Blind Spot Risk Collaborative Communication Method and Related Equipment Based on C-V2X and Roadside Sensing Fusion

Abstract

This invention relates to the field of vehicle-to-everything (V2X) communication technology, specifically to a blind spot risk collaborative communication method and related equipment based on the fusion of C-V2X and roadside perception. The method includes: acquiring perception data from multiple roadside units, basic safety messages broadcast by vehicle-mounted units, and electronic map data; performing fusion processing through spatiotemporal alignment and fusion algorithms to generate a global situation map; constructing a dynamic risk field model and generating a risk field probability distribution map; analyzing real-time vehicle status information and calculating predicted vehicle trajectories; performing spatiotemporal intersection analysis between the predicted vehicle trajectories and the risk field probability distribution map; calculating the correlation score between the vehicle and the risk field; classifying vehicles by risk level and performing hierarchical communication scheduling; and achieving seamless relay of early warnings through a token passing mechanism when a risk target moves between the coverage areas of multiple roadside units. This invention can overcome single-vehicle perception blind spots, optimize communication resource utilization, and achieve accurate differentiated early warnings and continuous collaborative services.

Description

Technical Field

[0001] This invention relates to the field of vehicle-to-everything (V2X) communication technology, and in particular to a blind spot risk collaborative communication method and related equipment based on the fusion of C-V2X and roadside perception. Background Technology

[0002] With the rapid development of intelligent connected vehicle technology, vehicle-to-everything (C-V2X) communication has become a key technology for improving road traffic safety. In complex traffic scenarios, blind spot risk warning is an important function to ensure driving safety, especially for sudden risk events such as blind spot intrusions (e.g., unexpected pedestrians suddenly appearing from behind a blind spot), which require beyond-line-of-sight perception and real-time warning.

[0003] Existing blind spot warning solutions in the field of intelligent connected vehicles are mainly divided into three categories, all of which have significant shortcomings: 1. Single-vehicle perception solution: relies on onboard cameras, radar and other sensors.

[0004] The core pain point is the existence of physical perception blind spots, making it impossible to detect targets that are completely obscured (such as "ghost peeking out"), which is a fundamental perception failure.

[0005] 2. Simple V2X broadcast solution: After the roadside unit (RSU) detects a risk, it broadcasts a warning message to all vehicles.

[0006] Key pain points: Inefficient communication: Unrelated vehicles also receive messages, leading to channel congestion and information overload.

[0007] The early warning system is crude: the message content is too simplistic and cannot provide accurate and differentiated early warnings based on the real-time relationship between vehicles and risks.

[0008] 3. Isolated roadside deployment: RSUs operate independently with limited coverage.

[0009] The core pain point is that when risk targets move between multiple RSUs, the early warnings are discontinuous and inconsistent, forming "information silos".

[0010] In summary, existing technological systems suffer from systemic flaws when addressing blind spot risks, including "invisibility, poor transmission, inaccuracy, and inability to connect." The root cause lies in the lack of a collaborative solution that integrates intelligent sensing, dynamic risk assessment, and efficient communication scheduling. Summary of the Invention

[0011] To address the aforementioned technical issues, this invention proposes a blind spot risk collaborative communication method and related equipment based on the fusion of C-V2X and roadside perception. By constructing a collaborative communication scheduling server and fusing perception data from multiple roadside units, dynamic risk modeling and hierarchical communication scheduling are achieved, systematically solving the problems of perception blind spots, communication congestion, inaccurate warnings, and collaborative interruptions in blind spot early warning.

[0012] On one hand, embodiments of the present invention provide a blind spot risk cooperative communication method based on the fusion of C-V2X and roadside perception, the method comprising the following steps: Acquire multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data; The multi-path side unit sensing data is input into the cooperative communication scheduling server, and the multi-path side unit sensing data is fused through a spatiotemporal alignment and fusion algorithm to generate a global situation map. Based on the movement status of the risk targets in the global situation map, a dynamic risk field model is constructed and a risk field probability distribution map is generated. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. Based on the basic safety message, the vehicle's real-time status information is parsed, and the vehicle's predicted trajectory is calculated in combination with the vehicle's real-time status information. The vehicle's predicted trajectory is then subjected to spatiotemporal intersection analysis with the risk field probability distribution map to calculate the correlation score between the vehicle and the risk field. Based on the correlation score, vehicles are classified into risk levels. For high-risk vehicles, unicast communication commands are generated and enhanced early warning messages are triggered. For medium-risk vehicles, multicast communication commands are generated and situational alert messages are triggered. For low-risk vehicles, no communication commands are generated. When a risk target moves between the coverage areas of multiple roadside units, the tracking identifier and historical trajectory data of the risk target are transferred from the original roadside unit to the target roadside unit through a token passing mechanism, and the warning source address of the vehicle communication link is updated simultaneously.

[0013] Optionally, the step of fusing the multi-side unit sensing data using a spatiotemporal alignment and fusion algorithm to generate a global situation map includes: Receive sensing data frames uploaded by multiple roadside units, wherein the sensing data frames include timestamps, roadside unit identifiers, target types, target location coordinates, and confidence levels; Based on the timestamp, the sensing data frames are time-synchronized and aligned, and sensing data frames collected at the same time are grouped into the same processing batch. Based on the roadside unit identifier, query the electronic map to obtain the spatial coordinates and sensing range of each roadside unit, and based on the spatial coordinates, transform the target location coordinates in the sensing data frame to a unified global coordinate system; Spatial correlation matching is performed on the target location coordinates from different roadside units within the same processing batch. When the distance between the target location coordinates reported by different roadside units is less than a preset threshold, they are determined to be the same risk target and the target trajectories are merged. The merged risk target trajectory is overlaid with the blind spot area of ​​the roadside unit perception, and a probabilistic interpolation algorithm is used to supplement the potential risk area in the blind spot area to generate a global situation map containing the complete distribution of risk targets.

