Intelligent networked automobile-oriented interchange exit ramp risk active prevention and control method

Abstract

The invention relates to an interchange exit ramp risk active prevention and control method for intelligent networked automobiles, which comprises a video monitoring unit, a road side unit (RSU) and an edge calculation unit, the edge calculation unit generates a graded prevention and control instruction by sensing an abnormal parking event of an exit ramp in real time, the RSU sends a customized control instruction to the intelligent networked automobiles at different positions, and the intelligent networked automobiles are controlled by the edge calculation unit. And issuing an instruction by adopting a point-to-point or broadcast communication mode according to the position of the vehicle in the road network, and triggering the emergency response of the intelligent networked vehicle. According to the method, the problem of response lag of traditional manual prevention and control is solved through vehicle-road cooperative accurate control, the risk active prevention and control of the interchange exit ramp is realized, and a secondary accident can be effectively prevented.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation and vehicle networking technology, specifically relating to a proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles. Background Technology

[0002] Interchanges are crucial hubs in highway networks. Their exit ramp areas experience reduced lanes, altered road alignments, and require vehicles to transition from high to low speeds, making them accident-prone zones. In these areas, vehicle malfunctions, accidents, or other abnormal stops can easily trigger serious rear-end collisions and secondary accidents, posing a significant threat to road safety and traffic efficiency.

[0003] To address the aforementioned risks, traditional prevention and control methods primarily rely on the observation and reaction of human drivers. When an anomaly is detected ahead, the driver must undergo a complex process of "perception-cognition-decision-execution," and their reaction time is affected by factors such as driving experience, fatigue, and distraction, exhibiting significant uncertainty and lag. Especially at high speeds, a reaction delay of just a few seconds can make an accident unavoidable. Furthermore, drivers of following vehicles often fail to notice stationary vehicles on ramps in time due to the terrain, further increasing the risk of accidents. With the development of intelligent connected vehicle technology, vehicles possess vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication capabilities, as well as assisted driving and autonomous driving systems. This provides a new technological path for proactive prevention and control of traffic accidents. However, existing technologies mostly focus on safety technologies for individual vehicles or roadside information prompts, lacking a systematic solution that can integrate roadside perception, real-time edge computing, and vehicle-to-infrastructure cooperative control to provide holistic, hierarchical, and precise risk prevention and control for ramp areas. How to utilize roadside facilities to detect risks in real time and provide differentiated control strategies based on the precise location of vehicles in the road network, thereby forming a closed loop of "perception-decision-control", is a technical problem that urgently needs to be solved. Summary of the Invention

[0004] To address the aforementioned issues, the present invention aims to provide a proactive risk prevention and control method for interchange exit ramps of intelligent connected vehicles. By constructing an integrated vehicle-road cooperative system encompassing "perception-assessment-decision-control," the method achieves rapid and accurate perception and hierarchical proactive prevention and control of abnormal parking events within the ramps. This effectively solves the problem of delayed response in traditional manual prevention and control, and significantly improves the traffic safety level of the ramp area.

[0005] To achieve the above objectives, the technical solution adopted by the present invention includes:

