A method, device, and medium for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification.

Abstract

This invention relates to a method, device, and medium for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification. The method includes: constructing a trajectory surface model based on the actual vehicle dimensions and lateral safety distance; establishing the trajectories of two vehicles to determine the conflict zone; calculating the improved intrusion time based on the temporal relationship between the rear of the preceding vehicle and the front of the following vehicle within the conflict zone to identify potential conflict events; assessing the severity of collision risk based on changes in vehicle momentum and kinetic energy during the conflict event; fusing the intrusion time and kinetic energy changes into a conflict risk index; and constructing a cooperative perception communication model to dynamically allocate communication bandwidth and perception resources according to detection priority, thereby prioritizing the perception needs of high-risk vehicles. Compared with existing technologies, this invention can guide cooperative perception decisions and communication resource allocation for multiple vehicles in complex traffic environments, improving target detection accuracy and traffic safety levels in high-risk areas such as intersections.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation and vehicle networking technology, and in particular to a method, device, and medium for dynamic allocation of vehicle collaborative perception resources based on conflict risk quantification. Background Technology

[0002] With the development of vehicle-to-everything (V2X) and multi-source sensing technologies, collaborative perception has become an important means for intelligent transportation systems to improve their environmental perception capabilities. Through vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication, cross-platform information sharing of road targets can be achieved, thereby expanding the perception field of a single vehicle and improving detection robustness. However, existing research generally assigns the same weight to all detected targets, failing to consider the differences in the safety risks of targets. This leads to delays or even the loss of information on critical targets in scenarios with limited bandwidth and computing resources, affecting the system's timely response to high-risk events.

[0003] In fact, at high-risk road sections such as intersections and ramps, different vehicles have different collision risks due to factors such as speed and travel path. A uniform detection weight not only causes low-risk vehicles to occupy too many resources, resulting in a waste of bandwidth and computing power, but may also prevent high-risk vehicles from receiving priority detection, increasing the probability of potential accidents and reducing overall driving safety.

[0004] The existing technology CN118116236A proposes a safety risk field method that acquires environmental information through millimeter-wave radar, GNSS, and lidar, then superimposes the risk fields of each vehicle at the macroscopic level and calculates the comprehensive conflict risk field value. However, its model only focuses on a binary judgment of whether an interaction will occur, and it does not distinguish the severity of collisions caused by differences in vehicle size and momentum when superimposing the risk fields. This results in the inability to prioritize high-risk targets in scenarios such as intersections where multiple vehicles are in conflict simultaneously, making it easy for system resources to be occupied by low-risk vehicles and for critical targets to be identified with delays. Publication number CN114987539A introduces personalized collision classification warnings in longitudinal collision avoidance in autonomous driving, setting three-level thresholds and triggering classified warnings through a comprehensive risk field index curve. However, its threshold generation process is independent of communication and perception resource constraints. When multiple vehicles need high-resolution data simultaneously, bandwidth is still allocated in a fixed proportion, causing critical vehicles to be unable to obtain information first, thereby reducing the overall collision avoidance success rate.

[0005] Therefore, considering the areas for improvement in existing technologies, there is an urgent need for a collaborative perception optimization method that can dynamically allocate detection resources based on the actual safety risks between vehicles. By introducing a risk assessment model based on the improved PET (Post Encroachment Time) index and the momentum changes of conflicting vehicles before and after the collision, the conflict risk between the vehicle and other vehicles in the region of interest is quantified, and the target detection priority is determined accordingly. This allows for prioritizing the perception and communication needs of high-risk vehicles under limited bandwidth conditions, thereby improving system resource utilization efficiency and traffic safety levels. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology by providing a method, device, and medium for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification. This method can guide the cooperative perception decision-making and communication resource allocation of multiple vehicles in complex traffic environments, thereby improving the target detection accuracy and traffic safety level in high-risk areas such as intersections.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] The first aspect of this invention provides a method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification, comprising the following steps:

[0009] A trajectory surface model is constructed based on the actual dimensions of the vehicles and the lateral safety distance, and the trajectories of the two vehicles are combined to determine the conflict area;

[0010] Based on the temporal relationship between the rear of the preceding vehicle and the front of the following vehicle within the conflict area, the improved intrusion time is calculated to identify potential conflict events.

[0011] Assess the severity of collision risk based on changes in the momentum and kinetic energy of vehicles during a conflict.

[0012] The intrusion time and kinetic energy change are combined into a conflict risk index, and the detection priority of each vehicle is determined accordingly.

[0013] A collaborative perception and communication model is constructed, and communication bandwidth and perception resources are dynamically allocated according to detection priority, so as to prioritize the perception needs of high-risk vehicles.

