Tunnel portal flexible traffic blocking management and control system

Abstract

The invention discloses a flexible traffic blocking management and control system for a tunnel portal. The flexible traffic blocking management and control system comprises a warning information issuing module, a graded early warning control module, a flexible blocking guide module, an intelligent equipment regulation and control module and an emergency linkage processing module, and the warning information issuing module is used for issuing warning information of abnormal conditions of the tunnel to a vehicle coming direction by utilizing an acousto-optic early warning warning unit and a roadside intelligent induction warning unit which are arranged at a tunnel portal and in front of the tunnel portal through information displayed by a variable information board and voice played by a directional loudspeaker, so that a driver is reminded to pay attention to potential dangers in front. According to the invention, the traffic safety and management and control efficiency of the tunnel portal are improved through multi-stage early warning, flexible blocking and optimization algorithms. Traffic conditions are analyzed through deep learning, equipment modes are dynamically adjusted, and accurate transmission of warning information is achieved. The virtual grating is generated by the intelligent laser and projection technology, and the traditional facility is replaced by flexible management and control. The distributed data acquisition and optimization algorithm reduces energy consumption and false triggering, and ensures efficient and economic traffic management.

Description

Technical Field

[0001] The invention relates to the technical field of tunnel opening traffic control, and in particular to a tunnel opening flexible blocking traffic control system. Background Art

[0002] The safety management of highway tunnels has always been an important topic in the field of traffic management, especially in the tunnel entrance area. Due to its special environment and frequent emergencies, this area has become a high-risk area for secondary accidents. At present, traditional means of traffic control at tunnel entrances mainly rely on fixed physical interception facilities, such as isolation piles, guardrails, etc., as well as basic sound and light warning equipment. Although these facilities can prevent vehicles from entering dangerous areas and provide warning information to drivers to a certain extent, they often seem powerless when dealing with emergencies. On the one hand, physical interception facilities require a lot of manpower and time to deploy and dismantle, making it difficult to respond to emergencies quickly; on the other hand, a single sound and light device lacks linkage, and its warning effect is weak. Drivers may fail to take effective risk avoidance measures due to untimely information reception or insufficient warning intensity.

[0003] In addition, existing technologies have significant deficiencies in the intelligence and flexibility of traffic control. Traditional systems are mostly independent devices with single functions, lacking overall coordination capabilities, and it is difficult to adjust warning and control strategies according to specific accident scenarios. Especially in the case of complex and changeable emergencies, existing equipment is difficult to quickly and accurately guide vehicles to slow down or stop, resulting in an increased risk of chain collisions in the tunnel entrance area.

[0004] In the face of these problems, an integrated, intelligent and efficient traffic control solution is urgently needed, which can not only respond quickly to emergencies, but also effectively avoid driver panic and reduce the possibility of secondary accidents through step-by-step enhanced early warning measures.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a flexible blocking traffic control system for tunnel entrances, which significantly improves the traffic safety and control efficiency of tunnel entrances through multi-level early warning mechanisms, flexible blocking technology and optimization algorithms. The system combines the "early warning-slowdown-stop" grading principle and deep learning algorithm to analyze traffic conditions in real time, dynamically adjust the equipment mode, and ensure that warning information is delivered accurately and in a timely manner. The intelligent laser projection system and high-definition ground projection generate dynamic virtual gratings to achieve flexible control and avoid the limitations of traditional physical facilities. The application of distributed data acquisition and multi-objective optimization algorithms reduces equipment energy consumption and false triggering, reduces the cognitive burden of drivers, and at the same time achieves efficient, economical and sustainable traffic management methods to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a flexible blocking traffic control system for a tunnel entrance, comprising a warning information release module, a graded warning control module, a flexible blocking guidance module, an intelligent device control module and an emergency linkage processing module;

[0008] The warning information release module uses the sound and light warning units installed at the tunnel entrance and in front of it and the roadside intelligent guidance warning units to display information through variable information boards and play voice through directional speakers to release warning information about abnormal tunnel conditions in the direction of oncoming vehicles, reminding drivers to pay attention to potential dangers ahead;

[0009] The hierarchical warning control module calculates the real-time distance and speed between the vehicle and the tunnel entrance based on the advanced path planning algorithm, and triggers the warning devices in the warning warning area, warning deceleration area and warning parking area in sequence, achieving a step-by-step warning effect and guiding the vehicle to gradually slow down;

[0010] The flexible blocking and guiding module uses an intelligent laser projection system to project a laser virtual grating at the tunnel entrance. At the same time, it uses industrial-grade high-definition ground projection to display "No Entry" or other dynamic warning signs on the ground. Combined with the significant red signal of the tunnel entrance lane indicator, a flexible blocking area is formed to guide vehicles to park in the designated area.

[0011] The intelligent equipment control module uses a deep learning-based traffic flow prediction algorithm to analyze the number and speed of incoming vehicles, adjust the working mode and warning level of each device in real time, and optimize the traffic control effect;

[0012] After receiving the trigger signal of an accident in the tunnel, the emergency linkage processing module links all traffic control equipment through the remote control platform, updates the warning information in real time, and synchronizes the accident-related data to the traffic management center for further emergency processing.

