An expressway special section active safety prevention and control method based on edge computing

Abstract

The application relates to an expressway special section active safety prevention and control method based on edge computing, which comprises the following steps: S1, collecting vehicle driving data and traffic accident data of an expressway; S2, judging whether an abnormal running state occurs based on the vehicle driving data and the traffic accident data: if the abnormal running state occurs, judging an accident type and an accident severity, introducing a result into a risk accumulation diffusion method, simultaneously judging whether the accident occurs in an accident-prone area in combination with an accident-prone point identification model, performing risk dynamic evaluation, obtaining an accident influence range model, and thus determining the accident influence range; if the abnormal running state does not occur, judging whether the accident dissipates: if the accident dissipates, returning to S1 to update data; and if the accident does not dissipate, returning to S1 to update data; S3, determining a release scheme according to the risk evaluation result, and releasing through each release end. Compared with the prior art, the application has the advantages of risk prediction accuracy and reliable release scheme.

Description

Technical Field

[0001] The present invention relates to the field of road safety control, and in particular to an active safety control method for special sections of highways based on edge computing. Background Art

[0002] The current road traffic safety situation in my country is still not optimistic. Further analysis of the environment in which traffic accidents occur reveals the following characteristics: (1) Special sections of highways, such as merge and diverge areas, tunnel entrances and exits, are areas where traffic accidents are more concentrated. Although their proportion of the total road mileage is relatively low, the number of accidents caused by rapid changes in vehicle speed and frequent lane changes is much higher than that of general sections. (2) Drivers have poor adaptability to special sections, resulting in the concentration of major and serious accidents in such areas with complex or sudden traffic environments and potential accident hazards. (3) Traffic management departments are still insufficient in their ability to respond to road traffic emergencies. Although many places in my country have deployed large-scale traffic information collection equipment, illegal photo-taking systems, real-time traffic information release systems and other modern traffic monitoring facilities, they are still weak in real-time analysis and judgment of emergencies, rapid decision-making and active guidance. Looking at the current situation, it can be found that one of the main reasons for the current contradictions in road traffic safety in my country is that drivers on the road cannot obtain road traffic and environmental information in the first place, and thus cannot make risk avoidance behaviors such as slowing down and changing lanes in advance. On the other hand, traffic management departments' processing of traffic data remains at the "process-driven, platform-regulated" stage, significantly reducing the timeliness of information release and the accuracy of warnings. Therefore, there is an urgent need to develop a roadside edge computing method that integrates information collection, analysis, and early warning. This method can obtain real-time vehicle status information on special road sections, conduct online analysis and prediction of road section risk levels, and provide drivers with proactive traffic safety intervention measures based on massive data. This can reduce the incidence of abnormal driving behavior on special road sections and ensure traffic safety on these sections. Summary of the Invention

[0003] The purpose of this invention is to provide an active safety prevention and control method for special sections of highways based on edge computing, which integrates information collection, analysis, and early warning. By real-time analysis of the risk level of special sections and matching multi-level release plans, the plan can better match the actual accident status and achieve effective and efficient protection of traffic safety in special sections.

[0004] The purpose of the present invention can be achieved by the following technical solutions:

[0005] An active safety control method for special sections of highways based on edge computing includes the following steps:

[0006] S1: The data acquisition module collects vehicle driving data and traffic accident data on the highway and sends them to the edge computing unit;

[0007] S2: The edge computing unit determines whether an abnormal running state occurs based on the vehicle running data and the traffic accident data:

[0008] If an abnormal running state occurs, the type and severity of the accident are determined, and the results are introduced into the risk accumulation and diffusion method, and the accident is determined to occur in an accident-prone area by combining the accident-prone point identification model to perform dynamic risk assessment to obtain an accident influence range model, thereby determining the accident influence range.

[0009] If an abnormal running state does not occur, it is determined whether the accident has dissipated. If the accident has dissipated, S3 is performed, and if the accident has not dissipated, the data is updated and returned to S1.

[0010] S3: The edge computing unit determines a release scheme based on the risk assessment results and sends the release scheme to each release end through a scheme release module for release.