[0014] Furthermore, the specific process of constructing a trajectory surface model based on the actual vehicle dimensions and lateral safety distance, and establishing the trajectory boundaries of the two vehicles to determine the conflict area, includes:

[0015] The vehicle passage width is defined as the sum of the vehicle body width and the safety distances on both sides, forming a trajectory surface model;

[0016] Based on the trajectory surface model, corresponding trajectory surface inner and outer boundary parameter equations are generated according to the vehicle's inlet and outlet lanes;

[0017] By simultaneously solving the outer and inner boundary equations of the trajectory surfaces of the two vehicles, the set of intersection points can be obtained.

[0018] The set of intersection points forms a closed region, which serves as the actual conflict area between the two vehicles.

[0019] Furthermore, when the vehicle's trajectory is a curve, the trajectory surface model is constructed using an ellipse equation, and the construction process specifically includes:

[0020] Determine the coordinates of the ellipse center based on the vehicle's import direction:

[0021] Vehicles turning left at the south entrance have the origin of the coordinate system as the center of the ellipse.

[0022] For vehicles turning left at the east entrance, the center of the ellipse is a point whose horizontal coordinate is an integer multiple of the lane radius and whose vertical coordinate is 0.

[0023] For vehicles turning left at the west entrance, the center of the ellipse is a point with an abscissa of 0 and a ordinate offset from the lane radius that is an integer multiple of the lane radius.

[0024] For vehicles turning left from the north entrance, the center of the ellipse is a point whose horizontal and vertical coordinates are both offset from the lane radius by an integer multiple;

[0025] Based on the vehicle's center position and actual occupied width, elliptical equations are established for the outer boundary, inner boundary, and center trajectory line, respectively.

[0026] Furthermore, (1) if trajectory r is a straight line, and if vehicle i is a straight trajectory from the east entrance, based on the vehicle's current position (x i ,y i The equations for the vehicle's inner and outer boundaries and center trajectory are:

[0027]

[0028] If vehicle i follows a straight trajectory from the west entrance, the equations of the vehicle's inner and outer boundaries and center trajectory are:

[0029]

[0030] If vehicle i follows a straight trajectory from the south entrance, the equations of the vehicle's inner and outer boundaries and center trajectory are:

[0031]

[0032] If vehicle i follows a straight trajectory from the north entrance, the equations for the vehicle's inner and outer boundaries and center trajectory are:

[0033]

[0034] (2) If the trajectory r is a curve, we use the equation of an ellipse to represent the vehicle's trajectory. When vehicle i is a left-turning vehicle at the south entrance, according to the vehicle's current position (x... i ,y i Given the center of the ellipse (0,0), the equations of the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0035]

[0036] When vehicle i is a vehicle turning left from the east entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (2n·d) r From the equations of the vehicle's inner and outer boundaries and center trajectory line, we obtain:

[0037]

[0038] When vehicle i is a vehicle turning left from the west entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (0, 2n·d) r The equations for the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0039]

[0040] When vehicle i is a vehicle turning left from the north entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (2n·d) r ,2n·d r The equations for the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0041]

[0042] Furthermore, based on the temporal relationship between the rear of the preceding vehicle and the front of the following vehicle within the conflict area, the specific process for calculating the improved intrusion time to identify potential conflict events includes:

[0043] The order in which vehicles entered the conflict zone determines the order of the preceding and following vehicles.

[0044] Calculate the travel time of the rear vehicle's front end from its current position to the entrance of the conflict zone, and the travel time of the front vehicle's rear end from its current position to the exit of the conflict zone;

[0045] The improved intrusion time is obtained by subtracting the time when the rear of the preceding vehicle completely leaves the conflict zone from the time when the front of the following vehicle enters the conflict zone.

[0046] If the improved intrusion time is positive and less than the set threshold, then a potential conflict event is determined to exist.

[0047] Furthermore, the specific process for assessing the severity of collision risk based on changes in the momentum and kinetic energy of vehicles during a conflict includes:

[0048] The mass and velocity of each pair of conflicting vehicles are obtained separately, and the momentum is decomposed into lateral and longitudinal components based on the driving angle of each conflicting vehicle entering the conflict zone.

[0049] The resultant momentum of the system after a collision is calculated based on the principle of conservation of momentum, and the change in the total kinetic energy of the system before and after the collision is obtained accordingly. This change in kinetic energy is used as a quantitative indicator of the severity of the collision.

[0050] Furthermore, the calculation process for the change in total kinetic energy of the system before and after the collision includes:

[0051] Based on the driving angle, momentum is decomposed into two mutually perpendicular directional components;

[0052] Calculate the sum of the momentum vectors of the two vehicles in the lateral and longitudinal directions before the collision;

[0053] According to the law of conservation of momentum, the magnitude of the velocity vector of the system after the collision can be deduced from the resultant momentum in the horizontal and vertical directions.

[0054] The degree of energy release is quantified by comparing the difference between the total kinetic energy before the collision and the system's kinetic energy after the collision.