[0013] Preferably, vehicle detection sensors are used to obtain real-time traffic condition data at the tunnel entrance and in front of it, including oncoming vehicle speed, traffic density and ambient light intensity; based on the traffic condition data, a dynamic warning information optimization algorithm is used to adjust the output mode of the sound and light early warning warning unit and the roadside intelligent induction warning unit, including dynamically changing the warning information display frequency of the variable information board according to the vehicle speed, and controlling the playback interval and content intensity of the directional speaker voice information, to ensure the accuracy and adaptability of information transmission.

[0014] Preferably, multiple distributed sound and light warning units and roadside intelligent induction warning units are set up, and a communication unit is embedded in each unit to enable real-time data sharing between the units through wireless communication protocols; an integrated controller is used to coordinate the linkage work of adjacent units based on the vehicle's moving direction and position, forming variable information board warning information and directional horn voice warnings that are progressively displayed from near to far, gradually strengthening the driver's perception of danger and guiding him to take corresponding measures.

[0015] Preferably, the dynamic warning information optimization algorithm is one of a fuzzy logic algorithm, a genetic algorithm, and a deep reinforcement learning algorithm.

[0016] Preferably, the real-time distance and speed between the vehicle and the tunnel entrance are calculated based on the advanced path planning algorithm, and the vehicle is guided to gradually slow down. The specific steps are as follows:

[0017] Through high-precision laser radar and visual sensor array, the initial position P0 (x0, y0) and velocity vector of the approaching vehicle are collected in real time Combined with the location of the tunnel entrance P T (x T ,y T ) and the current traffic environment parameters to establish a model, and the vehicle motion equation is expressed as:

[0018]

[0019] , where: P t (x t ,y t ) is the predicted position of the vehicle, Δs is the prediction interval, is the acceleration vector of the vehicle, which is calculated by the vehicle driving recorder and sensors;

[0020] Based on the A* path planning algorithm and combined with traffic flow constraints, the optimal driving path L = {P0, P1, ..., P T}, calculate the vehicle's current position P t To the tunnel entrance P T The dynamic distance d(P t , P T ), the expression for calculation is:

[0021]

[0022] , i represents the index of a discrete point on the path, which is used to identify the specific point position on the path, and n represents the total number of all discrete points on the path;

[0023] Calculating target speed using the extended speed-distance control model The expression to be calculated is:

[0024]

[0025] , where: g is the acceleration of gravity, μ is the road friction coefficient;

[0026] According to the dynamic distance d(P t , P T ) and target speed The warning area where the vehicle is located is determined by segmented threshold logic, and the determination expression is:

[0027]

[0028] , where D1 and D2 are the boundary distance thresholds. After judgment, the warning equipment in the corresponding area is triggered, including the sound and light warning unit, the roadside guidance unit and the laser projection system, to achieve a step-by-step enhanced warning effect. Region represents the type of warning area where the vehicle is currently located. It is a logical variable or classification result, which is used to judge the real-time status of the vehicle and trigger the corresponding warning equipment.

[0029] Preferably, the intelligent laser projection system generates a dynamically changing virtual grating, the intensity and spacing of which are dynamically adjusted by analyzing real-time traffic data to adapt to different traffic flow densities and vehicle speeds; industrial-grade high-definition ground projection simultaneously displays scrolling warning signs, including dynamic "no entry" signs and additional accident information.

[0030] Preferably, the intelligent laser projection system forms a three-dimensional blocking effect through a multi-layer grating design, the first layer displays a red warning line, the second layer displays dynamic optical text information, and the third layer simulates the physical barrier effect;

[0031] Combining industrial-grade high-definition ground projection and tunnel entrance lane indicators, the three are synchronized to make the grating, ground lighting signs and "stop" signals change dynamically in a timely manner, gradually enhancing the blocking effect and clarifying the parking area.

[0032] Preferably, the number and speed of incoming vehicles are analyzed by a traffic flow prediction algorithm based on deep learning, and the working mode and warning level of each device are adjusted in real time. The specific steps are as follows:

[0033] Using high-resolution cameras and radar sensors in the road monitoring system, real-time traffic flow data, including the number of vehicles N t , speed distribution of each vehicle and vehicle spacing distribution The traffic flow state matrix is ​​constructed through the deep feature extraction model based on the convolutional neural network. The construction expression time of the traffic flow state matrix is:

[0034]

[0035] , where: S t is the real-time traffic flow feature matrix, which is used to describe the multi-dimensional state of the current traffic flow. t is an additional environmental parameter, f CNN It is a convolutional neural network model used to extract implicit spatial patterns and temporal features in traffic status;

[0036] Based on the traffic flow characteristic matrix S t , use the deep reinforcement learning algorithm to optimize the mode of each warning device and define the reward function R(S t , A t ) is:

[0037]

[0038] , where: i' represents the index of the i'th vehicle in the current traffic flow, N t represents the total number of all vehicles in the traffic flow at the current time point t, A t Select the action set for the device mode, N stopped is the number of vehicles that have completed safe parking. is the target speed calculated by the path planning algorithm, E penalty are energy consumption and delay penalty terms, λ1, λ2, λ3 are weight parameters that reflect the priorities of different goals, and the optimal strategy π is solved by deep reinforcement learning algorithm. * , the expression to be solved is:

[0039]

[0040] , where γ is a discount factor used to balance current and future rewards.