[0006] A proactive risk prevention and control method for interchange exit ramps of intelligent connected vehicles includes the following steps: S1, a video surveillance unit A is installed at the entrance of the interchange exit ramp, and also includes a roadside unit RSU and an edge computing unit B installed on the same pole as the video surveillance unit A. The roadside unit RSU communicates with the vehicle. S2, the vehicle sends a Basic Safety Message (BSM) to the Roadside Unit (RSU) within the communication range and subscribes to the service. After the RSU parses the Basic Safety Message (BSM), it adds the vehicle ID to the monitoring list and synchronizes it to the Edge Computing Unit (B). The Basic Safety Message (BSM) includes at least the vehicle's latitude and longitude, heading angle, speed, acceleration, and vehicle size. S3, video monitoring unit A collects video data of vehicles added to the monitoring list and sends it to edge computing unit B. After decoding the video data, edge computing unit B obtains the vehicle bounding box, the coordinates of the bounding box center point, and the motion trajectory through a vehicle tracking algorithm; the video data covers the main line and ramps. S4. Based on the vehicle bounding box, the coordinates of the bounding box center point and the motion trajectory, determine whether the displacement change D of the vehicle bounding box center point within the time threshold Δt is less than the preset threshold ε. The edge calculation unit B then determines whether the vehicle is abnormally parked. S5, after determining that the vehicle is abnormally parked, the edge computing unit B analyzes the remaining effective passage space of the current ramp and outputs the passage assessment result of "not passable" or "partially passable". S6, Edge computing unit B generates differentiated control strategies and determines communication methods based on traffic assessment results and the real-time location of abnormally parked vehicles; S7, the edge computing unit B sends the control policy to the target vehicle via the roadside unit (RSU), triggering the target vehicle to automatically respond and report to the road monitoring and control center; the target vehicle is other vehicles within the communication range of the RSU, including vehicles on the main line, vehicles that have entered the deceleration lane, and vehicles that have entered the ramp. The target vehicle's automatic response is the onboard system performing safe operations according to the control instructions, including at least stopping, changing lanes, and replanning the route.

[0007] Preferably, the formula for calculating the threshold ε in S4 is:

[0008] in, and These are the width and height pixel values ​​of the vehicle bounding box in the current frame, respectively; and Here are the pixel coordinates of the center point of the bounding box; H and W represent the height and width of the image captured by video surveillance unit A, respectively; k is the basic scaling factor, ranging from 0.5 to 0.7; α is the vertical position influence coefficient, ranging from 0.2 to 0.4; β is the horizontal position influence coefficient, ranging from 0.1 to 0.2.

[0009] Preferably, the analysis of the remaining effective passage space of the current ramp in S5 is as follows: S5.1 Based on the current image frame collected by video monitoring unit A, identify and segment the lane areas of the ramp, and number each lane. S5.2, obtain the bounding box pixel coordinates of all vehicles determined to be abnormally parked by S4; S5.3 For each abnormally parked vehicle, determine the spatial relationship between its bounding box and the lane lines identified in S5.1; if the abnormally parked vehicle is completely located in a single lane, then determine that lane is impassable; if the abnormally parked vehicle occupies multiple lanes at the same time, then execute S5.4. S5.4 Calculate the area ratio occupied by the abnormally parked vehicle in each lane it occupies. If the area ratio of any lane exceeds a preset threshold, the lane is determined to be impassable. S5.5 integrates the traffic status determination results of all lanes and all abnormal vehicles. If at least one lane is passable, the evaluation result of "partially passable" is output; otherwise, the evaluation result of "not passable" is output.

[0010] Preferably, the identification and segmentation of each lane area of ​​the ramp in S5.1 specifically includes: S5.11 performs grayscale conversion and edge detection on the image frame, and extracts the contour pixel set of the left solid line, the middle dividing line and the right solid line; S5.12 uses a curve fitting algorithm to fit the contour pixel set of each lane line, generating a boundary function representing the boundary of each lane. = This yields the lane areas of each lane on the ramp; among which... Given pixel coordinates, the fitting coefficients are solved using the least squares method. .

[0011] Preferably, in S5.4, the calculation of the area ratio occupied by the abnormally parked vehicle in each lane it encroaches upon is based on a pixel area calculation method using numerical integration, including: S5.41, determine the integration interval based on the range of the vehicle bounding box's horizontal coordinates, and discretize it into multiple sub-intervals; S5.42, For each sub-interval, calculate the vertical overlap height between the vehicle profile function and the lane boundary function at that location; S5.43, multiply the overlap height of each sub-interval by the width of the sub-interval to obtain the area micro-element, and sum the area micro-element of all sub-intervals to obtain the total area occupied by the vehicle in the lane. S5.44, divide the total area occupied by the vehicle in the lane by the total area of ​​the complete lane area to obtain the proportion of area occupied by the abnormally parked vehicle in each lane it occupies.