[0055] Furthermore, the specific process of integrating the improved intrusion time and kinetic energy change into a conflict risk indicator, and determining the detection priority of each vehicle accordingly, includes:

[0056] The improved intrusion time is converted into the probability of collision occurrence, where the shorter the improved intrusion time, the higher the probability.

[0057] The change in kinetic energy is converted into the severity of the collision, where a larger change in kinetic energy corresponds to a higher severity.

[0058] The obtained collision probability and collision severity are combined to form a conflict risk indicator;

[0059] Based on the aforementioned conflict risk indicators, all potential conflict vehicle pairs are sorted in descending order to form a detection priority sequence for resource allocation.

[0060] Furthermore, the specific process of constructing a collaborative perception and communication model, and dynamically allocating communication bandwidth and perception resources based on detection priority to prioritize the perception needs of high-risk vehicles includes:

[0061] Using the detection priority sequence as input, a communication topology is generated by treating connected vehicles and their communicable vehicles as nodes;

[0062] Under the constraint of total bandwidth at the intersection, bandwidth allocation is directly and positively correlated with priority; the higher the priority, the larger the share of bandwidth allocated.

[0063] For each communication pair, the maximum amount of information that can be transmitted is determined based on the channel conditions, and the actual amount of transmission is simultaneously limited to the maximum value and the amount of sensing information that the paired sensor can provide.

[0064] A bandwidth allocation optimization model is established with the goal of maximizing the perceived safety gain obtained by vehicles through communication.

[0065] By solving the bandwidth allocation optimization model, a real-time bandwidth and computing power allocation scheme is obtained, and high-resolution perception data is prioritized for high-risk vehicles, thereby completing dynamic resource matching.

[0066] A second aspect of the present invention provides an electronic device, including a memory and a processor, wherein the processor is configured to execute a program in the memory to implement the above-described method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification.

[0067] A third aspect of the present invention provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, is used to execute the above-described method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] By introducing an improved PET index and changes in kinetic energy before and after a vehicle collision, potential high-risk targets can be dynamically and accurately identified, enabling risk classification of the detected objects. Based on this, communication and sensing resources are allocated according to the target's risk level, prioritizing bandwidth and computing power support for high-risk targets, significantly improving the timeliness and accuracy of key target detection. Simultaneously, it avoids over-allocation of resources to low-risk targets, effectively reducing system resource waste and improving overall collaborative sensing efficiency. Furthermore, by integrating time distance and kinetic energy changes for risk assessment, compared to traditional single-index methods, it can more comprehensively reflect the severity of potential safety conflicts, enhancing the reliability of system perception decisions and safety assurance capabilities. Attached Figure Description

[0070] Figure 1 This is a general flowchart of the specific implementation of the present invention;

[0071] Figure 2 A comparison of the overall perception accuracy and safety gain at intersections for three different scenarios;

[0072] Figure 3 The results show the perception accuracy and safety gain of a single vehicle with 1, 3, and 5 sensors. Detailed Implementation

[0073] Overall, this invention discloses a dynamic allocation method for vehicle cooperative perception resources based on conflict risk quantification, aiming to improve the detection efficiency and traffic safety of key targets in intelligent transportation systems. The method first constructs a risk assessment model based on improved post-intrusion time (PET) and momentum changes before and after a collision between conflicting vehicles. This model comprehensively quantifies the potential collision risk between the vehicle and other vehicles within the region of interest, thereby determining the priority of detection targets. Based on this, a perception and communication model for multi-vehicle cooperation is established, dynamically allocating bandwidth and computing resources according to the detection priority of each target, achieving priority perception and communication assurance for high-risk targets. This method effectively avoids the averaging problem in resource allocation, improves the timeliness and accuracy of key target detection, and reduces the probability of potential accidents in high-risk scenarios, demonstrating good practical value and promising prospects for wider application.

[0074] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Component models, material names, connection structures, circuit structures, control methods, algorithms, and other features not explicitly described in this technical solution are considered common technical features disclosed in the prior art.

[0075] Example 1

[0076] Existing collaborative perception research generally neglects target detection priority, assuming all targets have the same detection weight. This easily leads to resource waste and lag in the identification of key targets, reducing perception efficiency and driving safety. To address this, this invention proposes a target detection priority definition method based on PET safety indicators and risk assessment indicators of momentum changes before and after collisions between conflicting vehicles. By quantifying the collision risk between the vehicle and other vehicles within the region of interest, its detection priority is determined. Then, based on the defined target detection priority, a collaborative perception communication model between vehicles is established.

[0077] See Figure 1 This embodiment includes the following core processes:

[0078] Dynamic information acquisition: After startup, it acquires dynamic parameters such as position, speed, acceleration, and direction of the vehicle and surrounding vehicles in real time;

[0079] Two-factor risk quantification: The potential collision time window is calculated by improving the PET value, and the collision energy intensity is analyzed based on momentum change at the same time, and the conflict risk level is generated by fusion.