[0041] Preferably, dynamic data of the traffic environment, including vehicle speed, traffic density, ambient brightness and weather conditions, are collected through a distributed sensor network, and the collected dynamic data is transmitted to the central control system; the central control system dynamically adjusts the output mode using a multi-objective optimization algorithm, wherein the light intensity and flashing frequency of the sound and light warning unit are optimized according to the ambient brightness and the speed of the oncoming vehicle, and the voice playback content and volume of the roadside intelligent induction unit are adjusted in real time according to the traffic flow density and ambient noise, to ensure that the driver can obtain clear warning information in different environments.

[0042] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0043] The present invention significantly improves the traffic safety of tunnel entrances through the design of a multi-level warning mechanism, combined with a dynamic warning information optimization algorithm and flexible blocking technology. The warning system gradually transmits warning information to the driver from the long-distance warning area to the short-distance parking area according to the hierarchical principle of "warning-deceleration-parking", guiding him to slow down and park safely in time. At the same time, through the real-time analysis of the speed and number of incoming vehicles by the traffic flow prediction algorithm based on deep learning, the system can accurately judge the traffic conditions, and trigger the corresponding equipment mode and warning level according to the actual situation to ensure the timeliness and pertinence of information transmission. This scientific classification and intelligent warning method effectively avoids the driver's misjudgment and panic caused by untimely or overloaded information reception, reduces the probability of secondary accidents in the tunnel entrance area, and improves the safety of overall traffic control.

[0044] The present invention adopts flexible blocking technology, generates dynamic virtual gratings through intelligent laser projection systems and high-definition ground projection equipment, and combines the linkage effect of tunnel entrance lane indicators to achieve flexible control of tunnel entrances. This flexible blocking method can not only be quickly deployed and adjusted, but also avoids the physical damage to vehicles or long-term impacts on traffic flow caused by traditional physical blocking facilities. In addition, the system dynamically adjusts the output mode of the sound and light warning unit and the roadside intelligent induction unit through a multi-objective optimization algorithm, so that the equipment can adapt to different traffic environments and flow conditions, thereby enabling the present invention to quickly respond to complex and changeable traffic scenes, significantly improving the efficiency and accuracy of tunnel traffic control.

[0045] The present invention effectively reduces resource consumption and misoperation in traffic control by combining distributed data collection and optimization algorithms. The sensor network collects multi-dimensional data such as ambient brightness, oncoming vehicle speed, traffic density and background noise in real time, and analyzes these data through deep learning and multi-objective optimization algorithms to dynamically adjust the working mode of the equipment. In addition, algorithm optimization can also avoid unnecessary warnings caused by redundant information or false triggering of the equipment, thereby reducing the cognitive burden of the driver and ensuring the efficient transmission of the warning effect. By reducing resource consumption and misoperation, a more economical and sustainable traffic management method is achieved while ensuring safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0047] Figure 1 It is a module schematic diagram of the flexible blocking traffic control system at the tunnel entrance of the present invention. DETAILED DESCRIPTION

[0048] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0049] The present invention provides Figure 1 The tunnel entrance flexible blocking traffic control system shown includes a warning information release module, a graded warning control module, a flexible blocking guidance module, an intelligent device control module, and an emergency linkage processing module;

[0050] The warning information release module uses the sound and light warning units installed at the tunnel entrance and in front of it and the roadside intelligent guidance warning units to display information through variable information boards and play voice through directional speakers to release warning information about abnormal tunnel conditions in the direction of oncoming vehicles, reminding drivers to pay attention to potential dangers ahead;

[0051] Use vehicle detection sensors to obtain real-time traffic data at the tunnel entrance and in front of it, including oncoming vehicle speed, traffic density, and ambient light intensity;

[0052] According to traffic condition data, a dynamic warning information optimization algorithm is used to adjust the output mode of the sound and light warning unit and the roadside intelligent guidance warning unit, including dynamically changing the warning information display frequency of the variable information board according to the vehicle speed, and controlling the playback interval and content intensity of the directional speaker voice information to ensure the accuracy and adaptability of information transmission;

[0053] By collecting data through sensors and adjusting the output mode through optimization algorithms, it can dynamically adapt to different traffic environments, improve the visibility and audibility of warning information, enhance the driver's attention, and avoid misjudgment caused by information overload or insufficient information.

[0054] Set up multiple distributed sound and light warning units and roadside intelligent guidance warning units, and embed communication units in each unit to enable real-time data sharing between units through wireless communication protocols;

[0055] The integrated controller coordinates the adjacent units to work together based on the vehicle's moving direction and position, forming variable information board warning information and directional horn voice warnings that are displayed progressively from near to far, gradually strengthening the driver's perception of danger and guiding him to take corresponding measures;

[0056] The distributed linkage design forms a consistent and progressive warning effect through wireless communication and coordinated control, which helps the driver to understand the degree of danger in a timely manner and gradually take correct driving behavior, significantly improving the transmission effect of warning information and the overall safety of tunnel management.