[0014] Optionally, the step of constructing a dynamic risk field model and generating a risk field probability distribution map based on the movement state of the risk targets in the global situation map includes: Extract the location coordinates, velocity vector, and direction angle of the risk target from the global situation map; The predicted displacement of the risk target within a preset time window is calculated based on the motion velocity vector, and the main motion axis of the risk target is determined based on the motion direction angle. With the position coordinates as the center, an elliptical risk field core region is constructed along the main motion axis. The length of the major axis of the elliptical risk field core region is proportional to the displacement prediction, and the length of the minor axis is proportional to the lateral component of the motion velocity vector. A ring-shaped attenuation buffer is constructed around the core region of the elliptical risk field, and the probability value of the ring-shaped attenuation buffer decreases from the boundary of the core region outwards in a Gaussian function. The uncertainty covariance matrix is ​​calculated based on the roadside unit perception confidence level. The uncertainty covariance matrix is ​​then superimposed onto the risk field probability distribution map to generate a dynamic risk field model that includes uncertainty quantification.

[0015] Optionally, the step of parsing the vehicle's real-time status information based on the basic safety message, calculating the vehicle's predicted trajectory based on the vehicle's real-time status information, performing spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculating the correlation score between the vehicle and the risk field includes: Parse the basic safety messages to obtain vehicle identifier, vehicle position coordinates, vehicle speed, heading angle, and vehicle acceleration; Based on the vehicle speed, heading angle and vehicle acceleration, a uniform acceleration motion model is used to calculate the vehicle's position sequence in a preset prediction time domain and generate the vehicle's predicted trajectory. The vehicle prediction trajectory is spatiotemporally registered with the risk field probability distribution map, and the distance between each point on the vehicle prediction trajectory and the boundary of the core region of the risk field is calculated. Based on the distance value, query the probability distribution map of the risk field to obtain the probability value of the corresponding location, and integrate the probability value along the vehicle's predicted trajectory to obtain the cumulative probability value of the trajectory crossing the risk field. The collision time estimate is calculated based on the cumulative probability value, vehicle speed, and the speed of the risk target. The weighted combination of the cumulative probability value and the collision time estimate is used as the correlation score between the vehicle and the risk field.

[0016] Optionally, the step of classifying vehicles into risk levels based on the correlation score, generating unicast communication commands and triggering enhanced early warning messages for high-risk vehicles, and generating multicast communication commands and triggering situational awareness messages for medium-risk vehicles includes: Set a high-risk threshold and a medium-risk threshold, and compare the correlation score with the high-risk threshold and the medium-risk threshold respectively; When the correlation score is greater than or equal to the high-risk threshold, the vehicle is determined to be of high risk level, and a unicast communication command containing the target vehicle identifier, roadside unit identifier, and message priority is generated, with the message priority set to the highest level. An enhanced warning message is assembled according to the unicast communication command. The enhanced warning message includes the risk target type, risk target location, predicted collision time, suggested braking deceleration and lane departure warning, and is unicast to the target vehicle through the C-V2X PC5 interface. When the correlation score is less than the high-risk threshold and greater than or equal to the medium-risk threshold, the vehicle is determined to be at a medium-risk level, and a multicast communication command containing a geofence area identifier is generated. Assemble a situational awareness message according to the multicast communication instructions. The situational awareness message includes the location of the risk area, an overview of the risk level, and observation suggestions. Send the message via multicast to vehicles within the geofence area through the C-V2X PC5 interface.

[0017] Optionally, when a risky target moves between the coverage areas of multiple roadside units, transferring the target's tracking identifier and historical trajectory data from the original roadside unit to the target roadside unit via a token passing mechanism includes: Monitor the location coordinates of risk targets in the global situation map. When the location coordinates enter the overlapping boundary zone of the coverage areas of two roadside units, trigger the token passing preparation process. The original roadside unit generates a token data packet, which includes the tracking identifier of the risk target, historical trajectory sequence, current motion state vector, dynamic risk field model parameters, and a list of related vehicles; The token data packet is sent to the cooperative communication scheduling server, which predicts the target roadside unit to which the risk target belongs in the next moment based on the direction of movement of the risk target, and forwards the token data packet to the target roadside unit. The target roadside unit receives and parses the token data packet, inherits the tracking identifier of the risk target, performs trajectory association and fusion with the historical trajectory sequence and local sensing data, and updates the risk target attribution identifier in the global situation map; A warning source switching notification message is sent to the vehicles in the relevant vehicle list. The warning source switching notification message contains the new roadside unit identifier and communication parameters. The vehicles update their communication links according to the warning source switching notification message to maintain the continuity of the warning.

[0018] Optionally, the method further includes: The system receives message delivery confirmation information from the roadside unit, and calculates the warning success rate of vehicles with different risk levels based on the message delivery confirmation information. When the warning success rate of high-risk vehicles is lower than a preset threshold, the system dynamically adjusts the high-risk threshold and the sending frequency of unicast communication commands.

[0019] On the other hand, embodiments of the present invention provide a blind spot risk cooperative communication device based on the fusion of C-V2X and roadside perception, comprising: The first module is used to acquire multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data. The second module is used to input the multi-side unit sensing data into the cooperative communication scheduling server, and to perform fusion processing on the multi-side unit sensing data through a spatiotemporal alignment and fusion algorithm to generate a global situation map. The third module is used to construct a dynamic risk field model and generate a risk field probability distribution map based on the movement status of the risk targets in the global situation map. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. The fourth module is used to parse the real-time vehicle status information based on the basic safety message, calculate the vehicle's predicted trajectory based on the real-time vehicle status information, perform spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculate the correlation score between the vehicle and the risk field. The fifth module is used to classify vehicles into risk levels based on the correlation scores, generate unicast communication instructions for high-risk vehicles and trigger the sending of enhanced early warning messages, generate multicast communication instructions for medium-risk vehicles and trigger the sending of situation alert messages, and not generate communication instructions for low-risk vehicles. The sixth module is used to transfer the tracking identifier and historical trajectory data of the risk target from the original roadside unit to the target roadside unit through a token passing mechanism when the risk target moves between the coverage areas of multiple roadside units, and to simultaneously update the warning source address of the vehicle communication link.

[0020] On the other hand, embodiments of the present invention provide a blind spot risk cooperative communication system based on the fusion of C-V2X and roadside perception, comprising: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0021] On the other hand, embodiments of the present invention provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the above-described method.