[0080] Dynamic priority decision-making: All vehicles are sorted in descending order based on risk level, with high-risk targets receiving the highest detection priority;

[0081] Bandwidth constraint modeling: Establish collaborative communication rules under the priority framework, and integrate the perception task allocation mechanism with bandwidth upper limit constraints;

[0082] Precise resource allocation: The optimized algorithm allocates the maximum bandwidth share to high-priority vehicles to ensure real-time, high-resolution transmission of their perception data.

[0083] In specific implementation, the vehicle cooperative perception resource dynamic allocation method based on conflict risk quantification in this embodiment includes the following steps:

[0084] Step 1: In traditional conflict detection, vehicles are simplified to point masses, ignoring their actual physical dimensions, which can easily lead to errors in conflict area identification. To improve spatial accuracy, this technical solution abstracts the vehicle as a rectangular trajectory surface with width and sets a lateral safety distance. By combining the trajectory boundaries of the two vehicles, the actual conflict area, including the influence of vehicle dimensions, is accurately constructed.

[0085] Step 2: Based on the known conflict area, extract vehicle trajectory data and identify vehicles traveling sequentially within the same conflict area. By determining the time when the rear of the preceding vehicle completely leaves the conflict area and the time when the front of the following vehicle enters the area, calculate the Post Encroachment Time (PET) considering the size effect to determine whether a spatiotemporal conflict exists.

[0086] Step 3: For the identified conflict events, the principles of momentum conservation and kinetic energy change are further introduced. Combining the vehicle's mass, speed, and relative motion direction, the total change in the system's kinetic energy before and after the collision is calculated. This value reflects the degree of energy release that may occur during the conflict and can serve as a physical indicator of the conflict's severity.

[0087] Step 4: Finally, establish a collaborative perception and communication model. By integrating multi-source perception information from vehicle-mounted terminals and roadside units, efficient sharing of vehicle status and environmental elements is achieved. Based on this, and considering communication resource constraints, a bandwidth allocation optimization model is constructed to dynamically adjust the communication resource allocation strategy for different vehicles or areas. This maximizes overall perception accuracy and conflict risk detection capabilities, thereby comprehensively improving traffic safety at intersections.

[0088] During implementation, step 1 specifically includes the following steps:

[0089] Step 11: First, considering vehicle driving safety, the lateral width occupied during actual vehicle operation includes not only the vehicle body width but also the safety distance on both sides of the vehicle body. Therefore, this invention utilizes the lateral safety distance of the vehicle body to improve the vehicle's driving trajectory, thereby improving the definition of the vehicle conflict zone. Considering that a safety distance d should be reserved on one side of the vehicle body... s The actual width d occupied by vehicles for passage o Defined as vehicle width d w Including the safety distances reserved on both sides, as shown in formula (1):

[0090] d o =d w +2·d s (1)

[0091] Step 12: To determine the conflict zone between vehicles, it is necessary to first determine the vehicle trajectories. Given the vehicle's entrance and exit lanes, a trajectory for each vehicle within the intersection can be determined. Considering the width of the vehicles, the trajectory will not be a line, but a surface. Modeling the vehicle trajectories within the intersection yields the parametric equations for the outer and inner boundaries of each trajectory surface. Simultaneously establish the equations for the outer boundaries F of the two vehicle trajectories. r,1 ,F r,2 and inner boundary f r,1 ,f r,2 The parametric equations are used to find several intersection points of the outer and inner boundaries of the two vehicle trajectories. The intersection points are the vertices of the conflict area of ​​the two vehicles. The set of vertices is denoted by Q, where Q = {q1,q2,q3,q4}. The area enclosed by the vertices is the conflict area of ​​the two vehicles.

[0092] During implementation, step 2 specifically includes the following steps:

[0093] Step 21: Calculate the time it takes for the vehicle to enter and exit the conflict zone. First, it is necessary to identify the order in which the vehicle and the target vehicle enter the conflict zone, defining the preceding vehicle i and the following vehicle j. Then, identify the point at which the preceding vehicle leaves the conflict zone. and the point where the following vehicle entered the conflict zone To calculate the time a vehicle enters and leaves the conflict zone, the distance traveled by the following vehicle j from its current position to the conflict zone must first be determined, based on the time it enters and leaves the grid. and the distance that the vehicle in front i has moved from its current position away from the conflict zone. Therefore, the time it takes for the following vehicle j to enter the conflict zone can be calculated. The time it takes for the car in front to leave the conflict zone As shown in formulas (2)-(3) respectively. Where, u i ,u j Let i and j represent the speeds of vehicle i and vehicle j, respectively.

