A traffic situation regulation method and system for a highway risk point

Abstract

The application discloses a kind of expressway risk point traffic situation regulation and control method and system, it is related to expressway risk monitoring technical field, including obtaining the multi-source perception information of expressway along line.The application realizes the quantitative space-time modeling of risk state by fusing meteorological, pavement, traffic flow multidimensional perception information, simultaneously introduces driver behavior response constraint and double-layer game mechanism, fully considers the actual traffic characteristics in the process of control instruction landing, can output the optimal regulation and control scheme of matching for dynamic changing risk state.Compared with traditional static control scheme, the regulation and control method of the application can realize the active closed-loop dynamic control of expressway risk point situation, both can effectively reduce the collision accident probability of risk point area, also can avoid unnecessary traffic efficiency loss caused by excessive control, give consideration to traffic safety and road network operation efficiency.

Description

Technical Field

[0001] This invention relates to the field of highway risk monitoring technology, and in particular to a method and system for controlling traffic conditions at highway risk points. Background Technology

[0002] In recent years, my country's expressway network construction has continued to advance, with the total mileage constantly increasing. Simultaneously, the number of motor vehicles has maintained steady growth, leading to a continuous increase in daily traffic flow on expressways. This has increased the pressure on the road network and consequently, a rise in various potential safety risks. Expressways are affected by factors such as terrain, weather, and traffic conditions, resulting in numerous inherent risk points prone to accidents, such as sections prone to fog, long downhill sections, and blind spots around curves. Furthermore, temporary construction areas and accident sites can also create dynamic risk points. These risk points experience rapidly changing traffic conditions and numerous uncertainties. If timely dynamic control and traffic congestion mitigation are not implemented, secondary accidents or widespread traffic jams can easily occur, seriously threatening traffic safety and the efficiency of the road network.

[0003] Currently, the management and control of risk points on highways mainly relies on static early warning and manual dispatch. Some management and control schemes fail to achieve multi-dimensional real-time perception of the traffic situation at risk points, and are also difficult to automatically match differentiated control strategies according to the dynamic changes in risk levels. This easily leads to problems such as delayed early warnings and mismatch between the control intensity and the actual risk. It is also impossible to coordinate and allocate traffic resources upstream and downstream of risk points at the overall level, making it difficult to achieve proactive and precise control of the traffic situation at risk points, and failing to meet the current development needs of intelligent safety management and control of highways. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing risk control methods in terms of poor overall coordination and allocation, and to propose a traffic situation control method and system for highway risk points.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a traffic situation control method for highway risk points, the control method comprising: Acquire multi-source sensing information along the highway; Based on multi-source sensing information along the expressway, assess the road condition information of each section of the expressway and obtain the comprehensive risk status of each section; By analyzing past driving behavior data, we can predict the driver's actual response to control commands under different risk scenarios and determine the driver's behavioral response constraint information. Based on multi-source sensing information along the highway, and combined with the comprehensive risk status of each section and driver behavior response constraints, a spatiotemporal risk field along the highway is constructed. Based on the spatiotemporal risk field along the highway, a two-layer game control framework is constructed between upper-level managers and lower-level drivers, outputting control instructions that include variable speed limits and ramp adjustment rates. The two-layer game-theoretic control architecture includes: upper-level managers formulating control strategies with the goal of minimizing the expected risk under the worst-case scenario; and lower-level driver groups responding to speed based on the control strategies to achieve group equilibrium. Based on the control instructions, the lane dynamics and emergency resources along the expressway are dynamically reconfigured and dispatched to obtain the real-time control status of the expressway. Based on the real-time control status along the highway and combined with multi-source sensing information along the highway, the spatiotemporal risk field along the highway is corrected for different control time sequence states. Once the risk is eliminated, a second-order recovery strategy with variable damping is adopted to gradually lift the speed limit control.

[0006] In one feasible approach, obtaining the comprehensive risk status of each road segment includes: A meteorological-road multi-physics coupled model is constructed, which includes the spatiotemporal dynamic evolution equation of visibility and / or the spatiotemporal degradation equation of road adhesion coefficient; Specifically, the spatiotemporal dynamic evolution equation of visibility is: Equation 1; In Equation 1, For spatial location At any moment The spatiotemporal visibility field, For partial differential operators, For spatial gradient operators, For wind speed vectors, The turbulent diffusion coefficient is... For the Laplace operator, The droplet settling coefficient is given. The number density of fog droplets. The density of the aerosol. This is a source of localized visibility deterioration caused by vehicle dust. in, Equation 2; In Equation 2, Dust coefficient, For traffic flow, Location of the accident site. To affect the radius, is the base of the natural logarithm; The spatiotemporal degradation equation of the road surface adhesion coefficient is: Equation 3; In Equation 3, For spatial location At any moment The spatiotemporal road surface adhesion coefficient, The coefficient of adhesion for dry road surfaces. This represents the maximum decrease in adhesion coefficient caused by debris coverage. Let be the viscosity-time constant of the mud spreading on the road surface. The location of the landslide center. The radius is affected by frictional degradation. The erosion coefficient of the adhesion coefficient due to rainfall. Rainfall intensity, This represents the cumulative rainfall.