[0022] The embodiments of the present invention have the following beneficial effects: This invention constructs a closed-loop system from perception and decision-making to communication by deploying a collaborative communication scheduling server. First, it integrates perception data from multiple roadside units to form a global situational map, overcoming the physical limitations of single-vehicle perception and achieving reliable, all-weather perception of traditional blind spots (especially in intersection "ghost peek" scenarios), eliminating perception blind spots at their source. Second, by constructing a dynamic risk field model and calculating the correlation score between vehicles and the risk field, it transforms the vague "risky" into precise "who is at risk and how much risk is there," achieving accurate tiered early warning. Third, based on the correlation score, it executes tiered directional communication scheduling, changing the blind "one-to-all" broadcast to precise "one-to-one" or "one-to-many" communication, greatly reducing the redundant load of the C-V2X channel and effectively alleviating channel congestion caused by broadcast storms. This makes the overall system communication more reliable and has lower latency in high-density vehicle scenarios. Finally, through a token passing mechanism, it achieves seamless relay of early warnings between multiple roadside units, ensuring continuous, stable, and uninterrupted early warnings for moving risk targets, avoiding confusion caused by interrupted or repeated alarms, and providing users with a continuous and consistent connected service experience. This invention can systematically solve the defects of existing technologies such as "invisibility, poor transmission, inaccuracy, and inability to connect", and fundamentally optimize channel efficiency and early warning continuity while improving security, thus achieving a leap in system-level performance. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating the steps of a blind spot risk collaborative communication method based on the fusion of C-V2X and roadside perception provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the system architecture and data flow provided in the embodiments of the present invention; Figure 3 This is a flowchart of multi-source perception fusion and global situation map generation provided in an embodiment of the present invention; Figure 4 This is a timing diagram of the multi-side unit collaborative relay mechanism provided in an embodiment of the present invention; Figure 5 This is a structural block diagram of a blind spot risk collaborative communication device based on the fusion of C-V2X and roadside perception, provided in an embodiment of the present invention. Detailed Implementation

[0025] 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 and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0026] It should be noted that although the device diagram shows a modular division and the flowchart illustrates a logical order, in some cases, the steps shown or described may be performed in a different order than the modular division in the device or the order shown in the flowchart. The terms "first," "second," etc., used in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0028] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a full understanding of embodiments of the invention. However, those skilled in the art will recognize that the technical solutions of the invention can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the invention.

[0029] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0030] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0031] The key terms involved in this invention are explained as follows: C-V2X PC5 interface: refers to the direct communication interface for vehicle-to-everything (V2X) based on cellular technology. It operates in the 5.9GHz intelligent transportation system frequency band and supports low-latency direct communication between vehicles, roads, and people. It is the communication foundation for the real-time early warning system of this invention.

[0032] MEC (Multi-access Edge Computing): This technology extends cloud computing capabilities to the network edge (such as the base station). The scheduling server in this invention can be deployed in MEC to achieve extremely low decision-making and instruction issuance latency (typically <20ms), meeting the real-time requirements of highly dynamic traffic scenarios.

[0033] SAE J2735 standard: A vehicle-to-everything (V2X) messaging standard defined by the Society of Automotive Engineers (SAE). The "Enhanced Warning Message (E-HWM)" of this invention can be based on the Road Side Alert or Personal Safety Message framework in this standard, with custom data frame extensions to carry innovative information such as dynamic risk scenarios and suggested actions.

[0034] Dynamic risk field: This is a probabilistic model. It represents the spatial probability distribution of a risky target that may appear in the near future, with a high probability in the core area and a low probability in the peripheral area. Its shape and size are jointly determined by the target dynamic equation and the uncertainty (the covariance matrix in Kalman filtering).

[0035] like Figure 1 As shown, Figure 1 A blind spot risk cooperative communication method based on C-V2X and roadside perception fusion is provided for embodiments of the present invention. The method includes the following steps: S100 acquires multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data; S200, the multi-path side unit sensing data is input to the cooperative communication scheduling server, and the multi-path side unit sensing data is fused through a spatiotemporal alignment and fusion algorithm to generate a global situation map; S300, Based on the movement status of the risk targets in the global situation map, a dynamic risk field model is constructed and a risk field probability distribution map is generated. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. S400: Based on the basic safety message, parse the real-time vehicle status information, calculate the vehicle's predicted trajectory in combination with the real-time vehicle status information, perform spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculate the correlation score between the vehicle and the risk field. S500: Based on the correlation score, the vehicle risk level is classified. For high-risk vehicles, a unicast communication command is generated and an enhanced early warning message is triggered. For medium-risk vehicles, a multicast communication command is generated and a situational alert message is triggered. For low-risk vehicles, no communication command is generated. S600, when a risk target moves between the coverage areas of multiple roadside units, transfers the tracking identifier and historical trajectory data of the risk target from the original roadside unit to the target roadside unit through a token passing mechanism, and simultaneously updates the warning source address of the vehicle communication link.

[0036] This invention proposes a blind spot risk collaborative communication method and related equipment based on the fusion of C-V2X and roadside perception. By constructing a collaborative communication scheduling server as the system decision center, it organically combines multi-source perception fusion, dynamic risk modeling, correlation scoring and hierarchical communication scheduling to form a complete blind spot risk collaborative communication link. The method employs a multi-source perception fusion approach, integrating perception data from multiple roadside units through spatiotemporal alignment and fusion algorithms. This overcomes the limitations of single sensors or vehicle line-of-sight occlusion, generating a global situational map containing the complete distribution of risk targets. Dynamic risk field modeling constructs a probability distribution model based on the motion state of risk targets, with the core region expanding with speed and the edge region decaying with uncertainty, achieving a refined description of the risk space. Correlation scoring quantifies the degree of association between vehicles and risks by calculating the spatiotemporal intersection of the predicted vehicle trajectory and the risk field probability distribution map, providing a decision-making basis for hierarchical communication. Hierarchical communication scheduling executes differentiated communication strategies based on correlation scores: unicasting enhanced warning messages to high-risk vehicles, multicasting situational alerts to medium-risk vehicles, and not sending information to low-risk vehicles, significantly improving communication efficiency. Multi-roadside unit collaborative relay achieves seamless transfer of risk target tracking identifiers and historical data through a token passing mechanism, ensuring continuous warnings for moving risk targets. This method achieves low-latency direct communication via the C-V2XPC5 interface, meeting the real-time requirements of highly dynamic traffic scenarios.

