Expressway emergency response blocking device, system and method

Abstract

The invention relates to the technical field of traffic safety management, in particular to an expressway emergency response blocking device, system and method, which are arranged on a preset road section, and comprises the following steps: a road detection system monitors data of key indexes of vehicles and road conditions in real time; the remote control system receives the data, transmitted by the data transmission system, of the road detection system, judges the road condition according to the data, activates corresponding emergency measures automatically or through manual instructions according to the road condition, and dispatches and guides starting and stopping of the warning system and the physical interception system; the guiding and warning system is used for guiding and warning; when the physical interception system is activated, an entity or virtual barrier is formed to block a road; the road section information feedback system carries out information transmission with an external user system and information transmission between the expressway emergency response blocking devices of different road sections. According to the scheme, the safety and the emergency response efficiency of the expressway are greatly improved, the accident rate is reduced, and secondary accidents are minimized.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic safety management, and in particular to a highway emergency response blocking device, system and method. Background Art

[0002] As a transportation hub in modern society, the safe operation of highways is of vital importance. However, with the intensification of climate change and the complexity and changeability of the geographical environment, highways are facing more and more unexpected safety challenges. Traffic accidents and road collapse not only endanger driving safety, but also pose a threat to public safety, which has aroused social concern. Especially in areas with complex terrain and harsh climatic conditions, the operation safety of highways is more prominent.

[0003] With the development of social economy and the increase in the number of vehicles, the traditional highway monitoring and emergency response mechanism can no longer fully meet the needs of road safety management. The lag in information transmission will cause a large number of vehicles to be stranded near the accident site, which not only aggravates the congestion at the accident site, but also brings major challenges to subsequent road repair and rescue work.

[0004] Therefore, how to monitor and respond quickly to emergencies in real time, especially when emergencies occur in dangerous sections, it is very necessary to have devices that can intercept and warn in time. This ensures that when major safety incidents such as the Meida Expressway occur, we can respond quickly, provide accurate information to the public in a timely manner, effectively guide traffic, and minimize the impact on social public safety and traffic order. This has become an urgent problem that needs to be solved. Summary of the invention

[0005] In response to the above problems, one of the purposes of the present invention is to provide a highway emergency response blocking device that integrates advanced monitoring and control technologies, can realize real-time monitoring of dangerous sections of highways, and automatically or manually activate the guidance warning system when abnormal conditions are detected, and through the comprehensive use of physical and virtual blocking barriers, provide drivers with immediate and effective warning information, so as to greatly improve the safety and emergency response efficiency of highways, reduce the accident rate, and minimize secondary accidents.

[0006] The present invention provides a basic solution 1: a highway emergency response blocking device, which is arranged at a preset section of the road, comprising:

[0007] A highway detection system is used to monitor the data of key indicators of vehicle and road conditions in real time; the highway detection system extracts features from the data, and performs anomaly analysis based on the features to obtain data on abnormal vehicle behavior; the feature extraction uses a neural network model, and dynamically sets convolution parameters for feature extraction based on the functional relationship between the preset data and the convolution parameters in the neural network model;

[0008] A data transmission system, used for data transmission;

[0009] A remote control system is used to receive data monitored by the highway detection system transmitted by the data transmission system, judge the highway conditions based on the data, and activate corresponding emergency measures automatically or through manual instructions according to the highway conditions, and dispatch the opening and closing of the warning system and the physical interception system;

[0010] A guidance warning system, used for providing guidance warnings;

[0011] Physical blocking systems, which, when activated, form physical or virtual barriers to close the road;

[0012] The road section information feedback system is used to transmit information between external user systems and between emergency response blocking devices on different sections of highways.

[0013] Further, it also includes: a door frame perpendicular to the ground;

[0014] The highway detection system comprises:

[0015] Video image acquisition module, including: several video monitoring devices and external lights;

[0016] The video monitoring equipment is distributed along the gantry crossbeam, and the external lights are set above the two sides of the video monitoring equipment to provide light sources for the video monitoring equipment when the video monitoring equipment collects video image information such as road conditions, vehicle behavior and traffic flow in real time;

[0017] Intelligent video analysis module, used to receive real-time video image information transmitted by the video image acquisition module, and use image processing and artificial intelligence algorithms to perform in-depth analysis on the video image information;

[0018] The road surface condition detection module adopts a non-contact road surface detector. The non-contact road surface detector faces the road surface and is used to detect road icing, snow accumulation, and collapse.

[0019] Furthermore, the data transmission system comprises:

[0020] A wireless transmission module is used to receive data from the highway detection system and send the data wirelessly to the remote control system;

[0021] Internet transmission module, used to assist the wireless transmission module to perform remote data transmission via the Internet infrastructure;

[0022] The remote control system comprises:

[0023] A real-time monitoring module is used to receive data from the highway detection system transmitted by the data transmission system and display it on the monitoring interface, and is also used to generate real-time visual content based on the data;

[0024] The automatic dispatching module is used to receive the data transmitted by the real-time monitoring module, and when the data reaches the predetermined crisis condition, generate a dispatching decision and automatically activate the guidance warning system;

[0025] The manual scheduling module is used to input decisions and override the decisions of the automatic scheduling module with the input decisions;

[0026] A control module, used to manage and adjust the operating status of equipment in the guidance warning system and the physical interception system according to scheduling;

[0027] The data storage module is used to store all data transmitted from the highway detection system and the operation records of the remote control system.

[0028] Furthermore, the guidance and warning system comprises:

[0029] LED display screens are installed on both sides of the road, with a preset distance between each LED display screen, used to display text and image warning information on the electronic screen;

[0030] A voice announcement unit, used to issue a voice warning when the guidance warning system is activated;

[0031] A flashing light unit, which is used to emit a strong light warning when the guidance warning system is activated;

[0032] The physical interception system comprises:

[0033] physical barriers to stop vehicles;

[0034] A motor drive module for lowering the physical rail from a suspended state to the ground;

[0035] The laser emission module is installed at the center of the gantry beam and is started simultaneously with the motor drive module to generate a virtual wall curtain on the road through a powerful laser;

[0036] The road section information feedback system comprises:

[0037] The system interconnection module establishes real-time communication between the road information feedback systems of different road sections through network connection;

[0038] The user interconnection module is used to provide road condition information to external user systems in real time and receive road condition feedback from external user systems.

[0039] Further, it also includes: physical blocking system;

[0040] The physical blocking system includes: a blocking net, a hydraulic system and a control unit. The blocking net is suspended above the road surface in an inactive state, and the hydraulic system is installed on both sides of the gantry. The blocking net is connected to the hydraulic system through steel wire ropes at both ends; the control unit is used to identify abnormalities or emergency situations, and quickly lower the blocking net to the level of the physical railing through hydraulic control, and lock it at a fixed height not lower than a preset distance from the ground;

[0041] When a vehicle enters the warning area away from the physical barrier, the video image acquisition module captures images, identifies the type and speed of the vehicle, and transmits the identified information to the control unit for data analysis and processing to obtain the initial state of the preset barrier net. At the same time, sensors are embedded in the physical barrier to obtain the impact force when the vehicle hits the physical barrier, and dynamically adjust the blocking force of the barrier net.