[0057] Dynamic warning information optimization algorithms may include but are not limited to dynamic warning optimization based on fuzzy logic, genetic algorithm optimization control, and warning mode adaptive adjustment algorithms based on deep reinforcement learning. The fuzzy logic algorithm makes real-time adjustments to the warning information output parameters (such as frequency, brightness, and volume) by comprehensively analyzing fuzzy data variables such as vehicle speed, traffic flow, and ambient lighting; the genetic algorithm continuously optimizes the warning information output scheme based on historical data through the bionic evolution mechanism to find the optimal combination for different scenarios; the algorithm based on deep reinforcement learning uses large-scale traffic data for model training and automatically learns how to switch to the most appropriate warning mode according to the real-time scenario, such as adjusting the information display rhythm or choosing a more efficient warning method, so as to ensure that the driver quickly receives and correctly understands the warning information.

[0058] The hierarchical warning control module calculates the real-time distance and speed between the vehicle and the tunnel entrance based on the advanced path planning algorithm, and triggers the warning devices in the warning warning area, warning deceleration area and warning parking area in sequence, achieving a step-by-step warning effect and guiding the vehicle to gradually slow down;

[0059] The real-time distance and speed between the vehicle and the tunnel entrance are calculated based on the advanced path planning algorithm, and the vehicle is guided to gradually slow down. The specific steps are as follows:

[0060] Through high-precision laser radar and visual sensor array, the initial position P0 (x0, y0) and velocity vector of the approaching vehicle are collected in real time Combined with the location of the tunnel entrance P T (x T ,y T ) and current traffic environment parameters (such as road friction coefficient) to establish a model. According to the vehicle's motion equation expression:

[0061]

[0062] , where: P t (x t ,y t ) is the predicted position of the vehicle, Δs is the prediction interval, is the acceleration vector of the vehicle, which is calculated by the vehicle driving recorder and sensors;

[0063] The purpose of this step is to provide real-time and accurate dynamic input data for the subsequent path planning algorithm, ensure that the calculated distance and speed are consistent with the actual situation, and lay the foundation for switching the warning level.

[0064] Based on the A* path planning algorithm and combined with traffic flow constraints, the optimal driving path L = {P0, P1, ..., P T}, calculate the vehicle's current position P t To the tunnel entrance P T The dynamic distance d(P t , P T ), the expression for calculation is:

[0065]

[0066] , i represents the index of a discrete point on the path, which is used to identify the specific point position on the path, n represents the total number of all discrete points on the path, and the target speed is calculated using the extended speed-distance control model The expression to be calculated is:

[0067]

[0068] , where: g is the acceleration of gravity, μ is the road friction coefficient;

[0069] The purpose of this step is to plan the best driving path for the vehicle and guide the vehicle to an appropriate speed range by dynamically adjusting the target speed so as to trigger the corresponding warning area.

[0070] According to the dynamic distance d(P t , P T ) and target speed The warning area where the vehicle is located is determined by segmented threshold logic, and the determination expression is:

[0071]

[0072] , where D1 and D2 are the boundary distance thresholds. After the judgment, the warning equipment in the corresponding area is triggered, including the sound and light warning unit, the roadside guidance unit and the laser projection system, to achieve a step-by-step enhanced warning effect. Region represents the type of warning area the vehicle is currently in. It is a logical variable or classification result, which is used to judge the real-time status of the vehicle and trigger the corresponding warning equipment;

[0073] The purpose of this step is to gradually convey danger signals to the driver and guide him to stop the car through area judgment and equipment triggering, thereby achieving efficient and safe traffic control at the tunnel entrance.

[0074] The flexible blocking and guiding module uses an intelligent laser projection system to project a laser virtual grating at the tunnel entrance. At the same time, it uses industrial-grade high-definition ground projection to display "No Entry" or other dynamic warning signs on the ground. Combined with the significant red signal of the tunnel entrance lane indicator, a flexible blocking area is formed to guide vehicles to park in the designated area.

[0075] The intelligent laser projection system uses a multi-layer grating design to create a three-dimensional blocking effect. The first layer displays a red warning line, the second layer displays dynamic optical text information, and the third layer simulates the physical obstacle effect.

[0076] Combining industrial-grade high-definition ground projection and tunnel entrance lane indicators, the three are synchronized to make the grating, ground lighting signs and "stop" signals change dynamically in a timely manner, gradually enhancing the blocking effect and clarifying the parking area.

[0077] Through the three-dimensional presentation of multi-layer gratings and the time linkage of multiple devices, dynamic guidance and multi-level blocking effects of the flexible blocking area can be achieved, which helps to quickly attract the driver's attention in an emergency, improve parking response speed and tunnel area management safety.