[0007] In one feasible approach, determining the driver's behavioral response constraint information includes: Acquire past driving behavior data, perform driver behavior analysis on the past driving behavior data, and obtain the driver's driving status and performance. Based on the overall risk status of each road segment and the driver's driving performance, the driver's actual expected speed is determined. Based on the driver's actual expected speed and combined with the comprehensive risk status of each road segment, the probability of a collision is predicted when the driver drives at the actual expected speed. Based on the probability of a collision when the driver is driving at the actual expected speed, the driver's actual expected speed is constrained to obtain information on the driver's behavioral response constraints.

[0008] In one feasible approach, determining the driver's actual desired speed includes: Equation 4; In Equation 4, For the driver at all times Lower spatial position The actual expected speed For the published variable speed limit, For the speed deviation amplitude, Let be the driver's perceived weighting function for the collision probability. To help drivers perceive the probability of a collision. Due to the inherent impatience level of drivers, The influence coefficient of vehicle-road cooperative information. This measures drivers' level of trust in roadside information.

[0009] In one feasible approach, the construction of the spatiotemporal risk field along the highway includes: The area along the highway is divided into a continuous spatial domain. and time domain ,in, , An anisotropic risk field is constructed along the total length of the highway. : Equation 5; In Equation 5, For spatial location At any moment The spatiotemporal risk field value, The rate of change of the risk field over time. The dynamic diffusion coefficient is... For the spatial gradient of the risk field, The second spatial derivative of the risk field. For the risk movement velocity vector, For the collapse excitation function, For traffic accident excitation functions, This is the risk dissipation coefficient. This represents the risk field value at the current spatiotemporal location.

[0010] In one feasible approach, the upper-level manager formulates a control strategy with the goal of minimizing the expected risk under the worst-case scenario, including: based on Distance constructs a fuzzy set of driver response distribution And with the goal of minimizing the expected risk in the worst case, a distributed robust optimization model is constructed: Formula 6; In Equation 6, For the control strategy vector Find the minimum value, including the variable speed limit value. and ramp adjustment rate , For distribution set Seeking the ultimate limit, For the driver's response probability distribution, To predict the time domain, For the first The overall risk field value at any given moment. To control the cost weight matrix, Indicates distribution The expectations below Let be the control strategy vector at time i. To control the cost weight matrix; In addition, the distributed robust optimization model satisfies the chance constraint: ; ; In the formula, In distribution Find the infimum (worst-case scenario). As a probability measure, The actual speed at which the driver operates; , These are the lower and upper bounds of the velocity, respectively. This represents the maximum permissible risk value. The probability of allowing constraint violation.

[0011] In one feasible approach, the lower-level driver group responds to speed based on the control strategy and achieves group equilibrium, including: Constructing the variational inequality conditions for equilibrium among the lower-level driver group: , Equation 7; In Equation 7, Let be the driver's speed distribution vector. For the equilibrium velocity distribution vector, For vector difference transpose, For the velocity interaction matrix, For the velocity preference vector, To control the input matrix, For Lagrange multipliers, For matrix transpose, For the feasible speed set, This represents the velocity distribution vector of all drivers that meet the feasible conditions.

[0012] One feasible approach involves dynamic lane reconfiguration along the highway, including: Macro-traffic flow under lane dynamic reconstruction Extended model: Equation 8; In Equation 8, Traffic flow density, For actual execution speed, For the first The open / closed status of each lane. The total number of lanes For the source term to be merged into the ramp, For the Dirac function, Location of the accident site; The lane is blocked.

[0013] One feasible solution involves emergency resource allocation along the highway, including: Equation 9; Formula 10; In Equation 9, For resources Assigned to accident Decision variables, For the quantity of emergency resources, The number of accident sites. For travel time cost, For distance, To ensure the speed of rescue vehicles, To allow preparation time, Risk exposure weighting coefficient; For arrival time, For the time required to complete the clearance, Accident point At any moment The overall risk field value.

[0014] In a second aspect, the present invention also provides a traffic situation control system for highway risk points, employing the control method as described in any one of the first aspects, wherein the control system further includes: The data acquisition module is used to collect multi-source sensing information along the highway. The risk assessment module is used to calculate the comprehensive risk status of each road segment based on multi-source sensing information and construct the spatiotemporal risk field of the highway. The behavior analysis module is used to mine driver response patterns under different risk scenarios based on historical driving data and generate behavior response constraint information. A two-layer game control module is used to construct an upper and lower two-layer game architecture to solve for the optimal control command that includes the variable speed limit value and the ramp adjustment rate. The dynamic correction module is used to dynamically correct the spatiotemporal risk field by combining real-time updated sensing data and update the control strategy. The gradual recovery module is used to gradually restore the speed limit through a variable damping second-order recovery strategy after the risk is eliminated, thus completing the control exit.

