A freeway variable speed limit hierarchical cooperative control method considering bottleneck capacity reduction

Abstract

The application discloses a kind of expressway variable speed limit hierarchical collaborative control method considering bottleneck capacity drop, belong to intelligent traffic control field, first, according to expressway real-time detection data to establish traffic state model;Second, the traffic flow propagation prediction model considering the influence of variable speed limit and the effect of capacity drop is established, the future time domain traffic state evolution is predicted;Then, the bottleneck capacity drop prediction model is established, the bottleneck actual traffic capacity change and congestion propagation trend are predicted;Further establish network layer collaborative optimization control model, with network total travel time minimum as target optimization each link target outflow rate;Subsequently, the local layer variable speed limit control model is established, with the minimum outflow rate tracking error as target optimization each link speed limit value;Finally, the rolling prediction and optimization mechanism is used to output control scheme in real time.The application can identify bottleneck capacity drop risk in advance, actively adjust traffic demand, reduce congestion degree, improve expressway operation efficiency and stability.

Description

Technical Field

[0001] This invention relates to the field of intelligent traffic control and transportation system optimization technology, specifically to a hierarchical collaborative control method for variable speed limits on expressways that takes into account bottleneck capacity reduction. Background Technology

[0002] With the accelerating pace of urbanization and the continuous growth of motor vehicle ownership, expressways have gradually become an important component of urban transportation systems. Expressways are characterized by high capacity, high speed, and wide service coverage, playing a vital role in meeting the demand for medium- and long-distance urban travel. However, during morning and evening rush hours, large-scale events, and after traffic accidents, expressway networks frequently experience traffic demand exceeding road capacity, leading to and spreading traffic congestion.

[0003] Studies show that expressway congestion typically originates at bottleneck locations such as merging zones, lane reduction zones, ramp merging areas, and accident-affected zones. When traffic flow density continuously increases and exceeds a critical density, vehicle speeds decrease rapidly, and the following distance between vehicles decreases, causing the actual capacity of the bottleneck area to fall below its maximum capacity under normal operating conditions. This phenomenon is known as capacity drop. Capacity drop is a common nonlinear traffic characteristic in expressway traffic operation. When capacity drop occurs at a bottleneck, even if traffic demand remains constant, the number of vehicles that can pass through the bottleneck will still decrease significantly, leading to a rapid increase in upstream queues and the congestion spreading upstream. Numerous studies have shown that capacity drop is one of the important causes of expressway congestion formation and spread, and also a significant factor affecting expressway operating efficiency.

[0004] To alleviate traffic congestion on expressways, scholars both domestically and internationally have proposed various active traffic management strategies, including ramp control, dynamic path guidance, lane control, and variable speed limits. Among these, variable speed limit (VSL) is considered a crucial technical means to alleviate expressway congestion and improve traffic safety by dynamically adjusting road speed limits to guide vehicles to slow down in advance and improve traffic flow. Existing variable speed limit control methods mainly include rule-based control, feedback control, and model predictive control. Rule-based control methods typically set speed limit strategies based on traffic flow, speed, or occupancy, which are simple to implement but difficult to adapt to complex traffic environments. Feedback control methods adjust speed limits based on real-time traffic conditions, possessing a certain degree of adaptability but lacking the ability to predict future traffic conditions. Model predictive control methods can use traffic flow models to predict future traffic conditions and obtain the optimal control strategy through optimization algorithms, thus attracting widespread attention.