[0094]

[0095] Step 22: Calculate the PET between the two vehicles. The improved PET takes into account the actual physical dimensions of the vehicles and provides a more detailed description of the time proximity between the two vehicles. If vehicle i enters the conflict zone before vehicle j, use σ. i,j =1 indicates that the PET index between the preceding vehicle i and the following vehicle j is represented by δ. i,jThis means that σ is equal to the time it takes for the front of vehicle j to enter the conflict zone minus the time it takes for the rear of vehicle i to exit the conflict zone, as shown in formula (4). If vehicle j is the front vehicle, then σ j,i =1, the PET index is calculated as shown in formula (5).

[0096]

[0097] Where, σ i,j For a bivariate parameter, σ i,j =1 indicates that vehicle i entered the intersection before vehicle j, σ j,i =1 indicates that vehicle j enters the intersection before vehicle i.

[0098] In practice, step 3 includes the following steps:

[0099] Step 31: Determine the vehicle's driving angle in the conflict zone. The vehicle's trajectory surface consists of numerous trajectory lines, which are the paths traversed by a certain point on the vehicle body. The center trajectory line is the path traversed by the vehicle's center point. When determining the vehicle's driving angle in the conflict zone, this invention uses the entry and exit points of the vehicle's center trajectory line within that conflict zone for calculation. Based on the vehicle's current position and driving direction, the vehicle's trajectory equation is calculated:

[0100] (1) If trajectory r is a straight line, and vehicle i is traveling straight from the east entrance, based on the vehicle's current position (x i ,y i The equations for the vehicle's inner and outer boundaries and center trajectory are:

[0101]

[0102] If vehicle i follows a straight trajectory from the west entrance, the equations of the vehicle's inner and outer boundaries and center trajectory are:

[0103]

[0104] If vehicle i follows a straight trajectory from the south entrance, the equations of the vehicle's inner and outer boundaries and center trajectory are:

[0105]

[0106] If vehicle i follows a straight trajectory from the north entrance, the equations for the vehicle's inner and outer boundaries and center trajectory are:

[0107]

[0108] (2) If the trajectory r is a curve, we use the equation of an ellipse to represent the vehicle's trajectory. When vehicle i is a left-turning vehicle at the south entrance, according to the vehicle's current position (x... i ,y iGiven the center of the ellipse (0,0), the equations of the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0109]

[0110] When vehicle i is a vehicle turning left from the east entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (2n·d) r From the equations of the vehicle's inner and outer boundaries and center trajectory line, we obtain:

[0111]

[0112] When vehicle i is a vehicle turning left from the west entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (0, 2n·d) r The equations for the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0113]

[0114] When vehicle i is a vehicle turning left from the north entrance, according to the vehicle's current position (x) i ,y i ) and the center of the ellipse (2n·d) r ,2n·d r The equations for the vehicle's inner and outer boundaries and center trajectory are obtained as follows:

[0115]

[0116] Step 32: Based on the trajectory equations calculated in Step 31, the equations of the center trajectory of the current vehicle i and the inner and outer trajectory boundary equations of the conflicting vehicle j can be combined to obtain the entry and exit points of vehicle i in the conflict area. The entry point is... Exit point is Connect the entry point and the exit point to obtain the straight line Y. r =K r X+C r The slope K of the straight line r The calculation is shown in formula (14).

[0117]

[0118] straight line Y r The slope is K r Define the travel angle of trajectory r in the conflict zone. For the line Y r The angle between the x-axis and the x-axis. The calculation is shown in formula (15).

[0119]

[0120] Step 33: According to the law of conservation of momentum, the momentum of the two vehicles is conserved before and after the collision, as shown in formulas (16)-(17). First, determine the driving angle and speed of the two vehicles. Then, decompose the momentum values ​​of the two vehicles in the horizontal and vertical directions on the coordinate axis, and calculate the vector sum of the horizontal and vertical momentum of the two vehicles respectively, so as to obtain the horizontal and vertical momentum values ​​of the system after the collision. The total kinetic energy change caused by the collision of the two vehicles in the conflict area is shown in formulas (18)-(19):

[0121] m i u i sinθ i +m j u j sinθ j =(m i +m j )vsinθ i,j (16)

[0122] m i u i cosθ i +m j u j cosθ j =(m i +m j vcosθ i,j (17)

[0123]

[0124] In formulas (16)-(19), m i ,m j The masses of vehicle i and vehicle j are respectively; u i ,u j θ represents the speeds of vehicle i and vehicle j, respectively; i and θ j Vehicles i and j enter the conflict zone respectively. i,j Angle; △k i,j This indicates that vehicles i and j are in the conflict zone △ i,j Changes in total kinetic energy before and after the collision.