[0012] Preferably, the vehicle contour function in S5.42 includes an upper boundary function V1(x) and a lower boundary function V2(x), which are obtained in the following way: based on the vertex coordinates of the bounding box of the abnormally parked vehicle, its upper boundary vertex and lower boundary vertex are determined, and the upper boundary function V1(x) and lower boundary function V2(x) are generated by linear interpolation, where x is the pixel horizontal coordinate.

[0013] Preferably, the lane boundary function in S5.42 includes the inner edge function of the right solid line. The right edge function of the middle separator line and left edge function and the inner edge function of the solid line on the left , where x is the x-coordinate of a pixel.

[0014] Preferably, the generation of differentiated control strategies and determination of communication methods in S6 specifically include: for vehicles whose real-time location is on the main line but have not entered the deceleration lane, a broadcast communication method is used; for vehicles that have entered the deceleration lane and have entered the ramp, a point-to-point communication method is used. Scenario 1: When the assessment result is "not passable": a) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "the ramp ahead is closed, exit at the next exit"; b) For vehicles that have entered the deceleration lane, the control strategy field content is "the current ramp is closed, route replanning"; c) For vehicles that have entered the ramp, the control strategy field content is "the vehicle ahead is stopped and blocking the lane, cannot pass, stop immediately". Scenario 2: When the assessment result is "partially passable": d) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "the lane ahead is occupied, proceed with caution"; e) For vehicles that have entered the deceleration lane, the control strategy field content is "the current ramp lane is occupied, proceed with caution"; f) For vehicles that have entered the ramp, the control strategy field content is "abnormal parking ahead, proceed with caution".

[0015] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles disclosed in this application.

[0016] A computer program product includes a computer program / instructions, which, when executed by a processor, implement the proactive risk prevention and control method for interchange exit ramps of intelligent connected vehicles disclosed in this application.

[0017] Compared with the prior art, the advantages of the present invention are: (1) The present invention provides an active risk prevention and control method for interchange exit ramps for intelligent connected vehicles. It actively detects risks through roadside sensing devices and provides warnings or control commands before or immediately after a vehicle enters a dangerous area, transforming passive response into active prevention and control. Through accurate assessment of traffic space and real-time tracking of vehicle position, it provides more accurate prevention and control strategies for vehicles in different positions. By utilizing the communication capabilities between intelligent connected vehicles and roadside facilities, control commands are directly sent to the vehicle control system, realizing a rapid automated response from perception to control. This completely overcomes the problems of slow reaction and long decision-making time in manual driving, greatly shortens the risk response time, and effectively avoids secondary accidents caused by abnormal parking on ramps. At the same time, through lane-level guidance, it improves regional traffic efficiency while ensuring safety.

[0018] (2) The proactive risk prevention and control method for interchange exit ramps of the present invention for intelligent connected vehicles creatively couples and analyzes the real-time vehicle location (mainline not yet in the deceleration lane, already in the deceleration lane, already in the ramp) with the ramp traffic space assessment results ("not passable", "partially passable"). Based on this two-dimensional decision matrix, the system dynamically generates control instructions of different levels from "suggest detour", "route replanning" to "proceed with caution" and "stop immediately"; it provides a graded and differentiated prevention and control strategy based on accurate location and traffic status, thereby improving the accuracy and safety of control.

[0019] (3) The proactive risk prevention and control method for interchange exit ramps of the present invention for intelligent connected vehicles constructs a vehicle-road cooperative automated closed loop of "perception-assessment-decision-control". The structured control commands generated by the edge computing unit B are directly and reliably sent to the OBU of the target vehicle via the RSU and trigger the automatic execution of the on-board system. This process forms an end-to-end automated link from roadside perception to vehicle execution, eliminating the uncertainty of manual intervention and improving the determinism and reliability of system response. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a scenario for an active risk prevention and control system and method for interchange exit ramps for intelligent connected vehicles; Figure 2 This is a graph showing the integral calculation and analysis of the area occupied by abnormally stopped vehicles.

[0021] Figure 3This is a flowchart of a proactive risk prevention and control method for interchange exit ramps in one embodiment of the present invention. Detailed Implementation

[0022] The invention is not limited to the specific embodiments described below. All equivalent modifications made based on the technical solutions of this application fall within the protection scope of this invention. Unless otherwise specified, all components and devices in this invention utilize components and devices known in the prior art.