[0037] In this embodiment, a collaborative communication scheduling server is first deployed. The collaborative communication scheduling server can be deployed on MEC (Multi-access Edge Computing) nodes to achieve extremely low decision-making and instruction issuance latency (typically <20ms). Multiple intelligent roadside units (RSUs) are deployed at key locations at intersections, and each RSU establishes a network connection with the collaborative communication scheduling server. The vehicle is equipped with an on-board unit (OBU) that conforms to the C-V2X standard and periodically broadcasts basic safety messages (BSM).

[0038] The technical solution of this invention is a closed-loop system consisting of a data input layer, an intelligent decision-making layer, and a collaborative communication layer. Its linkage relationships and core processes are as follows: Figure 2 As shown: (I) System Architecture and Data Flow The entire system consists of the following core entities, and their interactions clearly define the division of responsibilities:

[0039] Architecture Explanation: The collaborative communication scheduling server is the brain of the system, responsible for centrally processing all information and making intelligent decisions. The intelligent RSU cluster is the system's eyes and local execution unit, responsible for perception and directional communication. The on-board unit (OBU) is the system's tentacles and endpoint, reporting its own status and receiving alerts.

[0040] (II) Step-by-step technical solutions and the mechanism of effect generation Combination Figure 2 and Figure 3 The steps work together to produce the desired effect, as follows: Step 1: Multi-source sensing fusion (solving the "invisibility" problem); The process corresponds to: from the data input layer (A1) to the intelligent decision-making layer (B1).

[0041] Solution Details: Data from multiple RSUs, perceived from different perspectives, is uploaded to the server. The server then uses spatiotemporal alignment and fusion algorithms to generate a "global situation map." For example, RSU1's view may be obstructed by a bus, but RSU2 may see a pedestrian from the side. After fusion, the pedestrian's complete trajectory can be reconstructed.

[0042] Results: Breaking through the blind spots of single-vehicle / single-point perception, enabling all-weather, beyond-line-of-sight tracking of high-risk targets, laying the data foundation for safety early warning.

[0043] In some embodiments, the step of fusing the multi-path side unit sensing data using a spatiotemporal alignment and fusion algorithm to generate a global situation map includes: S210, receive sensing data frames uploaded by multiple roadside units, wherein the sensing data frames include timestamps, roadside unit identifiers, target types, target location coordinates, and confidence levels; S220, based on the timestamp, perform time synchronization and alignment of the sensing data frames, and group the sensing data frames collected at the same time into the same processing batch. S230, based on the roadside unit identifier, query the electronic map to obtain the spatial coordinates and sensing range of each roadside unit, and based on the spatial coordinates, transform the target location coordinates in the sensing data frame to a unified global coordinate system; S240: Spatial correlation matching is performed on the target location coordinates from different roadside units within the same processing batch. When the distance between the target location coordinates reported by different roadside units is less than a preset threshold, they are determined to be the same risk target and the target trajectories are merged. S250 overlays the merged risk target trajectory with the blind spot area of ​​the roadside unit perception, and uses a probabilistic interpolation algorithm to supplement the potential risk area in the blind spot area, generating a global situation map containing the complete distribution of risk targets.

[0044] The multi-source sensing fusion process is as follows: Each roadside unit (RSU) uploads its sensing data frames to the cooperative communication scheduling server. The server first extracts the timestamp field from the data frames and performs time synchronization and alignment based on Network Time Protocol (NTP) or GPS time to eliminate clock deviations between RSUs. Then, it queries the pre-stored electronic map database based on the roadside unit identifier to obtain the precise spatial coordinates (longitude, latitude, and elevation) and sensing range parameters (sensing distance and field of view) of each roadside unit. A coordinate transformation matrix is ​​then used to transform the target location coordinates in each RSU's local coordinate system to a unified global coordinate system (such as the WGS-84 coordinate system). For the same processing... Within a batch, target location coordinates from different RSUs are spatially correlated and matched using the nearest neighbor algorithm or the joint probabilistic data association (JPDA) algorithm. When the Euclidean distance between target location coordinates reported by different RSUs is less than a preset threshold (e.g., 2 meters), they are determined to be the same risk target, their trajectories are merged, and the target state estimate is updated. Finally, the merged risk target trajectory is overlaid with the blind spot area of ​​the roadside unit perception (determined by the location of buildings and obstructions in the electronic map). The blind spot area is supplemented with a probabilistic interpolation algorithm (e.g., Kriging interpolation or Gaussian process regression) to supplement the potential risk area, generating a global situation map containing the complete distribution of risk targets.

[0045] Steps 2 and 3: Dynamic modeling and correlation determination (solving the "uncertainty"); Corresponding processes: Internal processes of the intelligent decision-making layer (B1, B2, and B3 in sequence).

[0046] Solution Details: The server constructs a "dynamic risk field" (such as a probability cloud that changes with the target's movement and speed) for each risk target. At the same time, it receives the BSM (A2) of all vehicles and calculates the spatiotemporal interaction between the predicted trajectory of each vehicle and the "risk field", outputting a quantified "correlation score (CS)".

[0047] The effect is that the vague concept of "risky" is transformed into a precise understanding of "who faces the risk and how great the risk is." This allows the system to scientifically distinguish between extremely high-risk vehicles (whose trajectories pass directly through the core risk area) and low-risk vehicles (who merely pass through the edge), achieving accurate classification of warnings.

