A comprehensive control method for merging bottleneck areas based on dynamic division of lane-level cells

Abstract

The present invention discloses a comprehensive control method for a merging bottleneck area based on dynamic division of lane-level cells, including: setting cells of dynamic adaptive size for each lane in the merging bottleneck area; obtaining the initial distribution information of the traffic flow of each cell, and constructing a CTM traffic flow model for the merging bottleneck area; based on the CTM traffic flow model, comprehensively controlling the traffic flow in the merging bottleneck area. The present invention implements lane-level dynamic division of cell size, optimizes and improves the CTM traffic flow model, optimizes and solves multiple controls and their coordinated strategies in the MPC framework, provides a feasible idea for the control strategy of the merging bottleneck section in the intelligent network environment, and can effectively alleviate the merging bottleneck problem.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation systems and traffic control technology, and specifically relates to a comprehensive control method for a merging bottleneck area based on dynamic division of lane-level cells. Background Art

[0002] In the field of highways, the intelligent network environment has brought great significance to the coordinated control of highways. First, the intelligent network environment can improve the traffic efficiency and smoothness of highways. Through information sharing and interaction between vehicles, it can optimize road condition prediction, traffic scheduling and road flow control, alleviate the congestion problem of highways, and improve road capacity. Secondly, the intelligent network traffic environment can improve the safety of highways. Through information exchange between vehicles, driver behavior monitoring and accident warning, traffic accidents can be avoided, and casualties and property losses can be reduced. In addition, the intelligent network environment can improve the environmental quality of highways. By controlling the speed and road conditions of vehicles and optimizing routes, automobile exhaust emissions can be reduced and environmental pollution can be reduced. Finally, the development of the intelligent network environment has also promoted the coordinated control of highways. Cooperative control can realize the intelligent management of highways through information exchange between vehicles, roads and the Internet, and improve the safety, efficiency and environmental protection of highways.

[0003] Traffic control on ramps and their adjacent mainline areas is one of the important means to solve the congestion problem on highways. By effectively managing the traffic on ramps and their adjacent mainline areas, the congestion on highways can be alleviated. Reasonable planning of traffic flow on ramps and adjacent mainline areas can improve the road's traffic capacity and reduce congestion. By strengthening the regulation and management of ramps entering highways and mainline entrances and exits, the pressure on highways can be relieved and traffic congestion can be avoided. For example, measures such as setting diversion lanes and time-limited flow control at ramp entrances can be adopted. In addition, taking flow control measures in the mainline area can also alleviate ramp and mainline congestion. For example, setting speed limits and intelligent lane guidance systems in the mainline area can improve vehicle traffic efficiency and reduce congestion. In order to achieve the synergistic effect of these control measures and achieve a control effect that is better than a single or independent control strategy, it is necessary to further promote and improve the construction of an intelligent network environment.

[0004] The invention with patent publication number CN115600410A discloses a method for simulating and predicting cellular transmission of mixed traffic flow in urban high-density road networks. The invention simulates and predicts mixed traffic flow based on three types of cells: section cells, interlaced cells and intersection cells, but cannot achieve comprehensive control at the lane level.

[0005] The invention with patent publication number CN113160569A discloses a coordinated control system and method for the main line and entrance ramp of a merging area of ​​an urban expressway. The method divides a road section into multiple cells, each of which contains a main road and an entrance ramp. However, the cell division accuracy of the invention is low, and each cell contains a main road and an entrance ramp. The invention is more inclined to regulate the entire main line section, and the overall regulation capability of the merging bottleneck area is poor.