[0042] Furthermore, when a vehicle enters the warning area away from the physical barrier, the video image acquisition module acquires images, identifies the type and speed of the vehicle, and transmits the identified information to the control unit for data analysis and processing, obtains the initial state of the preset blocking net, and embeds a sensor in the physical barrier to obtain the impact force when the vehicle hits the physical barrier, and dynamically adjusts the blocking force of the blocking net, including:

[0043] Real-time data collection steps:

[0044] Video surveillance data acquisition: When a vehicle enters the warning area, the video image acquisition module collects the image of the vehicle and uses image recognition technology to determine the type and speed of the vehicle;

[0045] Sensor data acquisition: Strain gauges embedded in the physical guardrail detect the impact force when the vehicle collides with the guardrail;

[0046] Initial state preset steps:

[0047] The control unit receives the information recognized by the video image acquisition module and adjusts the length of the steel wires on both sides of the barrier net in advance according to the stress characteristics of the barrier net:

[0048] E k =1 / 2mv 2 ;

[0049] Among them, E k is the vehicle kinetic energy, m is the vehicle mass, and v is the vehicle speed;

[0050] L=(E k +T×s) / (E×A);

[0051] Among them, L is the length of the steel wire that needs to be preset and adjusted in order to effectively buffer the impact force, T is the actual tension of the barrier net, s is the distance of deformation of the barrier net corresponding to the actual tension, E is the elastic modulus of the steel wire, and A is the cross-sectional area of ​​the steel wire;

[0052] Dynamic stopping force adjustment steps:

[0053] Data analysis: The control unit analyses the data measured by the strain gauges and calculates the impact force of the colliding vehicle and the stopping force that needs to be adjusted:

[0054] ε=R0×GF / ΔR;

[0055] Among them, ΔR is the change in resistance, R0 is the resistance value before deformation, GF is the sensitivity factor of the strain gauge, and ε is the strain;

[0056] σ=E×ε;

[0057] Among them, σ is stress and E is elastic modulus;

[0058] F=Σ(σ i ×A i );

[0059] Among them, F is the impact force, and A is the action area corresponding to the strain gauge;

[0060] ΔL=(FE k -T×s)×(1-α) / (E×A) (6)

[0061] Among them, ΔL is the length of the steel wire that needs to be changed, and α is the adjustment coefficient;

[0062] Hydraulic system action: The hydraulic system adjusts the oil flow and pressure provided by the hydraulic pump through the calculation and instructions of the control unit, uses the hydraulic control valve to control the hydraulic oil flowing to the hydraulic cylinder, pushes the piston to move to accurately adjust the length of the steel wire, and performs the retraction and extension action:

[0063] VF p =M hc ×dv p / d t ;

[0064] WF p =Δp×A s ;

[0065] T = ΔL × E × A;

[0066] Q p =V p ×A s ;

[0067] Q f =C d×A f ×Δp×√(2Δp / ρ);

[0068] Where, T is the tension of the wire rope, F p is the force of the buffer acting on the piston rod, M hc is the equivalent mass of the buffer moving parts; dv p / d t is the acceleration of the buffer moving parts. The buffer moving parts in the hydraulic system include: piston, piston rod, movable pulley, A s is the cross-sectional area of ​​the piston, Q p V is the flow regulation of the hydraulic pump. p is the speed of the buffer component, Q f For flow control of hydraulic control valve, C d is the flow coefficient, A f is the valve opening area, ρ is the liquid density; set Q p Equal to Q f According to the formula, the movement speed of the buffer component is obtained, the length change of the steel wire is controlled by controlling the movement speed of the piston, and the PID algorithm is used to control the opening of the hydraulic control valve to dynamically adjust the blocking force.

[0069] The second purpose of the present invention is to provide a highway emergency response blocking method, which can realize real-time monitoring of dangerous sections of highways, and automatically or manually activate the guidance warning system when abnormal conditions are detected, and provide drivers with instant and effective warning information through the comprehensive use of physical and virtual blocking barriers, so as to greatly improve the safety and emergency response efficiency of highways, reduce the accident rate, and minimize secondary accidents.

[0070] The present invention provides a second basic solution: a highway emergency response blocking method, which uses the above-mentioned highway emergency response blocking device, including:

[0071] S1. Data collection: The video image acquisition module and road surface condition detection module of the highway detection system monitor vehicle behavior and road surface conditions in real time, and collect video image information and road surface condition information of the road surface and vehicles;

[0072] S2. Analysis of data anomalies: Using edge computing technology, the collected video image information and road condition information are analyzed and processed at the device layer of the video image acquisition module and the road condition detection module. The highway detection system sends the analysis result data to the remote control system in real time through the data transmission system;

[0073] S3, Remote information decision control: The remote control system receives data detected by the highway detection system or data fed back by the road section information feedback system. If there is an abnormal situation, it will dispatch and control the opening and closing of the guidance warning system and the physical interception system;

[0074] S4. After the control is completed, the road section information feedback system feeds back the relevant data to the remote control system of the subsequent road section;

[0075] S5. When the abnormal situation is resolved, the guidance warning system and the physical interception system are released accordingly, and normal road condition monitoring is restored.

[0076] Further, the S2 includes:

[0077] S201, extracting features from the video image information and the road surface status information at the device layer, obtaining image features and detector features, and transmitting them to the next layer, including: using the CNN-lite model to process the video image information in real time on the video monitoring device of the video image acquisition module, and dynamically adjusting the parameters of the CNN-lite model according to the weather conditions; wherein the CNN-lite model is a modified convolutional neural network model;

[0078] The process steps include:

[0079] Establish a standard reference image library: Establish a standard reference image library to store standard images of clarity levels corresponding to each weather condition. The clarity of sunny images is 4, the clarity of light rain and fog is 3, the clarity of moderate rain and fog is 2, and the clarity of heavy rain and fog is 1;

[0080] Image acquisition: extract the real-time image from the video image information captured by the video surveillance device within the preset time, and import the real-time image into the Matlab processing module;

[0081] Clarity calculation: The clarity of the real-time image is evaluated by using a standard reference image library clarity evaluation method. The specific standard reference image library clarity evaluation method includes: based on the real-time image, using the Canny algorithm to obtain image edge information, calculate the average intensity of edge pixels, and then calculate the image contrast to obtain the mean of the histogram contrast, and then perform weighted average calculation on the average intensity of edge pixels and the mean of contrast;

[0082] Determine clarity: Use the K-nearest neighbor machine learning optimization algorithm to find the K images that are most similar to the input image, then vote to determine the weather conditions and determine the clarity under the corresponding weather conditions;

[0083] Dynamically adjust CNN parameters: Establish a functional relationship between the image clarity, convolution kernel size and sliding step size. For images with a clarity of 4, select a 7×7 convolution kernel and a step size of 3; for images with a clarity of 3, select a 5×5 convolution kernel and a step size of 2; for images with a clarity of 2, select a 3×3 convolution kernel and a step size of 1; for images with a clarity of 1, select a 2×2 convolution kernel and a step size of 1; set an automatic adjustment mechanism for environmental factors based on the clarity and functional relationship;

[0084] Apply the adjusted parameters for feature extraction: According to the parameter adjustment decision, the CNN-lite model is adjusted in real time, and the adjusted CNN-lite model is used to extract features from the real-time image, and the extracted image features are transferred to the next layer, i.e., the fog layer;

[0085] On the non-contact road surface detector, a lightweight autoencoder is used to process the road surface state information collected by the non-contact road surface detector;

[0086] The process steps include:

[0087] Data synchronization: Use the network time protocol to align the road surface status information from different non-contact road surface detectors;

[0088] Data preprocessing: Use a Butterworth filter with smooth frequency response characteristics to clean the road surface status information;

[0089] Feature extraction and fusion: The autoencoder is used to extract detector features from the data of each non-contact road detector, and the features are fused to form a comprehensive feature set using the average weighted method. The extracted detector features are then transmitted to the next layer, which is the fog layer.