[0078] The intelligent equipment control module uses a deep learning-based traffic flow prediction algorithm to analyze the number and speed of incoming vehicles, adjust the working mode and warning level of each device in real time, and optimize the traffic control effect;

[0079] The traffic flow prediction algorithm based on deep learning is used to analyze the number and speed of incoming vehicles, and adjust the working mode and warning level of each device in real time. The specific steps are as follows:

[0080] Using high-resolution cameras and radar sensors in the road monitoring system, real-time traffic flow data, including the number of vehicles N t , speed distribution of each vehicle and vehicle spacing distribution The traffic flow state matrix is ​​constructed by using a deep feature extraction model based on a convolutional neural network (CNN). The construction expression time of the traffic flow state matrix is:

[0081]

[0082] , where: S t is the real-time traffic flow feature matrix, which is used to describe the multi-dimensional state of the current traffic flow. t For additional environmental parameters, such as road friction coefficient, light intensity and weather conditions, f CNN It is a convolutional neural network model used to extract implicit spatial patterns and temporal features in traffic status;

[0083] Through deep feature extraction of high-dimensional traffic flow data, dynamic and complex traffic conditions are converted into easy-to-analyze feature matrices, providing accurate input for subsequent equipment mode optimization.

[0084] Based on the traffic flow characteristic matrix S t , use the deep reinforcement learning (DRL) algorithm to optimize the mode of each warning device and define the reward function R(S t , A t ) is:

[0085]

[0086] , where: i' represents the index of the i'th vehicle in the current traffic flow, N t represents the total number of all vehicles in the traffic flow at the current time point t, A t Select the action set for the device mode, N stopped is the number of vehicles that have completed safe parking. is the target speed calculated by the path planning algorithm, E penalty are energy consumption and delay penalty terms, λ1, λ2, λ3 are weight parameters that reflect the priorities of different goals, and the optimal strategy π is solved by deep reinforcement learning algorithm. * , the expression to be solved is:

[0087]

[0088] , where γ is a discount factor used to balance current and future rewards;

[0089] By optimizing the device mode and warning level in real time, we can ensure that each device works efficiently under different traffic flow conditions, while reducing unnecessary energy consumption and false triggering, significantly improving the overall effect of traffic control.

[0090] The emergency linkage processing module, upon receiving the trigger signal of an accident in the tunnel, links all traffic control equipment through the remote control platform, updates the warning information in real time, and synchronizes the accident-related data to the traffic management center for further emergency processing;

[0091] Collect dynamic data of the traffic environment, including vehicle speed, traffic density, ambient brightness and weather conditions, through a distributed sensor network and transmit this data to a central control system;

[0092] The central control system uses a multi-objective optimization algorithm to dynamically adjust the output mode. The light intensity and flashing frequency of the sound and light warning unit are optimized according to the ambient brightness and the speed of the oncoming vehicle. The voice playback content and volume of the roadside intelligent induction unit are adjusted in real time according to the traffic flow density and ambient noise, thereby ensuring that the driver can receive clear warning information in different environments.

[0093] Through distributed data collection and dynamic adjustment of multi-objective optimization algorithms, the system can provide appropriate sound and light information in complex and changeable traffic environments, avoiding panic caused by excessive information stimulation and misjudgment caused by insufficient information, thereby effectively improving warning effects and driving safety.

[0094] The specific steps of the multi-objective optimization algorithm to dynamically adjust the output mode include:

[0095] First, real-time traffic and environmental data, such as ambient brightness, oncoming vehicle speed, traffic density, and ambient noise, are collected and input into a multi-objective optimization model;

[0096] Then, the optimization objective function is defined, such as maximizing the visibility and audibility of the warning information while minimizing energy consumption and the possibility of driver distraction;

[0097] Next, the importance of different objectives is adjusted through weight allocation, and a genetic algorithm or a particle swarm optimization algorithm is used to iteratively solve in the search space to generate the optimal parameter combination, such as the light intensity and flashing frequency of the sound and light unit and the volume and speech content of the induction unit;

[0098] Finally, the central control system dynamically adjusts the equipment output mode according to the optimization results, and continuously updates the optimization parameters through the feedback loop to adapt to changes in the traffic environment in real time.

[0099] Implementation method 1: Distributed data collection and intelligent decision-making;

[0100] In order to achieve efficient traffic management, the system deploys multiple distributed sensor units at the tunnel entrance and in front of it. These units are composed of light sensors, vehicle detection radars, noise detectors, and temperature and humidity sensors. The role of the light sensor is to detect the ambient brightness to adapt to different lighting conditions during the day, night, and in bad weather (such as fog or heavy rain); the vehicle detection radar is used to capture the speed, distance, and traffic density data of the oncoming vehicle; the noise detector is used to perceive the noise level of the surrounding environment in real time, especially the sound interference in construction, high wind speed or heavy rain. In addition, the temperature and humidity sensors can detect weather conditions and further assist in determining whether there are conditions that affect driving safety, such as icy or slippery roads.

[0101] The collected multi-dimensional data is transmitted to the central control system through the wireless communication module. The system processes and classifies the data in real time based on the deep learning model, such as analyzing the number of vehicles and speed distribution, so as to determine the current traffic flow status. At the same time, the system will also combine historical data to predict future traffic changes and determine whether there are emergencies (such as abnormal congestion or a sharp increase in vehicle traffic). For example, when the system detects that the light level is reduced, the vehicle speed is higher than the safety threshold, and the traffic density is high, the emergency warning mode will be triggered.