[0023] Example This embodiment discloses a proactive risk prevention and control method for interchange exit ramps of intelligent connected vehicles, characterized by the following steps: S1, a video surveillance unit A is installed at the entrance of the interchange exit ramp, and also includes a roadside unit RSU and an edge computing unit B installed on the same pole as the video surveillance unit A. The roadside unit RSU communicates with the vehicle. The video surveillance unit in this embodiment has a coverage range of ≥500 meters, supports 1920×1080P resolution and a frame rate of 25 frames per second, and can transmit video data in real time via the RTSP protocol, accurately covering the entire ramp area; the roadside unit (RSU) has a coverage radius R ≥500 meters, and its transmission power is configured based on the IEEE 802.11p standard. By calculating the distance between the vehicle and the RSU, V2X communication is established when the distance is ≤ the coverage radius; the edge computing unit preferably has ≥200 TOPS AI computing power, supports hard decoding of video streams (latency ≤1ms / frame), is compatible with the TCP / IP protocol, and is equipped with a gigabit Ethernet interface and a 5G / Ethernet interface, enabling data interaction with the video surveillance unit and RSU, and meeting the needs of real-time decision-making and control strategy generation.

[0024] S2, the vehicle sends a basic safety message (BSM) to the roadside unit (RSU) within the communication range and subscribes to the service. After the roadside unit (RSU) parses the basic safety message (BSM), it adds the vehicle ID to the monitoring list and synchronizes it to the edge computing unit (B). The Basic Safety Message (BSM) should include at least the vehicle's latitude and longitude, heading angle, speed, acceleration, and vehicle dimensions. In this embodiment, the subscription service specifically involves the vehicle actively requesting to receive directional information related to proactive ramp risk prevention and control pushed by the RSU: After receiving the roadside information (RSI) broadcast by the RSU, the vehicle identifies the "ramp risk prevention and control service" identifier and, in conjunction with its own navigation plan (it is about to enter the ramp), initiates a subscription request; Within its effective communication range, the RSU periodically broadcasts Roadside Information (RSI) messages. Connected vehicles entering the RSU's communication range continuously send relevant information to the RSU, including Basic Safety Messages (BSMs). BSMs contain at least information such as vehicle latitude and longitude, heading angle, speed, acceleration, and vehicle size. Upon receiving an RSI message, a connected vehicle sends a service request message to the RSU to subscribe to the service. The RSU receives and parses the vehicle's BSM and service request messages, adds the vehicle IDs within its communication range to the monitoring list, and synchronously sends the monitoring list to the edge computing unit B.

[0025] S3, video monitoring unit A collects video data of vehicles added to the monitoring list and sends it to edge computing unit B. After decoding the video data, edge computing unit B obtains the vehicle bounding box, the coordinates of the bounding box center point, and the motion trajectory through a vehicle tracking algorithm; the video data covers the main line and ramps. In this embodiment, edge computing unit B decodes the video data to obtain a sequence of image frames per second, and performs vehicle target detection frame by frame. For each frame, the detection algorithm outputs the bounding box information of all vehicles. For example, if a vehicle is detected, the coordinates of the four vertices of its bounding box are (x, y, y). i y i (where i = 1, 2, 3, 4), the coordinates of the center point P are (500, 300), and the height h t =40 pixels, width w t =80 pixels. The vehicle tracking algorithm associates the same vehicle in consecutive frames, assigns and maintains a unique tracking ID for the vehicle (e.g., Track_05), and generates continuous motion trajectory information corresponding to the vehicle ID.

[0026] The vehicle tracking algorithm in this embodiment is a known technology, and can be found in CN118537635A, a detection-based multi-vehicle tracking method and system.