[0005] However, most existing variable speed limit control methods still have the following shortcomings: First, most studies use traditional cellular transmission models or simplified traffic flow models to describe the traffic state evolution process, making it difficult to accurately characterize the traffic flow propagation law and congestion formation mechanism under capacity reduction conditions, resulting in insufficient accuracy in traffic state prediction. Second, existing variable speed limit control methods usually treat capacity reduction as a fixed parameter or ignore the dynamic changes in capacity reduction, making it difficult to reflect the actual characteristics of the dynamic evolution of bottleneck capacity with changes in traffic state, resulting in the control strategy failing to respond to bottleneck state changes in a timely manner. Third, most existing control methods adopt a centralized optimization framework. When the scale of the expressway network increases, the optimization variables and constraints increase significantly, and the computational complexity grows rapidly, making it difficult to meet the real-time control requirements. In addition, most variable speed limit control methods directly optimize the speed limit value without fully considering the coordination relationship between the overall network operation goal and the local control execution, which easily leads to inconsistencies between local optimization and global optimization, thereby reducing the control effect. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing variable speed limit control methods for expressways, such as difficulty in accurately describing the dynamic evolution of capacity decline, difficulty in predicting bottleneck capacity changes, and insufficient network-level collaborative control capabilities. This invention proposes a hierarchical collaborative control method for variable speed limits on expressways that considers bottleneck capacity decline. This invention establishes a traffic flow propagation prediction model that considers the impact of variable speed limits and capacity decline effects to predict the future traffic state changes of expressways in the time domain; constructs a bottleneck capacity decline prediction model to dynamically identify bottleneck locations and their capacity change processes; and further establishes a network-level and local-level collaborative optimization control framework to achieve network-level traffic state coordination optimization and segment-level collaborative control of variable speed limits.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A hierarchical cooperative control method for variable speed limits on expressways that considers bottleneck capacity reduction includes the following steps:

[0009] S1. Establish a traffic status model of the expressway network based on real-time traffic detection data of the expressway;

[0010] S2. Establish a traffic flow propagation prediction model that considers the effects of variable speed limits and capacity reduction, and predict the future traffic state evolution process of expressways in the time domain.

[0011] S3. Based on the traffic condition prediction results, establish a bottleneck capacity reduction prediction model to predict the change process of the actual bottleneck capacity in the future time domain and the state of congestion propagation.

[0012] S4. Based on the bottleneck capacity reduction prediction results, establish a network layer collaborative optimization control model, with the goal of minimizing the total travel time of the expressway network, and optimize the outflow rate of each link.

[0013] S5. Based on the network layer optimization results, establish a local layer variable rate limiting control model, with the minimum tracking error between the actual outflow rate and the target outflow rate as the optimization objective, and optimize the variable rate limiting value of each link.

[0014] S6. Adopt a rolling prediction and optimization mechanism to update traffic status in real time and output a variable speed limit control scheme.

[0015] Furthermore, in step S1, real-time traffic operation information is obtained by using geomagnetic detectors, video detectors, radar detectors, and floating car data acquisition equipment deployed along the expressway. The traffic operation information includes traffic flow, average vehicle speed, traffic density, occupancy rate, and travel time. Based on the connection relationship between the expressway mainline, ramps, and connecting sections, an expressway network topology model is established, representing the expressway network as a directed traffic network composed of multiple links and nodes.

[0016] Furthermore, in step S2, a traffic flow propagation prediction model is established for each link in the expressway network. For any link i, the cumulative number of inbound vehicles is defined. and cumulative number of vehicles leaving , representing the cumulative number of vehicles entering and leaving the link up to time k, respectively. The cumulative number of inbound vehicles is updated as follows:

[0017]

[0018] in, Indicates the inflow rate of the link. This indicates the control time step.

[0019] The internal vehicle storage capacity of the link is represented as:

[0020]

[0021] Further, a link receiving capacity model is established. Link receiving capacity represents the maximum number of vehicles a link can accept at any given moment, and its magnitude is affected by the remaining space in the link and the congestion wave propagation process. According to congestion wave propagation theory, the maximum allowable cumulative inflow of the link is expressed as:

[0022]

[0023] in, Link length; Blocking density; This represents the maximum cumulative inflow after considering congestion wave propagation. Based on the above relationship, the maximum allowable inflow rate of the link is obtained as follows:

[0024]

[0025] Establish a link transmission capacity model, where the expected cumulative outflow of the link is represented as:

[0026]

[0027] in, This represents the number of discrete time steps corresponding to the free flow propagation time; This represents the time interpolation coefficient. To describe the impact of variable speed limit control on traffic flow propagation, a link free-flow velocity variable is introduced:

[0028]

[0029] in, The current speed limit value is displayed. The driver is subject to a deviation coefficient. Link capacity is calculated using free-flow velocity.

[0030]

[0031] Where w represents the congestion wave propagation speed. This establishes a dynamic mapping relationship between variable speed limits and link capacity. As the speed limit changes, the free-flow speed and link capacity change synchronously, enabling the prediction of the impact of variable speed limits on traffic state evolution.