[0006] The invention with patent publication number CN116798240A discloses a collaborative control system and method for expressway entrance ramp traffic lights and intelligent connected vehicles. Similarly, the invention is mainly used to reduce vehicle conflicts in ramp merging areas and improve the traffic capacity of main line sections, but the overall control capability of merging bottleneck areas is relatively poor. Summary of the invention

[0007] Technical problem solved: The present invention discloses a comprehensive control method for merging bottleneck areas based on dynamic division of lane-level cells. In the ramp control strategy, the inner and outer lanes of the main line section of the merging area are divided, and the feedback control of the outer lane and its reduction coefficient is considered; in the variable speed limit control, the target section density value updated in real time is adopted to consider the influence of the ramp merging flow; in the lane management strategy, a differentiated division of inner and outer lanes is provided for ramp control and main line traffic flow, and an early lane change suggestion is proposed upstream of the merging area with the merging of the outer lane traffic flow and the entrance ramp traffic flow as the variable; finally, in the collaborative control, various control methods are judged by threshold values, and the control distribution method and control parameters under different traffic conditions are released to maximize the optimization benefit of the merging area bottleneck.

[0008] Technical solution:

[0009] A method for comprehensive control of a merging bottleneck area based on dynamic division of lane-level cells, the method comprising the following steps:

[0010] Step A: According to the high-precision map of the highway merging bottleneck area, real-time lane-level traffic information, and the control strategy of the inner and outer lanes, a dynamically adaptive-sized cell is set for each lane in the merging bottleneck area, including an entrance ramp cell and a main lane cell, wherein the main lane cell includes a main inner cell, a main middle cell, and a main outer cell;

[0011] Step B, obtaining the initial distribution information of the traffic flow of each cell, and obtaining the improved CTM traffic flow model of the confluence bottleneck area; calculating the maximum inflow traffic flow, the maximum outflow traffic flow, and the transfer flow between adjacent cells of each cell of the CTM traffic flow model in the current time period, obtaining the traffic flow state information of each cell in the next time period, and updating the CTM traffic flow model;

[0012] Step C, adopting the feedback collaborative method, constructing the collaborative control framework under the intelligent network environment, based on the model predictive control algorithm, combined with the improved CTM traffic flow model and the collaborative control framework under the intelligent network environment, comprehensively controls the traffic flow in the merging bottleneck area; specifically, it includes the following sub-steps:

[0013] Step C1, obtaining the traffic flow status information of each cell of the main line outer lane in the merging bottleneck area in real time, and calculating the traffic flow of the main line outer lane;

[0014] Step C2, determining whether the traffic flow of the outer lane of the main line reaches the outer lane flow threshold, if not, proceeding to step C1, if reached, proceeding to step C3;

[0015] Step C3, in real time, obtain the traffic flow status information of each cell of the entrance ramp of the weaving area and each cell of the outer lane of the main line, calculate the weaving area traffic volume of the merging bottleneck area, and judge whether the weaving area traffic volume of the merging bottleneck area reaches the traffic volume threshold. If not, adopt the lane management strategy to reduce the level of the inner lane upstream of the main line in the merging bottleneck area, induce the upstream vehicles of the main line to concentrate on the inner lane, and then go to step C1; if it reaches, go to step C4;

[0016] Step C4, according to the traffic flow state information of each cell of the entrance ramp of the merging bottleneck area, the queue length of the entrance ramp of the merging bottleneck area is obtained in real time. If the queue length of the entrance ramp does not reach the queue length threshold, the lane management strategy and the ramp control strategy are adopted at the same time until the traffic volume in the weaving area is less than the traffic volume threshold, and then the process goes to step C1; the ramp control strategy refers to: inputting the entrance ramp merging restriction control instruction, and reducing the traffic volume in the weaving area by reducing the entrance ramp merging rate; if the queue length of the entrance ramp reaches the queue length threshold, the lane management strategy, the ramp control strategy and the variable speed limit control strategy are adopted at the same time until the queue length of the entrance ramp is less than the queue length threshold, and then the process goes to step C1; the variable speed limit control strategy refers to: inputting the speed limit control instruction of the upstream main line section of the merging bottleneck area, and reducing the traffic volume in the weaving area and the queue length of the entrance ramp by reducing the vehicle speed of the upstream main line section of the merging bottleneck area.