[0027] S4. The motion state of the vehicle is determined by analyzing the changes in the pixel positions of the vehicle in the image frame sequence. Specifically, based on the vehicle bounding box, the coordinates of the bounding box center point, and the motion trajectory, it is determined whether the displacement change D of the vehicle bounding box center point within the time threshold Δt is less than the preset threshold ε. The edge calculation unit B determines whether the vehicle has stopped abnormally. If the displacement change D of the vehicle's pixel position within the time threshold Δt is less than the preset threshold ε, it is determined that the vehicle has stopped abnormally. The abnormal stopping state is continuously tracked and confirmed based on subsequent image frame sequences to eliminate false judgments.

[0028] The formula for calculating the threshold ε is:

[0029] in, and These are the width and height pixel values ​​of the vehicle bounding box in the current frame, respectively; and Here are the pixel coordinates of the center point of the bounding box; H and W represent the height and width of the image captured by video surveillance unit A, respectively. t and w t This represents the height and width of the vehicle's bounding box; k is the basic scaling factor, ranging from 0.5 to 0.7; α is the vertical position influence factor, ranging from 0.2 to 0.4; β is the horizontal position influence factor, ranging from 0.1 to 0.2.

[0030] S5, after determining that the vehicle is abnormally parked, the edge computing unit B analyzes the remaining effective passage space of the current ramp and outputs a passage assessment result of "not passable" or "partially passable".

[0031] In this embodiment S5, the analysis of the remaining effective passage space of the current ramp is specifically as follows: S5.1, based on the current image frame acquired by video monitoring unit A, identifies and segments the lane areas of the ramp, and numbers each lane, such as L1, L2; the identification and segmentation of the lane areas of the ramp in S5.1 specifically includes: S5.11 performs grayscale conversion and edge detection on the image frame, and extracts the contour pixel set of the left solid line, the middle dividing line and the right solid line; S5.12 uses a curve fitting algorithm to fit the contour pixel set of each lane line, generating a boundary function representing the boundary of each lane. = This yields the lane areas of each lane on the ramp; in Given pixel coordinates, the fitting coefficients are solved using the least squares method. .

[0032] S5.2, obtain the bounding box pixel coordinates of all vehicles determined to be abnormally parked by S4; a. Lane line contour extraction: The system performs grayscale processing on the image frames, converting the color image to a grayscale image. At this point, the grayscale value of each pixel ranges from 0 to 255. Based on the significant grayscale difference between the lane lines and the road surface, a grayscale threshold range is set […]. , This is used to identify potential lane line pixels. The system employs an edge detection algorithm, calculating the grayscale gradient of pixels in different directions to accurately locate the lane line edges. For the middle dashed dividing line, discrete line segment pixels are connected into a complete contour point set through region connectivity analysis. The final output is the complete contour pixel set for the three lane lines: the left solid line, the middle dividing line, and the right solid line.

[0033] b. Lane line fitting: The system uses the least squares method to perform cubic curve fitting on the extracted lane line contour pixel set. For each lane line, the equation of the fitted curve is: = ,in These are pixel coordinates. The least squares method is used to fit the curve, and the fitting coefficients are obtained. .

[0034] S5.3 For each abnormally parked vehicle, determine the spatial relationship between its bounding box and the lane lines identified in S5.1; if the abnormally parked vehicle is completely located in a single lane, then determine that lane is impassable; if the abnormally parked vehicle occupies multiple lanes at the same time, then execute S5.4.

[0035] S5.4 Calculate the area ratio occupied by the abnormally parked vehicle in each lane it occupies. If the area ratio of any lane exceeds a preset threshold, the lane is determined to be impassable. S5.4 calculates the area ratio occupied by the abnormally parked vehicle in each lane it occupies using a pixel area calculation method based on numerical integration, including: S5.41, determining the integration interval based on the horizontal coordinate range of the vehicle bounding box, and discretizing it into multiple sub-intervals; S5.42, for each sub-interval, calculating the vertical overlap height between the vehicle contour function and the lane boundary function; S5.43, multiplying the overlap height of each sub-interval by the width of the sub-interval to obtain an area micro-element, and summing the area micro-element of all sub-intervals to obtain the total area occupied by the vehicle in that lane; S5.44, dividing the total area occupied by the vehicle in that lane by the total area of ​​the complete lane region of that lane to obtain the area ratio occupied by the abnormally parked vehicle in each lane it occupies.