[0011] The traffic accident data includes accident point coordinates, accident point to collection end distance, accident lane, accident type, and accident lane number.

[0012] The accident-prone point identification model matches the accident-prone point road section based on the vehicle's road section and determines the accident-prone point level of the vehicle's location based on the vehicle's coordinates, and adjusts the parameter values of the risk accumulation and diffusion method according to the accident-prone point level, wherein the accident-prone point level includes four levels: first-level accident-prone point, second-level accident-prone point, third-level accident-prone point, and non-accident-prone point.

[0013] The risk accumulation and diffusion method is established based on the Gaussian plume model, which is represented as:

[0014]

[0015] In the formula, C'(x', y', z', H') represents the concentration of toxic gas at point (x', y', z'); H' represents the diffusion source height; q' is the release rate of the diffusion source; u' is the average wind speed outside; σ' y , σ' z are the diffusion parameters in the horizontal and vertical directions, respectively; x' is the distance of the spatial point to the source in the wind direction axis; y' is the distance of the spatial point to the source in the vertical direction of the wind direction axis; and z' is the height of the spatial point.

[0016] The influencing factors of the Gaussian plume model are combined with the influencing factors of the accident influence range to replace the parameters of the Gaussian plume model to obtain an accident influence range model suitable for special road sections of expressways:

[0017]

[0018] Among them, X represents the accident impact range of a special section of the highway, ξ represents the traffic diffusion ratio of the accident point, which is determined by the accident type, P represents the potential energy of the accident source point, a represents the influencing parameter of the accident occupation ratio, and C d It represents the degree of traffic impact at the location with the largest accident impact, b represents the adjustment parameter, and h is the parameter describing the level of accident-prone points.

[0019] The parameter h describing the accident-prone point level is initially set to 1;

[0020] When the accident-prone point is a level 1 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.8 times the initial value;

[0021] When the accident-prone point is a level 2 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.5 times the initial value;

[0022] When the accident-prone point level is a level 3 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.2 times the initial value;

[0023] When the accident-prone point level is non-accident-prone point, the parameter h in the accident impact range model remains unchanged at the initial value.

[0024] The types of accidents mentioned include spilled objects, continuous lane changes, speeding, driving out of the designated lane, pedestrian intrusion, illegal parking, driving against traffic, and vehicle collisions;

[0025] When the accident type is spilled objects, the value of the diffusion traffic ratio ξ at the accident point is 0.25;

[0026] When the accident type is continuous lane change, speeding, or driving in the wrong lane, the value of the diffusion traffic ratio ξ at the accident point is 0.5;

[0027] When the accident type is pedestrian intrusion, illegal parking or wrong-way driving, the value of the diffusion traffic ratio ξ at the accident point is 0.75;

[0028] When the accident type is a vehicle collision, the value of the accident point diffusion traffic ratio ξ is 1.

[0029] When the distance between the accident point and the entrance and exit is less than or equal to 5 meters, the potential energy P of the accident source point is adjusted to twice the initial value;

[0030] When the distance between the accident point and the entrance and exit is more than 5 meters but less than or equal to 10 meters, the potential energy P of the accident source point is adjusted to 1.5 times the initial value;

[0031] When the distance between the accident point and the entrance and exit is more than 10 meters but less than or equal to 20 meters, the potential energy P of the accident source point is adjusted to 1.2 times the initial value;

[0032] When the distance between the accident point and the entrance is more than 20 meters, the potential energy P of the accident source point remains unchanged.

[0033] The accident influence range model divides the output into a plurality of regions with different influence degrees by adjusting the parameter b.

