Road traffic emergency broadcasting method, device and system

Abstract

The invention discloses a road traffic emergency broadcasting method, device and system, and relates to the technical field of intelligent traffic, and the method comprises the steps: collecting road traffic environment data in real time through a multi-source sensing device; carrying out fusion processing on the collected multi-source road traffic environment data, identifying emergencies, and determining the types, positions, influence ranges and emergency degrees of the emergencies; dynamically allocating 5G network slice resources according to the type, the influence range and the emergency degree of the event, and generating an optimal broadcast strategy; and based on the optimal broadcast strategy, through a vehicle wireless communication V2X interface and / or a 5G cellular network interface, cooperatively broadcasting emergency early warning information in a standardized format to vehicle terminals within the influence range. According to the invention, millisecond-level response from event sensing to information broadcasting is realized, the problems of response delay, non-uniform coverage, poor protocol compatibility and the like in the prior art are effectively solved, and the road traffic safety level in an intelligent network connection environment is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical fields of intelligent transportation systems, vehicle-to-everything (V2X), 5G communication, edge computing and artificial intelligence, and more particularly to a road traffic emergency broadcasting method, device and system. BACKGROUND

[0002] At present, in the intelligent networked transportation environment, the intelligent vehicle still has the following main problems in the perception and response to emergencies:

[0003] 1. Emergency response delay, with potential safety hazards;

[0004] The existing traffic information system often relies on manual reporting, central processing and redistribution to announce traffic events. There is a delay of tens of seconds to minutes from the occurrence of the event to the information broadcast, which may lead to secondary collisions, chain rear-end collisions and other serious consequences under high-speed driving conditions.

[0005] 2. Uncertain coverage range of information broadcast, and late acquisition by following vehicles;

[0006] In the car-to-car communication or local broadcasting mode, there is a lack of effective broadcast range control mechanism, resulting in that some vehicles fail to receive critical information in time, and there is a risk of blind area, especially in complex urban traffic environment or elevated, tunnel and other scenarios.

[0007] 3. Inefficient processing of non-uniform broadcast information structure;

[0008] The existing broadcast information has not been standardized and encapsulated, and there are protocol compatibility problems between different manufacturer devices, which makes it difficult to achieve fast analysis and response, and is not conducive to the formation of a unified and coordinated intelligent transportation ecosystem.

[0009] 4. Poor broadcast reliability, with information redundancy or loss;

[0010] When multiple vehicles broadcast or multiple sources detect the same event, information redundancy, conflict or packet loss may occur, which not only wastes bandwidth resources, but also affects the accuracy of judgment. There is currently a lack of a reliable multi-source coordination mechanism to integrate information and unify broadcast content.

[0011] 5. Lack of low-delay and large-bandwidth communication support mechanism;

[0012] Traditional wireless communication methods such as DSRC have been difficult to meet the communication needs of large-scale, low-latency and high-reliability in complex road scenarios, and have failed to effectively combine with 5G and edge computing technologies, resulting in low overall response speed and scheduling efficiency.

[0013] With the development of intelligent networked vehicle technology, vehicle-road cooperation has become an important means to ensure driving safety. However, the existing road emergency information notification mechanism generally has problems such as delayed response, insufficient information coverage, and non-standardized notification content. In particular, in high-density, highway or urban trunk road scenarios, information lag can lead to a chain of accidents. The development of V2X and 5G technologies provides a basis for implementing low-latency, high-reliability event information broadcasting, so it is necessary to propose an efficient broadcasting method based on the above technologies. SUMMARY

[0014] Therefore, the present application provides a road traffic emergency broadcasting method, device and system, which at least partially solves the problems of long response delay, insufficient coverage, and non-uniform protocols in the prior art.

[0015] To achieve the above purpose, the present application adopts the following technical solutions:

[0016] In a first aspect, the present application provides a road traffic emergency broadcasting method, comprising the following steps:

[0017] The perception step: real-time collection of road traffic environment data by deploying multi-source sensing devices at the roadside and / or vehicle end;

[0018] The processing step: fusion processing of the collected multi-source road traffic environment data based on an edge computing platform, identifying the emergency event, and determining the type, location, impact range and emergency degree of the emergency event;

[0019] The decision step: dynamically allocating 5G network slice resources according to the type, impact range and emergency degree of the emergency event, and generating an optimal broadcasting strategy; the optimal broadcasting strategy includes the broadcasting range, broadcasting priority and communication resource allocation;

[0020] The broadcasting step: based on the optimal broadcasting strategy, through the vehicle wireless communication V2X interface and / or 5G cellular network interface, the standardized format of the emergency warning information of the emergency event is broadcasted to the vehicle terminals within the impact range.