[0036] The vehicle contour function in S5.42 of this embodiment includes an upper boundary function V1(x) and a lower boundary function V2(x), which are obtained as follows: based on the vertex coordinates of the bounding box of the abnormally parked vehicle, its upper and lower boundary vertices are determined, and the upper boundary function V1(x) and lower boundary function V2(x) are generated respectively through linear interpolation, where x is the pixel horizontal coordinate. The lane boundary function in S5.42 includes the inner edge function of the right solid line. The right edge function of the middle separator line and left edge function and the inner edge function of the solid line on the left , where x is the x-coordinate of a pixel.

[0037] This embodiment calculates the pixel area occupied by abnormal vehicles in each lane. The method uses numerical integration for calculation, and the specific process is as follows: a. Determine the integration interval and discretization: Use the minimum and maximum x-coordinates of the bounding box of the abnormal vehicle in the image coordinate system. min and x max The integration interval is defined by pixels as the step size Δx.

[0038] b. Area Calculation: For each sub-interval, its central x-coordinate is x. i =x min +(i+0.5) Δx; Calculate the vertical overlap height between the vehicle profile and the boundaries of each lane at this location: The effective height within the L1 lane is: max

[0039] The effective height within the L2 lane is: max

[0040] Then the infinitesimal areas occupied by this sub-interval within lanes L1 and L2 are respectively: and Vehicles in the lane and The total area occupied by the inner interval is obtained by summing the area infinitesimals of all subintervals: A_L1 = A_L2= .

[0041] This embodiment calculates the area ratio of each lane: Ratio_L i =A_L i / A_total_L i The total lane area is A_total_L. i The numerical integration method described above can also be used to calculate it within its complete contour. The system presets a dynamic adjustment threshold θ based on different positions of the ramp in the image frame. If Ratio_L i If Ratio_L ≥ θ, then the lane is deemed impassable. i If θ < 0, the ramp is deemed partially passable.

[0042] S5.5 integrates the traffic status determination results of all lanes and all abnormal vehicles. If at least one lane is passable, the evaluation result of "partially passable" is output; otherwise, the evaluation result of "not passable" is output.

[0043] S6, Edge computing unit B generates differentiated control strategies and determines communication methods based on traffic assessment results and the real-time location of abnormally parked vehicles.

[0044] In this embodiment S6, generating differentiated control strategies and determining communication methods specifically includes: for vehicles whose real-time location is on the main line but have not entered the deceleration lane, a broadcast communication method is used.

[0045] In this embodiment, the edge computing unit B encapsulates the control strategy and interaction information into a new structured safety message format—the Ramp Risk Control Message (RRCM), and sends it to the target vehicle via the RSU in a point-to-point or broadcast communication manner. RRCM messages are a V2I message format specifically designed for proactive risk prevention and control on ramps. Their message structure includes the following fields: 1) Message type field, identifying the RRCM message type; 2) Timestamp field, used to identify the precise time the message was generated; 3) Geographic range label field, used to define the geographic area range to which the message applies; 4) Event Type field, which identifies the specific type of abnormal parking event; 5) Event Risk Level field, which is used to dynamically assess the risk level based on the type of stopped vehicle and the status of the interacting vehicle; 6) Vehicle ID field, which is used to identify the target vehicle during point-to-point communication; 7) Abnormal Event Location field, which provides the precise coordinate information of the abnormal vehicle, represented by latitude and longitude coordinates; 8) Control Strategy field, which provides specific operation instructions for vehicle control.

[0046] For vehicles that have entered the deceleration lane or the ramp, point-to-point communication is used. Scenario 1: When the assessment result is "not passable": a) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "The ramp ahead is closed, exit at the next exit"; b) For vehicles that have entered the deceleration lane, the control strategy field content is "The current ramp is closed, route replanning"; c) For vehicles that have entered the ramp, the control strategy field content is "Vehicle ahead is stopped and blocking the lane, unable to pass, stop immediately". Scenario 2: When the assessment result is "partially passable": d) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "The lane ahead is occupied, proceed with caution"; e) For vehicles that have entered the deceleration lane, the control strategy field content is "The current lane ahead is occupied, proceed with caution"; f) For vehicles that have entered the ramp, the control strategy field content is "Abnormal parking ahead, proceed with caution".