[0017] Furthermore, in step A, the real-time lane-level traffic information includes vehicle flow, vehicle average speed and microscopic vehicle motion information acquired in real time.

[0018] Furthermore, in step B, under the restriction of ramp control strategy, the traffic volume of ramp cell traffic flow merging into the main line cell is:

[0019] σ m (k) = min[δ j+1 (k), r m (k)]

[0020] Among them, r m (k) is the traffic volume that can be merged into the main line by ramp cell m in the kth time interval under the restriction of ramp control strategy; σ m (k) is the maximum number of vehicles that flow from ramp cell m to the outer cell j+1 of the main line during the kth interval, which is recorded as the sending capacity of ramp cell m; δ j+1 (k) is the maximum number of vehicles that the ramp cell m is allowed to flow into by the mainline outer cell j+1 during the kth interval, which is recorded as the acceptance capacity of the mainline outer cell j+1. The sending capacity and receiving capacity of the cell are calculated through the basic traffic flow diagram.

[0021] Furthermore, in step B, under the variable speed limit control strategy, the speed limit value of the upstream main line lane in the kth time interval is set to v sl,h (k), the density of lanes when the speed limit is taken is ρ vsl,h (k), the maximum traffic flow of this lane is q vsl,h (k), the sending and receiving capacity of the main lane cell under variable speed limit control is obtained as:

[0022] σ h (k) = min{v h (k)·ρ h (k)·n h , ρ vsl,h (k)·v sl,h (k)·n h}·Δt

[0023] δ h (k) = min{w·(ρ jam -ρ h (k))·n h , ρ vsl,h (k)·v sl,h (k)·n h}·Δt

[0024] in:

[0025]

[0026] Where: h (k), v h (k) are the traffic density and speed of the traffic flow of cell h in the main lane in time period k; ρ jam is the blocking density; n h is the number of lanes in the main lane cell h; w is the wave speed; the interval time Δt is the time it takes for a vehicle to pass through the main lane cell at free flow speed;

[0027] In the next time interval, the speed of each main lane cell in the variable speed limit control strategy is determined by the traffic flow density of the cell and the speed limit value in the previous time interval:

[0028]

[0029] Among them, v f It is the free-flowing speed, that is, the average speed when the traffic density approaches zero and vehicles travel unimpeded.

[0030] Furthermore, in step B, in the case of a bottleneck, the traffic flow density ρ i Greater than the critical density ρ m ′, the traffic evacuation flow at the bottleneck head will decrease linearly with a fixed slope w′; the maximum number of vehicles that the bottleneck area can accept in the modified cellular model is calculated as:

[0031]

[0032] Furthermore, in step C4, considering the different effects of inner and outer lanes on the ramp control scheme, the improved ALINEA algorithm is used to convert the lane flow; the calculation formula for the entrance ramp rate is:

[0033]

[0034] Where: R(k) is the entrance ramp rate in the kth cycle after improvement. The entrance ramp rate refers to the difference between the cross-sectional occupancy rate downstream of the ramp entrance and its expected value; r(k-1) is the entrance ramp rate in the k-1th cycle; (α u ) u represents the reduction factor of lane u; o obj represents the set occupancy rate of the main traffic flow section; o(k-1) represents the occupancy rate of the main traffic flow section in the k-1th cycle; N v With N v+1 Respectively represent the number of lanes in the vth and v+1th intervals; K R To adjust the parameters.