[0021] Further, the perception step comprises:

[0022] Road environment data is collected by roadside sensors, vehicle terminals, weather sensors and / or electronic police devices, and uploaded to edge computing nodes through 5G uplink, PC5 direct communication, optical fiber or wireless link.

[0023] Further, the processing step specifically comprises:

[0024] The edge computing platform performs time synchronization, spatial calibration and noise filtering processing on road traffic environment data from different sources;

[0025] Based on the processed data, dynamic and static targets on the road are identified by a target detection and tracking algorithm;

[0026] According to a preset rule and / or a machine learning model, the type of the emergency event is determined, and the determination result is cross-verified by multi-source data;

[0027] In combination with the type of the emergency event, the road topology, and the environmental parameters, the influence range of the event and the set of affected vehicles are calculated.

[0028] Further, in the decision step, the emergency event is divided into three levels according to the degree of urgency; and different quality of service 5G network slice resources are dynamically allocated to emergency events of different levels; wherein

[0029] The first-level event is allocated an ultra-reliable low-latency communication slice;

[0030] The second-level event is allocated an enhanced mobile broadband slice;

[0031] The third-level event is allocated a massive machine type communication slice;

[0032] The delay of the first-level event is less than that of the second-level event, and the delay of the second-level event is less than that of the third-level event.

[0033] Further, in the broadcasting step, the following is further included:

[0034] According to the density of vehicle terminals within the influence range, a broadcast mode is adaptively selected:

[0035] When the density of vehicle terminals is higher than a first threshold value, direct communication broadcasting is preferentially performed using a V2X PC5 interface;

[0036] When the density of vehicle terminals is lower than a second threshold value, cellular network broadcasting is performed using a 5G Uu interface.

[0037] Further, the broadcast mode further includes a hybrid mode in which the V2X PC5 interface broadcasting and the 5G Uu interface broadcasting are simultaneously enabled.

[0038] Further, the standardized format of the emergency event warning information at least includes: event type code, precise position information, timestamp, geofencing range, and recommended response action, and all are subjected to encryption processing.

[0039] In a second aspect, the embodiments of the present application also provide a road traffic emergency event broadcasting device applied to an edge computing node, and the device comprises:

[0040] A collection and perception module collects road traffic environment data in real time by using multi-source sensing devices deployed at road sides and / or vehicle ends;

[0041] An event processing module is configured to fuse the collected multi-source road traffic environment data, identify a sudden event, and determine a type, a location, an influence range and an emergency level of the sudden event;

[0042] A strategy generating module is configured to dynamically allocate 5G network slice resources and generate an optimal broadcast strategy according to the type, the influence range and the emergency level of the sudden event; the optimal broadcast strategy includes a broadcast range, a broadcast priority and communication resource allocation;

[0043] An information broadcast module is configured to cooperatively broadcast sudden event early warning information in a standardized format to vehicle terminals within the influence range based on the optimal broadcast strategy through a vehicle wireless communication (V2X) interface and / or a 5G cellular network interface.

[0044] In a third aspect, an embodiment of the present application provides a road traffic sudden event broadcast system, which comprises the broadcast device of the second aspect.

[0045] A roadside sensing unit comprising one or more of a camera and a radar is configured to collect roadside traffic environment data;

[0046] A vehicle terminal is arranged on a vehicle and is configured to collect vehicle-end data, receive early warning information and perform human-computer interaction;

[0047] An auxiliary monitoring terminal is configured to collect road traffic related information based on a meteorological sensor, an electronic police and a UAV inspection device;

[0048] A communication network comprising a 5G base station and a roadside unit (RSU) is configured to provide a data transmission channel for the broadcast device, the roadside sensing unit, the vehicle terminal and the auxiliary monitoring terminal.

[0049] Further, the system supports linkage with a traditional traffic management system and is compatible with vehicle terminals and mobile devices of different manufacturers.

[0050] The description of the second aspect to the third aspect of the present application can refer to the detailed description of the first aspect; and the beneficial effects of the description of the second aspect to the third aspect can refer to the beneficial effect analysis of the first aspect, which will not be repeated here.