[0125] Step 34: Modeling the improved PET index and the change in total kinetic energy before and after the collision. This invention uses the ConflictIndex (CI) index to model the conflict risk between vehicles i and j. The CI index can simultaneously reflect the probability of a collision and the expected severity. If the two vehicles collide, the probability of the collision is estimated based on the improved PET between the two vehicles; that is, the smaller the PET, the greater the probability of a collision. The severity of the collision is evaluated based on the change in kinetic energy before and after the collision as the two vehicles pass through the conflict area together, i.e., Δk. i,j A larger CI value indicates a greater severity of the collision; therefore, a larger CI value indicates a greater risk of conflict between vehicles i and j, as shown in formula (20). This invention uses the CI value as a weight value for target detection priority. A larger weight value indicates that conflicting vehicles should be detected more frequently, and a higher level of perception accuracy is required.

[0126]

[0127] In formula (20), CI(i,j) represents the conflict index value between vehicle i and vehicle j; △k i,j δ represents the total change in kinetic energy of vehicles i and j before and after the collision in the conflict zone; i,j The PET value represents the time difference between the rear of vehicle i leaving the conflict zone and the front of vehicle j entering the conflict zone.

[0128] In practice, step 4 includes the following steps:

[0129] Step 41: Define the set of vehicles as I, i∈I, and the set of vehicles that each vehicle can communicate with as J. i ,j∈J i Maximum transmission bandwidth B i,j Communication bandwidth is allocated to different vehicle-to-vehicle (V2V) pairs. It's important to note that only networked vehicles can perform cooperative sensing communication. Let the total available bandwidth at the intersection be B, and the communication bandwidth allocated between networked vehicle i and networked vehicle j be denoted as Bi. i,j All communication pairs share the total bandwidth B, as shown in equation (27):

[0130] ∑ i∈I ∑ j∈I B i,j ≤B (27)

[0131] Next, the amount of perceived information transmitted by each communication pair is calculated using equations (28)–(30). Equation (28) limits the amount of information q transmitted between vehicle i and vehicle j. i,j The maximum transmission rate must not be exceeded. According to Shannon's theorem, the maximum transmission rate for V2V (vehicle-to-vehicle) communication is given by the right side of the equation. Equation (29) indicates that the amount of information transmitted cannot exceed the maximum amount of sensed information that can be obtained after communicating with all paired sensors.

[0132]

[0133] In equation (28), G represents the sensor's transmit power. i,j This represents the channel power gain of the V2V communication pair. The channel gain of all communication pairs can be pre-calculated based on the vehicle and sensor locations and used as input parameters to reduce the real-time computational burden on the model. σ 2 This indicates the noise power at the receiving end.

[0134] Step 42: Perceived safety gain Ω for each vehicle i i As shown in formulas (31)-(32). Ω i,j p represents the perceived safety gain for vehicle j after vehicle i has communicated with all paired vehicles. i,j This represents the maximum amount of information that vehicles i and j can sense:

[0135]

[0136]

[0137] Finally, with the goal of maximizing the safety gain obtained by the vehicle, a cooperative perception and communication model is established, and the objective function is as follows:

[0138] max∑ i∈I Ω i (33)

[0139] Verification Example 1

[0140] A six-lane, two-way intersection in Nanjing, Jiangsu Province, was selected as a case study to verify the effectiveness of the proposed model. To reflect the actual traffic flow characteristics, 20 frames of vehicle distribution maps inside the intersection were extracted from surveillance videos during peak and off-peak hours as the basis for sensor placement. The intersection was modeled in a Cartesian coordinate system with the lower left corner as the origin, the west entrance straight-through direction as the x-axis, and the south entrance straight-through direction as the y-axis. The intersection area was discretized into 144 square grids with a side length of 1.75m. To simplify the trajectory analysis, elliptic curves were used to fit left-turning vehicles from the south (S), north (N), west (W), and east (E) directions, with the ellipse centers set to (0,0), (21,21), (0,21), and (21,0) respectively. The lengths of the major and minor semi-axes were determined by the distance from the center to the corresponding entrance / exit lane.

[0141] To verify the importance of considering the priority of perception message transmission in cooperative perception, we compared cooperative perception models that considered and did not consider information transmission priority. Random pairing was used when information transmission priority was not considered. Furthermore, to verify the necessity of using the intersection safety index (CI) as the target to guide inter-vehicle information transmission in this technical solution, we also compared an information transmission model that aimed to maximize safety with one that aimed to maximize perception accuracy. The intersection safety perception gain and perception accuracy of the three schemes are as follows: Figure 2 As shown.