[0102] In the warning mode, the central control system will dynamically adjust the output parameters of each unit according to the results of data analysis. For example, when the ambient brightness is low, the light intensity of the sound and light warning unit will automatically increase, and the flashing frequency will also increase to attract the driver's attention; in a high-noise environment, the voice playback volume of the roadside intelligent induction unit will increase, and key prompts may be repeated to ensure that the driver can clearly receive the warning information. In addition, the system can also determine whether the voice content needs to be dynamically adjusted through algorithm optimization, such as adding more specific guidance information, such as "The tunnel ahead is closed, please park 200 meters away." This distributed data collection and intelligent decision-making mode can not only adapt to complex and changing traffic environments, but also significantly improve the accuracy and timeliness of warnings, thereby effectively reducing the risk of traffic accidents in tunnel areas.

[0103] Implementation method 2: linkage between flexible blocking and dynamic warning;

[0104] At the entrance of the tunnel, the system deploys an intelligent laser projection system, industrial-grade high-definition ground projection equipment, and tunnel entrance lane indicators. These devices operate in conjunction to form a highly intelligent flexible blocking area. The intelligent laser projection system can generate a dynamic virtual grating, which achieves a flexible blocking effect through multiple layers of dynamic changes. For example, the first layer of grating displays a fixed red warning line, the second layer simulates the physical obstacle effect (such as vertically moving light beams) through dynamically changing optical graphics, and the third layer displays dynamic text warning information (such as "No passage" or "Slow down ahead of the accident"). This multi-level optical design can greatly enhance the visual impact and guide the driver to take correct driving behavior.

[0105] Industrial-grade high-definition ground projection equipment further enhances the warning effect. The projection content includes the words "No Entry" and graphic signs such as dynamic arrows, which are displayed on the ground at the entrance of the tunnel, forming an obvious traffic control signal. When it is detected that the approaching vehicle is close and the speed is high, the projection content will automatically adjust to a more eye-catching dynamic flashing mode, and add a countdown sign when necessary to remind the driver that he is about to enter the flexible blocking area. In addition, the lane indicator at the entrance of the tunnel strengthens the warning effect through a prominent red "Stop" prompt, allowing the driver to gradually slow down and stop safely under the guidance of multiple visual signals.

[0106] This linkage design of flexible blocking and dynamic warning also has adaptive functions. When the traffic flow is high, the interval between the laser gratings will automatically decrease to form a denser optical blocking area; and when the ambient light is strong, the ground projection device will increase the brightness to ensure visibility. Through this multi-device linkage and real-time adjustment method, the system can not only effectively prevent vehicles from entering the tunnel by mistake, but also minimize the damage and inconvenience that may be caused by traditional physical blocking methods, thereby significantly improving the traffic control capabilities of the tunnel entrance.

[0107] Implementation method 3: Enhanced warning effect based on multi-objective optimization;

[0108] In the central control system, a multi-objective optimization algorithm is used to adjust the output mode of the warning device in real time to ensure the maximum warning effect under different traffic conditions. The optimization goals include: improving the visibility and audibility of the warning information, reducing the information processing burden on the driver, reducing energy consumption, and minimizing the possibility of false triggering. The system first collects traffic environment data, such as traffic density, oncoming vehicle speed, ambient light intensity, and background noise, and inputs this data into the optimization algorithm model.

[0109] The optimization algorithm determines the importance of each target through a weight distribution mechanism. For example, in low-light environments, the system will give priority to improving the brightness and flashing frequency of the sound and light unit; in high-noise conditions, the system will pay more attention to improving the volume and clarity of voice playback. Specifically, when the light sensor detects that the ambient brightness drops below the set threshold, the light intensity of the sound and light warning unit will increase to the maximum value, and the flashing frequency will increase from 1 per second to 3 times per second; and when the noise detector detects that the background noise exceeds 70 decibels, the volume of the roadside intelligent induction unit will increase to 1.5 times the original, and the warning voice content will be played repeatedly. In addition, the algorithm will dynamically adjust the content of the voice playback and add more guidance information in high-traffic scenarios, such as "Be careful to slow down and keep a safe distance."

[0110] In order to further improve the optimization effect, the system will use the real-time feedback mechanism to iteratively adjust the optimization parameters according to the actual control effect. For example, when the proportion of vehicles that successfully stop before entering the tunnel entrance is less than 90%, the system will increase the density of the laser grating and the dynamic change frequency of the projection logo to enhance the visual warning effect. Through this dynamic adjustment based on multi-objective optimization, the system can continue to provide efficient and accurate warning information in a complex and changing traffic environment, effectively reduce misjudgments caused by insufficient warnings or information overload, and improve traffic safety in the tunnel entrance area.