[0051] According to the above technical solution, compared with the prior art, the present application has the following technical advantages:

[0052] 1. Millisecond-level event detection and broadcast response are achieved, which significantly improves the level of road traffic safety;

[0053] 2. Multi-source data fusion is used to improve the event identification accuracy and the influence range evaluation accuracy;

[0054] 3. Support dynamic broadcast strategy and multi-mode cooperative broadcast, expand coverage and eliminate blind area;

[0055] 4. Good compatibility and scalability, support multi-terminal and multi-protocol access;

[0056] 5. Reduce dependence on central system, improve system robustness and response efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the provided drawings.

[0058] Figure 1 The road traffic emergency broadcast method flow chart provided by the present application.

[0059] Figure 2 The overall principle diagram of the road traffic emergency broadcast provided by the present application.

[0060] Figure 3 The structure block diagram of the road traffic emergency broadcast device provided by the present application.

[0061] Figure 4 The structure block diagram of the road traffic emergency broadcast system provided by the present application. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0063] Embodiment 1:

[0064] The embodiment of the present application discloses a road traffic emergency broadcast method, as shown in Figure 1 The steps S1~S4 are included:

[0065] S1, sensing step: through the multi-source sensing device deployed at the roadside and / or the vehicle end, real-time collection of road traffic environment data. For example, using roadside sensors (cameras, millimeter wave radars, laser radars), vehicle terminals (OBU), weather stations, electronic police and other devices to collect real-time road environment data.

[0066] S2, processing step: based on the edge computing platform, the collected multi-source road traffic environment data are fused and processed, a sudden event is identified, and the type, location, influence range and emergency degree of the sudden event are determined;

[0067] Through real-time fusion analysis of multi-source data by a 5G edge computing node (MEC), the type of the sudden event (such as collision, anchor throwing, road collapse), the location of the event, the influence range (such as the number of affected lanes and the duration) and the emergency degree are identified.

[0068] S3, decision step: according to the type, influence range and emergency degree of the sudden event, 5G network slice resources are dynamically allocated, and an optimal broadcast strategy is generated; the optimal broadcast strategy includes a broadcast range, a broadcast priority and a communication resource allocation.

[0069] Based on the emergency degree and the influence range of the event, 5G network slice resources such as a dedicated low-latency slice are dynamically allocated, and an optimal broadcast strategy is generated, and PC5 direct communication, Uu cellular broadcast or a hybrid mode is selected.

[0070] S4, broadcast step: based on the optimal broadcast strategy, through a vehicle wireless communication V2X interface and / or a 5G cellular network interface, standardized format sudden event warning information is cooperatively broadcast to vehicle terminals within the influence range. For example, through V2X (PC5 interface) and 5G cellular (Uu interface) cooperative broadcast, it is ensured that the sudden event information covers all intelligent connected vehicles within a certain range, whether or not they are equipped with a specific manufacturer OBU, and supports linkage with traditional traffic management systems.

[0071] The method is characterized by constructing a closed-loop process of "perception-fusion-decision-broadcast". The method collects data through multi-source sensors deployed at the roadside and the vehicle end; uses edge computing nodes (MEC) to fuse, synchronize and calibrate multi-source data, and uses AI models and rule engines to quickly identify and verify events; according to the emergency degree of the event, 5G network slice resources such as URLLC and eMBB slice are dynamically allocated, and a broadcast strategy is generated; finally, through the adaptive broadcast mode of V2X (PC5) and 5G cellular (Uu) cooperation, standardized and encrypted warning information is sent to vehicles in the affected area. The invention can realize millisecond-level sudden event detection, information coding standardization, multi-source cooperative broadcast, regional broadcast optimization, chain vehicle-to-vehicle linkage response and other functions, thereby improving the overall safety and traffic efficiency of the road traffic system.

[0072] For example, the method of the present application specifically includes the following steps:

[0073] Step 1: Real-time collection and uploading of multi-source data;

[0074] Roadside sensors such as cameras, radars upload raw data such as video, point cloud to edge computing nodes (MEC) through 5G uplink (Uu interface) in real time, upload delay ≤10ms;

[0075] On-board terminals (OBU) upload their status such as position, speed, acceleration and surrounding perception data such as distance to the front obstacle to MEC through PC5 direct communication or Uu interface;

[0076] Weather sensors, electronic police and other devices upload environmental data such as visibility, road slipperiness to MEC through wired (such as optical fiber) or wireless (NB-IoT) links.

[0077] Step 2: Multi-source data fusion and emergency identification;

[0078] Data preprocessing: MEC performs time synchronization on raw data based on 5G network time synchronization protocol, accuracy ≤10μs; For example, IEEE 1588v2 (PTP) precision time protocol is adopted to achieve nanosecond-level synchronization accuracy, a joint synchronization algorithm based on 5G air interface timing signal (5G-S-TMSI) is introduced to solve the synchronization problem of heterogeneous networks, a time sequence consistency verification mechanism is established, and the correctness of data time sequence is ensured through timestamp reverse lookup.