[0142] Depend on Figure 2 It can be seen that (1) the PSG value obtained by the information transmission model with the goal of maximizing safety is about 19.06% and 32.56% higher than the other two schemes on average. This shows the effectiveness of the technical scheme model in improving intersection safety. (2) The information transmission model with the goal of maximizing perception accuracy can achieve the highest perception accuracy, which is about 3.82% and 16.53% higher than the other two models. However, compared with the model that maximizes safety, its perception accuracy only increases by 3.82%, while the safety gain decreases by about 32.56%. This shows that simply pursuing perception accuracy while ignoring safety benefits will lead to a significant loss of safety performance and is not conducive to the robust operation of the traffic system. (3) The optimized communication model has significant advantages in both PSG and perception accuracy, with an average minimum improvement of 11.28% and 12.00% respectively, further verifying the superiority of the model proposed in this technical solution in terms of collaborative perception and information sharing. (4) Compared with the method of not deploying roadside sensors and relying only on vehicle perception and V2V communication, the model of this technical solution has an additional improvement of 11.48%-26.63% in PSG value, indicating that reasonable deployment of sensors plays an irreplaceable and important role in enhancing intersection safety. In summary, the information transmission and sensor deployment joint optimization model constructed in this study can provide an efficient and reliable technical solution for traffic intersection safety management while taking into account both perception accuracy and safety gain.

[0143] Figure 3 The comparison of safety metrics and sensing accuracy under three schemes with 1, 3, and 5 sensors is further demonstrated. Figure 3 As shown in (a), (b), and (c), the average SPG value acquired by each vehicle gradually decreases, demonstrating the effectiveness of the proposed technical solution model. Furthermore, the significant difference between the maximum and minimum safety values ​​acquired per vehicle indicates that the model effectively considers the heterogeneity of traffic conflict risks for each vehicle when optimizing information transmission, thereby achieving more refined safety assurance. When the number of sensors is 3 and 5, the average safety values ​​acquired per vehicle are similar. This is because vehicles have an upper limit to their perception accuracy; simply increasing the number of roadside sensors without considering the actual needs of vehicles would be a waste of resources. Figure 3 Figures (d), (e), and (f) show that, regardless of whether the safety-first or accuracy-first model is used, the average perception accuracy of vehicles is similar, but the safety gains differ significantly due to the different optimization objectives. This indicates that the communication objective determines the level of safety benefits at the intersection. Under the random pairing scheme, both perception accuracy and safety gain are the lowest, indicating low bandwidth allocation efficiency and resource waste.

[0144] Example 2

[0145] This embodiment provides an electronic device, including a memory and a processor. The processor executes a program in the memory to implement the vehicle cooperative perception resource dynamic allocation method based on conflict risk quantification as described above. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The memory may include Random Access Memory (RAM) and may also include Non-Volatile Memory, such as at least one disk storage device. The memory can be an internal memory of the Random Access Memory (RAM) type. The processor and memory can be integrated into one or more independent circuits or hardware, such as an Application Specific Integrated Circuit (ASIC). It should be noted that when the computer program in the aforementioned memory 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, or the part that contributes to the prior art, or a 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 several instructions to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention.

[0146] Example 3

[0147] This embodiment provides a storage medium containing computer-executable instructions. When executed by a computer processor, the storage medium with computer-executable instructions is used to perform the vehicle cooperative perception resource dynamic allocation method based on conflict risk quantification as described above. The storage medium can be an electronic medium, magnetic medium, optical medium, electromagnetic medium, infrared medium, or semiconductor system or propagation medium. The storage medium may also include semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory (RAM), read-only memory (ROM), hard disk, and optical disk. Optical disks may include optical disc-read-only memory (CD-ROM), optical disc-read / write (CD-RW), and DVD.

[0148] The above description of the embodiments is provided to enable those skilled in the art to understand and use the invention. It will be apparent to those skilled in the art that various modifications can be made to these embodiments, and the general principles described herein can be applied to other embodiments without inventive effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the invention should be within the protection scope of the present invention.

Claims

1. A method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification, characterized in that, Includes the following steps: A trajectory surface model is constructed based on the actual dimensions of the vehicles and the lateral safety distance, and the trajectories of the two vehicles are combined to determine the conflict area; Based on the temporal relationship between the rear of the preceding vehicle and the front of the following vehicle within the conflict area, the improved intrusion time is calculated to identify potential conflict events. Assess the severity of collision risk based on changes in the momentum and kinetic energy of vehicles during a conflict. The intrusion time and the changes in momentum and kinetic energy are integrated into a conflict risk index, and the detection priority of each vehicle is determined accordingly. Construct a collaborative perception and communication model, and dynamically allocate communication bandwidth and perception resources according to detection priority, so as to prioritize the perception needs of high-risk vehicles. Based on the temporal relationship between the rear of the preceding vehicle and the front of the following vehicle within the conflict area, the specific process for calculating the improved intrusion time to identify potential conflict events includes: The order in which vehicles entered the conflict zone determines the order of the preceding and following vehicles. Calculate the travel time of the rear vehicle's front end from its current position to the entrance of the conflict zone, and the travel time of the front vehicle's rear end from its current position to the exit of the conflict zone; The improved intrusion time is obtained by subtracting the time when the rear of the preceding vehicle completely leaves the conflict zone from the time when the front of the following vehicle enters the conflict zone. If the improved intrusion time is positive and less than the set threshold, then a potential conflict event is determined to exist; The specific process for assessing the severity of collision risk based on changes in the momentum and kinetic energy of vehicles during a conflict includes: The mass and velocity of each pair of conflicting vehicles are obtained separately, and the momentum is decomposed into lateral and longitudinal components based on the driving angle of each conflicting vehicle entering the conflict zone. The resultant momentum of the system after the collision of the conflicting vehicles is calculated based on the principle of conservation of momentum, and the change in the total kinetic energy of the system before and after the collision is obtained accordingly. This change in kinetic energy is used as a quantitative indicator of the severity of the collision. The specific process of integrating the improved intrusion time and kinetic energy change into a conflict risk index, and determining the detection priority of each vehicle accordingly, includes: The improved intrusion time is converted into the probability of collision occurrence, where the shorter the improved intrusion time, the higher the probability. The change in kinetic energy is converted into the severity of the collision, where a larger change in kinetic energy corresponds to a higher severity. The obtained collision probability and collision severity are combined to form a conflict risk indicator; Based on the aforementioned conflict risk indicators, all potential conflict vehicle pairs are sorted in descending order to form a detection priority sequence for resource allocation.