[0090] S202: Performing spatiotemporal relationship analysis and combining the fusion characteristics of the fog layer to analyze the probability of abnormal vehicle behavior or abnormal road conditions, including:

[0091] Time series feature extraction: Time series data is processed in the fog layer, and a recurrent neural network model is used to capture the features of the time series data, thereby extracting vehicle behavior features and identifying abnormal road conditions; the time series data includes: image features and detector features;

[0092] Feature fusion: The weighted method is used for multi-modal feature fusion, Fi = wv × V + wp × P, Fi is the fusion feature, V and P are image features and detector features respectively, wv and wp are weights, which can be set according to the functions to be achieved. When identifying abnormal vehicle behavior, wv is 0.8 and wp is 0.2; when identifying abnormal road conditions, wv is 0.2 and wp is 0.8;

[0093] S203, performing a spatiotemporal relationship analysis, and combining the fusion characteristics of the fog layer to analyze the probability of abnormal vehicle behavior or abnormal road conditions, including:

[0094] Time-space graph update: Use a complete graph neural network model to update the time-space relationship graph, helping the system to fully understand the movement patterns of vehicles in a specified time and space, and providing a basis for abnormal behavior identification;

[0095] Abnormal behavior recognition: Apply the abnormal behavior recognition model based on the attention mechanism, combine the fusion characteristics of the fog layer and the spatiotemporal relationship analysis of the cloud layer, calculate the probability of abnormal vehicle behavior or abnormal road conditions, and conduct real-time analysis of abnormal vehicle behavior and road conditions.

[0096] Further, the S3 includes: S301, remote command and data collection: the remote control system sends a command to start the physical barrier, the motor drive module receives the command to start the physical barrier and starts working, and at the same time, obtains the clarity under the corresponding weather conditions during this period stored in the remote control system;

[0097] S302, equipment adjustment: according to the clarity under the corresponding weather conditions, the control module adjusts the brightness and color of the LED display screen and the laser emission module. When the clarity value is 1, it is adjusted to strong brightness and red. When the clarity value is 2, it is adjusted to medium brightness and yellow. When the clarity is 3, it is adjusted to medium intensity and green. When the clarity is 4, it is adjusted to weak intensity and blue.

[0098] S303, physical railing startup and placement: the motor drive module is activated, and starts to drive the railing down to the ground, and the folded railing is stretched and unfolded to form a physical barrier;

[0099] S304, laser wall curtain activation: when the railing has been stably placed on the ground, the laser emission module starts to work and generates a layer of virtual wall curtain;

[0100] S305, stop process: When the remote control system sends a stop command or detects that there is no need to block the front, the motor drive module automatically retracts the railing and the laser wall curtain is closed synchronously.

[0101] The third purpose of the present invention is to provide a highway emergency response blocking system, which can realize real-time monitoring of dangerous sections of highways, and automatically or manually activate the guidance warning system when abnormal conditions are detected, and provide drivers with instant and effective warning information through the comprehensive use of physical and virtual blocking barriers, so as to greatly improve the safety and emergency response efficiency of highways, reduce the accident rate, and minimize secondary accidents.

[0102] The present invention provides a third basic solution: a highway emergency response blocking system, which adopts the above-mentioned highway emergency response blocking method.

[0103] Beneficial effects: This solution monitors and analyzes road sections and driving conditions in real time, and activates the guidance warning system and physical interception system when abnormal or dangerous conditions are detected, guiding the driver to change the driving track or stop; with the help of automatic and manual dispatch modules, it can efficiently and accurately manage and dispatch road resources to respond to emergencies of different types and levels. In addition, through real-time information sharing and feedback, information islands are eliminated, and the overall traffic system's responsiveness and resource allocation efficiency are improved.

[0104] This solution integrates advanced monitoring and control technologies, which can achieve real-time monitoring of dangerous sections of highways. When abnormal conditions are detected, the guidance warning system is automatically or manually activated. Through the comprehensive use of physical and virtual barriers, it provides drivers with immediate and effective warning information, thereby greatly improving the safety and emergency response efficiency of highways, reducing the accident rate, and minimizing secondary accidents.

[0105] In particular, in this solution, when monitoring the data of key indicators of vehicle and road conditions in real time, feature extraction will be performed on the data, and abnormal analysis will be performed based on the features to obtain data on abnormal vehicle behavior; and feature extraction uses a neural network model, and according to the functional relationship between the preset data and the convolution parameters in the neural network model, the convolution parameters are dynamically set for feature extraction, so that it can adapt to different data and adjust the network model according to the targeted application, so that the extracted features are more accurate, and then the results of subsequent processing based on the features are more accurate, ensuring the accuracy of abnormal analysis, which is higher than traditional abnormal analysis;

[0106] The physical blocking system set in this scheme adapts to the setting of the hydraulic system in the physical blocking system when adjusting the wire, analyzes the relationship between the change of the force in the hydraulic system and the tension of the wire rope, and performs reasonable dynamic regulation according to the speed of the car. Compared with the traditional method of adjusting the rope length only according to the speed of the car, it is more targeted and more accurate.

[0107] In addition, the physical barriers provide the first layer of physical blocking, which serves to warn vehicles to slow down and effectively prevent vehicles from deviating from the road or running off the road. At the same time, the blocking system can provide additional protection based on real-time conditions, such as automatically initiating blocking measures that dynamically adjust the blocking force in an emergency, thereby providing double protection for the vehicle and its occupants. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] Figure 1 It is a logic block diagram of the first embodiment of the highway emergency response blocking device of the present invention;

[0109] Figure 2 This is a schematic diagram of the door frame arrangement in the first embodiment of the highway emergency response blocking device of the present invention;

[0110] Figure 3 This is a schematic diagram of the installation of a non-contact road surface detector in Embodiment 1 of the highway emergency response blocking device of the present invention;

[0111] Figure 4 It is a flow chart of Embodiment 1 of the highway emergency response blocking method of the present invention. DETAILED DESCRIPTION

[0112] The following is further described in detail through specific implementation methods:

[0113] The figure marks in the drawings of the specification include: video surveillance equipment 1, external light 2, laser emission module 3, door frame 4, physical railing 5, voice broadcast unit 6, LED display screen 7, flash light unit 8, non-contact road surface detector 9, guidance warning system 10.

[0114] Embodiment 1

[0115] The embodiment is basically as shown in the attached Figure 1 As shown: the highway emergency response blocking device is set in a preset section, which is a dangerous section of the highway in this embodiment, and includes a gantry 4 perpendicular to the ground, such as Figure 2 As shown, it also includes:

[0116] Highway detection system, which is used to monitor key indicators of vehicle and road conditions in real time to ensure that accidents can be detected and responded to quickly when they occur;

[0117] Specifically, the highway detection system includes:

[0118] The video image acquisition module includes: a plurality of video monitoring devices 1 and an external light 2;

[0119] The specific number of video monitoring devices 1 is determined according to the number of lanes. When the video monitoring devices are set, they are distributed along the crossbeam of the gantry 4. The external lights 2 are set above the two sides of the video monitoring devices 1 to provide light sources for the video monitoring devices 1 when the video monitoring devices collect video image information such as road conditions, vehicle behavior and traffic flow in real time. The video image acquisition module can automatically adjust according to the ambient light and weather conditions through the external lights 2, which is more accurate and timely than traditional manual adjustment, and can effectively improve the quality of video image information in bad weather conditions;

[0120] Intelligent video analysis module, used to receive real-time video image information transmitted by the video image acquisition module, and use image processing and artificial intelligence algorithms to conduct in-depth analysis of the video image information to detect road congestion, high-speed parking, and traffic accidents;

[0121] The road surface state detection module adopts a non-contact road surface detector 9, such as Figure 3 As shown, the guidance and warning system 10 is installed on the pillars of the LED display screen 7 on both sides of the road, and the non-contact road surface detector 9 faces the road surface and is used to detect road ice, snow and collapse. The non-contact road surface state detector uses remote sensing technology to avoid damage to the road, so that there will be no interference to traffic caused by installing a road weather station. At the same time, multi-spectral measurement technology enables it to accurately detect ice, snow and collapse on the road surface.