[0032] Further, in step S3, potential bottleneck locations in the expressway network are first identified, including mainline merging areas, lane reduction areas, and ramp merging areas. For any bottleneck link i, the upstream traffic demand is calculated. and downstream transportation supply capacity When the following conditions are met:

[0033]

[0034] The bottleneck is identified as entering a congested state, triggering a capacity reduction. The capacity reduction coefficient is defined as follows:

[0035]

[0036] in, This represents the link capacity under normal conditions. This represents the actual capacity after the capacity reduction. Therefore, the actual bottleneck capacity is expressed as:

[0037]

[0038] Furthermore, the capacity degradation factor is dynamically updated based on changes in link density. When the link density is higher than the critical density, the capacity decline gradually increases; when the link density is lower than the critical density, the capacity gradually recovers. By combining the traffic flow propagation prediction model, the capacity change process and actual traffic capacity at each bottleneck location in the future time domain are recursively calculated to obtain the capacity decline evolution state and congestion propagation trend.

[0039] Furthermore, in step S4, the network layer controller optimizes the entire expressway network, reducing traffic demand in bottleneck areas and suppressing further capacity degradation by coordinating the outflow rate distribution of each link. The control objective is to minimize the total network travel time within the predicted time domain, which is obtained by summing the vehicle storage capacity of each link within the predicted time domain.

[0040]

[0041] in, Optimize the objective function for the network layer; For link At any moment Vehicle storage capacity; To control the time step, the outflow rate of each link is used as the control variable. Simultaneously, traffic flow conservation constraints, link capacity constraints, and capacity reduction constraints are established, with the actual bottleneck capacity used as the constraint boundary for the capacity reduction constraint. After solving the network layer optimization model, the sequence of target outflow rates for each link in the future prediction time domain is obtained, serving as the reference trajectory input for the local layer variable speed limit controller.

[0042] Further, in step S5, the local layer controller uses a single link or a local control area as the control object, and achieves active traffic flow control by adjusting the display value of the variable speed limit sign. The local layer control objective is to minimize the tracking error between the actual outflow rate and the target outflow rate of the link.

[0043]

[0044] in, This represents the actual outflow rate of the link. The target outflow rate is given for the network layer. To avoid frequent changes in the speed limit affecting driver acceptance, this invention further introduces a speed limit smoothing control term. The local layer objective function is composed of the outflow rate tracking error and the speed limit change:

[0045]

[0046] in, This indicates the outflow rate tracking error; This indicates the penalty for changes in speed limits; This represents the weighting coefficient. The local layer control variable is the link's variable rate limiting value. Simultaneously satisfying:

[0047]

[0048]

[0049] in, and These represent the minimum and maximum permissible speed limits, respectively. This represents the maximum allowable speed limit change in a single control cycle.

[0050] When a capacity reduction is predicted to occur in a bottleneck area, the controller lowers the speed limit of the upstream link in advance to reduce the speed at which vehicles reach the bottleneck, thereby reducing peak traffic demand and delaying the capacity reduction. As the bottleneck area gradually returns to normal operation, the controller gradually increases the speed limit to restore traffic flow to normal operation.

[0051] Furthermore, in step S6, let the prediction time domain length be... The control time domain length is ,in Greater than After each optimization, only the first control action in the current optimization sequence is executed, i.e., the variable speed limit value corresponding to the current control cycle. After the control cycle ends, real-time traffic status information of the expressway is collected again, the state variables of the traffic flow propagation prediction model and the capacity reduction prediction model are updated, and the optimization model is reconstructed and solved to achieve rolling prediction and rolling optimization.

[0052] This invention has the following advantages and beneficial effects:

[0053] 1. High accuracy in capacity reduction prediction: This invention establishes a traffic flow propagation prediction model that considers the effects of variable speed limits and capacity reduction, as well as a bottleneck capacity reduction prediction model. It can dynamically identify the capacity reduction state based on the relationship between traffic demand and traffic supply, and update the actual capacity of the bottleneck in real time through the capacity reduction coefficient. It can accurately predict the bottleneck capacity change process and congestion propagation trend, overcoming the shortcomings of existing methods that treat capacity reduction as a fixed parameter or ignore its dynamic changes.