[0048] In some embodiments, constructing a dynamic risk field model and generating a risk field probability distribution map based on the movement state of the risk targets in the global situation map includes: S310, Extract the position coordinates, velocity vector, and direction angle of the risk target from the global situation map; S320, calculate the predicted displacement of the risk target within a preset time window based on the motion velocity vector, and determine the main motion axis of the risk target based on the motion direction angle; S330, with the position coordinates as the center, an elliptical risk field core region is constructed along the main motion axis. The length of the major axis of the elliptical risk field core region is proportional to the displacement prediction, and the length of the minor axis is proportional to the lateral component of the motion velocity vector. S340, a ring-shaped attenuation buffer is constructed around the core region of the elliptical risk field, and the probability value of the ring-shaped attenuation buffer decreases from the boundary of the core region outwards in a Gaussian function. S350, calculate the uncertainty covariance matrix based on the roadside unit perception confidence level, and superimpose the uncertainty covariance matrix onto the risk field probability distribution map to generate a dynamic risk field model containing uncertainty quantification.

[0049] The schematic diagram of the dynamic risk field model is as follows: The predicted displacement within a preset time window (e.g., 2 seconds) is calculated based on the motion velocity vector. An elliptical core region of the risk field is constructed along the main motion axis, centered on the current position of the risk target. A ring-shaped attenuation buffer zone is constructed around the ellipse, with the probability value decreasing as a Gaussian function from the boundary of the core region outwards. The uncertainty covariance matrix is ​​calculated based on the roadside unit perception confidence level. This covariance matrix is ​​then superimposed onto the risk field probability distribution map to quantify the uncertainty.

[0050] In some embodiments, the step of parsing the vehicle's real-time status information based on the basic safety message, calculating the vehicle's predicted trajectory based on the vehicle's real-time status information, performing spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculating the correlation score between the vehicle and the risk field includes: S410, parse the basic safety message to obtain vehicle identifier, vehicle position coordinates, vehicle speed, heading angle and vehicle acceleration; S420, based on the vehicle speed, heading angle and vehicle acceleration, a uniform acceleration motion model is used to calculate the vehicle's position sequence in a preset prediction time domain, and a vehicle prediction trajectory is generated. S430, perform spatiotemporal registration of the vehicle prediction trajectory with the risk field probability distribution map, and calculate the distance values ​​between each point on the vehicle prediction trajectory and the boundary of the core area of ​​the risk field. S440, based on the distance value, query the probability distribution map of the risk field to obtain the probability value of the corresponding position, and integrate the probability value along the vehicle's predicted trajectory to obtain the cumulative probability value of the trajectory crossing the risk field; S450, calculate the collision time estimate based on the cumulative probability value, vehicle speed and the speed of movement of the risk target, and use the weighted combination of the cumulative probability value and the collision time estimate as the correlation score between the vehicle and the risk field.

[0051] The correlation score calculation and hierarchical communication scheduling process is as follows: Parse the BSM message to obtain vehicle state parameters; use a uniform acceleration motion model to calculate the position sequence in the prediction time domain; perform spatiotemporal registration between the vehicle's predicted trajectory and the risk field probability distribution map; calculate the distance from each point on the trajectory to the boundary of the core area of ​​the risk field; and retrieve the corresponding probability value from the risk field probability distribution map based on the distance value. Integrate the probability value along the trajectory to obtain the cumulative probability value. Divide the minimum distance between the trajectory and the risk target by the relative speed between the vehicle and the risk target to calculate the collision time estimate. Calculate the correlation score by weighted summing the reciprocal of the collision time estimate and the cumulative probability value.

[0052] Step 4: Hierarchical communication scheduling (solving the problems of "poor transmission" and "inaccurate communication"); The process corresponds to the intelligent decision-making layer (B4) and the collaborative communication layer (C).

[0053] Solution Details: Based on the CS score, the server generates differentiated communication commands: For high CS vehicles (C1): Instruct the RSU to send "Enhanced Warning Message (E-HWM)" via "unicast", with specific content and high frequency.

[0054] For CS vehicles (C2): Instruct the RSU to send a "Situation Alert Message (SAM)" via multicast. The message is brief and sent infrequently.

[0055] For low CS vehicles (C3): No communication commands are generated.

[0056] Effect produced: Communication efficiency is greatly improved: the whole network broadcast is changed to targeted communication, which directly reduces more than 70% of redundant messages and fundamentally alleviates channel congestion.

[0057] A leap in early warning effectiveness: Drivers receive only high-value information that is highly relevant to them and includes clear action recommendations, avoiding alert fatigue and improving trust and response speed.

[0058] In some embodiments, the step of classifying vehicles into risk levels based on the correlation score, generating unicast communication commands and triggering enhanced early warning message transmission for high-risk vehicles, and generating multicast communication commands and triggering situational awareness message transmission for medium-risk vehicles includes: S510, Set a high-risk threshold and a medium-risk threshold, and compare the correlation score with the high-risk threshold and the medium-risk threshold respectively; S520, when the correlation score is greater than or equal to the high-risk threshold, the vehicle is determined to be of high-risk level, and a unicast communication command containing the target vehicle identifier, roadside unit identifier and message priority is generated, wherein the message priority is set to the highest level. S530, an enhanced warning message is assembled according to the unicast communication instruction. The enhanced warning message includes the risk target type, risk target location, predicted collision time, suggested braking deceleration and lane departure warning, and is unicast to the target vehicle through the C-V2XPC5 interface. S540, when the correlation score is less than the high-risk threshold and greater than or equal to the medium-risk threshold, the vehicle is determined to be at a medium-risk level, and a multicast communication command containing a geofence area identifier is generated. S550, according to the multicast communication command, assembles a situational awareness message, the situational awareness message includes the location of the risk area, an overview of the risk level and observation suggestions, and sends it via C-V2X PC5 interface to vehicles within the geofence area via multicast.

[0059] Enhanced Risk Warning Messages (E-HWM) are based on the Road Side Alert or Personal Safety Message framework in the SAE J2735 standard, with custom data frame extensions to carry innovative information such as dynamic risk fields and suggested actions. The message format includes: message identifier, timestamp, risk target identifier, risk target type (pedestrian / non-motorized vehicle / vehicle), risk target location (latitude, longitude, and elevation), predicted collision time (TTC), suggested braking deceleration (m / s²), lane departure warning (left / right / none), and message priority (0-7, 7 being the highest).

[0060] Step 5: Multi-RSU collaborative relay (solving the "connection failure" issue) Corresponding process: Intelligent decision-making layer (B5) coordinates the management of the roadside subsystem.