[0111] The present invention significantly improves the traffic safety of tunnel entrances through the design of a multi-level warning mechanism, combined with a dynamic warning information optimization algorithm and flexible blocking technology. The warning system gradually transmits warning information to the driver from the long-distance warning area to the short-distance parking area according to the hierarchical principle of "warning-deceleration-parking", guiding him to slow down and park safely in time. At the same time, through the real-time analysis of the speed and number of incoming vehicles by the traffic flow prediction algorithm based on deep learning, the system can accurately judge the traffic conditions, and trigger the corresponding equipment mode and warning level according to the actual situation to ensure the timeliness and pertinence of information transmission. This scientific classification and intelligent warning method effectively avoids the driver's misjudgment and panic caused by untimely or overloaded information reception, reduces the probability of secondary accidents in the tunnel entrance area, and improves the safety of overall traffic control.

[0112] The present invention adopts flexible blocking technology, generates dynamic virtual gratings through intelligent laser projection systems and high-definition ground projection equipment, and combines the linkage effect of tunnel entrance lane indicators to achieve flexible control of tunnel entrances. This flexible blocking method can not only be quickly deployed and adjusted, but also avoids the physical damage to vehicles or long-term impacts on traffic flow that traditional physical blocking facilities may cause. In addition, the system dynamically adjusts the output mode of the sound and light warning unit and the roadside intelligent induction unit through a multi-objective optimization algorithm, so that the equipment can adapt to different traffic environments and traffic conditions, such as adjusting the light intensity and flashing frequency to cope with low light conditions, or increasing the voice playback volume to overcome high noise interference. The combination of intelligence and flexibility enables the present invention to quickly respond to complex and changeable traffic scenarios, significantly improving the efficiency and accuracy of tunnel traffic control.

[0113] The present invention effectively reduces resource consumption and misoperation in traffic control through the combination of distributed data collection and optimization algorithms. The sensor network collects multi-dimensional data such as ambient brightness, oncoming vehicle speed, traffic density and background noise in real time, analyzes these data through deep learning and multi-objective optimization algorithms, and dynamically adjusts the working mode of the equipment. For example, in an environment with low traffic or low noise, the system can reduce the light intensity and volume of the sound and light unit, thereby reducing the energy consumption of the equipment; in an emergency, the system quickly switches to a high-intensity warning mode to ensure that the driver receives accurate warning information. In addition, algorithm optimization can also avoid unnecessary warnings caused by redundant information or false triggering of the equipment, thereby reducing the cognitive burden of the driver and ensuring the efficient transmission of the warning effect. By reducing resource consumption and misoperation, the present invention achieves a more economical and sustainable traffic management method while ensuring safety.

[0114] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. Tunnel entrance flexible blocking traffic control system, characterized by: It includes warning information release module, graded warning control module, flexible blocking guidance module, intelligent equipment control module and emergency linkage processing module; The warning information release module uses the sound and light warning units installed at the tunnel entrance and in front of it and the roadside intelligent guidance warning units to display information through variable information boards and play voice through directional speakers to release warning information about abnormal tunnel conditions in the direction of oncoming vehicles, reminding drivers to pay attention to potential dangers ahead; The hierarchical warning control module calculates the real-time distance and speed between the vehicle and the tunnel entrance based on the advanced path planning algorithm, and triggers the warning devices in the warning warning area, warning deceleration area and warning parking area in sequence, achieving a step-by-step warning effect and guiding the vehicle to gradually slow down; The flexible blocking and guiding module uses an intelligent laser projection system to project a laser virtual grating at the tunnel entrance. At the same time, it uses industrial-grade high-definition ground projection to display "No Entry" or other dynamic warning signs on the ground. Combined with the significant red signal of the tunnel entrance lane indicator, a flexible blocking area is formed to guide vehicles to park in the designated area. The intelligent equipment control module uses a deep learning-based traffic flow prediction algorithm to analyze the number and speed of incoming vehicles, adjust the working mode and warning level of each device in real time, and optimize the traffic control effect; After receiving the trigger signal of an accident in the tunnel, the emergency linkage processing module links all traffic control equipment through the remote control platform, updates the warning information in real time, and synchronizes the accident-related data to the traffic management center for further emergency processing.

2. The tunnel entrance flexible blocking traffic control system according to claim 1 is characterized in that: Vehicle detection sensors are used to obtain real-time traffic condition data at the tunnel entrance and in front of it, including oncoming vehicle speed, traffic density and ambient light intensity. Based on the traffic condition data, a dynamic warning information optimization algorithm is used to adjust the output mode of the sound and light warning unit and the roadside intelligent induction warning unit, including dynamically changing the warning information display frequency of the variable information board according to the vehicle speed, and controlling the playback interval and content intensity of the directional speaker voice information to ensure the accuracy and adaptability of information transmission.

3. The tunnel entrance flexible blocking traffic control system according to claim 1 is characterized in that: Multiple distributed sound and light warning units and roadside intelligent guidance warning units are set up, and a communication unit is embedded in each unit to enable real-time data sharing between units through wireless communication protocols; an integrated controller is used to coordinate the linkage work of adjacent units based on the vehicle's movement direction and position, forming variable information board warning information and directional horn voice warnings that are progressively displayed from near to far, gradually enhancing the driver's perception of danger and guiding him to take corresponding measures.