[0015] The beneficial effects of this invention are as follows: This invention achieves quantitative spatiotemporal modeling of risk states by integrating multi-dimensional perception information from meteorology, road surface, and traffic flow. It also introduces driver behavior response constraints and a two-layer game mechanism, fully considering the actual traffic characteristics during the implementation of control commands. This allows it to output optimally matched control schemes for dynamically changing risk states. Compared to traditional static control schemes, this invention's control method enables proactive closed-loop dynamic control of highway risk points, effectively reducing the probability of collisions in risk areas while avoiding unnecessary traffic efficiency losses due to over-control, thus balancing traffic safety and road network operational efficiency. Attached Figure Description

[0016] Figure 1 This is an overall flowchart of a traffic situation control method for highway risk points provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the overall process of a traffic situation control method for highway risk points provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of multi-source sensing data acquisition for a traffic situation control method for highway risk points provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a feasible control command scheduling method for traffic situation control at highway risk points provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the meaning of "and / or" throughout the text includes three parallel solutions; for example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0019] Reference Figures 1 to 4 As shown, in this embodiment of the invention, to address the shortcomings of existing risk control methods in terms of overall coordination and allocation, a traffic situation control method for highway risk points is provided. This method achieves quantitative spatiotemporal modeling of risk states by integrating multi-dimensional perception information from meteorology, road surface, and traffic flow. Simultaneously, it introduces driver behavior response constraints and a two-layer game mechanism, fully considering the actual traffic characteristics during the implementation of control instructions. This allows it to output the optimal control scheme for dynamically changing risk states. Compared to traditional static control schemes, the control method of this invention enables proactive closed-loop dynamic control of highway risk point situations. It effectively reduces the probability of collisions in risk point areas while avoiding unnecessary traffic efficiency losses due to over-control, thus balancing traffic safety and road network operational efficiency.

[0020] like Figure 1 , Figure 2 and Figure 3 As shown, specifically, a traffic situation control method for highway risk points includes: acquiring multi-source sensing information along the highway; this can be achieved by deploying multi-source sensing devices (such as roadside communication units (e.g., roadside radar), video surveillance equipment (e.g., cameras), visibility sensors, wind speed and direction sensors, road surface condition sensors, and roadside communication units) along the highway, and then using these devices to collect real-time traffic flow parameters, meteorological parameters, road surface parameters, communication parameters, and driver behavior parameters. It should be noted that the traffic flow parameters include: vehicle speed, position, spacing, lane occupancy, and flow rate for each road segment; the meteorological parameters include visibility, wind speed and direction, rainfall intensity, and temperature; the road surface parameters include road surface adhesion coefficient, water film thickness, and debris coverage; the communication parameters include V2X communication latency and packet loss rate; and the driver behavior parameters include historical emergency braking frequency and following distance distribution.

[0021] Reference Figure 2 and Figure 3As shown, based on multi-source sensing information along the highway, the road condition information of each section of the highway is assessed to obtain the comprehensive risk status of each section. The comprehensive risk status includes at least one of the following: physical blockage and road surface friction degradation risk caused by landslides, visibility reduction risk caused by dense fog, and sudden drop in traffic flow speed and traffic saturation risk. It should be noted that in this embodiment, a meteorological-road multi-physics coupling model can be constructed to assess the spatiotemporal evolution of visibility and the degradation of the road surface adhesion coefficient, thereby obtaining the comprehensive risk status of each section. The meteorological-road multi-physics coupling model includes a spatiotemporal dynamic evolution equation for visibility and / or a spatiotemporal degradation equation for the road surface adhesion coefficient. That is, in this embodiment, the spatiotemporal dynamic evolution equation for visibility is used to evaluate the degree of visibility reduction at different locations and times along the highway; the spatiotemporal degradation equation for the road surface adhesion coefficient describes the decay process of the road surface friction coefficient under scenarios such as rainfall and landslide debris coverage. This allows for accurate acquisition of road condition risk parameters at any spatiotemporal node, avoiding the shortcomings of traditional static assessment models that cannot capture the dynamic evolution characteristics of risks. After calculating the risk parameters for a single scenario, a weighted fusion algorithm can be used to fuse multi-dimensional risk parameters to obtain a comprehensive risk assessment value for the corresponding spatiotemporal location. This allows for the delineation of different levels of comprehensive risk status, providing an accurate risk basis for subsequent regulation strategy formulation. As an example, the method for constructing a meteorological-road multiphysics coupled model includes: The spatiotemporal dynamic evolution equation for visibility is: Equation 1; In Equation 1, For spatial location At any moment The spatiotemporal visibility field, For partial differential operators, For spatial gradient operators, For wind speed vectors, The turbulent diffusion coefficient is... For the Laplace operator, The droplet settling coefficient is given. The number density of fog droplets. The density of the aerosol. This is a source of localized visibility deterioration caused by vehicle dust. in, Equation 2; In Equation 2, Dust coefficient, For traffic flow, Location of the accident site. To affect the radius, This is the base of the natural logarithm. In this embodiment, Equations 1 and 2 can be used to perform real-time dynamic calculations of local visibility under different scenarios, accurately capturing the dynamic changes caused by sudden fog outbreaks and vehicle dust spikes leading to a sharp drop in visibility, thus solving the problem that traditional static risk assessments cannot adapt to the dynamic evolution of local weather. This allows for the assessment of road condition information along highway sections, obtaining the comprehensive risk status of each section.

[0022] The spatiotemporal degradation equation of the road surface adhesion coefficient is: Equation 3; In Equation 3, For spatial location At any moment The spatiotemporal road surface adhesion coefficient, The coefficient of adhesion for dry road surfaces. This represents the maximum decrease in adhesion coefficient caused by debris coverage. Let be the viscosity-time constant of the mud spreading on the road surface. The location of the landslide center. The radius is affected by frictional degradation. The erosion coefficient of the adhesion coefficient due to rainfall. Rainfall intensity, The cumulative rainfall is used. In this embodiment, Equations 1 and 2 can accurately calculate the degree of visibility attenuation at different locations and times along the highway, while Equation 3 can accurately describe the dynamic attenuation process of the road surface friction coefficient under scenarios such as rainfall erosion and landslide debris coverage. The combination of these three can accurately obtain the road condition risk parameters at any time and space node, avoiding the shortcomings of traditional static assessment models that cannot capture the dynamic evolution characteristics of risks.