[0122] Data transmission system, used for remote data transmission, to achieve efficient and low-cost remote data transmission;

[0123] Specifically, the data transmission system includes:

[0124] The wireless transmission module is used to receive data from the highway detection system and send the data to the remote control system. In this process, the seamless reception and transmission of data depends on wireless communication protocols and standards to ensure the integrity and accuracy of the data.

[0125] The Internet transmission module is used to assist the wireless transmission module to perform remote data transmission through the Internet infrastructure; specifically, based on the wireless transmission module, the Internet transmission module plays an extended role and is used to work in conjunction with the wireless transmission module to achieve remote data transmission using the existing Internet infrastructure;

[0126] A remote control system for receiving data from the highway detection system transmitted by the data transmission system, determining highway conditions based on the data, and activating corresponding emergency measures automatically or by manual command based on the highway conditions;

[0127] Specifically, the remote control system includes:

[0128] The real-time monitoring module is used to receive the data of the highway detection system transmitted by the data transmission system and display it on the monitoring interface. It is also used to generate real-time traffic situation maps, dynamic weather maps and other visual content based on the data, providing more intuitive information support so that managers can understand the operation status of the highway at the first time;

[0129] The automatic dispatch module is the decision center of the system, which is used to receive the data transmitted by the real-time monitoring module, and when the data reaches the predetermined crisis condition, it generates a dispatch decision and automatically activates the guidance warning system 10;

[0130] The manual dispatch module is used to input decisions and override the decisions of the automatic dispatch module with the input decisions; specifically, in certain circumstances, the operator can override the decisions of the automatic dispatch module through the manual dispatch module to implement the start-up guidance warning system 10 and the physical blocking system, which provides greater flexibility for the system and ensures that the highway conditions can still be effectively managed under various unpredictable circumstances;

[0131] A control module, used to manage and adjust the operating status of the equipment in the guidance warning system 10 and the physical interception system according to the scheduling;

[0132] The data storage module is used to store all data transmitted from the highway detection system and the operation records of the remote control system, and can provide persistent data support for subsequent analysis and review; all data transmitted from the highway detection system include: clarity under different weather conditions at different times;

[0133] The guidance warning system 10 is used to provide guidance warnings, including: using flashing lights, display screens, and voice broadcasts to guide drivers to take detours or reduce speed to prevent secondary accidents caused by emergencies;

[0134] Specifically, the guidance warning system 10 includes:

[0135] The voice broadcast unit 6 is installed above the LED display screen 7 and is used to issue a voice warning when the guidance warning system 10 is activated, prompting the driver to slow down, change lanes or slow down to stop;

[0136] LED display screens 7 are installed on both sides of the road, with each LED display screen 7 spaced 1000 meters apart, for displaying text and image warning information on the electronic screen;

[0137] The flash light unit 8 is installed beside the language broadcasting device and is used to emit a strong light warning when the guidance warning system 10 is activated.

[0138] A physical blocking system that, when activated, forms a physical or virtual blocking barrier to close a roadway, and that can be quickly activated to form a physical or virtual blocking barrier to temporarily close a roadway when necessary;

[0139] Specifically, the physical interception system includes:

[0140] Physical barrier 5, used to intercept vehicles;

[0141] The motor drive module is used to lower the physical barrier 5 from a suspended state to the ground to achieve physical blocking of the traffic flow; specifically, after receiving the signal, the motor drive module starts immediately to lower the suspended physical barrier 5 to the ground; the physical barrier 5 adopts a folding design and a honeycomb structure, and a color-changing LED light strip is installed on the top of the barrier. The outside of the barrier is wrapped with high-strength rubber, which has the functions of occupying a small space when stored, adaptive warning, and safety buffering, but it can effectively block the passage of vehicles after being lowered to the ground. After the physical barrier 5 is lowered to the ground, it begins to effectively block the traffic flow, forcing vehicles to stop or change routes, and preventing vehicles from entering dangerous areas or areas that need to be temporarily kept clear;

[0142] The laser emission module 3 is installed at the center of the crossbeam of the gantry 4 and is started simultaneously with the motor drive module to generate a virtual wall curtain on the road through a powerful laser. The virtual wall curtain is a pattern containing warning information to enhance the visual warning effect on the driver.

[0143] The road section information feedback system is used to transmit information between external user systems and between emergency response blocking devices on different sections of highways, that is, to transmit information between the road section system and the user system, and to achieve instant information sharing and feedback between the road section systems and the user systems;

[0144] Specifically, the road section information feedback system includes:

[0145] System interconnection module, which establishes real-time communication between information systems of different road sections through high-speed network connection, and is used for real-time communication and data sharing between information systems of different road sections;

[0146] The user interconnection module is equipped with mobile Internet access, vehicle communication system access or other forms of user equipment access, which is used to provide road condition information to the user side in real time and receive sudden road conditions feedback from the user side, so as to realize two-way flow and interactivity of information.

[0147] Highway emergency response blocking methods, such as Figure 4 As shown, the above-mentioned highway emergency response blocking device is adopted, including the following contents:

[0148] S1. Data collection: Collecting video image information of the road surface and vehicles and road surface status information;

[0149] The video image acquisition module and road surface condition detection module of the highway detection system monitor vehicle behavior and road surface conditions in real time, and collect video image information and road surface condition information of the road surface and vehicles;

[0150] S2. Data anomaly analysis: Using edge computing technology, the collected video image information and road condition information (reflecting vehicle behavior and road condition anomalies) are analyzed and processed at the device layer of the video image acquisition module and the road condition detection module. The highway detection system sends the analysis result data to the remote control system in real time through the data transmission system;

[0151] Specifically include:

[0152] S201, specific implementation of the device layer, extracting features of the video image information and road state information at the device layer, obtaining image features and detector features, and transmitting them to the next layer (fog layer), including:

[0153] On the video monitoring device 1 of the video image acquisition module, a CNN-lite model is used to process the video image information in real time, and the parameters of the CNN-lite model are dynamically adjusted according to the weather conditions; wherein the CNN-lite model is a modified convolutional neural network model;

[0154] The specific process steps are as follows:

[0155] Establishing a standard reference image library: Establishing a standard reference image library to store standard images of clarity levels corresponding to each weather condition. In this embodiment, the clarity of a sunny day image is 4, the clarity of a light rainy and foggy day image is 3, the clarity of a moderate rainy and foggy day image is 2, and the clarity of a heavy rainy and foggy day image is 1;

[0156] Image acquisition: extract the real-time image from the video image information captured by the video surveillance device within the preset time, and import the real-time image into the Matlab processing module;

[0157] Clarity calculation: The clarity of the real-time image is evaluated by using a standard reference image library clarity evaluation method. The specific standard reference image library clarity evaluation method includes: based on the real-time image, using the Canny algorithm to obtain image edge information, calculate the average intensity of edge pixels, and then calculate the image contrast to obtain the mean of the histogram contrast, and then perform weighted average calculation on the average intensity of edge pixels and the mean of contrast;

[0158] Determine clarity: Use the K-nearest neighbor machine learning optimization algorithm to find the K images most similar to the input image, then vote to determine the weather conditions and determine the clarity under the corresponding weather conditions; the implementation image is used as the input image;

[0159] Dynamically adjust CNN parameters: Establish a functional relationship between the image clarity, convolution kernel size and sliding step size. For images with a clarity of 4, select a 7×7 convolution kernel and a step size of 3; for images with a clarity of 3, select a 5×5 convolution kernel and a step size of 2; for images with a clarity of 2, select a 3×3 convolution kernel and a step size of 1; for images with a clarity of 1, select a 2×2 convolution kernel and a step size of 1; set an automatic adjustment mechanism for environmental factors based on the clarity and functional relationship;

[0160] Apply the adjusted parameters for feature extraction: According to the parameter adjustment decision, the CNN-lite model is adjusted in real time, the adjusted CNN-lite model is used to extract features from the real-time image, and the extracted image features are transferred to the next layer (fog layer).