[0034] The S3 comprises the following steps:

[0035] S31: calculating the coordinates of each risk section: according to the accident influence range and the position of the accident point, the position of the accident point is subtracted from each accident influence range to obtain the coordinates of each risk section, and the relative positions of the coordinates of each risk section are obtained by comparing the coordinates of each risk section with the coordinates of each release end;

[0036] S32: according to the coordinates of the risk section and the coordinates of the release end, a hierarchical classification release scheme is formulated, and the coverage range of the scheme release end covers all release ends within 1km upstream of the section where the farthest influence range is located downstream of the accident occurrence;

[0037] S33: after obtaining the grade of each release end scheme, it is matched with the reserved release scheme, wherein the reserved release scheme includes the release terminal and the release information content used by each type of accident at different levels, and the release information content reserves key fields that need to be filled with information received from the edge computing unit, the key fields include lane and license plate information, and the complete release scheme, i.e., the terminal and the specific information of the release scheme used by each release end, is obtained by combining the grade of each release end scheme and the reserved release scheme;

[0038] S34: each release end releases the corresponding scheme.

[0039] The release end is realized by different combination and setting modes of various information release devices arranged on the roadside, and the information release devices include variable information boards, roadside broadcasts, variable speed limit signs and lane indicator lights.

[0040] Compared with the prior art, the present application has the following beneficial effects:

[0041] (1) The present application starts from the type of traffic accident, determines the influence range upstream of the special section of the expressway after the accident occurs, and improves the traditional Gaussian smoke plume model in combination with the characteristics of the special section of the expressway and the accident-prone point, so as to determine the diffusion law of the accident influence range in the special section, and according to the difference of the influence radiation range caused by the position of the accident, the road state and other factors, a traffic accident influence diffusion model for the special section of the expressway is constructed, and a range division method of the hierarchical influence region is proposed, which can provide reliable and useful judgment basis for the risk release scheme.

[0042] (2) The risk release scheme of the present invention is dynamically adjusted according to actual conditions, covers a wide range of influencing factors, and can adapt to various situations.

[0043] (3) The release method of the release scheme of the present invention takes ergonomics into consideration, can achieve a safe, efficient and comfortable intervention effect, and the scheme is highly feasible.

[0044] (4) The present invention opens up data connectivity between the collection end and the publishing end through the connection between the collection end, the edge computing unit and the publishing end, which has a fast computing speed, timely solution publishing, and improved rapid response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of the method of the present invention;

[0046] Figure 2 A flow chart of the method for the accident impact range model;

[0047] Figure 3 A method flow chart for the program release process;

[0048] Figure 4 A schematic diagram showing the classification of release schemes. DETAILED DESCRIPTION

[0049] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0050] This embodiment first provides an active safety prevention and control system for special sections of highways based on edge computing, including:

[0051] (1) Data acquisition module

[0052] Smart cameras and radars detect vehicle driving data and traffic accident data, and transmit all collected data to the edge computing unit.

[0053] (2) Edge computing unit

[0054] The edge computing unit is used to build an accident impact range model to determine the accident impact range. The release plan is then determined based on the accident impact range, the accident location, and the location of the release terminal.

[0055] (3) Solution release module

[0056] After obtaining the relevant results of the warning information, the edge computing unit transmits the calculation results to the control platform, which determines whether to issue the relevant warning information and which devices should issue it. If the warning information needs to be issued, the release end is determined according to the release plan. For example, lane-level control includes devices such as variable speed limit signs, lane lights, and road studs, while section control includes devices such as variable information boards, roadside broadcasts, and water curtains.

[0057] Secondly, this embodiment also provides an active safety control method for special sections of highways based on edge computing, such as Figure 1 As shown, the following steps are included:

[0058] S1: The data acquisition module collects vehicle driving data and traffic accident data on the highway and sends them to the edge computing unit.

[0059] In this embodiment, traffic accident data includes the coordinates of the accident point, the distance from the accident point to the collection end, the lane where the accident occurred, the type of accident, and the number of lanes occupied by the accident; vehicle driving data includes traffic information such as vehicle speed, license plate number, and the distance between the vehicle and the monitoring section.