[0061] Solution details: When the server (MEC) detects that a risky target has moved from the RSU1 jurisdiction to the RSU2 jurisdiction, it automatically transfers the target's "tracking token" and all context information to RSU2 and notifies the relevant vehicles that the warning source has been switched.

[0062] Results: It enables seamless and smooth switching of early warning responsibilities, ensures continuous, stable, and uninterrupted early warning of moving targets, and provides a seamless connected service experience.

[0063] In some embodiments, the step of transferring the tracking identifier and historical trajectory data of the risk target from the original roadside unit to the target roadside unit via a token passing mechanism when the risk target moves between the coverage areas of multiple roadside units includes: S610, monitor the location coordinates of the risk target in the global situation map, and trigger the token passing preparation process when the location coordinates enter the overlapping boundary zone of the coverage area of ​​two roadside units. S620, the original roadside unit generates a token data packet, which includes the tracking identifier of the risk target, historical trajectory sequence, current motion state vector, dynamic risk field model parameters and related vehicle list; S630, the token data packet is sent to the cooperative communication scheduling server, which predicts the target roadside unit to which the risk target belongs in the next moment based on the direction of movement of the risk target, and forwards the token data packet to the target roadside unit; S640, the target roadside unit receives and parses the token data packet, inherits the tracking identifier of the risk target, performs trajectory association and fusion with the historical trajectory sequence and local perception data, and updates the risk target attribution identifier in the global situation map; S650, a warning source switching notification message is sent to the vehicles in the relevant vehicle list. The warning source switching notification message includes the new roadside unit identifier and communication parameters. The vehicles update their communication links according to the warning source switching notification message to maintain warning continuity.

[0064] refer to Figure 4 The multi-roadside unit collaborative relay mechanism has the following timing sequence: Assume that the coverage areas of RSU1 and RSU2 have overlapping boundary zones. When a risk target moves from the RSU1 area to the RSU2 area and enters the overlapping boundary zone, the collaborative communication scheduling server triggers the token passing process. RSU1 generates a token data packet, which includes: the risk target tracking ID (globally unique identifier), historical trajectory sequence (position coordinate sequence of the past N seconds), current motion state vector (position, velocity, acceleration), dynamic risk field model parameters (ellipse major axis, minor axis, orientation angle, probability distribution parameters), and a list of related vehicles (a list of vehicle identifiers that have received the warning). Based on the risk target's current motion direction vector and position coordinates, the server predicts the next moment's position, determines that the roadside unit to which this position belongs is the target roadside unit RSU2, and forwards the token data packet to RSU2. After receiving the token, RSU2 inherits the risk target's tracking ID, performs trajectory association and fusion with the historical trajectory sequence and local sensing data, smooths the trajectory using Kalman filtering or particle filtering algorithms, and updates the risk target's attribution identifier in the global situation map to RSU2. RSU2 sends a warning source switching notification message to vehicles in the relevant vehicle list. The message includes the new roadside unit identifier RSU2 and the communication parameters of RSU2 (IP address, port number, and transmission power of the PC5 interface). After receiving the message, the vehicle OBU updates its local communication link configuration and switches the message receiving source from RSU1 to RSU2 to maintain the continuity of warnings. The switching process has a delay of less than 50ms.

[0065] This technical solution is not a mere collection of isolated modules, but rather an organic whole that is data-driven, intelligent in decision-making, and precise in execution. Accurate data input (fusion perception) ensures the correctness of decision-making.

[0066] Intelligent central decision-making (modeling and judgment) ensures efficient communication and targeted content.

[0067] Coordinated communication execution (hierarchical scheduling and relay) ultimately delivers the right information to the right recipient at the right time.

[0068] Ultimately, this closed-loop linkage system comprehensively generated four core beneficial effects: "no blind spots in perception, high efficiency in communication, high accuracy in early warning, and continuous service," systematically solving all the pain points of existing technologies.

[0069] Specific implementation example (taking the "ghost peek" scenario at an urban intersection as an example): Scene setting: At a traffic light intersection, the east-west direction has a green light. A bus is stopped near the intersection (to pick up or drop off passengers), completely obstructing a pedestrian who is about to run across the road on the right side of the crosswalk. A private car (HV) is approaching the intersection at normal speed in the adjacent lane to the bus, posing an extremely high risk of colliding with the suddenly appearing pedestrian.

[0070] System Configuration: Roadside equipment: One smart RSU (RSU_N, RSU_E, RSU_S, RSU_W) is deployed at each of the four corners of the intersection, and all of them are connected to the collaborative communication scheduling server deployed on the edge cloud.

[0071] Vehicle: The main vehicle (HV) is equipped with an OBU that conforms to the C-V2X standard and periodically broadcasts a BSM.

[0072] Workflow sequence details: Process Phase Analysis: 1. Risk Perception and Fusion (Phase 1): Pedestrians were detected on the north side (RSU_N), but not on the west side (RSU_W) due to obstruction by a bus. The server combined the information from both sides (RSU_N has data, RSU_W has no data) with the electronic map to infer that there is a potentially high-risk area behind the bus that is obstructed.

[0073] 2. Dynamic Modeling and Judgment (Phase Two): Based on the pedestrian's running speed and direction, the server generates a "risk cloud map" that rapidly expands from the rear of the bus eastward (in the direction of entering the lane). Simultaneously, the predicted trajectory of the main vehicle's HV (straight-ahead) is calculated, revealing a significant spatiotemporal overlap between the two in approximately 1.5 seconds, with a correlation score (CS) as high as 0.85, thus qualifying for the highest warning level.

[0074] 3. Tiered Communication Scheduling (Phase Three): The server does not broadcast to all vehicles. It precisely instructs the RSU_W, which is closest to the risk and responsible for communicating with the HV, to unicast only an enhanced warning message containing specific action recommendations ("Immediate Braking") to the HV. At this time, vehicles in the oncoming lane will not receive any interference information.

[0075] 4. Collaborative Relay (Phase Four): When a pedestrian runs past the front of the bus and enters the direct line of sight of RSU_W, the server automatically and smoothly transfers the warning responsibility from RSU_N to RSU_W, and synchronizes all historical data. The warning for HV is seamlessly connected at the millisecond level, without any sense of interruption or restart.