[0079] Spatial calibration, positioning accuracy ≤10cm and noise filtering; For example, Kalman filtering. Specifically, a multi-modal sensor joint calibration algorithm can be used to establish a unified coordinate system conversion model, implement dynamic calibration technology based on feature point matching, and real-time correct sensor pose deviation. Through the introduction of RTK-GNSS+IMU+laser SLAM multi-source fusion positioning, centimeter-level positioning accuracy is achieved.

[0080] Noise filtering processing, adaptive Kalman filtering combination algorithm can be used to dynamically adjust filtering parameters according to signal-to-noise ratio, or a noise recognition model based on deep learning can be designed to effectively distinguish between real signals and noise, and multi-scale wavelet transform denoising can also be implemented to retain signal detail features while suppressing noise.

[0081] Target detection and tracking: deep learning models are used to fuse and analyze camera video and radar point cloud to identify vehicles, pedestrians, obstacles on the road and track their motion trajectories; attention mechanism guided feature fusion network can be developed to weight and fuse visual and radar features; point cloud-image fusion architecture based on PointPainting can be designed to improve target detection accuracy; a multi-sensor complementarity evaluation model can be established to dynamically select the optimal sensor combination; multi-hypothesis tracking (MHT) algorithm can be implemented to handle target tracking in dense scenarios.

[0082] Event identification and verification: Determine the risk of collision through the rule engine, such as "the relative speed of two vehicles > 80 km / h and the distance < 50 m", identify the event type through machine learning models, such as random forest classifier, such as collision, anchor throwing, road collapse, pedestrian intrusion, etc., and cross-verify through multi-source data, such as camera image + radar to confirm the authenticity of the obstacle;

[0083] Impact range assessment: Combine event type, road topology (such as number of lanes, speed limit), environmental parameters (such as visibility, road friction coefficient) to calculate event impact range and affected vehicle set, and filter OBU within the range through V2X positioning data.

[0084] The entire calculation process can be divided into the following three core stages, and the process can be summarized as:

[0085] Stage one: Calculate the basic physical impact range;

[0086] This is the starting point of the calculation, according to the type and location of the event, determine an initial impact area.

[0087] Determine the event center point:

[0088] Input: Precise event positioning from processing steps (latitude and longitude coordinates, accurate to lane level). Map the event coordinates to a high-precision map.

[0089] Define the initial impact radius:

[0090] Input: Event type and severity, match based on pre-set rule base. For example: I-level event, such as multi-vehicle collision: initial radius R_initial set to 500 meters. II-level event, such as single vehicle anchor throwing: initial radius R_initial set to 200 meters. III-level event, such as wet road: initial radius R_initial set to 100 meters.

[0091] Generate basic geometric range: Take the event center point as the center and R_initial as the radius to generate a circular area. At the same time, combined with the road topology of the high-precision map, such as road direction, ramp, isolation belt, clip the circular area to strictly limit it on the road where the event occurs and the adjacent associated roads, forming an irregular polygon geographic fence. This avoids sending warning information to the opposite lane or irrelevant elevated road.

[0092] Stage two: Multi-factor dynamic correction and impact range optimization;

[0093] The system will simulate the propagation of event impact and dynamically adjust the basic range of the first stage.

[0094] Road topology factor correction (ƒ_road):

[0095] Inputs: Number of lanes, Speed limit value, Road curvature, Slope.

[0096] Calculation Process:

[0097] Speed Limit: The higher the speed limit, the longer the required safety braking distance, and the influence range needs to be expanded. Correction factor ƒ_speed = (Actual_Speed_Limit / 60) 2 , assuming 60 km / h as the baseline speed.

[0098] Number of Lanes: The more lanes, the more complex the chain reaction that may be triggered by the incident, such as lane changing to avoid, and the influence range needs to be expanded. Correction factor ƒ_lane = 1 + (Number_Of_Lanes - 2) * 0.2; assuming 2 lanes as the baseline.

[0099] Comprehensive Road Factor Correction: R_road = R_initial * ƒ_speed * ƒ_lane

[0100] Environmental Parameter Factor Correction (ƒ_env):

[0101] Inputs: Visibility, Road Friction Coefficient (μ), Weather Conditions (Rain / Snow / Fog).