[0035] Furthermore, in step C4, the process of simultaneously adopting the lane management strategy, the ramp control strategy and the variable speed limit control strategy includes the following sub-steps:

[0036] The variable speed limit control strategy is used as the center of mainline traffic flow control, and the MTFC-VSL algorithm is improved. The critical density in the algorithm is updated in real time and the vehicle speed difference rate is considered to optimize the variable speed limit control:

[0037] Assume that the traffic flow and density satisfy:

[0038] q(k)=w(k)ρ(k)+b(k)

[0039] Where: w(k) is the slope, b(k) is the constant term, q(k) and ρ(k) are the vehicle flow and density in the merging bottleneck area respectively;

[0040] By using the least squares method, w(k) is obtained as:

[0041]

[0042] By setting the threshold β + With β - And introduce the parameter α to the critical density ρ cr To update:

[0043]

[0044] In the formula, ρ cr (k) is the critical density to be updated in the merging bottleneck area in the kth cycle; ρ(k) is the vehicle density in the merging bottleneck area in the kth cycle;

[0045] The relationship between the vehicle speed difference rate and the speed limit display value in the merging bottleneck area is as follows:

[0046]

[0047] Where: V sl (k) is the variable speed limit value of the merging bottleneck area in the kth cycle; V c (k) is the optimal variable speed limit value output by the control strategy in the kth cycle; b is the variable speed limit; v f For smooth travel speed; is the average speed of the merging bottleneck section in the k-1th cycle; α c is the difference rate.

[0048] Furthermore, in step C, the total travel time J is selected as the benefit indicator from the aspect of traffic efficiency. TTT , total queue time J TWT and the total time spent J TTS Construct the collaborative control objective function of the highway merging bottleneck area control scenario:

[0049] J TTS =J TTT +J TWT

[0050]

[0051]

[0052] In the formula, the total travel time J TTTIt is expressed as the running time of all vehicles in the road network during the entire simulation cycle, and the total queue time J TWT It is expressed as the waiting time of all vehicles in the queue in the road network during the entire simulation cycle, and the total time spent is J TTS It is represented by the running time of all vehicles in the road network and the waiting time of all queued vehicles in the entire simulation cycle; T is the sampling period; I all is the sequence number of the expressway merging area and its upstream and downstream sections; ρ h (k) is the density of cell h in the kth sampling period; L u is the length of road section u; u is the number of lanes included in road section u; δ r , δ v With r o (l) are the penalty factors in the entry ramp rate and variable speed limit penalty items, which are positive values; θ(k) is the ramp queue length in the kth sampling period; α queue is the entrance ramp queue weight coefficient;

[0053] The constraints are:

[0054] λ u ρ h-1 (k)v h-1 (k)≤q m

[0055] |v h-1 (k)-v h (k)|≤10

[0056] |v h (k+1)-v h (k)|≤10

[0057] θ(k)<θ max

[0058]

[0059] where q m is the maximum traffic flow in the bottleneck section; θ(k) is the ramp queue length; θ max is the maximum queue length of the ramp.

[0060] Furthermore, a genetic algorithm is introduced to optimize the objective function of the highway merging bottleneck area control scenario.

[0061] Beneficial effects:

[0062] First, the comprehensive control method for merging bottleneck areas based on lane-level dynamic cell division of the present invention is different from the traditional uniform division. It implements lane-level dynamic division for cell size and optimizes and improves the CTM traffic flow model. It designs the framework and process of feedback collaborative control in an intelligent network environment, that is, a collaborative control strategy that combines ramp control, variable speed limit control and refined lane management. At the same time, based on the MPC algorithm, the genetic algorithm is introduced to optimize and solve the objective function, and the optimization effect is significant.

[0063] Second, the comprehensive control method of the merging bottleneck area based on the dynamic division of lane-level cells of the present invention analyzes the merging bottleneck section of the highway ramp, improves the control algorithm and traffic flow model, optimizes and solves multiple controls and their collaborative strategies in the MPC framework, and compares and analyzes the optimization effects. It provides a feasible idea for the control strategy of the merging bottleneck section in the intelligent network environment, which can effectively alleviate the merging bottleneck problem. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of a comprehensive control method for a merging bottleneck area based on dynamic division of lane-level cells according to an embodiment of the present invention;

[0065] Figure 2 is the basic diagram of traffic flow with variable speed limit;

[0066] Figure 3 This is a graph showing the decrease in traffic capacity;

[0067] Figure 4 Flowchart for cell size partitioning;