[0054] 2. Excellent Layered Cooperative Control Performance: This invention employs a two-layer cooperative optimization control architecture with network and local layers. The network layer optimizes the outflow rate of each link to minimize the total travel time, achieving network-level global coordination. The local layer optimizes variable speed limits to minimize the tracking error of the outflow rate, achieving segment-level precise control. This two-layer architecture balances overall network operating efficiency with local control execution effectiveness, avoiding inconsistencies between local and global optimization.

[0055] 3. High computational efficiency: This invention adopts a hierarchical control structure to decompose the large-scale network optimization problem into two sub-problems: network layer outflow rate optimization and local layer speed limit optimization. This significantly reduces the optimization dimensionality and computational complexity, making it more suitable for the real-time control requirements of large-scale expressway networks compared to centralized optimization frameworks.

[0056] 4. Strong proactive regulation capability: Based on the rolling prediction and rolling optimization mechanism of model predictive control, this invention can identify the risk of bottleneck capacity reduction in advance, and proactively adjust the upstream speed limit value before the capacity reduction occurs to reduce the peak traffic demand, delay the occurrence of capacity reduction and shorten its duration, effectively reduce the degree of congestion and vehicle delays, and improve the operating efficiency and stability of the expressway network. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the directed graph structure of the expressway network according to a specific embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of the bottleneck location of the expressway according to a specific embodiment of the present invention;

[0059] Figure 3 This is a schematic diagram of a rapid transit network structure including multiple entrance ramps and exit ramps, as described in a specific embodiment of the present invention.

[0060] Figure 4 This is a comparison result diagram of the variable speed limit hierarchical predictive control described in a specific embodiment of the present invention;

[0061] Figure 5 This is an overall flowchart of the method of the present invention. Detailed Implementation

[0062] The present invention will be further described below with reference to the embodiments. It should be noted that these are merely examples and descriptions of the inventive concept. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in the claims, they should all be considered to fall within the protection scope of the present invention.

[0063] like Figure 5 As shown, the present invention proposes a hierarchical collaborative control method for variable speed limits on expressways that considers bottleneck capacity reduction, which is implemented through the following steps.

[0064] Step S1: Establish a traffic state model for the expressway network

[0065] Real-time traffic information is acquired using geomagnetic detectors, video detectors, radar detectors, and floating car data acquisition equipment deployed along the expressway. This traffic information includes traffic flow, average vehicle speed, traffic density, occupancy rate, and travel time. An expressway network topology model is established based on the connectivity between the expressway mainline, ramps, and connecting sections. The expressway network is represented as a directed traffic network composed of multiple links and nodes, where links represent the expressway mainline, ramps, or connecting sections, and nodes represent divergence points, merging points, and expressway entrances / exits. Figure 1 As shown.

[0066] The current traffic status of each link is determined based on real-time detection data, including link traffic, link speed, and link density, and this is used as the initial state for subsequent traffic flow prediction models.

[0067] Step S2: Establish a traffic flow propagation prediction model

[0068] For each link in the expressway network, a traffic flow propagation prediction model is established, considering the effects of variable speed limits and capacity reduction. For any link i, the cumulative number of inbound vehicles is defined. and cumulative number of vehicles leaving , , represent the cumulative number of vehicles entering and leaving the link up to time k. The cumulative number of inbound vehicles is updated according to equation (1), and the vehicle storage capacity within the link is calculated according to equation (2). As traffic demand increases, the vehicle storage capacity within the link gradually increases. When the storage capacity approaches the link's storage capacity, congestion occurs.

[0069] A link receiving capacity model is established, which represents the maximum number of vehicles that the link can accept at the current moment. Its size is affected by the remaining space of the link and the congestion wave propagation process. According to the congestion wave propagation theory, the maximum allowable cumulative inflow of the link is calculated according to equation (3), and the maximum allowable inflow rate of the link is calculated according to equation (4), thereby realizing the dynamic prediction of the link receiving capacity.