[0102] Calculation Process:

[0103] Visibility (Vis): The lower the visibility, the shorter the driver's reaction time after discovering the warning, and the influence range needs to be expanded. Correction factor ƒ_vis = 1000 / Visibility (Visibility unit is meters, set the maximum upper limit value).

[0104] Road Friction Coefficient (μ): The more slippery the road surface (the lower the μ value), the longer the braking distance, and the influence range needs to be expanded. Correction factor ƒ_μ = 0.7 / μ (assuming dry road surface μ=0.7 as the baseline).

[0105] Comprehensive Environmental Factor Correction: R_env = R_road * ƒ_vis * ƒ_μ;

[0106] Traffic Flow State Factor Correction (ƒ_traffic):

[0107] Inputs: Real-time average speed, Traffic density.

[0108] Calculation Process: Obtain real-time traffic flow data within the current influence range from the edge computing platform.

[0109] Average vehicle speed (v): the higher the vehicle speed, the larger the impact range. Correction factor ƒ_flow = (v / 60) 2 .

[0110] Final dynamic impact radius (R_final): R_final = R_env * ƒ_flow;

[0111] Output: a final irregular polygon geographic fence G_final that has been dynamically corrected by multiple factors.

[0112] Phase three: accurate identification and screening of the affected vehicle set; based on the final determined geographic fence G_final, accurately lock the vehicles that need to receive information.

[0113] Data source access:

[0114] Through the V2X interface (PC5) and the 5G network (Uu), the edge computing platform receives basic safety messages (BSM) from all vehicles equipped with OBU within its coverage range in real time. BSM contains the precise position, speed, heading angle, vehicle ID, and other key data of the vehicle.

[0115] Spatial relationship matching:

[0116] Calculation process: the system performs spatial geometric calculations (such as the ray method) on the coordinates of each vehicle reporting the position (x_vehicle, y_vehicle) and the final geographic fence G_final to determine whether it is within the polygon range.

[0117] Output: generate a preliminary "potential affected vehicle ID list".

[0118] Motion trend screening:

[0119] Calculation process: to further improve the broadcast efficiency and avoid sending information to irrelevant vehicles such as vehicles that have left or vehicles in the opposite lane, the system will combine the vehicle's motion vector, position + speed + heading angle, for secondary screening.

[0120] Rule example: only keep those vehicles whose heading angle points to the event point and are expected to enter the G_final range within the next 60 seconds.

[0121] Output: generate an accurate "final affected vehicle set V_set" and send it to the broadcast module.

[0122] From static to dynamic: the impact range is no longer a fixed value, but an intelligent calculation result based on real-time multi-source data, event attributes, roads, environment, and traffic flow dynamic evolution.

[0123] Physical and communication mapping: perfectly combine road traffic safety theory, such as braking distance model, traffic flow theory, with communication resource scheduling, such as 5G network slicing, V2X broadcast, to realize the real "sense-communication-computing" integration.

[0124] Through the two-level screening mechanism of "geometric fence + motion situation", it not only ensures that all affected vehicles can be covered, but also maximizes the reduction of communication resource waste and irrelevant vehicle information interference.

[0125] The above fine calculation model is the core technical guarantee to realize millisecond-level response and improve road traffic safety level. Its multi-factor fusion and dynamic correction architecture is significantly different from traditional static or semi-static judgment methods.

[0126] In this embodiment, by fusing visual, radar, sensor data, combining rule engine and machine learning, the event recognition accuracy is greatly improved. The specific structure of the above deep learning model and machine learning model is not limited, as long as it can identify relevant factors of the emergency.

[0127] Step 3: Dynamic priority scheduling and resource allocation;

[0128] Event classification: According to the emergency degree and potential harm of the event, the event is divided into three levels (Table 1):

[0129]

[0130] Network slice allocation: 5G core network (5GC) allocates dedicated network slices for different levels of events:

[0131] Class I event: allocate "Ultra-Reliable Low-Latency Communication (URLLC) slice", guarantee end-to-end delay ≤ 50ms, reliability ≥ 99.999%;

[0132] Class II event: allocate "Enhanced Mobile Broadband (eMBB) slice", guarantee delay ≤ 100ms, bandwidth ≥ 100Mbps;

[0133] Class III event: share "Massive Machine Type Communication (mMTC) slice", delay ≤ 200ms.

[0134] Edge computing resource scheduling: MEC dynamically allocates computing resources according to the event level, and prioritizes the fusion and decision of Class I events.

[0135] In this step, the synergistic optimization of 5G network slicing and edge computing dynamically allocates URLLC / eMBB / mMTC slices, and processes data locally through MEC, compressing the total delay from sensing to broadcasting to within 50-200ms, while the traditional solution is generally greater than 500ms.