[0023] In this embodiment, to predict the responses of different drivers to different traffic conditions, driving behavior data of the driver can be collected through the vehicle terminal. Then, the past driving behavior data is analyzed to predict the driver's actual response to control commands under different risk scenarios, thus determining the driver's behavioral response constraint information. Specifically, determining the driver's behavioral response constraint information includes: acquiring past driving behavior data, performing driver behavior analysis on the past driving behavior data to obtain the driver's driving state execution status; wherein, the past driving behavior data includes: emergency braking frequency, following distance distribution, acceleration behavior data, deceleration behavior data, and historical speed limit command execution deviation data; then, based on the comprehensive risk status of each road segment and combined with the driver's driving state execution status, the driver's actual expected speed is determined; subsequently, based on the driver's actual expected speed and combined with the comprehensive risk status of each road segment, the collision probability of the driver driving at the actual expected speed can be predicted; then, based on the collision probability of the driver driving at the actual expected speed, the driver's actual expected speed is constrained to obtain the driver's behavioral response constraint information. In this embodiment, by statistically analyzing driver behavior under different risk scenarios, the deviation distribution pattern between actual driving speed and speed limit commands can be determined. This transforms the driver's actual response characteristics into probabilistic constraints on speed ranges, preventing the control strategy from becoming disconnected from actual traffic conditions due to the assumption that the driver completely obeys control commands. The resulting behavioral response constraints are then input into the subsequent game theory optimization model to ensure that the generated control commands conform to driving behavior patterns in real traffic scenarios, thereby improving the actual control effect after the strategy is implemented. To facilitate understanding of the above process, the following example illustrates the process: Determining the driver's actual desired speed includes: Equation 4; In Equation 4, For the driver at all times Lower spatial position The actual expected speed For the published variable speed limit, For the speed deviation amplitude, Let be the driver's perceived weighting function for the collision probability. To help drivers perceive the probability of a collision. Due to the inherent impatience level of drivers, The influence coefficient of vehicle-road cooperative information. This involves assessing the driver's trust in roadside information. In this embodiment, the driver's expected speed at the corresponding spatiotemporal location is first calculated using Equation 4. Then, a collision probability prediction is performed based on this expected speed, and the expected speed can be adjusted accordingly. Specifically, in one feasible embodiment, the actual expected speed, current road segment risk level, and headway are used as input parameters from the real-time comprehensive risk status output by the constructed meteorological-road multiphysics coupling model to calculate the collision probability under the current driving behavior. This collision probability is then used to constrain and correct the initially calculated expected speed: if the calculated collision probability is lower than a preset safety threshold, the currently calculated expected speed is retained as a response parameter; if the collision probability is higher than the preset safety threshold, the expected speed is proportionally reduced according to the percentage of the collision probability exceeding the threshold, resulting in a corrected expected speed. This corrected value is then used as the final behavioral response constraint output, providing accurate driver behavior response data for subsequent spatiotemporal risk field construction and control strategy formulation.

[0024] In this embodiment, to facilitate spatiotemporal linkage analysis along the highway, a spatiotemporal risk field along the highway can be constructed based on multi-source sensing information along the highway, combined with the comprehensive risk status of each road segment and driver behavior response constraints. This enables spatiotemporal linkage analysis along the highway and allows for rapid capture of risk points along the highway. Specifically, constructing the spatiotemporal risk field along the highway includes: The area along the highway can be divided into a continuous spatial domain. and time domain ,in, , An anisotropic risk field is constructed along the total length of the highway. : Equation 5; In Equation 5, For spatial location At any moment The spatiotemporal risk field value, The dynamic diffusion coefficient is... For spatial gradient operators, For the Laplace operator, For the risk movement velocity vector, For the collapse excitation function, For traffic accident excitation functions, This is the risk dissipation coefficient. This represents the risk field value at the current spatiotemporal location. It should be noted that the collapse excitation function satisfies: ; In the formula, For blocking equivalent, The location of the landslide center. To influence the radius. That is, in this embodiment, it can be achieved through... The difference scheme numerically solves the spatiotemporal risk field of Equation 5 above. The central difference scheme is used to process the diffusion term to ensure the calculation accuracy. Finally, the risk field value of all grid cells in the entire road segment at each time step is output, thus completing the construction of the spatiotemporal risk field along the highway and realizing the dynamic quantitative characterization of the risk diffusion and evolution process with traffic flow.