[0070] A link transmission capacity model is established, where link transmission capacity represents the link's ability to release vehicles downstream, and its magnitude is constrained by vehicle speed and link capacity. Vehicles can only leave the link after completing free-flow propagation, so the expected cumulative outflow of the link is calculated according to equation (5).

[0071] To describe the impact of variable speed limit control on traffic flow propagation, a link free-flow speed variable is introduced. The actual free-flow speed is calculated according to equation (6) based on the displayed speed limit value and the driver compliance deviation coefficient. The link capacity is calculated using the free-flow speed according to equation (7), establishing a dynamic mapping relationship between variable speed limit and link capacity. As the speed limit value changes, the link free-flow speed and link capacity change synchronously, thereby realizing the prediction of the impact of variable speed limit on the traffic state evolution process.

[0072] Traffic conditions are updated based on link receiving capacity, transmitting capacity, and vehicle conservation relationships. The changes in link traffic, speed, density, and vehicle storage capacity in the future time domain are predicted, ultimately obtaining the future traffic condition prediction results for the expressway network. This provides traffic condition input for subsequent capacity reduction prediction and hierarchical collaborative control.

[0073] Step S3: Establish a bottleneck capacity decline prediction model

[0074] Based on the traffic condition prediction results obtained in step S2, a bottleneck capacity reduction prediction model is established. First, potential bottleneck locations in the expressway network are identified, including key areas such as mainline merging zones, lane reduction zones, and ramp merging zones. Figure 2 As shown.

[0075] For any bottleneck link i, calculate the upstream traffic demand. and downstream transportation supply capacity When equation (8) is satisfied, i.e., upstream traffic demand exceeds downstream traffic supply, the bottleneck is determined to enter a congested state and a capacity reduction is triggered.

[0076] The capacity reduction coefficient is defined according to equation (9), and the actual bottleneck capacity is calculated according to equation (10). The capacity reduction coefficient is further dynamically updated based on changes in link density. When the link density is higher than the critical density, the capacity decline gradually increases; when the link density is lower than the critical density, the capacity gradually recovers. Combining the traffic flow propagation prediction model established in step S2, the capacity change process and actual traffic capacity of each bottleneck location in the future time domain are recursively calculated to obtain the capacity decline evolution state and congestion propagation trend. The prediction results are then used as input to the network layer collaborative optimization control model.

[0077] Step S4: Establish a network layer collaborative optimization control model

[0078] Based on the bottleneck capacity reduction prediction results obtained in step S3, a network-layer collaborative optimization control model is established. The network-layer controller takes the entire expressway network as the optimization object, and by coordinating the outflow rate distribution of each link, it reduces traffic demand in bottleneck areas, suppresses further deterioration of capacity reduction, and thus improves the overall operating efficiency of the expressway network.

[0079] First, define the network layer state variables, including the vehicle storage capacity of each link, the cumulative number of inbound vehicles, the cumulative number of outbound vehicles, and the actual capacity of the bottleneck. Based on the traffic flow propagation prediction model established in step S2 and the capacity reduction prediction model established in step S3, obtain the traffic state change process and bottleneck capacity change process of each link in the future prediction time domain.

[0080] The control objective is to minimize the total network travel time within the prediction time domain. The total network travel time is obtained by summing the vehicle storage capacity of each link within the prediction time domain according to equation (11). Since the vehicle storage capacity of a link can reflect the queuing situation within the link, minimizing the vehicle storage capacity is equivalent to minimizing the total network travel time.

[0081] The target outflow rate of each link is used as the control variable, where the target outflow rate represents the outflow level that the network layer controller expects the link to reach in the future time domain. At the same time, traffic flow conservation constraints, link capacity constraints, and capacity decline constraints are established. Among them, the capacity decline constraint uses the actual bottleneck capacity obtained in step S3 as the constraint boundary to ensure that the optimization result meets the actual operating capacity requirements of the bottleneck.

[0082] After solving the network layer optimization model, the target outflow rate sequence of each link in the future prediction time domain is obtained. This target outflow rate sequence is not directly used for traffic control, but is used as the reference trajectory input for the local layer variable speed limit controller. When a capacity reduction is predicted to occur in the bottleneck area, the network layer controller reduces the target outflow rate of the upstream link of the bottleneck while maintaining a high release capacity of the downstream link, thereby slowing down the speed at which vehicles gather in the bottleneck area.