[0068] Figure 5 Set up a map to differentiate cell positions for confluence;

[0069] Figure 6 A collaborative control flow chart combining multiple control methods according to an embodiment of the present invention;

[0070] Figure 7 This is a flow chart of early lane change control according to an embodiment of the present invention;

[0071] Figure 8 This is a diagram of a collaborative control structure based on MPC according to an embodiment of the present invention;

[0072] Fig. 9 4 is a comparison chart of objective function values ​​of various control scenarios in an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The following examples will enable those skilled in the art to more fully understand the present invention, but are not intended to limit the present invention in any way.

[0074] This embodiment discloses a method for comprehensive control of a merging bottleneck area based on dynamic division of lane-level cells. The method for comprehensive control of a merging bottleneck area includes the following steps:

[0075] Step A: According to the high-precision map of the highway merging bottleneck area, real-time lane-level traffic information, and the control strategy of the inner and outer lanes, a dynamically adaptive-sized cell is set for each lane in the merging bottleneck area, including an entrance ramp cell and a main lane cell, wherein the main lane cell includes a main inner cell, a main middle cell, and a main outer cell;

[0076] Step B, obtaining the initial distribution information of the traffic flow of each cell, and obtaining the improved CTM traffic flow model of the confluence bottleneck area; calculating the maximum inflow traffic flow, the maximum outflow traffic flow, and the transfer flow between adjacent cells of each cell of the CTM traffic flow model in the current time period, obtaining the traffic flow state information of each cell in the next time period, and updating the CTM traffic flow model;

[0077] Step C, adopting the feedback collaborative method, constructing the collaborative control framework under the intelligent network environment, based on the model predictive control algorithm, combined with the improved CTM traffic flow model and the collaborative control framework under the intelligent network environment, comprehensively controls the traffic flow in the merging bottleneck area; specifically, it includes the following sub-steps:

[0078] Step C1, obtaining the traffic flow status information of each cell of the main line outer lane in the merging bottleneck area in real time, and calculating the traffic flow of the main line outer lane;

[0079] Step C2, determining whether the traffic flow of the outer lane of the main line reaches the outer lane flow threshold, if not, proceeding to step C1, if reached, proceeding to step C3;

[0080] Step C3, in real time, obtain the traffic flow status information of each cell of the entrance ramp of the weaving area and each cell of the outer lane of the main line, calculate the weaving area traffic volume of the merging bottleneck area, and judge whether the weaving area traffic volume of the merging bottleneck area reaches the traffic volume threshold. If not, adopt the lane management strategy to reduce the level of the inner lane upstream of the main line in the merging bottleneck area, induce the upstream vehicles of the main line to concentrate on the inner lane, and then go to step C1; if it reaches, go to step C4;

[0081] Step C4, according to the traffic flow state information of each cell of the entrance ramp of the merging bottleneck area, the queue length of the entrance ramp of the merging bottleneck area is obtained in real time. If the queue length of the entrance ramp does not reach the queue length threshold, the lane management strategy and the ramp control strategy are adopted at the same time until the traffic volume in the weaving area is less than the traffic volume threshold, and then the process goes to step C1; the ramp control strategy refers to: inputting the entrance ramp merging restriction control instruction, and reducing the traffic volume in the weaving area by reducing the entrance ramp merging rate; if the queue length of the entrance ramp reaches the queue length threshold, the lane management strategy, the ramp control strategy and the variable speed limit control strategy are adopted at the same time until the queue length of the entrance ramp is less than the queue length threshold, and then the process goes to step C1; the variable speed limit control strategy refers to: inputting the speed limit control instruction of the upstream main line section of the merging bottleneck area, and reducing the traffic volume in the weaving area and the queue length of the entrance ramp by reducing the vehicle speed of the upstream main line section of the merging bottleneck area.