[0060] S2: If Figure 2 As shown, the edge computing unit determines whether an abnormal operating state occurs based on vehicle driving data and traffic accident data:

[0061] If an abnormal operating state occurs, the type and severity of the accident are determined, and the results are introduced into the risk accumulation and diffusion method. At the same time, the accident frequent point identification model is combined to determine whether the accident occurred in an accident-prone area. A dynamic risk assessment is performed to obtain an accident impact range model, thereby determining the accident impact range;

[0062] If no abnormal operating state occurs, determine whether the accident has dissipated. If the accident has dissipated, go to S3. If the accident has not dissipated, return to S1 to update the data.

[0063] This method identifies the impact range upstream of specific highway sections after an accident, based on the type of traffic accident. By integrating the characteristics of these sections with high-accident locations, the traditional Gaussian plume model is improved to determine the diffusion pattern of the impact range within these sections. Based on the differential impact of factors such as accident location and road conditions on the impact radius, a model for the impact range of specific highway sections is constructed, and a method for categorizing the impact range into hierarchical levels is proposed.

[0064] As a toxic gas diffusion model, the Gaussian plume model is often used to study the diffusion laws of toxic gases. When constructing a Gaussian plume model, it is necessary to consider various factors, such as diffusion properties, light intensity, ambient temperature, and external wind force. The Gaussian plume model is expressed as:

[0065]

[0066] Where: C'(x', y', z', H') represents the concentration of toxic gas at point (x', y', z'); H' represents the height of the diffusion source; q' is the release rate of the diffusion source; u' is the average wind speed outside; σ' y ,σ′ z are the diffusion parameters in the horizontal and vertical directions respectively; x′ is the distance from the spatial point to the source on the wind direction axis; y′ is the distance from the spatial point to the source in the direction perpendicular to the wind direction axis; and z′ is the height of the spatial point.

[0067] Previous research has successfully applied plume models to predict traffic impact areas and port hinterland areas, demonstrating the accuracy and practicality of three-dimensional plume models even in low-dimensional scenarios. Theoretically, the propagation of the impact of accidents on specific sections of highways within a road network is comparable to the diffusion of toxic gases. For example, the impact of an accident on upstream traffic is similarly confined to a limited space and is influenced by factors such as the location and type of accident.

[0068] The above analysis shows that the application of the Gaussian smoke-rain model to define the impact range of traffic accidents on special sections of highways is feasible and has practical value.

[0069] When constructing a Gaussian plume model, it is necessary to consider various factors, such as diffusion properties, light intensity, ambient temperature, and external wind force. These parameters are combined with factors affecting the accident range and adjusted appropriately. The adjusted parameter variables are used to calculate the accident impact range and convert them to specific values. After parameter substitution, an accident impact range model suitable for special sections of highways is obtained:

[0070]

[0071] Where X represents the accident impact range of a special section of the highway, ξ represents the diffusion traffic ratio of the accident point (the value range is between 0 and 1), which is determined by the accident type, P represents the potential energy of the accident source point, a represents the influencing parameter of the accident road occupation ratio, and C d Indicates the degree of traffic impact at the largest area affected by the accident (C d close to 0, which can be calculated based on actual data), b represents the adjustment parameter, and h is the parameter describing the level of accident-prone points.

[0072] The accident-prone spot identification model matches the vehicle's road section with the most likely accident-prone road section. It also determines the vehicle's accident-prone spot level based on the vehicle's coordinates. The parameter h in the risk accumulation and diffusion method is adjusted based on the accident-prone spot level. In this embodiment, the accident-prone spot levels include level 1, level 2, level 3, and level 4.

[0073] In this embodiment, the parameter h describing the level of accident-prone locations is initially set to 1;

[0074] When the accident-prone point is a level 1 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.8 times the initial value;

[0075] When the accident-prone point is a level 2 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.5 times the initial value;

[0076] When the accident-prone point level is a level 3 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.2 times the initial value;

[0077] When the accident-prone point level is non-accident-prone point, the parameter h in the accident impact range model remains unchanged at the initial value.

[0078] In this embodiment, accident types include spilled objects, continuous lane changes, speeding, driving in the wrong lane, pedestrian intrusion, illegal parking, driving the wrong way, and vehicle collision. Alternatively, spilled objects, continuous lane changes, speeding, driving in the wrong lane, pedestrian intrusion, illegal parking, and driving the wrong way can be defined as traffic incidents, while vehicle collisions can be defined as traffic accidents. These different definitions do not affect the essence of this method.