[0136] Step 4: V2X and 5G cellular cooperative broadcasting;

[0137] Broadcast strategy generation: Based on event location, impact range, and surrounding terminal density, the optimal broadcast mode is selected through the location information reported by OBU:

[0138] Dense scenario, such as OBU density > 100 vehicles / km2: PC5 direct communication broadcast is adopted, taking advantage of the high reliability of V2X short-range communication, with a packet loss rate < 0.1%, covering all OBUs within 300 meters;

[0139] Sparse scenario, such as OBU density < 10 vehicles / km2: Uu cellular broadcast is adopted, sending broadcast messages to all OBU / mobile terminals in the designated area through 5G base stations, covering an area up to 1 km;

[0140] Mixed scenario: PC5 and Uu broadcast are enabled simultaneously to ensure full coverage.

[0141] Broadcast mode adaptation based on terminal density can be achieved, dynamically switching between PC5 direct connection and Uu cellular broadcast according to OBU density, covering an area up to 1 km or more, solving the problem of traditional V2X "near-range reliability, long-range failure".

[0142] Broadcast message format: The message contains event type (standardized coding), location (latitude and longitude + lane-level coordinates), timestamp, impact range (polygon geofence), and recommended action, such as "slow down to 40 km / h" and "change lane to the right lane", encrypted using the SM4 algorithm to prevent tampering.

[0143] Cross-terminal compatibility: Broadcast messages support multi-protocol encapsulation to ensure that OBUs / car systems from different manufacturers can be parsed; such as supporting C-V2X, DSRC, and mobile APP message parsing, breaking down device barriers, and achieving "global vehicle reception".

[0144] Step 5: Terminal response and feedback

[0145] After receiving the broadcast message, the vehicle terminal (OBU) displays the warning information through the HMI human-machine interface, such as sound and light alarm, AR navigation prompt, and generates obstacle avoidance strategies such as automatic speed reduction and lane change based on its own positioning and path planning algorithm;

[0146] Surrounding vehicles without OBU, such as traditional fuel vehicles, can receive broadcast messages through mobile APP based on 5G cellular network, receive warnings, and prompt the driver to pay attention;

[0147] The terminal will feedback the response result, such as "has slowed down to 40 km / h", to the MEC for optimizing subsequent broadcast strategies and adjusting impact range evaluation.

[0148] The complete principle flow is shown with reference to Figure 2

[0149] 1. Event occurrence and perception: millimeter wave radars and cameras deployed on the roadside and other sensors capture real-time data on a "multi-vehicle chain collision" incident.

[0150] 2. Data upload and processing: perception data is uploaded to an edge computing node (MEC) via a 5G network (Uu interface) with ultra-low latency. The MEC platform fuses and processes multi-source data, uses AI models to quickly identify and verify the incident, assess the impact range, and generate the optimal broadcast strategy.

[0151] 3. Dynamic resource scheduling: MEC and 5G core network work together to dynamically allocate URLLC (Ultra-Reliable and Low-Latency Communication) network slices for this major incident, ensuring communication priority and resources.

[0152] 4. Cooperative broadcasting: broadcast instructions are sent to 5G base stations (gNB) and roadside units (RSU) simultaneously, starting the cooperative broadcasting mode of Uu interface cellular broadcasting and PC5 interface direct communication, forming a dead-angle-free coverage.

[0153] 5. Terminal reception and response: all vehicles within the coverage area can receive the warning.

[0154] Among them, connected cars: receive information through on-board terminals (OBU), display warnings on on-board AR navigation, or automatically perform actions such as deceleration and lane changes. Traditional cars: can receive voice or pop-up warnings through the owner's mobile phone APP, prompting the driver to handle manually.

[0155] This scenario perfectly embodies the core closed loop of "perception-fusion-decision-broadcast" and the overall architecture of vehicle-road-edge-cloud collaboration, highlighting its technical advantages of ultra-low latency, high reliability, wide coverage, and strong compatibility.