[0047] S7, the edge computing unit B sends the control policy to the target vehicle via the roadside unit (RSU), triggering the target vehicle to automatically respond and report to the road monitoring and control center; the target vehicle is other vehicles within the communication range of the RSU, including vehicles on the main line, vehicles that have entered the deceleration lane, and vehicles that have entered the ramp. The target vehicle's automatic response is the onboard system performing safe operations according to the control instructions, including at least stopping, changing lanes, and replanning the route.

[0048] In this embodiment, the target vehicle automatically responds to avoid the risk of abnormal parking on the ramp. For example, if the traffic assessment result is "not passable"; the vehicle is located on the main road section; and the control instruction is "the ramp ahead is closed, exit at the next exit"; then the vehicle's response is "the onboard navigation system automatically cancels the ramp exit plan and replans the route."

[0049] This embodiment also discloses a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles disclosed in this application.

[0050] This embodiment also discloses a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, it implements the proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles disclosed in this application.

[0051] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0052] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0053] Furthermore, the various implementation methods disclosed in this solution can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content invented by this disclosure.

Claims

1. A proactive risk prevention and control method for interchange exit ramps of intelligent connected vehicles, characterized in that, Includes the following steps: S1, a video surveillance unit A is installed at the entrance of the interchange exit ramp, and a roadside unit RSU and an edge computing unit B are also installed on the same pole as the video surveillance unit A. The roadside unit RSU communicates with the vehicle. S2, the vehicle sends a basic safety message (BSM) to the roadside unit (RSU) within the communication range and subscribes to the service. After the RSU parses the basic safety message (BSM), it adds the vehicle ID to the monitoring list and synchronizes it to the edge computing unit (B). The Basic Safety Message (BSM) includes at least the vehicle's latitude and longitude, heading angle, speed, acceleration, and vehicle dimensions. S3, video monitoring unit A collects video data of vehicles added to the monitoring list and sends it to edge computing unit B. Edge computing unit B decodes the video data and obtains the vehicle bounding box, the coordinates of the bounding box center point and the motion trajectory through a vehicle tracking algorithm. The video data covers the main line and ramps; S4. Based on the vehicle bounding box, the coordinates of the bounding box center point and the motion trajectory, determine whether the displacement change D of the vehicle bounding box center point within the time threshold Δt is less than the preset threshold ε. The edge calculation unit B then determines whether the vehicle is abnormally parked. S5, after determining that the vehicle is abnormally parked, the edge computing unit B analyzes the remaining effective passage space of the current ramp and outputs a passage assessment result of "not passable" or "partially passable". S6, the edge computing unit B generates a differentiated control strategy and determines the communication method based on the traffic assessment results and the real-time location of abnormally parked vehicles; S7, Edge computing unit B sends the control policy to the target vehicle via roadside unit RSU, triggering the target vehicle to automatically respond and report to the road monitoring and control center; The target vehicle refers to other vehicles within the RSU communication range, including vehicles on the main line, vehicles already in the deceleration lane, and vehicles already on the ramp. The target vehicle's automatic response refers to the onboard system performing safe operations based on control instructions, including at least stopping, changing lanes, and route replanning.

2. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 1, characterized in that, The formula for calculating the threshold ε mentioned in S4 is: in, and These are the width and height pixel values ​​of the vehicle bounding box in the current frame, respectively; and Here are the pixel coordinates of the center point of the bounding box; H and W represent the height and width of the image captured by video surveillance unit A, respectively; k is the basic scaling factor, ranging from 0.5 to 0.7; α is the vertical position influence coefficient, ranging from 0.2 to 0.4; β is the horizontal position influence coefficient, ranging from 0.1 to 0.

2.

3. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 1, characterized in that, The analysis of the remaining effective passage space of the current ramp in S5 is as follows: S5.1 Based on the current image frame collected by video monitoring unit A, identify and segment the lane areas of the ramp, and number each lane. S5.2, obtain the bounding box pixel coordinates of all vehicles determined to be abnormally parked by S4; S5.3, for each abnormally parked vehicle, determine the spatial relationship between its bounding box and the lane lines identified in S5.1; If an abnormally parked vehicle is entirely within a single lane, that lane is deemed impassable. If an illegally parked vehicle occupies multiple lanes simultaneously, then S5.4 will be executed; S5.4 Calculate the area ratio occupied by the abnormally parked vehicle in each lane it occupies. If the area ratio of any lane exceeds a preset threshold, the lane is determined to be impassable. S5.5 integrates the traffic status determination results of all lanes and all abnormal vehicles. If at least one lane is passable, the evaluation result of "partially passable" is output; otherwise, the evaluation result of "not passable" is output.

4. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 3, characterized in that, The specific lane areas identified and segmented in S5.1 include: S5.11 performs grayscale conversion and edge detection on the image frame, and extracts the contour pixel set of the left solid line, the middle dividing line and the right solid line; S5.12 uses a curve fitting algorithm to fit the contour pixel set of each lane line, generating a boundary function representing the boundary of each lane. = This yields the lane areas of each lane on the ramp; in Given pixel coordinates, the fitting coefficients are solved using the least squares method. .

5. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 4, characterized in that, S5.4 calculates the proportion of area occupied by the abnormally parked vehicle in each lane it encroaches upon using a pixel area calculation method based on numerical integration, including: S5.41, determine the integration interval based on the range of the vehicle bounding box's horizontal coordinates, and discretize it into multiple sub-intervals; S5.42, For each sub-interval, calculate the vertical overlap height between the vehicle profile function and the lane boundary function at that location; S5.43, multiply the overlap height of each sub-interval by the width of the sub-interval to obtain the area micro-element, and sum the area micro-element of all sub-intervals to obtain the total area occupied by the vehicle in the lane; S5.44, divide the total area occupied by the vehicle in the lane by the total area of ​​the complete lane area to obtain the proportion of area occupied by the abnormally parked vehicle in each lane it occupies.

6. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 5, characterized in that, The vehicle profile function in S5.42 includes the upper boundary function V1(x) and the lower boundary function V2(x), which are obtained in the following way: Based on the vertex coordinates of the bounding box of the abnormally parked vehicle, the upper and lower boundary vertices are determined, and the upper boundary function V1(x) and lower boundary function V2(x) are generated by linear interpolation, where x is the pixel x-coordinate.

7. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in claim 5, characterized in that, The lane boundary function in S5.42 includes the inner edge function of the right solid line. The right edge function of the middle separator line and left edge function and the inner edge function of the solid line on the left , where x is the x-coordinate of a pixel.

8. The proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in any one of claims 1-7, characterized in that, The generation of differentiated control strategies and determination of communication methods in S6 specifically include: for vehicles whose real-time location is on the main line but have not entered the deceleration lane, a broadcast communication method is used; for vehicles that have entered the deceleration lane and have entered the ramp, a point-to-point communication method is used. Scenario 1: When the assessment result is "not passable": a) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "The ramp ahead is closed, exit at the next exit"; b) For vehicles that have already entered the deceleration lane, the control policy field content is "Current ramp closed, route replanning"; c) For vehicles that have already entered the ramp, the control policy field content is "Vehicle ahead is stopped and blocking the lane, unable to pass, stop immediately"; Scenario 2: When the assessment result is "partially passable": d) For vehicles that have not entered the deceleration lane on the main line, the control strategy field content is "The lane ahead is occupied, proceed with caution"; e) For vehicles that have already entered the deceleration lane, the control strategy field content is "Current ramp lane is occupied, proceed with caution"; f) For vehicles that have already entered the ramp, the control policy field content is "Abnormal parking ahead, proceed with caution".

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in any one of claims 1-7.

10. A computer program product, characterized in that, It includes a computer program / instruction, which, when executed by a processor, implements the proactive risk prevention and control method for interchange exit ramps for intelligent connected vehicles as described in any one of claims 1-7.

Images