4. The tunnel entrance flexible blocking traffic control system according to claim 2 is characterized in that: The dynamic warning information optimization algorithm is one of the fuzzy logic algorithm, genetic algorithm, and deep reinforcement learning algorithm.

5. The tunnel entrance flexible blocking traffic control system according to claim 1 is characterized in that: The real-time distance and speed between the vehicle and the tunnel entrance are calculated based on the advanced path planning algorithm, and the vehicle is guided to gradually slow down. The specific steps are as follows: Through high-precision laser radar and visual sensor array, the initial position P0 (x0, y0) and velocity vector of the approaching vehicle are collected in real time Combined with the location of the tunnel entrance P T (x T ,y T ) and the current traffic environment parameters to establish a model, and the vehicle motion equation is expressed as: ， Where: P t (x t ,y t ) is the predicted position of the vehicle, Δs is the prediction interval, is the acceleration vector of the vehicle, which is calculated by the vehicle driving recorder and sensors; Based on the A* path planning algorithm and combined with traffic flow constraints, the optimal driving path L = {P0, P1, ..., P T }, calculate the vehicle's current position P t To the tunnel entrance P T The dynamic distance d(P t , P T ), the expression for calculation is: ， i represents the index of a discrete point on the path, which is used to identify the specific point position on the path, and n represents the total number of all discrete points on the path; Calculating target speed using the extended speed-distance control model The expression to be calculated is: ， Where: g is the acceleration of gravity, μ is the road friction coefficient; According to the dynamic distance d(P t , P T ) and target speed The warning area where the vehicle is located is determined by segmented threshold logic, and the determination expression is: ， Among them, D1 and D2 are the boundary distance thresholds. After judgment, the warning equipment in the corresponding area is triggered, including the sound and light warning unit, the roadside guidance unit and the laser projection system, to achieve a step-by-step enhanced warning effect. Region indicates the type of warning area where the vehicle is currently located, which is used to judge the real-time status of the vehicle and trigger the corresponding warning equipment.

6. The tunnel entrance flexible blocking traffic control system according to claim 1 is characterized in that: The intelligent laser projection system generates a dynamically changing virtual grating. The intensity and spacing of the grating are dynamically adjusted by analyzing real-time traffic data to adapt to different traffic flow densities and vehicle speeds. The industrial-grade high-definition ground projection simultaneously displays rolling warning signs, including dynamic "no passage" signs and additional accident information.

7. The tunnel entrance flexible blocking traffic control system according to claim 1 is characterized in that: The intelligent laser projection system uses a multi-layer grating design to create a three-dimensional blocking effect. The first layer displays a red warning line, the second layer displays dynamic optical text information, and the third layer simulates the physical obstacle effect. Combining industrial-grade high-definition ground projection and tunnel entrance lane indicators, the three are synchronized to make the grating, ground lighting signs and "stop" signals change dynamically in a timely manner, gradually enhancing the blocking effect and clarifying the parking area.

8. The tunnel entrance flexible blocking traffic control system according to claim 5 is characterized in that: The traffic flow prediction algorithm based on deep learning is used to analyze the number and speed of incoming vehicles, and adjust the working mode and warning level of each device in real time. The specific steps are as follows: Using high-resolution cameras and radar sensors in the road monitoring system, real-time traffic flow data, including the number of vehicles N t , speed distribution of each vehicle and vehicle spacing distribution The traffic flow state matrix is ​​constructed through the deep feature extraction model based on the convolutional neural network. The construction expression time of the traffic flow state matrix is: ， Where: S t is the real-time traffic flow feature matrix, which is used to describe the multi-dimensional state of the current traffic flow. t is an additional environmental parameter, f CNN It is a convolutional neural network model used to extract implicit spatial patterns and temporal features in traffic status; Based on the traffic flow characteristic matrix S t , use the deep reinforcement learning algorithm to optimize the mode of each warning device and define the reward function R(S t , A t ) is: ， Where: i' represents the index of the i'th vehicle in the current traffic flow, N t represents the total number of all vehicles in the traffic flow at the current time point t, A t Select the action set for the device mode, N stopped is the number of vehicles that have completed safe parking. is the target speed calculated by the path planning algorithm, E penalty are energy consumption and delay penalty terms, λ1, λ2, λ3 are weight parameters that reflect the priorities of different goals, and the optimal strategy π is solved by deep reinforcement learning algorithm. * , the expression to be solved is: ， Among them, γ is a discount factor used to balance current and future rewards.

9. The tunnel entrance flexible blocking traffic control system according to claim 1, characterized in that: Dynamic data of the traffic environment, including vehicle speed, traffic density, ambient brightness and weather conditions, is collected through a distributed sensor network and transmitted to the central control system. The central control system uses a multi-objective optimization algorithm to dynamically adjust the output mode, in which the light intensity and flashing frequency of the sound and light warning unit are optimized according to the ambient brightness and oncoming vehicle speed, and the voice playback content and volume of the roadside intelligent induction unit are adjusted in real time according to the traffic flow density and ambient noise, ensuring that drivers can receive clear warning information in different environments.

Images