[0079] When the accident type is spilled objects, the value of the diffusion traffic ratio ξ at the accident point is 0.25;

[0080] When the accident type is continuous lane change, speeding, or driving in the wrong lane, the value of the diffusion traffic ratio ξ at the accident point is 0.5;

[0081] When the accident type is pedestrian intrusion, illegal parking or wrong-way driving, the value of the diffusion traffic ratio ξ at the accident point is 0.75;

[0082] When the accident type is a vehicle collision, the value of the accident point diffusion traffic ratio ξ is 1.

[0083] Then, the weight of parameter P is updated by determining whether the accident point is located at a high-accident location or at the entrance or exit of the merge / diverge area. Whether the entrance or exit of the merge / diverge area is blocked is determined by the distance from the accident point to the entrance or exit and the lane it is in:

[0084] When the distance between the accident point and the entrance and exit is less than or equal to 5 meters, the potential energy P of the accident source point is adjusted to twice the initial value;

[0085] When the distance between the accident point and the entrance and exit is more than 5 meters but less than or equal to 10 meters, the potential energy P of the accident source point is adjusted to 1.5 times the initial value;

[0086] When the distance between the accident point and the entrance-exit is more than 10 meters but less than or equal to 20 meters, the point potential P of the accident source is adjusted to 1.2 times of the initial value;

[0087] When the distance between the accident point and the entrance-exit is more than 20 meters, the point potential P of the accident source remains unchanged.

[0088] The accident influence range model divides the output into multiple regions with different influence degrees by adjusting the parameter b. In this embodiment, the output of the accident influence range model is a first-level, a second-level and a third-level influence range.

[0089] S3: As shown in Figure 3 The edge computing unit determines a publishing scheme according to the risk assessment result, and sends the publishing scheme to each publishing terminal through a scheme publishing module for publishing.

[0090] S31: Calculate the coordinates of each risk section: according to the accident influence range and the position of the accident point, subtract the position of the accident point from each accident influence range to obtain the coordinates of each risk section, and compare the coordinates of each risk section with the coordinates of each publishing terminal to obtain the relative position of each risk section;

[0091] S32: According to the coordinates of the risk section and the coordinates of the publishing terminal, a classified publishing scheme is formulated, and the coverage range of the scheme publishing terminal covers all publishing terminals within 1 km upstream of the section where the farthest influence range is located downstream of the accident occurrence position;

[0092] S33: After obtaining the scheme level of each publishing terminal, it is matched with the reserved publishing scheme, wherein the reserved publishing scheme includes the publishing terminal and the publishing information content used by each type of accident at different levels, and the publishing information content reserves key fields that need to be filled with information received from the edge computing unit, including lane and license plate information. The complete publishing scheme, i.e., the terminal and specific information of the publishing scheme used by each publishing terminal, is obtained by combining the scheme level of each publishing terminal and the reserved publishing scheme;

[0093] S34: Publish the corresponding scheme through each publishing terminal.

[0094] This embodiment starts from the human factor theory, builds two special road section simulation scenes through simulation driving experiments, and sets up information publishing devices such as variable message boards, roadside broadcasts, variable speed limit signs and lane indicator lights in different combinations and settings on the roadside, tests the acceptance degree of drivers to the publishing information content, publishing means and combination mode of the roadside unit and the intervention effect on driving behavior under different abnormal operating states, determines the optimal information publishing mode under different special road section types and abnormal operating states, and formulates an active safety information publishing standard suitable for special road sections of expressways.