[0161] When the image clarity is high, a larger convolution kernel (such as 7×7) can cover a wider image range and capture more comprehensive features, which is suitable for extracting global features and structural information. On the contrary, a smaller convolution kernel (such as 3×3 or 2×2) is more suitable for situations with lower image clarity. They can focus on the local detail features of the image. Small convolution kernels help to enhance the model's ability to detect blurred or small-sized objects when the image quality is poor.

[0162] At the same time, using a larger step size will reduce the overlap of the model's receptive field coverage area, thereby reducing the amount of calculation. Using a large step size in a high-definition image can quickly process the image without losing important feature information; a smaller step size will increase the repeated coverage area and improve the model's sensitivity to features, which is more suitable for analyzing fine features in low-definition images.

[0163] Common convolution kernel sizes are 7×7, 5×5, 3×3, and 2×2, and common step sizes are 3, 2, and 1. Therefore, different convolution kernel sizes and sliding step sizes are dynamically used for different clarity. When the clarity is good, a larger convolution kernel with a larger step size can reduce the number of required calculations, thereby improving the computational efficiency of the entire network. At the same time, such efficiency improvement will not lead to the loss of important features. When the clarity is poor, a smaller convolution kernel and a smaller step size are required to capture more fine features.

[0164] On the non-contact road surface detector 9, a lightweight autoencoder is used to process the road surface state information collected by the non-contact road surface detector 9;

[0165] The specific process is as follows:

[0166] Data synchronization: using the network time protocol to align the road surface status information from different non-contact road surface detectors 9;

[0167] Data preprocessing: Considering the high-speed movement of vehicles on highways and environmental interference, a Butterworth filter with smooth frequency response characteristics is used to clean the road state information to ensure the accuracy of the data; in addition, data standardization can also be performed to ensure the consistency of subsequent data processing;

[0168] Feature extraction and fusion: The detector features are extracted from the data of each non-contact road detector 9 using an autoencoder, and the detector features are fused to form a comprehensive feature set using an average weighted method, and the extracted detector features are transmitted to the next layer (fog layer);

[0169] S202, a specific implementation of the fog layer, capturing the features of the image features and the detector features of the captured time series data, performing multi-modal feature fusion, and obtaining fusion features, including:

[0170] Time series feature extraction: Process time series data in the fog layer and use a recurrent neural network model to capture the features of time series data, thereby extracting vehicle behavior features and identifying abnormal road conditions;

[0171] The time series data processed by the fog layer in S202 includes: image features and detector features. Specifically, for the data on the video monitoring device 1 of the video image acquisition module: the video image acquisition module monitors and collects video image information of road conditions and vehicle behavior in real time, and processes the video image information through the intelligent video analysis module to extract feature information related to vehicle behavior, namely, the image features. The continuous changes in the video image information form time series data, which is used to reflect the behavior characteristics of the vehicle at different times;

[0172] Regarding the data on the non-contact road detector 9 in the road state detection module: the road state detection module monitors the road state through the non-contact road detector 9, including ice, snow or collapse, etc. The road state data collected by the non-contact road detector changes with time, forming the time series information of the road state, that is, the road state information. After the road state information is processed by the lightweight autoencoder, the detector features are extracted, and the extracted detector features are transmitted to the fog layer;

[0173] The time series data includes the vehicle behavior change data (i.e., image features) obtained from the video image acquisition module and the road condition change data (detector features) obtained from the road condition detection module. After preliminary analysis at the device layer, these data are aggregated to the fog layer, and the recurrent neural network model is used to further extract and analyze the time series features.

[0174] Feature fusion: The weighted method is used to perform multi-modal feature fusion, Fi = wv × V + wp × P, where Fi is the fusion feature, V and P are image features and detector features respectively, and wv and wp are weights, which can be set according to the functions to be implemented. In this embodiment, when identifying abnormal vehicle behavior, wv is 0.8 and wp is 0.2; when identifying abnormal road conditions, wv is 0.2 and wv is 0.8; by integrating the image features and detector features provided by the device layer, comprehensive vehicle behavior and road condition information can be obtained;

[0175] S203, specific implementation of the cloud layer, performing spatiotemporal relationship analysis, and combining the fusion characteristics of the fog layer to analyze the probability of abnormal vehicle behavior or abnormal road conditions, including:

[0176] Time-space graph update: Use a complete graph neural network model to update the time-space relationship graph, helping the system to fully understand the movement patterns of vehicles in a specified time and space, and providing a basis for abnormal behavior identification;

[0177] Abnormal behavior recognition: Apply the abnormal behavior recognition model based on the attention mechanism, combine the fusion characteristics of the fog layer and the spatiotemporal relationship analysis of the cloud layer, calculate the probability of abnormal vehicle behavior or abnormal road conditions, and realize real-time analysis of abnormal vehicle behavior and road conditions.

[0178] S3. Remote information decision control:

[0179] The remote control system receives the data detected by the highway detection system or the data fed back by the road section information feedback system. If there is an abnormal situation, it will dispatch and control the opening and closing of the guidance warning system 10 and the physical interception system;

[0180] Specifically, the remote control system receives the data of abnormal vehicle behavior sent by the highway detection system, and analyzes whether to start the dispatch according to the data of abnormal vehicle behavior. If so, the automatic dispatch module is activated, and the guidance warning system 10 is started. The guidance warning system 10 displays the abnormality of the road ahead and calls on the driver to slow down or change the route. The abnormal vehicle behavior data is the probability of abnormal vehicle behavior or abnormal road surface conditions, and the remote control system analyzes whether to start the dispatch according to the abnormal vehicle behavior data, in order to determine whether the probability of abnormal vehicle behavior or abnormal road surface conditions is higher than a preset threshold value. If so, the automatic dispatch module is activated.

[0181] When the remote control system receives instructions from the road section information feedback system, the remote control system makes decisions through the manual scheduling module, inputs the decisions through the manual scheduling module, and overrides the decisions of the automatic scheduling module with the input decisions, wherein the decisions include scheduling, controlling the opening and closing of the guidance warning system 10 and the physical interception system; specifically, the manager determines whether it is necessary to activate the physical interception system and the guidance warning system 10 based on the specific information content of the received instructions and the speculation of the on-site situation based on the real-time monitoring data. At the same time, the system will feedback the abnormal road conditions to the manager. After receiving the notification, the manager determines whether it is necessary to activate the physical interception system in combination with the real-time monitoring image information. If necessary, the manager activates the physical interception system through the manual scheduling module and changes the content of the guidance warning system 10 to prompt the driver to slow down and stop, and the road section ahead is about to be closed;

[0182] S301, remote command and data collection: the remote control system sends a command to start the physical barrier, the motor drive module receives the command to start the physical barrier and starts working, and at the same time, obtains the clarity under the corresponding weather conditions during this period stored in the remote control system;

[0183] S302, equipment adjustment: according to the clarity under the corresponding weather conditions, the control module in the remote control system adjusts the brightness and color of the LED display screen 7 and the laser emission module 3. When the clarity value is 1, it is adjusted to strong brightness and red. When the clarity value is 2, it is adjusted to medium-strong brightness and yellow. When the clarity is 3, it is adjusted to medium intensity and green. When the clarity is 4, it is adjusted to weak intensity and blue. The data storage module of the remote control system stores the clarity under the corresponding weather conditions during this period, and automatically adjusts the brightness and color of the LED display screen 7 and the laser emission module 3 according to different clarity values ​​through the received video clarity.