[0083] Step S5: Establish a local layer variable speed limiting control model

[0084] Based on the target outflow rate obtained in step S4, a local layer variable speed limit control model is established. The local layer controller takes a single link or a local control area as the control object and achieves active traffic flow control by adjusting the displayed value of the variable speed limit sign. The target outflow rate given by the network layer controller reflects the optimal operating state of the overall network, while the local layer controller is responsible for making the actual traffic flow state as close as possible to the network layer optimization result through speed limit control.

[0085] The tracking error between the actual outflow rate and the target outflow rate is minimized as the local layer control objective, and the tracking error is calculated according to equation (12). To avoid frequent changes in the speed limit affecting the driver's acceptance, a speed limit smoothing control term is further introduced. The local layer objective function is composed of the outflow rate tracking error and the speed limit change, and is calculated according to equation (13).

[0086] The local layer control variable is the link variable rate limit value. It simultaneously satisfies the upper and lower limits of the speed limit value shown in equation (14) and the maximum speed limit change in a single cycle shown in equation (15).

[0087] Based on traffic flow propagation predictions and the target outflow rate sequence, the optimal speed limit value in the future control time domain is obtained by solving a local layer optimization model. When a capacity decrease is predicted to occur in the bottleneck area, the controller lowers the speed limit value of the upstream link of the bottleneck in advance, reducing the speed at which vehicles arrive at the bottleneck, thereby reducing peak traffic demand and delaying the capacity decrease. As the bottleneck area gradually returns to normal operation, the controller gradually increases the speed limit value to restore the traffic flow to normal operation.

[0088] By combining network-layer target outflow rate optimization with local-layer variable speed limit tracking control, proactive regulation of traffic conditions in expressway networks can be achieved, improving operational efficiency in bottleneck areas and reducing the risk of congestion spread caused by capacity reduction.

[0089] Step S6: Rolling Prediction and Optimization Control

[0090] A rolling prediction and rolling optimization mechanism is used to update traffic conditions in real time and output a variable speed limit control scheme. First, the traffic condition information of the expressway network at the current control time is obtained, including traffic flow, vehicle speed, traffic density and bottleneck operation status of each link, and this information is used as the initial state for the traffic flow propagation prediction model and the capacity reduction prediction model.

[0091] Subsequently, the traffic flow propagation prediction model established in step S2 is used to predict the traffic state changes of each link in the future prediction time domain, and the bottleneck capacity reduction prediction model established in step S3 is combined to predict the bottleneck capacity changes and congestion propagation processes in the future time domain. After obtaining the future traffic state prediction results, the network layer collaborative optimization control model established in step S4 is executed to solve for the target outflow rate of each link in the future prediction time domain; further, the local layer variable speed limit control model established in step S5 is executed to solve for the optimal variable speed limit value of each link in the future control time domain.

[0092] Let the prediction time domain length be The control time domain length is ,in Greater than After each optimization is completed, only the first control action in the current optimization sequence is executed, that is, the variable speed limit value corresponding to the current control cycle, and the control results of subsequent time steps are not executed.

[0093] After the control cycle ends, real-time traffic status information of the expressway is collected again, and the state variables of the traffic flow propagation prediction model and capacity reduction prediction model are updated using the latest traffic status. Subsequently, the network layer optimization model and local layer optimization model are reconstructed, and the optimal control scheme in the future time domain is solved again. By continuously repeating the above process, dynamic coordinated control of variable speed limits on the expressway network is achieved.

[0094] Example

[0095] This embodiment selects a city's expressway network as the research object. This expressway network includes one mainline, two entrance ramps, and one mainline merging bottleneck, divided into eight continuous links, denoted as L1 to L8, as follows: Figure 3 As shown. The link length is 500-1000m, and the expressway mainline is equipped with variable speed limit signs, with a speed limit range of 40km / h to 80km / h. The system collects traffic status information every 30 seconds, including parameters such as traffic flow, average vehicle speed, traffic density, and occupancy. The network layer controller's prediction time domain is set to 10 control steps, corresponding to 300 seconds; the local layer controller's control time domain is set to 5 control steps, corresponding to 150 seconds.