[0095] Specifically, this embodiment sets the release plan content into the following categories based on the output results of different accident impact range models:

[0096] (1) Five-level release plan

[0097] like Figure 4 As shown, under this release plan category, the release plan is divided into five levels according to the three-level impact area, namely:

[0098] 1. Scattered objects

[0099] ① Level 1

[0100] · Message board: "There is spillage in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0101] Lane indicator: Close the lane where the incident occurred

[0102] ·Spiked: Red light

[0103] ② Level 2

[0104] · Message board: "There is spillage in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0105] Lane indicator: Close the lane where the incident occurred

[0106] Road studs: red and flashing

[0107] ③Level 3

[0108] · Message board: "There is spillage in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0109] Lane indicator: Close the lane where the incident occurred

[0110] ·Spiked: Yellow light

[0111] ④ Level 4

[0112] · Message board: "There is spillage in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0113] Lane indicator: Close the lane where the incident occurred

[0114] Road studs: flashing yellow light

[0115] ⑤Level 5

[0116] · Message board: "There is spillage in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0117] 2. Pedestrian intrusion

[0118] Level 1

[0119] · Information board: "Pedestrian crossing in {distance} meters, please drive carefully."

[0120] Variable speed limit: All lanes speed limit value is the current speed limit minus 20 km / h

[0121] Broadcast: "You have entered the highway, please leave quickly for your safety"

[0122] Lane indicator light

[0123] Level 2, Level 3, Level 4

[0124] · Information board: "Pedestrian crossing in {distance} meters, please drive carefully."

[0125] Variable speed limit: All lanes speed limit value is the current speed limit minus 20 km / h

[0126] Broadcast: "You have entered the highway, please leave quickly for your safety"

[0127] Lane indicator light

[0128] Level 5

[0129] · Information board: "Pedestrian crossing in {distance} meters, please drive carefully."

[0130] 3. Reverse

[0131] Level 1

[0132] · Information board: "Vehicle reverse in {distance} kilometers {lane} lane, please drive carefully."

[0133] Lane indicator light: Close the lane where the event occurs

[0134] Studs: Red light

[0135] Variable speed limit: The speed limit value of the adjacent lane is the current speed limit minus 20 km / h

[0136] Broadcast: "{plate}, you have reversed, please drive to the emergency lane"

[0137] Level 2

[0138] · Information board: "Vehicle reverse in {distance} kilometers {lane} lane, please drive carefully."

[0139] Lane indicator light: Close the lane where the event occurs

[0140] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0141] Road studs: red and flashing

[0142] ③Level 3

[0143] · Information board: "There is a vehicle traveling the wrong way in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0144] Lane indicator: Close the lane where the incident occurred

[0145] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0146] ·Spiked: Yellow light

[0147] ④ Level 4

[0148] · Information board: "There is a vehicle traveling the wrong way in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0149] Lane indicator: Close the lane where the incident occurred

[0150] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0151] Road studs: flashing yellow light

[0152] ⑤Level 5

[0153] Information board: "There is a vehicle traveling the wrong way in {distance} kilometers ahead, {lane} lane. Please drive with caution." 4. Illegal parking

[0154] ① Level 1

[0155] · Information board: "There is illegal parking in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0156] Lane indicator: Close the lane where the incident occurred

[0157] ·Spiked: Red light

[0158] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0159] Water curtain: only used at tunnel entrances and exits, projecting “stop”

[0160] ② Level 2

[0161] · Information board: "There is illegal parking in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0162] Lane indicator: Close the lane where the incident occurred

[0163] Road studs: red and flashing

[0164] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0165] ③Level 3

[0166] · Information board: "There is illegal parking in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0167] Lane indicator: Close the lane where the incident occurred

[0168] ·Speed ​​spikes: yellow light

[0169] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0170] ④ Level 4

[0171] · Information board: "There is illegal parking in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0172] Lane indicator: Close the lane where the incident occurred

[0173] Road studs: yellow and flashing

[0174] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0175] ⑤Level 5

[0176] · Information board: "There is illegal parking in {distance} kilometers ahead, {lane} lane. Please drive with caution."

[0177] Lane indicator: Close the lane where the incident occurred

[0178] 5. Vehicle collision

[0179] ① Level 1

[0180] · Information board: "Account accident in {lane} lane ahead {distance} kilometers. Please drive with caution."