[0076] 5. Risk Clearance (Phase Five): After the pedestrian has safely crossed, the system automatically terminates the warning and notifies HV that the risk has been cleared, thus avoiding continued tension for the driver.

[0077] Verification of the beneficial effects of this embodiment: Safety: HV drivers receive clear and strong audible and visual warnings and braking advice approximately 1.5 seconds before a pedestrian enters their field of vision, gaining valuable reaction time and greatly reducing the risk of accidents.

[0078] Communication efficiency: Throughout the entire process, only HV received high-frequency warnings. Dozens of other vehicles at the intersection were not affected by any irrelevant messages, resulting in a channel resource saving of over 90%.

[0079] Accuracy and Continuity: The warning content accurately corresponds to the emergency situation of HV, and the switching of the warning source is seamless during pedestrian movement, providing a continuous experience.

[0080] This invention achieves a leap in system-level performance by deeply synergizing dynamic risk quantification and intelligent scheduling of communication resources, while improving security and fundamentally optimizing channel efficiency and early warning continuity.

[0081] See Figure 5 This invention provides a blind spot risk cooperative communication device based on the fusion of C-V2X and roadside perception, comprising: The first module is used to acquire multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data. The second module is used to input the multi-side unit sensing data into the cooperative communication scheduling server, and to perform fusion processing on the multi-side unit sensing data through a spatiotemporal alignment and fusion algorithm to generate a global situation map. The third module is used to construct a dynamic risk field model and generate a risk field probability distribution map based on the movement status of the risk targets in the global situation map. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. The fourth module is used to parse the real-time vehicle status information based on the basic safety message, calculate the vehicle's predicted trajectory based on the real-time vehicle status information, perform spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculate the correlation score between the vehicle and the risk field. The fifth module is used to classify vehicles into risk levels based on the correlation scores, generate unicast communication instructions for high-risk vehicles and trigger the sending of enhanced early warning messages, generate multicast communication instructions for medium-risk vehicles and trigger the sending of situation alert messages, and not generate communication instructions for low-risk vehicles. The sixth module is used to transfer the tracking identifier and historical trajectory data of the risk target from the original roadside unit to the target roadside unit through a token passing mechanism when the risk target moves between the coverage areas of multiple roadside units, and to simultaneously update the warning source address of the vehicle communication link.

[0082] It is evident that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented in this device embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0083] This invention provides a blind spot risk cooperative communication system based on the fusion of C-V2X and roadside perception, comprising: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0084] It is evident that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0085] Furthermore, embodiments of the present invention also disclose a computer program product or computer program stored in a computer-readable storage medium. A processor of a computer device can read the computer program from the computer-readable storage medium, and the processor executes the computer program, causing the computer device to perform the described method. Similarly, the content of the above method embodiments is applicable to this storage medium embodiment. The specific functions implemented in this storage medium embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0086] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0087] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0088] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0089] It should be understood that in this invention, "at least one (item)" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0090] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0091] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0092] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0093] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0094] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of the present invention should be within the scope of the claims of the present invention.

Claims

1. A blind spot risk collaborative communication method based on the fusion of C-V2X and roadside perception, characterized in that, The method includes the following steps: Acquire multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data; The multi-path side unit sensing data is input into the cooperative communication scheduling server, and the multi-path side unit sensing data is fused through a spatiotemporal alignment and fusion algorithm to generate a global situation map. Based on the movement status of the risk targets in the global situation map, a dynamic risk field model is constructed and a risk field probability distribution map is generated. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. Based on the basic safety message, the vehicle's real-time status information is parsed, and the vehicle's predicted trajectory is calculated in combination with the vehicle's real-time status information. The vehicle's predicted trajectory is then subjected to spatiotemporal intersection analysis with the risk field probability distribution map to calculate the correlation score between the vehicle and the risk field. Based on the correlation score, vehicles are classified into risk levels. For high-risk vehicles, unicast communication commands are generated and enhanced early warning messages are triggered. For medium-risk vehicles, multicast communication commands are generated and situational alert messages are triggered. For low-risk vehicles, no communication commands are generated. When a risk target moves between the coverage areas of multiple roadside units, the tracking identifier and historical trajectory data of the risk target are transferred from the original roadside unit to the target roadside unit through a token passing mechanism, and the warning source address of the vehicle communication link is updated simultaneously.

2. The method according to claim 1, characterized in that, The step of fusing the multi-path side unit sensing data through a spatiotemporal alignment and fusion algorithm to generate a global situation map includes: Receive sensing data frames uploaded by multiple roadside units, wherein the sensing data frames include timestamps, roadside unit identifiers, target types, target location coordinates, and confidence levels; Based on the timestamp, the sensing data frames are time-synchronized and aligned, and sensing data frames collected at the same time are grouped into the same processing batch. Based on the roadside unit identifier, query the electronic map to obtain the spatial coordinates and sensing range of each roadside unit, and based on the spatial coordinates, transform the target location coordinates in the sensing data frame to a unified global coordinate system; Spatial correlation matching is performed on the target location coordinates from different roadside units within the same processing batch. When the distance between the target location coordinates reported by different roadside units is less than a preset threshold, they are determined to be the same risk target and the target trajectories are merged. The merged risk target trajectory is overlaid with the blind spot area of ​​the roadside unit perception, and a probabilistic interpolation algorithm is used to supplement the potential risk area in the blind spot area to generate a global situation map containing the complete distribution of risk targets.

3. The method according to claim 1, characterized in that, The step of constructing a dynamic risk field model and generating a risk field probability distribution map based on the movement state of risk targets in the global situation map includes: Extract the location coordinates, velocity vector, and direction angle of the risk target from the global situation map; The predicted displacement of the risk target within a preset time window is calculated based on the motion velocity vector, and the main motion axis of the risk target is determined based on the motion direction angle. With the position coordinates as the center, an elliptical risk field core region is constructed along the main motion axis. The length of the major axis of the elliptical risk field core region is proportional to the displacement prediction, and the length of the minor axis is proportional to the lateral component of the motion velocity vector. A ring-shaped attenuation buffer is constructed around the core region of the elliptical risk field, and the probability value of the ring-shaped attenuation buffer decreases from the boundary of the core region outwards in a Gaussian function. The uncertainty covariance matrix is ​​calculated based on the roadside unit perception confidence level. The uncertainty covariance matrix is ​​then superimposed onto the risk field probability distribution map to generate a dynamic risk field model that includes uncertainty quantification.