[0025] like Figure 4 As shown, after determining the driver behavior response constraints, a two-layer game optimization model can be constructed based on the established highway spatiotemporal risk field and behavior response constraints to solve for the optimal control scheme, thereby achieving risk regulation and resource scheduling of the spatiotemporal risk field along the highway. It should be noted that in the two-layer game optimization model, the upper-layer game is a game between the control entity and the overall traffic demand of the road network, with the goal of maximizing road network traffic efficiency while meeting safety constraints. The decision variable is the variable speed limit value of different road sections on the main line. The lower-layer game is a game between the control entity and the traffic flow at ramp entrances, with the goal of balancing the traffic risks of the main line and ramps. The decision variable is the ramp adjustment rate. That is, the two-layer game control architecture includes: upper-level managers formulating control strategies with the goal of minimizing the worst-case risk expectation; and lower-level driver groups responding with speed based on the control strategies to achieve group equilibrium. Examples are provided below: Specifically, the upper-level managers formulate control strategies with the goal of minimizing the expected risk under the worst-case scenario, including: based on Distance constructs a fuzzy set of driver response distribution And with the goal of minimizing the expected risk in the worst case, a distributed robust optimization model is constructed: Formula 6; In Equation 6, For the control strategy vector Find the minimum value, including the variable speed limit value. and ramp adjustment rate , For distribution set Seeking the ultimate limit, For the driver's response probability distribution, To predict the time domain, For the first The overall risk field value at any given moment. To control the cost weight matrix, Indicates distribution The expectations below Let be the control strategy vector at time i. To control the cost weight matrix, the driver response probability distribution can be obtained by fitting historical statistical data on driving behavior under different risk scenarios. The fuzzy set boundary of the distance is determined by the sample standard deviation of historical data. After solving the above distributed robust optimization model, the upper-level initial control strategy that meets the requirement of minimizing the worst-case scenario risk can be obtained. The lower-level driver group updates their respective expected driving speeds based on the initial control strategy. When the expected driving speeds of all drivers no longer change globally, that is, when the variance of the overall speed distribution of all road segments is lower than a preset threshold, it is determined that the lower-level group has reached Nash equilibrium, and at this time, the final control command is output. It should be noted that the distributed robust optimization model satisfies the chance constraint: ; ; In the formula, In distribution Find the infimum (worst-case scenario). As a probability measure, The actual speed at which the driver operates; , These are the lower and upper bounds of the velocity, respectively. This represents the maximum permissible risk value. This represents the permissible probability of constraint violation. By setting opportunity constraints, it can be ensured that even under the worst-case driver response distribution, the final control strategy still meets safety requirements, preventing secondary accidents caused by speed ranges or risk values ​​exceeding safety thresholds, thus further improving the robustness of the control scheme. After obtaining the convergent two-level game equilibrium solution, the complete optimal control command, including the variable speed limits for each road segment and the adjustment rates at each ramp entrance, can be output, completing this round of strategy formulation.

[0026] Specifically, the lower-level driver group responds to speed based on the control strategy and achieves group equilibrium, including: Constructing the variational inequality conditions for equilibrium among the lower-level driver group: , Equation 7; In Equation 7, Let be the driver's speed distribution vector. For the equilibrium velocity distribution vector, For vector difference transpose, For the velocity interaction matrix, For the velocity preference vector, To control the input matrix, For Lagrange multipliers, For matrix transpose, For the feasible speed set, This represents the speed distribution vector of all drivers satisfying the feasible conditions. In this embodiment, the alternating direction multiplier method can be used to iteratively solve the upper-level distributed robust optimization model and the lower-level variational inequality equilibrium problem until the change in the upper-level objective function value and the violation degree of the lower-level equilibrium constraint are both lower than the preset convergence threshold. This completes the collaborative solution of the two-level game problem, outputting the final variable speed limit value and ramp adjustment rate control command. It should be noted that the two-level game control model constructed in this embodiment is updated using a rolling time-domain optimization framework. Every preset time interval, the spatiotemporal risk field is dynamically corrected by combining the latest collected multi-source perception information, and the game model is updated again to obtain a new optimal control command, thereby adapting to the characteristics of the dynamic evolution of highway risk status over time.

[0027] After outputting control commands containing variable speed limits and ramp adjustment rates using the aforementioned two-layer game-theoretic control architecture, dynamic lane reconfiguration and emergency resource scheduling along the highway are then implemented based on these commands, thereby obtaining the real-time control status of the highway. To facilitate understanding of how control commands are used for dynamic lane reconfiguration and emergency resource scheduling along the highway, the following explanations are provided: Specifically, dynamic lane reconstruction along the highway includes: Macro-traffic flow under lane dynamic reconstruction Extended model: Equation 8; In Equation 8, Traffic flow density, For actual execution speed, For the first The open / closed status of each lane. The total number of lanes For the source term to be merged into the ramp, For the Dirac function, Location of the accident site; This represents a blocked lane condition. Specifically, in Equation 8, an indicator function is introduced. The number of passable lanes is dynamically represented, and when a corresponding lane needs to be closed due to a risk event... Setting the value directly to 0 resets the lane's contribution to the overall traffic flow of the road segment, thus accurately describing the traffic flow conservation relationship after dynamic lane opening and closing, avoiding the limitations of traditional fixed topology models that cannot adapt to dynamic lane reconfiguration scenarios. Based on this, and combining the risk distribution results of the spatiotemporal risk field obtained earlier, the opening and closing states of each lane are optimized with the dual objectives of minimizing the overall road network risk expectation and maximizing traffic efficiency. This yields the optimal lane reconfiguration scheme that satisfies the current risk scenario: for areas where the risk level exceeds the preset handling threshold, the corresponding lane's opening and closing value is directly reset. Setting the value to 0 achieves physical isolation of the risk area, preventing social vehicles from accidentally entering the risk area and causing secondary accidents; for the remaining passable lanes, the lane function division is adjusted in combination with the main line traffic demand and the emergency rescue priority principle, and a dedicated emergency rescue channel is opened to ensure that rescue vehicles can quickly reach the accident scene. At the same time, the passage rights of social vehicle lanes are optimized and allocated to maximize the use of existing road space to maintain the main line traffic.