[0184] S303, starting and placing the physical barrier 5: the motor drive module is activated, and starts to drive the barrier down to the ground, and the folded barrier is stretched and unfolded to form an effective physical barrier;

[0185] S304, laser wall curtain activation: when the railing has been stably placed on the ground, the laser emission module 3 starts to work, generating a layer of virtual wall curtain, the brightness and color of this wall curtain are consistent with the LED light strip, and it is close to the top of the physical railing 5, so as to further improve the visibility of the warning;

[0186] S305, stop process: When the remote control system sends a stop command or detects that there is no need to block the front, the motor drive module automatically retracts the railing and the laser wall curtain is closed synchronously.

[0187] S4, road section information feedback: After the regulation is completed, the road section information feedback system will feed back the relevant data to the remote control system of the subsequent road section so as to make timely response decisions;

[0188] S5. System returns to normal: When the abnormal situation is resolved, the guidance warning system 10 and the physical interception system are correspondingly released (closed) and normal road condition monitoring is restored.

[0189] Embodiment 2

[0190] This embodiment is basically the same as the above embodiment, except that: it also includes a physical blocking system;

[0191] The physical blocking system includes: a blocking net, a hydraulic system and a control unit. The blocking net is similar to a physical railing 5 in an inactive state and is suspended above the road surface, while the hydraulic system is installed on both sides of the gantry 4. At the same time, the blocking net is connected to the hydraulic system through steel wire ropes at both ends; the control unit is used to identify abnormalities or emergency situations, and quickly lower the blocking net to the same level as the physical railing 5 through hydraulic control, and lock it at a fixed height of not less than 3 meters from the ground; as the blocking net descends, the steel wire rope connecting the blocking net is tightened accordingly to ensure that the blocking net is in a natural state; the physical railing 5 is organically linked to the blocking net, and the physical railing 5 provides the first layer of physical blocking, which serves to warn vehicles to slow down, and also effectively prevents vehicles from deviating from the road or running off the road. At the same time, the blocking system can provide additional protection according to real-time conditions, such as automatically starting blocking measures that dynamically adjust the blocking force in an emergency, thereby providing double protection for the vehicle and occupants.

[0192] Specifically, when a vehicle enters the warning area away from the physical barrier 5, the video image acquisition module collects images, identifies the type and speed of the vehicle, and transmits the identified information to the control unit for data analysis and processing, so as to preset the initial state of the blocking net. At the same time, a sensor is embedded in the physical barrier 5 to obtain the impact force when the vehicle hits the physical barrier 5, so as to achieve dynamic blocking force adjustment. The blocking net uses the adjusted blocking force strength to effectively envelop and block the impacting vehicle, thereby achieving safety protection for the vehicle and minimizing damage to the facilities and the vehicle; wherein the sensor adopts a strain gauge;

[0193] The working process of the barrier net:

[0194] Real-time data collection steps:

[0195] Video surveillance data acquisition: When a vehicle enters the warning area, the video image acquisition module collects the image of the vehicle and uses image recognition technology to determine the type and speed of the vehicle.

[0196] Sensor data acquisition: The strain gauge embedded in the physical railing 5 detects the impact force when the vehicle collides with the railing. Since strain measurement can usually only provide local stress information, it is necessary to install multiple strain gauges at different locations of the physical railing 5 to obtain sufficiently accurate and comprehensive impact force information.

[0197] Initial state preset steps:

[0198] The control unit receives the information recognized by the video image acquisition module and adjusts the length of the steel wires on both sides of the barrier net in advance according to the stress characteristics of the barrier net to effectively buffer the upcoming impact force;

[0199] E k =1 / 2mv 2 (1)

[0200] Among them, E k is the vehicle kinetic energy, m is the vehicle mass (estimated based on vehicle type), and v is the vehicle speed;

[0201] L=(E k +T×s) / (E×A) (2)

[0202] Among them, L is the preset length of the steel wire that needs to be adjusted in order to effectively buffer the impact force, T is the actual tension of the barrier net, s is the deformation distance of the barrier net corresponding to the actual tension, E is the elastic modulus of the steel wire, and A is the cross-sectional area of ​​the steel wire.

[0203] Dynamic stopping force adjustment steps:

[0204] Data analysis: The control unit analyzes the data measured by the strain gauges and calculates the impact force of the colliding vehicle and the stopping force that needs to be adjusted. Specifically:

[0205] ε=R0×GF / ΔR (3)

[0206] Among them, ΔR is the change in resistance, R0 is the resistance value before deformation, GF is the sensitivity factor of the strain gauge, and ε is the strain;

[0207] σ=E×ε (4)

[0208] Among them, σ is stress and E is elastic modulus;

[0209] F=Σ(σ i ×A i ) (5)

[0210] Among them, F is the impact force, and A is the action area corresponding to the strain gauge;

[0211] ΔL=(FE k -T×s)×(1-α) / (E×A) (6)

[0212] Among them, ΔL is the length of the steel wire that needs to be changed, and α is the adjustment coefficient. According to the preliminary analysis of the stopping force variation law, if the stopping force generated by the stopping system is large, the vehicle will have a large deceleration acceleration, which may exceed the bearing capacity of the vehicle and the driver and passengers, and may cause damage to the vehicle or personal injury. At the same time, considering the errors in the collection of various parameters and the kinetic energy lost by the vehicle when colliding with the physical barrier 5, α is taken as 0.2;

[0213] The hydraulic system's action mechanism: The hydraulic system adjusts the oil flow and pressure provided by the hydraulic pump through the calculation and instructions of the control unit, and uses the hydraulic control valve to accurately control the hydraulic oil flowing to the hydraulic cylinder, pushing the piston to move to accurately adjust the length of the steel wire and realize the retraction and extension action. Specifically:

[0214] TF p =M hc ×dv p / d t (7)

[0215] F p =Δp×A s (8)

[0216] T=ΔL×E×A (9)

[0217] Q p =V p ×A s (10)

[0218] Q f =C d ×A f ×Δp×√(2Δp / ρ) (11)

[0219] Where, T is the tension of the wire rope, F p is the force of the buffer acting on the piston rod, M hc is the equivalent mass of the buffer moving parts; dv p / d t A is the acceleration of the moving parts of the buffer. The buffer parts in the hydraulic system are composed of a piston, a piston rod, and a movable pulley. s is the cross-sectional area of ​​the piston, Q p V is the flow regulation of the hydraulic pump. p is the speed of the buffer component, Q f For flow control of hydraulic control valve, C d is the flow coefficient, A f is the valve opening area, ρ is the liquid density, and under ideal conditions, Q p Equal to Q fTherefore, according to the combined solution of equations (7), (8), (9), (10), and (11), the movement speed of the buffer component, that is, the change of the length of the steel wire is achieved by controlling the movement speed of the piston. In addition, the PID algorithm is used to control the opening of the hydraulic control valve to achieve dynamic adjustment of the blocking force.