[0096] Step (1): Establish a traffic state model of the expressway network based on the traffic state data collected by the detectors. Use the mainline and ramp detectors to obtain the traffic flow and average speed of each link, and calculate the vehicle density of each link. Represent the expressway network as a directed traffic network composed of links and nodes, where links represent the mainline and ramp segments of the expressway, and nodes represent ramp merging points and mainline connection points.

[0097] Step (2): Establish a traffic flow propagation prediction model that considers the effects of variable speed limits and capacity reduction. For any link, calculate the cumulative number of inflowing vehicles, the cumulative number of outflowing vehicles, and the vehicle storage capacity in real time, and predict the future traffic state changes in the time domain based on the traffic flow conservation relationship. Dynamically correct the free-flow speed of the link using the variable speed limit value to obtain the predicted results of link flow, speed, and density in the future time domain.

[0098] Step (3): Establish a bottleneck capacity reduction prediction model. In this embodiment, link L5 is the main line merging bottleneck. When upstream traffic demand continues to increase, the traffic density in the bottleneck area gradually approaches the critical density. The bottleneck operation status is judged based on the relationship between traffic demand and traffic supply. When upstream demand exceeds downstream supply capacity, it is determined that the bottleneck has experienced a capacity reduction, and the actual throughput capacity after the capacity reduction is calculated. Simulation results show that under the condition of continuous increase in traffic demand during the morning peak, the maximum throughput capacity of the bottleneck area decreases from 7200veh / h to 6300veh / h, and the capacity reduction ratio is about 12.5%. As the capacity reduction occurs, the vehicle storage capacity of the upstream link increases rapidly, and a queue propagation phenomenon is formed.

[0099] Step (4): Establish a network layer collaborative optimization control model based on the predicted capacity reduction results. The network layer controller takes minimizing the total travel time of the expressway network as the optimization objective and the target outflow rate of each link as the decision variable. The optimal target outflow rate sequence of each link in the future time domain is obtained through optimization. In this embodiment, when the bottleneck capacity reduction risk is predicted, the network layer controller reduces the target outflow rate of the upstream links L3 and L4 of the bottleneck, while maintaining a high release capacity of the downstream links, thereby slowing down the speed at which vehicles gather in the bottleneck area.

[0100] Step (5): Establish a local layer variable speed limit control model. The local layer controller uses the target outflow rate output from the upper layer as a reference value, and takes minimizing the tracking error between the actual outflow rate and the target outflow rate as the optimization objective. Traffic flow control is achieved by adjusting the displayed value of the variable speed limit sign. For example, at the beginning of the control cycle, the displayed speed limit values ​​for links L3 and L4 are both 80 km / h. When it is predicted that the bottleneck will experience a capacity decrease in the future time domain, the controller adjusts the speed limit value of L3 to 60 km / h and the speed limit value of L4 to 50 km / h, thereby reducing the speed at which vehicles reach the bottleneck area. As traffic demand decreases and the bottleneck gradually returns to normal operation, the controller gradually restores the speed limit values ​​to 70 km / h and 80 km / h.

[0101] Step (6): A rolling prediction and rolling optimization mechanism is adopted throughout the control process. After each control cycle, the real-time traffic status information of the expressway is reacquired, and the traffic flow propagation prediction model and capacity reduction prediction model are updated. Then, the network layer optimization model and local layer optimization model are resolved, and a new variable speed limit control strategy is executed.

[0102] To verify the effectiveness of this invention, it was compared with a fixed speed limit control strategy and a traditional variable speed limit predictive control strategy (MPC-B). Experimental results are as follows: Figure 4 As shown, the Variable Speed ​​Limit Hierarchical Predictive Control (MPC-V) proposed in this invention can identify bottleneck capacity reduction risks in advance and proactively adjust traffic demand through hierarchical collaborative control. Compared with fixed speed limit control, the total network travel time is reduced by approximately 15%, and the average speed is increased by approximately 12%. Compared with traditional variable speed limit predictive control, the bottleneck queue length is reduced by approximately 18%, and the duration of capacity reduction is shortened by approximately 20%. The results show that this invention can effectively suppress traffic congestion caused by bottleneck capacity reduction and improve the operational efficiency and stability of the expressway network.