[0028] Specifically, emergency resource allocation along highways includes: Equation 9; Formula 10; In Equation 9, For resources Assigned to accident Decision variables, For the quantity of emergency resources, The number of accident sites. For travel time cost, For distance, To ensure the speed of rescue vehicles, To allow preparation time, Risk exposure weighting coefficient; For arrival time, For the time required to complete the clearance, Accident point At any moment The comprehensive risk field value is calculated. Specifically, in Equations 9 and 10, the total risk exposure after an accident and the total cost of rescue dispatch are incorporated into a unified objective function for optimization. Risk control priority is adjusted by setting risk exposure weight coefficients: high-risk accidents correspond to higher risk exposure weights, allowing for priority dispatch of nearby rescue resources to complete the response, reducing the impact of continued risk spread on the overall road network safety. Resource capacity constraints are also incorporated into the optimization process, meaning the total number of rescue resources that a single emergency station can dispatch does not exceed its own available reserves, and the type and quantity of resources obtained by each risk event point meet the response requirements of the corresponding risk level. Solving the above optimization model yields the optimal emergency resource allocation scheme for the current risk scenario, clarifying the type and quantity of resources that each emergency station needs to dispatch and the accident location to, thus achieving efficient and accurate emergency resource dispatch arrangements. Figure 4 As shown, in a feasible control and dispatch scheme, vehicles can be divided into civilian vehicles and resource dispatch vehicles (such as tow trucks, ambulances, and fire trucks). In the event of an accident, lanes can be closed at the accident site, allowing civilian vehicles to suspend passage and enabling resource dispatch vehicles to flow into the rescue lane. Resource dispatch vehicles can then use the rescue lane as a detour. Alternatively, when the accident site is under control, civilian vehicles can be allowed to use the rescue lane to avoid exacerbating lane congestion.

[0029] After receiving the initial control instructions, lane dynamic reconstruction and emergency resource scheduling are performed along the highway according to the instructions. Then, based on real-time updated multi-source perception information, the spatiotemporal risk field is dynamically corrected. While updating the risk state, the control strategy is re-solved to form a dynamic closed-loop management system. Specifically, the correction of the spatiotemporal risk field along the highway for different control time sequences includes: acquiring the latest multi-source perception update data along the highway, combining it with the actual traffic flow state parameters after real-time control, recalculating the actual risk value at the current spatiotemporal location, obtaining the error term by subtracting the actual risk value from the predicted risk value of the current spatiotemporal risk field, and using the Kalman filter algorithm to recursively update the diffusion coefficient and risk source parameters of the risk field, completing the closed-loop correction of the spatiotemporal risk field state, and providing an accurate risk assessment basis for the next round of control strategy updates. As an example, the following is a detailed explanation: Prediction steps: ; ; Update steps: ; ; In the above formula, For the first Predicted state at any given time For the first The calibration status at any given moment. This is the state transition function. To predict the state covariance matrix, Let Jacobian be the state transition matrix. The process noise covariance matrix is... For Kalman gain, To observe the Jacobian matrix, To observe the noise covariance matrix, For the first Actual roadside radar readings at any given time. For observation functions; For the first The correction status at any given time. In this embodiment, through the above prediction update steps, the recursive correction of the spatiotemporal risk field state parameters can be completed. This allows the real-time collected measured information to be integrated into the risk field update process, correcting prediction deviations caused by model assumptions, ensuring that the risk field state always remains consistent with the actual road conditions, and providing support for the accuracy of subsequent control strategies.

[0030] Once the risk is detected to be gradually eliminated, a variable damping second-order recovery strategy is adopted to gradually restore the speed limit, achieving a smooth exit from the control process and avoiding traffic disruptions and secondary risks caused by a one-time lifting of control measures. It should be noted that the damping ratio in the variable damping second-order recovery strategy can be dynamically adjusted according to the real-time risk field gradient. In the initial stage of recovery, the damping ratio is kept in an overdamped state to prevent speed overshoot. As the risk field gradient decreases, the damping ratio is reduced to an underdamped state to accelerate recovery to the target speed. To facilitate understanding of the above-mentioned gradual speed limit restoration using a variable damping second-order recovery strategy, the following example is provided: A variable damping second-order recovery system can be constructed: Equation 11; In Equation 11, To restore speed, The first derivative of the recovery velocity, The second derivative of the recovery velocity, The system's inherent frequency, For target speed; The damping ratio varies with time, and the damping ratio varies with time. : Equation 12; In Equation 12, As the reference damping ratio, The gradient sensitivity coefficient; For recovery point The comprehensive risk field gradient is used. In this embodiment, by binding the damping ratio to the real-time risk field gradient, the speed limit recovery rate can be dynamically adjusted according to changes in the risk state: when the risk has not completely subsided and the risk field gradient is still large, the damping ratio is increased to put the system into an overdamped state, and the speed limit recovery rate is controlled at a low level to prevent the driver from rapidly accelerating and causing traffic instability; when the risk is completely eliminated and the risk field gradient drops to near zero, the damping ratio falls back to the benchmark value, the system recovers to a near-critical damping state, and the speed limit can be quickly and smoothly restored to the target design value, balancing safety and traffic efficiency during the recovery process. The entire recovery process requires no manual intervention and can automatically adjust recovery parameters based on the real-time risk field output, achieving a smooth transition in the control exit phase and completely solving the problem that traditional one-time recovery strategies are prone to causing secondary accidents such as rear-end collisions.