[0220] Dynamic arresting force adjustment process: If the arresting force of the arresting system needs to be increased to cope with a greater impact force, the control unit instructs the hydraulic pump to increase the pressure in the hydraulic cylinder so that the piston pushes the arresting net to tighten; on the contrary, if the arresting force needs to be reduced to avoid excessive stretching or to better absorb the impact energy, the control unit instructs the hydraulic system to release part of the pressure so that the piston retracts and reduces the tension of the arresting net;

[0221] During the blocking process, the dynamic adjustment of the blocking force of the blocking system works together with the hydraulic energy absorber to convert and absorb the kinetic energy of the impacting vehicle.

[0222] Steps to complete the safety arrest:

[0223] As the vehicle's kinetic energy gradually dissipates, the vehicle eventually stops safely surrounded by the barrier net, completing the blocking process. At this time, the barrier net and hydraulic system reach the maximum deformation or pressure value, completing the absorption and conversion of the vehicle's kinetic energy.

[0224] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. Highway emergency response blocking device, set up in a preset section, characterized by: include: Highway monitoring systems, which monitor data on key indicators of vehicle and road conditions in real time; The highway detection system extracts features from the data, and performs anomaly analysis based on the features to obtain data on abnormal vehicle behavior; the feature extraction uses a neural network model, and dynamically sets convolution parameters for feature extraction based on the functional relationship between the preset data and the convolution parameters in the neural network model; A data transmission system, used for data transmission; A remote control system is used to receive data monitored by the highway detection system transmitted by the data transmission system, judge the highway conditions based on the data, and activate corresponding emergency measures automatically or through manual instructions according to the highway conditions, and dispatch the opening and closing of the warning system and the physical interception system; A guidance warning system, used for providing guidance warnings; Physical blocking systems, which, when activated, form physical or virtual barriers to close the road; The road section information feedback system is used to transmit information between external user systems and between emergency response blocking devices on different sections of highways.

2. The highway emergency response blocking device according to claim 1 is characterized in that: Also includes: A door frame perpendicular to the ground; The highway detection system comprises: Video image acquisition module, including: several video monitoring devices and external lights; The video monitoring equipment is distributed along the gantry crossbeam, and the external lights are set above the two sides of the video monitoring equipment to provide light sources for the video monitoring equipment when the video monitoring equipment collects video image information such as road conditions, vehicle behavior and traffic flow in real time; Intelligent video analysis module, used to receive real-time video image information transmitted by the video image acquisition module, and use advanced image processing and artificial intelligence algorithms to perform in-depth analysis on the video image information; The road surface condition detection module adopts a non-contact road surface detector. The non-contact road surface detector faces the road surface and is used to detect road icing, snow accumulation, and collapse.

3. The highway emergency response blocking device according to claim 1 is characterized in that: The data transmission system comprises: A wireless transmission module is used to receive data from the highway detection system and send the data wirelessly to the remote control system; Internet transmission module, used to assist the wireless transmission module to perform remote data transmission via the Internet infrastructure; The remote control system comprises: A real-time monitoring module is used to receive data from the highway detection system transmitted by the data transmission system and display it on the monitoring interface, and is also used to generate real-time visual content based on the data; The automatic dispatching module is used to receive the data transmitted by the real-time monitoring module, and when the data reaches the predetermined crisis condition, generate a dispatching decision and automatically activate the guidance warning system; The manual scheduling module is used to input decisions and override the decisions of the automatic scheduling module with the input decisions; A control module, used to manage and adjust the operating status of equipment in the guidance warning system and the physical interception system according to scheduling; The data storage module is used to store all data transmitted from the highway detection system and the operation records of the remote control system.

4. The highway emergency response blocking device according to claim 1, characterized in that: The guidance and warning system comprises: LED display screens are installed on both sides of the road, with a preset distance between each LED display screen, used to display text and image warning information on the electronic screen; A voice announcement unit, used to issue a voice warning when the guidance warning system is activated; A flashing light unit, which emits a strong light warning when the guidance warning system is activated; The physical interception system comprises: physical barriers to stop vehicles; A motor drive module for lowering the physical rail from a suspended state to the ground; The laser emission module is installed at the center of the gantry beam and is started simultaneously with the motor drive module to generate a virtual wall curtain on the road through a powerful laser; The road section information feedback system comprises: The system interconnection module establishes real-time communication between the road information feedback systems of different road sections through network connection; The user interconnection module is used to provide road condition information to external user systems in real time and receive road condition feedback from external user systems.

5. The highway emergency response blocking device according to claim 2 is characterized in that: Also includes: Physical blocking system; The physical blocking system includes: a blocking net, a hydraulic system and a control unit. The blocking net is suspended above the road surface in an inactive state, and the hydraulic system is installed on both sides of the gantry. The blocking net is connected to the hydraulic system through steel wire ropes at both ends; the control unit is used to identify abnormalities or emergency situations, and quickly lower the blocking net to the level of the physical railing through hydraulic control, and lock it at a fixed height not lower than a preset distance from the ground; When a vehicle enters the warning area away from the physical barrier, the video image acquisition module captures images, identifies the type and speed of the vehicle, and transmits the identified information to the control unit for data analysis and processing to obtain the initial state of the preset barrier net. At the same time, sensors are embedded in the physical barrier to obtain the impact force when the vehicle hits the physical barrier, and dynamically adjust the blocking force of the barrier net.

6. The highway emergency response blocking device according to claim 5, characterized in that: When a vehicle enters the warning area away from the physical barrier, the video image acquisition module acquires images, identifies the type and speed of the vehicle, and transmits the identified information to the control unit for data analysis and processing, obtains the initial state of the preset blocking net, and embeds a sensor in the physical barrier to obtain the impact force when the vehicle hits the physical barrier, and dynamically adjusts the blocking force of the blocking net, including: Real-time data collection steps: Video surveillance data acquisition: When a vehicle enters the warning area, the video image acquisition module collects the image of the vehicle and uses image recognition technology to determine the type and speed of the vehicle; Sensor data acquisition: Strain gauges embedded in the physical guardrail detect the impact force when the vehicle collides with the guardrail; Initial state preset steps: The control unit receives the information recognized by the video image acquisition module and adjusts the length of the steel wires on both sides of the barrier net in advance according to the stress characteristics of the barrier net: AND k =1 / 2mv 2 ; Among them, E k is the vehicle kinetic energy, m is the vehicle mass, and v is the vehicle speed; L=(E k +T×s) / (E×A); Among them, L is the length of the steel wire that needs to be preset and adjusted in order to effectively buffer the impact force, T is the actual tension of the barrier net, s is the distance of deformation of the barrier net corresponding to the actual tension, E is the elastic modulus of the steel wire, and A is the cross-sectional area of ​​the steel wire; Dynamic stopping force adjustment steps: Data analysis: The control unit analyses the data measured by the strain gauges and calculates the impact force of the colliding vehicle and the stopping force that needs to be adjusted: ε=R0×GF / ΔR; Among them, ΔR is the change in resistance, R0 is the resistance value before deformation, GF is the sensitivity factor of the strain gauge, and ε is the strain; σ=E×ε; Among them, σ is stress and E is elastic modulus; F=Σ(σ i ×A i ); Among them, F is the impact force, and A is the action area corresponding to the strain gauge; ΔL=(FE k -T×s)×(1-α) / (E×A) (6) Among them, ΔL is the length of the steel wire that needs to be changed, and α is the adjustment coefficient; Hydraulic system action: The hydraulic system adjusts the oil flow and pressure provided by the hydraulic pump through the calculation and instructions of the control unit, uses the hydraulic control valve to control the hydraulic oil flowing to the hydraulic cylinder, pushes the piston to move to accurately adjust the length of the steel wire, and performs the retraction and extension action: T-F p =M hc ×dv p / d t ; U-F p =Δp×A s ; T = ΔL × E × A; Q p =V p ×A s ; Q f =C d ×A f ×Δp×√(2Δp / ρ); Where, T is the tension of the wire rope, F p is the force of the buffer acting on the piston rod, M hc is the equivalent mass of the buffer moving parts; dv p / d t is the acceleration of the buffer moving parts. The buffer moving parts in the hydraulic system include: piston, piston rod, movable pulley, A s is the cross-sectional area of ​​the piston, Q p V is the flow regulation of the hydraulic pump. p is the speed of the buffer component, Q f For the flow control of the hydraulic control valve, C d is the flow coefficient, A f is the valve opening area, ρ is the liquid density; set Q p Equal to Q f According to the above formula, the movement speed of the buffer component is obtained, the length change of the steel wire is controlled by controlling the movement speed of the piston, and the PID algorithm is used to control the opening of the hydraulic control valve to dynamically adjust the blocking force.