2. The method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification according to claim 1, characterized in that, The specific process of constructing a trajectory surface model based on the actual vehicle dimensions and lateral safety distance, and establishing the trajectory boundaries of the two vehicles to determine the conflict area includes: The vehicle passage width is defined as the sum of the vehicle body width and the safety distances on both sides, forming a trajectory surface model; Based on the trajectory surface model, corresponding trajectory surface inner and outer boundary parameter equations are generated according to the vehicle's inlet and outlet lanes; By simultaneously solving the outer and inner boundary equations of the trajectory surfaces of the two vehicles, the set of intersection points can be obtained. The set of intersection points forms a closed region, which serves as the actual conflict area between the two vehicles.

3. The method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification according to claim 2, characterized in that, When the vehicle's trajectory is a curve, the trajectory surface model is constructed using an ellipse equation. The construction process specifically includes: Determine the coordinates of the ellipse center based on the vehicle's import direction: Vehicles turning left at the south entrance have the origin of the coordinate system as the center of the ellipse. For vehicles turning left at the east entrance, the center of the ellipse is a point whose horizontal coordinate is an integer multiple of the lane radius and whose vertical coordinate is 0. For vehicles turning left at the west entrance, the center of the ellipse is a point with an abscissa of 0 and a ordinate offset from the lane radius that is an integer multiple of the lane radius. For vehicles turning left from the north entrance, the center of the ellipse is a point whose horizontal and vertical coordinates are both offset from the lane radius by an integer multiple; Based on the vehicle's center position and actual occupied width, elliptical equations are established for the outer boundary, inner boundary, and center trajectory line, respectively.

4. The method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification according to claim 1, characterized in that, The calculation process for the change in the total kinetic energy of the system before and after the collision includes: Based on the driving angle, momentum is decomposed into two mutually perpendicular directional components; Calculate the sum of the momentum vectors of the two vehicles in the lateral and longitudinal directions before the collision; According to the law of conservation of momentum, the magnitude of the velocity vector of the system after the collision can be deduced from the resultant momentum in the horizontal and vertical directions. The degree of energy release is quantified by comparing the difference between the total kinetic energy before the collision and the system's kinetic energy after the collision.

5. The method for dynamic allocation of vehicle cooperative perception resources based on conflict risk quantification according to claim 1, characterized in that, The specific process of constructing a collaborative perception and communication model, and dynamically allocating communication bandwidth and perception resources based on detection priority to prioritize the perception needs of high-risk vehicles, includes: Using the detection priority sequence as input, a communication topology is generated by treating connected vehicles and their communicable vehicles as nodes; Under the constraint of total bandwidth at the intersection, bandwidth allocation is directly and positively correlated with priority; the higher the priority, the larger the share of bandwidth allocated. For each communication pair, the maximum amount of information that can be transmitted is determined based on the channel conditions, and the actual amount of transmission is simultaneously limited to the maximum amount of information that can be transmitted and the amount of sensing information that the paired sensor can provide. A bandwidth allocation optimization model is established with the goal of maximizing the perceived safety gain obtained by vehicles through communication. By solving the bandwidth allocation optimization model, a real-time bandwidth and computing power allocation scheme is obtained, and high-resolution perception data is prioritized for high-risk vehicles, thereby completing dynamic resource matching.

6. An electronic device, comprising a memory and a processor, characterized in that, The processor is used to execute the program in the memory to implement the vehicle cooperative perception resource dynamic allocation method based on conflict risk quantification as described in any one of claims 1 to 5.

7. A storage medium containing computer-executable instructions, characterized in that, When executed by a computer processor, the storage medium of the computer-executable instructions is used to execute the vehicle cooperative perception resource dynamic allocation method based on conflict risk quantification as described in any one of claims 1 to 5.

Images