[0031] In a second aspect, the present invention also provides a traffic situation control system for highway risk points. The control system includes: a data acquisition module, a risk assessment module, a behavior analysis module, a two-layer game-theoretic control module, a dynamic correction module, and a gradual recovery module. The data acquisition module is used to collect multi-source sensing information such as visibility, road surface condition, traffic flow, and vehicle speed along the highway. The risk assessment module is used to calculate the comprehensive risk state of each road segment based on the multi-source sensing information, constructing a spatiotemporal risk field for the highway. The behavior analysis module is used to mine driver response patterns under different risk scenarios based on historical driving data, generating behavioral response constraint information. The two-layer game-theoretic control module is used to construct an upper and lower layer two-layer game architecture to solve for the optimal control command including variable speed limits and ramp adjustment rates. The dynamic correction module is used to dynamically correct the spatiotemporal risk field by combining real-time updated sensing data and update the control strategy. The gradual recovery module is used to gradually restore the speed limit through a variable damping second-order recovery strategy after the risk is eliminated, completing the control exit. In this embodiment, the control system, through closed-loop collaboration among its modules, achieves fully automated operation from risk perception and strategy generation to status updates and control exit. It can dynamically manage highway risk points without human intervention, meeting the current automated management needs of smart highways. The control system interfaces with existing roadside control equipment and information dissemination platforms on highways, allowing for direct execution of generated variable speed limits and ramp adjustment commands. Deployment can be completed without additional modifications to existing control hardware, adapting to the management upgrade needs of highways of different sizes.

[0032] In one feasible solution, the control system further includes a roadside information dissemination unit and a dispatch and command unit. The roadside information dissemination unit is used to disseminate the final control instructions to passing drivers through variable information signs, vehicle-road cooperative information push, and other means. The dispatch and command unit then completes the dynamic dispatch of emergency resources and lane configuration adjustment according to the control instructions.

[0033] In some implementations, the control system can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0034] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), etc.

[0035] A third aspect of the present invention provides a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements a traffic situation control method for highway risk points as described in any one of the first aspects. The computer-readable medium in this embodiment can be written in one or more programming languages ​​or a combination thereof to perform computer program code for carrying out operations of some embodiments of the present disclosure. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on a user's computer, partially on a user's computer, as a standalone software package, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0036] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0037] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts.

[0038] A fourth aspect of the present invention provides an electronic device, comprising: one or more processors; a storage device storing one or more programs thereon; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement a traffic situation control method for highway risk points as described in the first aspect. The computer-readable medium may be included in the electronic device or may exist independently, i.e., not assembled into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, enable the electronic device to implement the traffic situation control method for highway risk points as described in the first aspect.

[0039] The fifth aspect of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements a traffic situation control method for highway risk points as described in the first aspect.

[0040] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for traffic situation control at highway risk points, characterized in that, include: Acquire multi-source sensing information along the highway; Based on multi-source sensing information along the expressway, assess the road condition information of each section of the expressway and obtain the comprehensive risk status of each section; By analyzing past driving behavior data, we can predict the driver's actual response to control commands under different risk scenarios and determine the driver's behavioral response constraint information. Based on multi-source sensing information along the highway, and combined with the comprehensive risk status of each section and driver behavior response constraints, a spatiotemporal risk field along the highway is constructed. Based on the spatiotemporal risk field along the highway, a two-layer game control framework is constructed between upper-level managers and lower-level drivers, outputting control instructions that include variable speed limits and ramp adjustment rates. The two-layer game-theoretic control architecture includes: upper-level managers formulating control strategies with the goal of minimizing the expected risk under the worst-case scenario; and lower-level driver groups responding to speed based on the control strategies to achieve group equilibrium. Based on the control instructions, the lane dynamics and emergency resources along the expressway are dynamically reconfigured and dispatched to obtain the real-time control status of the expressway. Based on the real-time control status along the highway and combined with multi-source sensing information along the highway, the spatiotemporal risk field along the highway is corrected for different control time sequence states. Once the risk is eliminated, a second-order recovery strategy with variable damping is adopted to gradually lift the speed limit control.

2. The traffic situation control method for highway risk points according to claim 1, characterized in that, The process of obtaining the comprehensive risk status of each road segment includes: A meteorological-road multi-physics coupled model is constructed, which includes the spatiotemporal dynamic evolution equation of visibility and / or the spatiotemporal degradation equation of road adhesion coefficient; Specifically, the spatiotemporal dynamic evolution equation of visibility is: , formula 1; In formula 1, for spatial position at time spatiotemporal visibility field, partial differential operator, spatial gradient operator, wind speed vector, turbulent diffusion coefficient, Laplacian operator, fog droplet settling coefficient, fog droplet number density, aerosol density, local visibility deterioration source term caused by vehicle dust in, , formula 2; In formula 2, is a dust-raising coefficient, is a traffic flow, is an accident point position, is an influence radius, is a base of a natural logarithm; The spatiotemporal degradation equation of the road surface adhesion coefficient is: , formula 3; In formula 3, For spatial position At time The space-time road adhesion coefficient, The dry road adhesion coefficient, The maximum adhesion coefficient attenuation caused by debris coverage, The mud spreading viscosity time constant on the road, The collapse center position, The friction degradation influence radius, The rainfall erosion coefficient on the adhesion coefficient, The rainfall intensity, The cumulative rainfall.