7. A highway emergency response blocking method, characterized in that: The highway emergency response blocking device according to any one of claims 1 to 6 comprises: S1. Data collection: The video image acquisition module and road surface condition detection module of the highway detection system monitor vehicle behavior and road surface conditions in real time, and collect video image information and road surface condition information of the road surface and vehicles; S2. Analysis of data anomalies: Using edge computing technology, the collected video image information and road condition information are analyzed and processed at the device layer of the video image acquisition module and the road condition detection module. The highway detection system sends the analysis result data to the remote control system in real time through the data transmission system; S3, Remote information decision control: The remote control system receives data detected by the highway detection system or data fed back by the road section information feedback system. If there is an abnormal situation, it will dispatch and control the opening and closing of the guidance warning system and the physical interception system; S4, road section information feedback: after completing the control, the road section information feedback system will feed back the relevant data to the remote control system of the subsequent road section; S5. System returns to normal: When the abnormal situation is resolved, the guidance warning system and the physical interception system are released accordingly, and normal road condition monitoring is restored.

8. The highway emergency response blocking method according to claim 7, characterized in that: The S2 comprises: S201, extracting features from the video image information and the road surface status information at the device layer, obtaining image features and detector features, and transmitting them to the next layer, including: using the CNN-lite model to process the video image information in real time on the video monitoring device of the video image acquisition module, and dynamically adjusting the parameters of the CNN-lite model according to the weather conditions; wherein the CNN-lite model is a modified convolutional neural network model; The process steps include: Establish a standard reference image library: Establish a standard reference image library to store standard images of clarity levels corresponding to each weather condition. The clarity of sunny images is 4, the clarity of light rain and fog is 3, the clarity of moderate rain and fog is 2, and the clarity of heavy rain and fog is 1; Image acquisition: extract the real-time image from the video image information captured by the video surveillance device within the preset time, and import the real-time image into the Matlab processing module; Clarity calculation: The clarity of the real-time image is evaluated by using a standard reference image library clarity evaluation method. The specific standard reference image library clarity evaluation method includes: based on the real-time image, using the Canny algorithm to obtain image edge information, calculate the average intensity of edge pixels, and then calculate the image contrast to obtain the mean of the histogram contrast, and then perform weighted average calculation on the average intensity of edge pixels and the mean of contrast; Determine clarity: Use the K-nearest neighbor machine learning optimization algorithm to find the K images that are most similar to the input image, then vote to determine the weather conditions and determine the clarity under the corresponding weather conditions; Dynamically adjust CNN parameters: Establish a functional relationship between the image clarity, convolution kernel size and sliding step size. For images with a clarity of 4, select a 7×7 convolution kernel and a step size of 3; for images with a clarity of 3, select a 5×5 convolution kernel and a step size of 2; for images with a clarity of 2, select a 3×3 convolution kernel and a step size of 1; for images with a clarity of 1, select a 2×2 convolution kernel and a step size of 1; set an automatic adjustment mechanism for environmental factors based on the clarity and functional relationship; Apply the adjusted parameters for feature extraction: According to the parameter adjustment decision, the CNN-lite model is adjusted in real time, and the adjusted CNN-lite model is used to extract features from the real-time image, and the extracted image features are transferred to the next layer, i.e., the fog layer; On the non-contact road surface detector, a lightweight autoencoder is used to process the road surface state information collected by the non-contact road surface detector; The process steps include: Data synchronization: Use the network time protocol to align the road surface status information from different non-contact road surface detectors; Data preprocessing: Use a Butterworth filter with smooth frequency response characteristics to clean the road surface status information; Feature extraction and fusion: The autoencoder is used to extract detector features from the data of each non-contact road detector, and the features are fused to form a comprehensive feature set using the average weighted method. The extracted detector features are then transmitted to the next layer, which is the fog layer. S202: Performing spatiotemporal relationship analysis and combining the fusion characteristics of the fog layer to analyze the probability of abnormal vehicle behavior or abnormal road conditions, including: Time series feature extraction: Time series data is processed in the fog layer, and a recurrent neural network model is used to capture the features of the time series data, thereby extracting vehicle behavior features and identifying abnormal road conditions; the time series data includes: image features and detector features; Feature fusion: The weighted method is used for multi-modal feature fusion, Fi = wv × V + wp × P, Fi is the fusion feature, V and P are image features and detector features respectively, wv and wp are weights, which can be set according to the functions to be achieved. When identifying abnormal vehicle behavior, wv is 0.8 and wp is 0.2; when identifying abnormal road conditions, wv is 0.2 and wp is 0.8; S203, performing a spatiotemporal relationship analysis, and combining the fusion characteristics of the fog layer to analyze the probability of abnormal vehicle behavior or abnormal road conditions, including: Time-space graph update: Use a complete graph neural network model to update the time-space relationship graph, helping the system to fully understand the movement patterns of vehicles in a specified time and space, and providing a basis for abnormal behavior identification; Abnormal behavior recognition: Apply the abnormal behavior recognition model based on the attention mechanism, combine the fusion characteristics of the fog layer and the spatiotemporal relationship analysis of the cloud layer, calculate the probability of abnormal vehicle behavior or abnormal road conditions, and conduct real-time analysis of abnormal vehicle behavior and road conditions.

9. The highway emergency response blocking method according to claim 8, characterized in that: The S3 includes: S301, remote command and data collection: the remote control system sends a command to start the physical barrier, the motor drive module receives the command to start the physical barrier and starts working, and at the same time, obtains the clarity under the corresponding weather conditions during this period stored in the remote control system; S302, equipment adjustment: according to the clarity under the corresponding weather conditions, the control module adjusts the brightness and color of the LED display screen and the laser emission module. When the clarity value is 1, it is adjusted to strong brightness and red. When the clarity value is 2, it is adjusted to medium brightness and yellow. When the clarity is 3, it is adjusted to medium intensity and green. When the clarity is 4, it is adjusted to weak intensity and blue. S303, physical railing startup and placement: the motor drive module is activated, and starts to drive the railing down to the ground, and the folded railing is stretched and unfolded to form a physical barrier; S304, laser wall curtain activation: when the railing has been stably placed on the ground, the laser emission module starts to work and generates a layer of virtual wall curtain; S305, stop process: When the remote control system sends a stop command or detects that there is no need to block the front, the motor drive module automatically retracts the railing and the laser wall curtain is closed synchronously.

10. Highway emergency response blocking system, characterized in that: A highway emergency response blocking method as described in any one of claims 7 to 9 is adopted.

Images