4. The method according to claim 1, characterized in that, The process involves parsing the vehicle's real-time status information based on the basic safety messages, calculating the vehicle's predicted trajectory using the real-time status information, performing a spatiotemporal intersection analysis between the predicted trajectory and the risk field probability distribution map, and calculating the correlation score between the vehicle and the risk field. This includes: Parse the basic safety messages to obtain vehicle identifier, vehicle position coordinates, vehicle speed, heading angle, and vehicle acceleration; Based on the vehicle speed, heading angle and vehicle acceleration, a uniform acceleration motion model is used to calculate the vehicle's position sequence in a preset prediction time domain and generate the vehicle's predicted trajectory. The vehicle prediction trajectory is spatiotemporally registered with the risk field probability distribution map, and the distance between each point on the vehicle prediction trajectory and the boundary of the core region of the risk field is calculated. Based on the distance value, query the probability distribution map of the risk field to obtain the probability value of the corresponding location, and integrate the probability value along the vehicle's predicted trajectory to obtain the cumulative probability value of the trajectory crossing the risk field. The collision time estimate is calculated based on the cumulative probability value, vehicle speed, and the speed of the risk target. The weighted combination of the cumulative probability value and the collision time estimate is used as the correlation score between the vehicle and the risk field.

5. The method according to claim 1, characterized in that, The step of classifying vehicles into risk levels based on the correlation score, generating unicast communication commands and triggering enhanced early warning messages for high-risk vehicles, and generating multicast communication commands and triggering situational awareness messages for medium-risk vehicles includes: Set a high-risk threshold and a medium-risk threshold, and compare the correlation score with the high-risk threshold and the medium-risk threshold respectively; When the correlation score is greater than or equal to the high-risk threshold, the vehicle is determined to be of high risk level, and a unicast communication command containing the target vehicle identifier, roadside unit identifier, and message priority is generated, with the message priority set to the highest level. An enhanced warning message is assembled according to the unicast communication command. The enhanced warning message includes the risk target type, risk target location, predicted collision time, suggested braking deceleration and lane departure warning, and is unicast to the target vehicle through the C-V2X PC5 interface. When the correlation score is less than the high-risk threshold and greater than or equal to the medium-risk threshold, the vehicle is determined to be at a medium-risk level, and a multicast communication command containing a geofence area identifier is generated. Assemble a situational awareness message according to the multicast communication instructions. The situational awareness message includes the location of the risk area, an overview of the risk level, and observation suggestions. Send the message via multicast to vehicles within the geofence area through the C-V2X PC5 interface.

6. The method according to claim 1, characterized in that, When a risky target moves between the coverage areas of multiple roadside units, the tracking identifier and historical trajectory data of the risky target are transferred from the original roadside unit to the target roadside unit through a token passing mechanism, including: Monitor the location coordinates of risk targets in the global situation map. When the location coordinates enter the overlapping boundary zone of the coverage areas of two roadside units, trigger the token passing preparation process. The original roadside unit generates a token data packet, which includes the tracking identifier of the risk target, historical trajectory sequence, current motion state vector, dynamic risk field model parameters, and a list of related vehicles; The token data packet is sent to the cooperative communication scheduling server, which predicts the target roadside unit to which the risk target belongs in the next moment based on the direction of movement of the risk target, and forwards the token data packet to the target roadside unit. The target roadside unit receives and parses the token data packet, inherits the tracking identifier of the risk target, performs trajectory association and fusion with the historical trajectory sequence and local sensing data, and updates the risk target attribution identifier in the global situation map; A warning source switching notification message is sent to the vehicles in the relevant vehicle list. The warning source switching notification message contains the new roadside unit identifier and communication parameters. The vehicles update their communication links according to the warning source switching notification message to maintain the continuity of the warning.

7. The method according to claim 1, characterized in that, The method further includes: The system receives message delivery confirmation information from the roadside unit, and calculates the warning success rate of vehicles with different risk levels based on the message delivery confirmation information. When the warning success rate of high-risk vehicles is lower than a preset threshold, the system dynamically adjusts the high-risk threshold and the sending frequency of unicast communication commands.

8. A blind spot risk collaborative communication device based on C-V2X and roadside perception fusion, characterized in that, include: The first module is used to acquire multi-side unit perception data, basic safety messages broadcast by the vehicle unit, and electronic map data. The second module is used to input the multi-side unit sensing data into the cooperative communication scheduling server, and to perform fusion processing on the multi-side unit sensing data through a spatiotemporal alignment and fusion algorithm to generate a global situation map. The third module is used to construct a dynamic risk field model and generate a risk field probability distribution map based on the movement status of the risk targets in the global situation map. The probability of the core area of ​​the dynamic risk field model expands as the movement speed of the risk targets increases, while the probability of the edge area decreases as uncertainty increases. The fourth module is used to parse the real-time vehicle status information based on the basic safety message, calculate the vehicle's predicted trajectory based on the real-time vehicle status information, perform spatiotemporal intersection analysis on the vehicle's predicted trajectory and the risk field probability distribution map, and calculate the correlation score between the vehicle and the risk field. The fifth module is used to classify vehicles into risk levels based on the correlation scores, generate unicast communication instructions for high-risk vehicles and trigger the sending of enhanced early warning messages, generate multicast communication instructions for medium-risk vehicles and trigger the sending of situation alert messages, and not generate communication instructions for low-risk vehicles. The sixth module is used to transfer the tracking identifier and historical trajectory data of the risk target from the original roadside unit to the target roadside unit through a token passing mechanism when the risk target moves between the coverage areas of multiple roadside units, and to simultaneously update the warning source address of the vehicle communication link.

9. A blind spot risk collaborative communication system based on the fusion of C-V2X and roadside perception, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a processor-executable program, characterized in that, The processor-executable program, when executed by the processor, is used to perform the method as described in any one of claims 1 to 7.

Images