3. The traffic situation regulation method for a highway risk point according to claim 1, characterized in that, The information used to determine the driver's behavioral response constraints includes: Acquire past driving behavior data, perform driver behavior analysis on the past driving behavior data, and obtain the driver's driving status and performance. Based on the overall risk status of each road segment and the driver's driving performance, the driver's actual expected speed is determined. Based on the driver's actual expected speed and combined with the comprehensive risk status of each road segment, the probability of a collision is predicted when the driver drives at the actual expected speed. Based on the probability of a collision when the driver is driving at the actual expected speed, the driver's actual expected speed is constrained to obtain information on the driver's behavioral response constraints.

4. The traffic situation regulation method for a highway risk point according to claim 3, characterized in that, Determining the driver's actual desired speed includes: , formula 4; In formula 4, is the actual expected speed of the driver at the moment is the spatial position of the vehicle, is the actual expected speed of the driver at the moment is the published variable speed limit value, is the speed deviation amplitude, is the driver's perception weight function of the collision probability, is the driver's perception of the collision probability, is the driver's inherent impatience coefficient, is the influence coefficient of car-road cooperation information, is the driver's trust in the roadside published information.

5. The traffic situation regulation method for a highway risk point according to claim 1, characterized in that, The construction of the spatiotemporal risk field along the highway includes: The highway along the line is divided into continuous spatial domain and time domain , , The total length of the highway along the line is constructed anisotropic risk field : , formula 5; In Equation 5, For spatial location At any moment The spatiotemporal risk field value, The rate of change of the risk field over time. The dynamic diffusion coefficient is... For the spatial gradient of the risk field, The second spatial derivative of the risk field. For the risk movement velocity vector, For the collapse excitation function, For traffic accident excitation functions, This is the risk dissipation coefficient. This represents the risk field value at the current spatiotemporal location.

6. The traffic situation regulation method for a highway risk point according to claim 1, characterized in that, The upper-level managers formulate control strategies with the goal of minimizing the expected risk under the worst-case scenario, including: Based on Fuzzy sets constructing driver response distributions And a distributed robust optimization model is constructed with the objective of minimizing the risk expectation in the worst case. , formula 6; In formula 6, is a control strategy vector is minimized, including variable speed limit values and ramp metering rates , is a set of distributions is supremized over, is a driver response probability distribution, is a prediction horizon, is a comprehensive risk field value at the moment, is a control cost weight matrix, denotes an expectation under the distribution , is a control strategy vector at the moment, is a control cost weight matrix; In addition, the distributed robust optimization model satisfies the chance constraint: ； ； In the formula, To distribute Lower bound (worst case) is calculated as Probability measure is Actual speed of the driver; , Lower and upper bounds of speed, respectively; Maximum allowed risk value; Allowed constraint violation probability.

7. The traffic situation regulation method for a highway risk point according to claim 6, characterized in that, The lower-level driver group responds to speed based on the control strategy and achieves group equilibrium, including: Constructing the variational inequality conditions for equilibrium among the lower-level driver group: , , Formula 7; In formula 7, is the driver speed profile vector, is the equalized speed profile vector, is the vector difference is the transpose of is the speed interaction matrix, is the speed preference vector, is the control input matrix, is the Lagrange multiplier, is the matrix is the transpose of is the feasible speed set, denotes all driver speed profile vectors that satisfy the feasibility conditions.

8. The traffic situation regulation method for a highway risk point according to claim 1, characterized in that, Dynamic lane reconstruction along the highway includes: Constructing macroscopic traffic flow under dynamic lane reconfiguration Extended model: , formula 8; In Equation 8, Traffic flow density, For actual execution speed, For the first The open / closed status of each lane. The total number of lanes For the source term to be merged into the ramp, For the Dirac function, Location of the accident site; The lane is blocked.

9. A traffic situation control method for highway risk points according to claim 1, characterized in that, Emergency resource allocation along the highway includes: Equation 9; Formula 10; In Equation 9, For resources Assigned to accident Decision variables, For the quantity of emergency resources, The number of accident sites. For travel time cost, For distance, To ensure the speed of rescue vehicles, To allow preparation time, Risk exposure weighting coefficient; For arrival time, For the time required to complete the clearance, Accident point At any moment The overall risk field value.

10. A traffic situation control system for highway risk points, characterized in that, The control system employs the control method as described in any one of claims 1 to 9, and further includes: The data acquisition module is used to collect multi-source sensing information along the highway. The risk assessment module is used to calculate the comprehensive risk status of each road segment based on multi-source sensing information and construct the spatiotemporal risk field of the highway. The behavior analysis module is used to mine driver response patterns under different risk scenarios based on historical driving data and generate behavior response constraint information. A two-layer game control module is used to construct an upper and lower two-layer game architecture to solve for the optimal control command that includes the variable speed limit value and the ramp adjustment rate. The dynamic correction module is used to dynamically correct the spatiotemporal risk field by combining real-time updated sensing data and update the control strategy. The gradual recovery module is used to gradually restore the speed limit through a variable damping second-order recovery strategy after the risk is eliminated, thus completing the control exit.

Images