the cloud platform. The vehicle-side data acquisition and transmission module is specifically used to extract the EDR, DVR, and vehicle-road message data that match the current request when receiving a vehicle-side accident extraction request from the cloud platform, form a corresponding vehicle-side accident dataset, and send it back to the cloud platform: Extract the corresponding accident collection time period from the current vehicle-side accident extraction request and use it as the current time period; The first EDR data records in the first EDR dataset that are in the current time period are extracted to form the corresponding EDR data sequence; The first DVR data records in the first DVR dataset that are in the current time period are extracted to form the corresponding DVR data sequence; and the first roadside message records in the first roadside message set that are in the current time period are extracted to form the corresponding roadside message sequence. The first vehicle-side message records in the first vehicle-side message set that are in the current time period are extracted to form the corresponding vehicle-side message sequence; and the corresponding vehicle-side accident dataset is formed by the EDR data sequence, the DVR data sequence, the roadside message sequence, and the vehicle-side message sequence obtained this time and sent back to the cloud platform.

6. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The roadside data acquisition and transmission module is specifically used when the corresponding road-cloud synchronized dataset is generated periodically based on the stored traffic participant, signal, and traffic event perception data and sent to the cloud platform: According to the preset roadside synchronization frequency, the current time is periodically used as the second end time, and the time point obtained by subtracting the preset second duration from the second end time is used as the second start time. The corresponding current most recent time period is composed of the second start time and the second end time. Then, extract the subset of first participants whose tracking trajectories have been updated in the current most recent time period from the first participant dataset to form the corresponding third synchronization dataset; And extract the first signal light subset from the first signal dataset whose first signal tracking trajectory has been updated in the current most recent time period to form the corresponding fourth synchronization dataset; The first event data records that are in the most recent time period in the first event dataset are extracted to form the corresponding fifth synchronization dataset; and the obtained third, fourth and fifth synchronization datasets are combined to form the corresponding cloud synchronization dataset and sent to the cloud platform. The roadside data acquisition and transmission module is specifically used to extract visual sensor data and vehicle-road message data that match the current request when receiving a roadside accident extraction request from the cloud platform, form a corresponding roadside accident dataset, and send it back to the cloud platform: Extract the corresponding accident collection time period from the current vehicle-side accident extraction request as the current time period; and extract the sensor data records in the multimodal dataset whose sensor type is camera and whose fifth data timestamp is within the current time period to form the corresponding roadside visual dataset; The second roadside message records in the second roadside message set that are in the current time period are extracted to form the corresponding roadside broadcast message set; and the second vehicle-end message records in the second vehicle-end message set that are in the current time period are extracted to form the corresponding roadside vehicle message set. The roadside accident dataset, composed of the roadside visual dataset, the roadside broadcast message set, and the roadside vehicle message set obtained in this study, is then sent back to the cloud platform.

7. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The cloud-based data acquisition and transmission module is specifically used when performing cloud-based vehicle data synchronization based on the vehicle-cloud synchronization dataset periodically uploaded by each of the vehicle-end data acquisition and transmission modules: The currently received vehicle-cloud synchronized dataset is taken as the current dataset; and the first and second synchronized datasets of the current dataset are taken as the corresponding current chassis dataset and current fault dataset; The first vehicle subset in the vehicle synchronization dataset that corresponds to the current dataset is taken as the current vehicle subset; And add all the first chassis data records of the current chassis dataset to the chassis synchronization dataset of the current vehicle subset; And add all the first fault data records of the current fault dataset to the fault synchronization dataset of the current vehicle subset; The cloud-based data acquisition and transmission module is specifically used when performing cloud-based roadside data synchronization based on the roadside cloud synchronization dataset periodically uploaded by each of the roadside data acquisition and transmission modules: The currently received road-cloud synchronization dataset is taken as the current dataset; and the third, fourth, and fifth synchronization datasets of the current dataset are taken as the corresponding current participant dataset, current signal dataset, and current event dataset; The first roadside subset in the roadside synchronization dataset that corresponds to the current dataset is taken as the current roadside subset; The participant synchronization dataset, the signal synchronization dataset, and the event synchronization dataset of the current roadside subset are used as the corresponding current synchronization dataset A, current synchronization dataset B, and current synchronization dataset C. And perform one round of traversal on all the first participant subsets of the current participant dataset; During this round of traversal, the first participant subset of the current traversal is taken as the corresponding current traversal subset; The subset of first participants whose identifiers match the first participant identifiers in the current synchronized dataset A is taken as the corresponding current matching subset; And it identifies whether the current matching subset is empty; If yes, add the currently traversed subset to the current synchronized dataset A; if no, replace the current matching subset based on the currently traversed subset. And perform one round of traversal on all the first traffic light subsets of the current signal dataset; During this round of traversal, the first subset of traffic lights currently being traversed is taken as the corresponding current traversal subset; The first traffic light subset in the current synchronized dataset B that matches the first traffic light identifier of the current traversed subset is taken as the corresponding current matching subset; And it identifies whether the current matching subset is empty; If yes, then add the currently traversed subset to the current synchronized dataset B; if no, then replace the current matching subset based on the currently traversed subset. Then, all the first event data records in the current event dataset are added to the current synchronization dataset C.

8. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The accident handling module is specifically used when monitoring traffic accidents on the roadside synchronous dataset and generating corresponding monitoring accident information based on the real-time monitoring results: The record addition operations of each event synchronization dataset in the roadside synchronization dataset are monitored in real time; The latest added first event data record in any of the event synchronization datasets is taken as the current new record; and when the first event type of the current new record is a traffic accident event, a corresponding monitoring accident information is composed of the sixth data timestamp of the current new record, the second roadside identifier, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane identifiers occupied by the accident site.

9. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The accident handling module is specifically used when, based on the various monitored accident information, meteorological and traffic data are collected from the road network meteorological dataset and the road network traffic dataset to obtain the corresponding cloud dataset: The sixth data timestamp and the second roadside identifier of the currently monitored accident information are used as the corresponding current time and current roadside identifier; the first region subset in the road network meteorological dataset corresponding to the current roadside identifier is used as the current region meteorological subset, and the second region subset in the road network traffic dataset corresponding to the current roadside identifier is used as the current region traffic subset; the first meteorological dataset of the first meteorological data record whose seventh data timestamp is closest to the current time in the current region meteorological subset is extracted to form the corresponding accident area meteorological data; the first traffic indicator set of the first traffic data record whose eighth data timestamp is closest to the current time in the current region traffic subset is extracted to form the corresponding accident area traffic data; and the accident area meteorological data and the accident area traffic data obtained this time are used to form the corresponding cloud dataset.

10. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The accident handling module is specifically used when, based on the various monitored accident information, vehicle-road data is collected from the vehicle synchronous dataset and the roadside synchronous dataset to obtain the corresponding vehicle-road dataset: The sixth data timestamp, the second roadside identifier, the coordinates of the center point of the accident site, and the set of traffic participants within the accident site of the current monitored accident information are extracted as the corresponding current time, current roadside identifier, current center point coordinates, and current set of participants; The current start time is obtained by subtracting the preset third duration from the current time, and the current end time is obtained by adding the third duration to the current time. The current start time and the current end time together form the corresponding current accident period. The system iterates through all traffic participant data of type motor vehicle in the current participant set; during this iteration, the traffic participant identifier of the currently iterated traffic participant data is used as the current vehicle identifier; and the first vehicle subset in the vehicle synchronization dataset that matches the first vehicle identifier with the current vehicle identifier is used as the current vehicle subset. Then, extract all the first chassis data records in the chassis synchronization dataset of the current vehicle subset that are in the current accident period to form the corresponding chassis data sequence, extract all the first fault data records in the fault synchronization dataset of the current vehicle subset that are in the current accident period to form the corresponding fault data sequence, and use the current vehicle identifier as a corresponding accident vehicle identifier; and use the accident vehicle identifier, the chassis data sequence, and the fault data sequence obtained in this round to form a corresponding vehicle-end data; and at the end of this round of traversal, use all the obtained vehicle-end data to form the corresponding vehicle-end dataset; And the participant synchronization dataset and the signal synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset are taken as the corresponding current participant dataset and current signal dataset; And perform a round of traversal on all the traffic participant data of the current participant set; and during this round of traversal, use the traffic participant identifier of the traffic participant data currently being traversed as the current participant identifier; The subset of first participants whose first participant identifier matches the current participant identifier in the current participant dataset is taken as the current subset; and the trajectory segment of the first participant tracking trajectory in the current subset that is in the current accident time period is extracted as the corresponding accident participant trajectory; The traffic participant identifier and the traffic participant type in the current participant data are used as a corresponding set of accident participant identifier and accident participant type; The participant trajectory data is composed of the participant identifier, participant type, and participant trajectory obtained in this iteration; and at the end of this round of traversal, the participant trajectory data obtained are combined to form the corresponding participant trajectory set. The first traffic light subset in the current signal dataset whose Euclidean distance between the first traffic light coordinates and the current center point coordinates does not exceed a preset first distance threshold is taken as the corresponding neighboring traffic light subset; the total number of the neighboring traffic light subsets obtained this time is identified; if the total number of subsets obtained this time is 0, the corresponding signal trajectory set is set to empty; if the total number of subsets obtained this time is greater than 0, the trajectory segments on the first signal tracking trajectory of each neighboring traffic light subset that are in the current accident period are extracted as the corresponding neighboring traffic light trajectory, and the second roadside identifier, the first traffic light identifier, and the first traffic light coordinates of each neighboring traffic light subset are used as a set of corresponding accident roadside identifier, neighboring traffic light identifier, and neighboring traffic light coordinates, and the set of accident roadside identifier, neighboring traffic light identifier, neighboring traffic light coordinates, and neighboring traffic light trajectory corresponding to each neighboring traffic light subset is used to form a corresponding traffic light trajectory data, and all the obtained traffic light trajectory data are used to form the corresponding signal trajectory set; The obtained vehicle-side dataset, participant trajectory set, and signal trajectory set constitute the corresponding vehicle-road dataset.

11. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The accident handling module is specifically used when, based on each of the vehicle-road datasets, it sends corresponding vehicle-end and roadside accident extraction requests to the vehicle-end and roadside devices related to the current accident and generates corresponding vehicle-road supplementary datasets based on the request-returned data: Each of the aforementioned vehicle-road datasets is used as the current vehicle-road dataset; And take the vehicle-side dataset of the current vehicle-road dataset as the corresponding current vehicle-side dataset; and take the sixth data timestamp and the second roadside identifier of the monitoring accident information corresponding to the current vehicle-road dataset as the corresponding current time and current roadside identifier; And the participant synchronization dataset of the first roadside subset corresponding to the current roadside identifier in the roadside synchronization dataset is taken as the corresponding current participant synchronization dataset; The current start time is obtained by subtracting the preset third duration from the current time, and the current end time is obtained by adding the third duration to the current time. The current start time and the current end time together form the corresponding accident collection period. And the vehicle-end device corresponding to each of the vehicle-end data in the current vehicle-end dataset is regarded as the accident vehicle device; And send the vehicle-side accident extraction request carrying the accident collection period to each of the accident vehicle devices; The vehicle-side accident dataset sent back by each of the accident vehicle devices is taken as the corresponding accident vehicle subset; and all the obtained accident vehicle subsets are combined to form the corresponding accident vehicle supplementary set. The first participant subset in the current participant synchronization dataset is defined as the first participant subset whose type is motor vehicle, whose trajectory time period of the first participant's tracking trajectory intersects with the accident collection time period, and whose first participant identifier does not intersect with all vehicle identifiers in the current vehicle dataset. The total number of the surrounding participant subsets obtained this time is identified; if the total number of the subsets obtained this time is 0, the corresponding surrounding vehicle supplementary set is set to empty; if the total number of the subsets obtained this time is greater than 0, the vehicle-end device corresponding to each of the surrounding participant subsets is taken as the surrounding vehicle device, and the vehicle-end accident extraction request carrying the accident collection period is sent to each of the surrounding vehicle devices, and the vehicle-end accident dataset returned by each of the surrounding vehicle devices is taken as the corresponding surrounding vehicle subset, and the corresponding surrounding vehicle supplementary set is composed of all the obtained surrounding vehicle subsets. The roadside device corresponding to the current roadside identifier is designated as the current device; the roadside accident extraction request carrying the accident collection period is sent to the current device; and the roadside accident dataset returned by the current device is designated as the corresponding roadside supplementary dataset. The corresponding vehicle-road supplementary dataset is composed of the accident vehicle supplementary set, the surrounding vehicle supplementary set, and the roadside supplementary dataset obtained in this study.

12. The traffic accident information processing system based on vehicle-road-cloud multi-dimensional fusion data according to claim 4, characterized in that, The accident handling module is specifically used to integrate multi-dimensional accident data from the vehicle-road dataset, the vehicle-road supplementary dataset, and the cloud dataset corresponding to each of the monitored accident information, and to refresh the road network accident dataset based on the integration result: Each of the aforementioned monitored incident information is used as the current incident information; The sixth data timestamp of the current accident information, the second roadside identifier, the traffic accident type, the coordinates of the center point of the accident site, the set of vertex coordinates of the accident range, the area of ​​the accident range, the set of traffic participants in the accident site, and the set of lane identifiers occupied by the accident site are used as a set of corresponding accident occurrence time, accident roadside identifier, accident type, center point of the accident area, set of vertex coordinates of the accident area, area of ​​the accident area, set of accident participants, and set of accident lanes to form a corresponding basic accident information; And extract the corresponding vehicle-side dataset, participant trajectory set, signal trajectory set, accident vehicle supplementary set, surrounding vehicle supplementary set, roadside supplementary dataset, accident area meteorological data and accident area traffic data from the vehicle-road dataset, vehicle-road supplementary dataset and cloud dataset corresponding to the current accident information; The first accident subset, composed of the basic accident information corresponding to the current accident information, the meteorological data of the accident area, the traffic data of the accident area, the vehicle-side dataset, the participant trajectory set, the signal trajectory set, the accident vehicle supplementary set, the surrounding vehicle supplementary set, and the roadside supplementary dataset, is added to the road network accident dataset.

Images