[0181] Lane indicator: Close the lane where the incident occurred

[0182] Road spike: red light

[0183] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0184] Announcement: "An accident has occurred on the current road section. Please drive carefully."

[0185] Water curtain: only used at tunnel entrances and exits, projecting “stop”

[0186] ② Level 2

[0187] · Information board: "Account accident in {lane} lane ahead {distance} kilometers. Please drive with caution."

[0188] Lane indicator: Close the lane where the incident occurred

[0189] Road studs: red and flashing

[0190] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0191] ③Level 3

[0192] · Information board: "Account accident in {lane} lane ahead {distance} kilometers. Please drive with caution."

[0193] Lane indicator: Close the lane where the incident occurred

[0194] ·Spiked: Yellow light

[0195] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0196] ④ Level 4

[0197] · Information board: "Account accident in {lane} lane ahead {distance} kilometers. Please drive with caution."

[0198] Lane indicator: Close the lane where the incident occurred

[0199] Road studs: flashing yellow light

[0200] Variable speed limit: The speed limit of the adjacent lane is the current speed limit minus 20 km / h

[0201] ⑤Level 5

[0202] Information board: "Accounting for an accident in {lane} lane, {distance} kilometers ahead. Please drive with caution." *Field description: {distance} is the distance from the current information board to the accident / incident.

[0203] {lane} is the lane where the accident / incident occurred;

[0204] {plate} is the license plate number

[0205] (2) First-level release plan

[0206] Under this release plan, this is a more urgent situation and only includes the first-level release plan.

[0207] 1. Speeding

[0208] Information board: "{plate}, you are speeding, please slow down"

[0209] Road spikes: glow red and flash

[0210] 2. Not driving in the designated lane

[0211] Information board: '{plate}, please drive in the designated lane'

[0212] Road spikes: glow red and flash

[0213] 3. Continuous Lane Change

[0214] Information board: '{plate}, please do not change lanes continuously. For your safety, please drive carefully'

[0215] Road spikes: glow red and flash

[0216] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A method for active safety control of special sections of highways based on edge computing, characterized in that: The following steps are involved: S1: The data acquisition module collects vehicle driving data and traffic accident data on the highway and sends them to the edge computing unit; S2: The edge computing unit determines whether an abnormal operating state occurs based on vehicle driving data and traffic accident data: If an abnormal operating state occurs, the type and severity of the accident are determined, and the results are introduced into the risk accumulation and diffusion method. At the same time, the accident frequent point identification model is combined to determine whether the accident occurred in an accident-prone area. A dynamic risk assessment is performed to obtain an accident impact range model, thereby determining the accident impact range; If no abnormal operating state occurs, determine whether the accident has dissipated. If the accident has dissipated, go to S3. If the accident has not dissipated, return to S1 to update the data. S3: The edge computing unit determines the release plan based on the risk assessment results and sends the release plan to each publishing end through the plan release module for release; The traffic accident data includes the coordinates of the accident point, the distance from the accident point to the collection end, the lane where the accident occurred, the type of accident, and the number of lanes occupied by the accident; The accident-prone point identification model matches the accident-prone point section based on the vehicle's road section, and determines the accident-prone point level of the vehicle's location based on the vehicle's coordinates. The parameter values ​​of the risk accumulation and diffusion method are adjusted according to the accident-prone point level, wherein the accident-prone point level includes four levels: level 1 accident-prone point, level 2 accident-prone point, level 3 accident-prone point, and non-accident-prone point. The risk accumulation and diffusion method is established based on the Gaussian plume model, which is expressed as: Where: C'(x', y', z', H') represents the concentration of toxic gas at point (x', y', z'); H' represents the height of the diffusion source; q' is the release rate of the diffusion source; u' is the average wind speed outside; σ y ′,σ z ' is the diffusion parameter in the horizontal and vertical directions respectively; x' is the distance from the spatial point to the source on the wind direction axis; y' is the distance from the spatial point to the source in the direction perpendicular to the wind direction axis; z' is the height of the spatial point; By combining the influencing factors of the Gaussian plume model with the influencing factors of the accident impact range and replacing the parameters of the Gaussian plume model, we can obtain the accident impact range model applicable to special sections of highways: Among them, X represents the accident impact range of a special section of the highway, ξ represents the traffic diffusion ratio of the accident point, which is determined by the accident type, P represents the potential energy of the accident source point, a represents the influencing parameter of the accident occupation ratio, and C d It represents the degree of traffic impact at the location with the largest accident impact, b represents the adjustment parameter, and h is the parameter describing the level of accident-prone points.