[0156] Embodiment 2:

[0157] Based on the same inventive concept, the present application also provides a road traffic emergency broadcasting device applied to an edge computing node, as shown with reference to Figure 3 The device comprises:

[0158] A collection and perception module that collects real-time road traffic environment data through multi-source sensing devices deployed on the roadside and / or vehicle end;

[0159] An event processing module for fusing and processing the collected multi-source road traffic environment data, identifying the incident, and determining the type, location, impact range, and urgency of the incident;

[0160] ​A strategy generation module is configured to dynamically allocate 5G network slice resources and generate an optimal broadcast strategy according to the type, influence range and emergency level of the emergency event; the optimal broadcast strategy includes a broadcast range, a broadcast priority and a communication resource allocation;

[0161] An information broadcast module is configured to cooperatively broadcast emergency event early warning information in a standardized format to vehicle terminals within the influence range based on the optimal broadcast strategy through a vehicle wireless communication (V2X) interface and / or a 5G cellular network interface.

[0162] The broadcast device is deployed at an edge computing node (MEC), which is mainly responsible for processing local tasks with high real-time and low delay requirements to achieve millisecond-level response. The cloud server plays a role of global management, macro coordination, data persistence and deep learning. The two constitute a complementary and edge-cloud integrated intelligent whole.

[0163] Edge device upload: the edge device periodically uploads the processing results of major events, such as event type, location, influence range, processing log and other metadata, to the cloud server. This is not the original massive sensor data, but structured abstract information after preliminary processing by the edge to save bandwidth.

[0164] Cloud-end global information: the cloud server integrates data from multiple edge nodes in the region to form a global traffic situation map, such as regional traffic flow state, large event information and global weather conditions. The edge device can pull or receive these over-the-horizon information pushed by the cloud server as needed to assist local decision-making. For example, the local can refer to the regional traffic flow prediction data provided by the cloud when evaluating the influence range of the event.

[0165] In addition, the cloud server can use massive historical data to perform offline training and version management of machine learning models such as event recognition models and influence range prediction models. After training, the optimized model parameters and algorithm rules are incrementally issued to each edge device.

[0166] The edge device uses these models for real-time inference locally, and feeds back the performance data of the inference results and new sample data to the cloud. The cloud uses this feedback data to continuously retrain and optimize the model, realizing the self-evolution of the algorithm. This forms a "cloud training-edge inference-feedback optimization" closed loop.

[0167] Embodiment 3:

[0168] Referring to Figure 4 The application further provides a road traffic emergency event broadcast system, comprising the device of embodiment 2;

[0169] Roadside sensing unit, including one or more of cameras, radars, for collecting roadside traffic environment data;

[0170] Vehicle terminal, provided on the vehicle, for collecting vehicle-end data, receiving early warning information and performing human-computer interaction;

[0171] Auxiliary monitoring terminal, based on meteorological sensors, electronic police, unmanned aerial vehicle inspection equipment, for collecting road traffic related information;

[0172] Communication network, including 5G base station, roadside unit RSU, for providing data transmission channel for broadcast device, roadside sensing unit, vehicle terminal and auxiliary monitoring terminal.

[0173] The system supports linkage with traditional traffic management systems and is compatible with vehicle terminals and mobile devices from different manufacturers.

[0174] The system is based on V2X communication, 5G network and edge computing platform, and builds a "vehicle-road-edge-cloud" collaborative sensing and broadcasting system, which has the following characteristics:

[0175] 1. Second-level sensing and millisecond-level broadcasting response

[0176] 1) Use multi-source sensors such as millimeter wave radars and roadside video analysis for real-time detection;

[0177] 2) Load lightweight AI model for local determination of event type and level;

[0178] 3) Broadcast information coding complies with V2X standard protocol format;

[0179] 2. Regional adaptive broadcasting mechanism

[0180] 1) Dynamically calculate the broadcast influence range according to the event location, road type, vehicle speed, traffic density and other factors, such as sector area, specified radius;

[0181] 2) Support setting broadcast duration, frequency and priority strategy to ensure that critical vehicles receive priority;

[0182] 3. Multi-source information fusion and deduplication mechanism

[0183] 1) Support vehicle-to-vehicle (V2V), vehicle-to-road (V2I) and vehicle-to-network (V2N) three path broadcasting;

[0184] 2) RSU and edge server have information fusion judgment ability to avoid repeated broadcasting and conflict;

[0185] 4. Broadcastable in disconnected network, upgradable in connected network

[0186] 1) In areas without public network signal coverage, local communication can be completed through V2V;

[0187] 2) Under 5G coverage, edge / cloud system can be accessed to realize large-scale broadcast or report the whole network.

[0188] The application can be widely applied in intelligent transportation system, Internet of Vehicles, smart city and other fields, and is suitable for various scenes such as expressway, urban trunk road, tunnel and bridge, and has significant practical value and popularization prospect.