[0103] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A hierarchical cooperative control method for variable speed limits on expressways that considers bottleneck capacity reduction, characterized in that, Includes the following steps: S1. Establish a traffic status model of the expressway network based on real-time traffic detection data of the expressway; S2. Establish a traffic flow propagation prediction model that considers the effects of variable speed limits and capacity reduction, and predict the future traffic state evolution process of expressways in the time domain. S3. Based on the traffic condition prediction results, establish a bottleneck capacity reduction prediction model to predict the change process of the actual bottleneck capacity and the congestion propagation status in the future time domain. S4. Based on the bottleneck capacity reduction prediction results, establish a network layer collaborative optimization control model, with the goal of minimizing the total travel time of the expressway network, and optimize the outflow rate of each link target. S5. Based on the network layer optimization results, establish a local layer variable rate limiting control model, with the goal of minimizing the tracking error between the actual outflow rate and the target outflow rate, and optimize the variable rate limiting values ​​of each link. S6. Adopt a rolling prediction and optimization mechanism to update traffic status in real time and output a variable speed limit control scheme.

2. The hierarchical collaborative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 1, characterized in that, In step S2, the link free-flow speed is dynamically updated according to the variable speed limit value, and the updated free-flow speed is used to correct the link transmission capacity and the link reception capacity; the link free-flow speed and the displayed speed limit value satisfy the correction relationship of the driver compliance deviation coefficient.

3. The hierarchical collaborative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 2, characterized in that, In step S2, the link capacity is determined by the congestion density, free flow velocity, and congestion wave propagation velocity, and changes dynamically with the variable speed limit value; the link receiving capacity is constrained by the remaining storage space of the link and the congestion wave propagation process, and the link transmitting capacity is constrained by the vehicle operating speed and the link capacity.

4. The hierarchical cooperative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 1, characterized in that, In step S3, the capacity decline state is determined based on the relationship between upstream traffic demand and downstream traffic supply. When upstream traffic demand exceeds downstream traffic supply, the bottleneck is determined to enter a capacity decline state.

5. The hierarchical collaborative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 4, characterized in that, In step S3, a capacity degradation coefficient is established, the actual throughput capacity of the bottleneck is dynamically calculated based on the capacity degradation coefficient, and the capacity degradation coefficient is dynamically updated according to the changes in link density. When the link density is higher than the critical density, the capacity degradation gradually increases; when the link density is lower than the critical density, the capacity gradually recovers.

6. The hierarchical collaborative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 1, characterized in that, In step S4, a network-layer collaborative optimization control model is established with the outflow rate of each link as the optimization variable and the minimum total network travel time as the optimization objective, while simultaneously satisfying traffic flow conservation constraints, link capacity constraints, and capacity reduction constraints; the capacity reduction constraint uses the actual throughput capacity of the bottleneck as the constraint boundary.

7. The hierarchical cooperative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 6, characterized in that, In step S4, when it is predicted that a capacity decrease will occur in the bottleneck area, the target outflow rate of the upstream link of the bottleneck is reduced, while the downstream link is kept at a high release capacity to slow down the speed at which vehicles gather in the bottleneck area.

8. The hierarchical collaborative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 1, characterized in that, In step S5, the variable speed limit value of each link is used as the optimization variable. The local layer objective function is composed of the outflow rate tracking error term and the speed limit change penalty term, while satisfying the upper and lower limit constraints of the speed limit value and the constraint of the maximum speed limit change in a single cycle.

9. The hierarchical cooperative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 8, characterized in that, In step S5, when it is predicted that a capacity decrease will occur in the bottleneck area, the speed limit value of the upstream link of the bottleneck is reduced in advance to reduce the speed at which vehicles reach the bottleneck; after the bottleneck area returns to normal operation, the speed limit value is gradually increased to restore traffic flow to normal.

10. The hierarchical cooperative control method for variable speed limits on expressways considering bottleneck capacity reduction as described in claim 1, characterized in that, In step S6, only the first control action in the current optimization sequence is executed after each optimization is completed. After the control cycle ends, traffic state information is collected again and the prediction model state is updated. Steps S2 to S5 are repeated to achieve rolling prediction and rolling optimization.

Images