2. The method for active safety control of special sections of highways based on edge computing according to claim 1 is characterized in that: The parameter h describing the accident-prone point level is initially set to 1; When the accident-prone point is a level 1 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.8 times the initial value; When the accident-prone point is a level 2 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.5 times the initial value; When the accident-prone point level is a level 3 accident-prone point, the parameter h in the accident impact range model is adjusted to 1.2 times the initial value; When the accident-prone point level is non-accident-prone point, the parameter h in the accident impact range model remains unchanged at the initial value.

3. The method for active safety control of special sections of highways based on edge computing according to claim 1 is characterized in that: The types of accidents mentioned include spilled objects, continuous lane changes, speeding, driving out of the designated lane, pedestrian intrusion, illegal parking, driving against traffic, and vehicle collisions; When the accident type is spilled objects, the value of the diffusion traffic ratio ξ at the accident point is 0.25; When the accident type is continuous lane change, speeding, or driving in the wrong lane, the value of the diffusion traffic ratio ξ at the accident point is 0.5; When the accident type is pedestrian intrusion, illegal parking or wrong-way driving, the value of the diffusion traffic ratio ξ at the accident point is 0.75; When the accident type is a vehicle collision, the value of the accident point diffusion traffic ratio ξ is 1.

4. The method for active safety control of special sections of highways based on edge computing according to claim 1 is characterized in that: When the distance between the accident point and the entrance and exit is less than or equal to 5 meters, the potential energy P of the accident source point is adjusted to twice the initial value; When the distance between the accident point and the entrance and exit is more than 5 meters but less than or equal to 10 meters, the potential energy P of the accident source point is adjusted to 1.5 times the initial value; When the distance between the accident point and the entrance and exit is more than 10 meters but less than or equal to 20 meters, the potential energy P of the accident source point is adjusted to 1.2 times the initial value; When the distance between the accident point and the entrance and exit is more than 20 meters, the potential energy P of the accident source point remains unchanged at the initial value.

5. The method for active safety control of special sections of highways based on edge computing according to claim 1 is characterized in that: The accident impact range model divides the output into multiple areas with different impact levels by adjusting the parameter b.

6. The method for active safety control of special highway sections based on edge computing according to claim 1 is characterized in that: The S3 includes the following steps: S31: Calculate the coordinates of each risk section: Based on the accident impact range and the location of the accident point, subtract the accident point location from the accident impact range to obtain the risk section coordinates, and compare the risk section coordinates with the coordinates of each publishing end to obtain the relative position of each risk section coordinate; S32: Develop a hierarchical and classified release plan based on the risk section coordinates and the release end coordinates. The release end coverage of the plan covers all release ends within 1 km upstream of the section closest to the downstream of the accident site and the farthest section in the impact range. S33: After obtaining the solution level of each publishing terminal, match it with the reserved publishing solution. The reserved publishing solution includes the publishing terminals and publishing information content used for different accident levels. The publishing information content reserves key fields that need to be filled with information received from the edge computing unit. The key fields include lane and license plate information. Combining the solution level of each publishing terminal and the reserved publishing solution, a complete publishing solution is obtained, that is, the terminal used by each publishing terminal and the specific information of the publishing solution. S34: Publish the corresponding solution through each publishing terminal.

7. The method for active safety control of special highway sections based on edge computing according to claim 1 is characterized in that: The publishing end is realized by various information publishing devices arranged on the roadside through different combinations and settings. The information publishing equipment includes variable information boards, roadside broadcasts, variable speed limit signs and lane indicator lights.

Images