[0189] The various embodiments are described in a progressive manner in the specification, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between the various embodiments can be referred to each other. For the device disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0190] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the application. Therefore, the application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A road traffic incident broadcasting method, characterized by, The method comprises the following steps: a sensing step: collecting real-time road traffic environment data through multi-source sensing devices deployed on the roadside and / or vehicle end; a processing step: fusing the collected multi-source road traffic environment data based on an edge computing platform, identifying the incident, and determining the type, location, impact range, and urgency of the incident; a decision-making step: dynamically allocating 5G network slice resources according to the type, impact range, and urgency of the incident, and generating an optimal broadcast strategy; the optimal broadcast strategy includes broadcast range, broadcast priority, and communication resource allocation; a broadcast step: based on the optimal broadcast strategy, through the vehicle wireless communication V2X interface and / or 5G cellular network interface, the standardized format of the incident warning information is broadcasted to the vehicle terminals within the impact range.

2. The road traffic incident broadcasting method of claim 1, wherein, The sensing step comprises: Collecting road environment data through roadside sensors, vehicle terminals, weather sensors, and / or electronic police devices, and uploading the data to the edge computing node through 5G uplink, PC5 direct communication, optical fiber, or wireless link.

3. The road traffic emergency broadcasting method of claim 1, wherein, The processing step specifically comprises: The edge computing platform performs time synchronization, spatial calibration, and noise filtering processing on the road traffic environment data from different sources; Based on the processed data, the dynamic and static targets on the road are identified through target detection and tracking algorithms; According to the preset rules and / or machine learning model, the type of the incident is determined, and the determination result is cross-verified through multi-source data; Combined with the incident type, road topology, and environmental parameters, the impact range of the incident and the affected vehicle set are calculated.

4. The road traffic emergency broadcasting method of claim 1, wherein, In the decision-making step, the incident is divided into three levels according to the urgency; and different quality of service 5G network slice resources are dynamically allocated for incidents of different levels; Among them The super-reliable low-latency communication slice is allocated for level I incidents; The enhanced mobile broadband slice is allocated for level II incidents; The massive machine type communication slice is allocated for level III incidents; The delay of level I incident < the delay of level II incident < the delay of level III incident.

5. The road traffic emergency broadcasting method of claim 1, wherein, In the broadcast step, it also includes: Adaptively selecting the broadcast mode according to the density of vehicle terminals within the impact range: When the density of vehicle terminals is higher than the first threshold, direct communication broadcast through V2X PC5 interface is preferred; When the density of vehicle terminals is lower than the second threshold, cellular network broadcast through 5G Uu interface is adopted.

6. The road traffic incident broadcasting method of claim 5, wherein, The broadcast mode also includes a hybrid mode in which V2X PC5 interface broadcast and 5G Uu interface broadcast are enabled simultaneously.

7. The road traffic emergency broadcasting method of claim 1, wherein, The standardized format of the incident warning information at least includes: incident type code, precise location information, timestamp, geofencing range, and recommended response action, and all are encrypted.

8. A road traffic emergency broadcasting device characterized by comprising: The device applied to the edge computing node comprises: A collection and sensing module that collects real-time road traffic environment data through multi-source sensing devices deployed on the roadside and / or vehicle end; An incident processing module for fusing the collected multi-source road traffic environment data, identifying the incident, and determining the type, location, impact range, and urgency of the incident; A policy generation module is configured to dynamically allocate 5G network slice resources and generate an optimal broadcast policy according to the type, impact range, and emergency level of the emergency event; the optimal broadcast policy includes a broadcast range, a broadcast priority, and a communication resource allocation; An information broadcast module is configured to broadcast the emergency event warning information in a standardized format to vehicle terminals within the impact range based on the optimal broadcast policy through a vehicle wireless communication (V2X) interface and / or a 5G cellular network interface.

9. A road traffic incident broadcasting system, characterized by The system comprises: The broadcast device according to claim 8; A roadside sensing unit comprising one or more of a camera and a radar, configured to collect roadside traffic environment data; A vehicle terminal arranged on a vehicle, configured to collect vehicle-end data, receive warning information, and perform human-computer interaction; An auxiliary monitoring terminal configured to collect road traffic related information based on a meteorological sensor, an electronic police, and a UAV inspection device; A communication network comprising a 5G base station and a roadside unit (RSU), configured to provide a data transmission channel for the broadcast device, the roadside sensing unit, the vehicle terminal, and the auxiliary monitoring terminal.

10. A road traffic incident broadcasting system as claimed in claim 9, wherein, The system supports linkage with a traditional traffic management system and is compatible with vehicle terminals and mobile devices of different manufacturers.

Images