[0082] First, this embodiment is based on the high-precision map of the highway merging bottleneck area and real-time lane-level traffic information, while considering the impact of various control strategies on the inner and outer lanes, and optimizing and improving the CTM traffic flow model (including setting cells of dynamic adaptive size for each lane, etc.); then, the framework of collaborative control in an intelligent network environment is designed, and a feedback collaborative method is adopted to consider the interaction between ramp control and mainline control. When the mainline traffic flow is given priority, only ramp control is adopted, and ramp control and mainline control are adopted simultaneously after the threshold is reached; finally, based on the MPC control method, the above collaborative control framework is applied, and the above-mentioned improved traffic flow model is used as the prediction model, and a genetic algorithm is introduced to optimize and solve the objective function of highway collaborative control.

[0083] Different from the traditional uniform division, the present invention implements dynamic lane-level division of cell size based on the high-precision map of the highway merging bottleneck area and real-time lane-level traffic information, such as real-time traffic flow, average speed, microscopic vehicle motion information, etc.

[0084] Considering the influence of ramp control, the number of vehicles merging from ramp cells in the weaving area into the main line cells is restricted, so the traffic volume of ramp cell traffic flow merging into the main line cells is:

[0085] σ m (k) = min[δ j+1 (k), r m (k)]

[0086] Among them, r m (k) is the traffic volume that can be merged into the main line by ramp cell m in the kth time interval under the restriction of ramp control strategy; σ m (k) is the maximum number of vehicles that flow from ramp cell m to the outer cell j+1 of the main line during the kth interval, which is recorded as the sending capacity of ramp cell m; δj+1 (k) is the maximum number of vehicles that the ramp cell m is allowed to flow into by the mainline outer cell j+1 during the kth interval, which is recorded as the acceptance capacity of the mainline outer cell j+1. The sending capacity and receiving capacity of the cell are calculated through the basic traffic flow diagram.

[0087] In the CTM model, when determining the traffic flow transfer of each cell segment, it is necessary to first obtain the initial value of the CTM model, that is, the initial distribution of traffic flow of each cell, and at the same time set the inflow of the starting cell traffic flow and the outflow of the ending cell as well as the inflow and outflow of the ramp traffic flow in the overall model. After that, it is necessary to calculate the maximum inflow and outflow traffic flow of each cell in the current time period and calculate the transfer flow between adjacent cells based on this, so as to obtain the traffic flow status information of each cell in the next time period and update the overall CTM model status.

[0088] Considering the influence of variable speed limit control, the basic diagram of traffic flow changes as follows Figure 2 As shown, let the speed limit of the upstream main line lane in the kth time interval be v sl,h (k), the density of lanes when the speed limit is taken is ρ vsl,h (k), the maximum traffic flow of this lane is q vsl,h (k), the sending and receiving capacity of the main lane cell under variable speed limit control is obtained as:

[0089] σ h (k) = min{v h (k)·ρ h (k)·n h , ρ vsl,h (k)·v sl,h (k)·n h}·Δt

[0090] δ h (k) = min{w·(ρ jam -ρ h (k))·n h , ρ vsl,h (k)·v sl,h (k)·n h}·Δt

[0091] in:

[0092]

[0093] Where: h (k), v h (k) are the traffic density and speed of the traffic flow of cell h in the main lane in time period k; ρ jam is the blocking density; n his the number of lanes in the main lane cell h; w is the wave speed; the interval time Δt is the time it takes for a vehicle to pass through the main lane cell at the free flow speed. The main lane cells here include the main line inner cells, the main line middle cells, and the main line outer cells; the purpose of setting the main line middle cells is to implement a more detailed division of the lanes, and to more accurately calculate the flow transferred between adjacent cells of adjacent lanes (when the vehicle changes lanes).

[0094] In the next time interval, the speed of each main lane cell in the variable speed limit control strategy is determined by the traffic flow density of the cell and the speed limit value in the previous time interval:

[0095]

[0096] Among them, v f It is the free-flowing speed, that is, the average speed when the traffic density approaches zero and vehicles travel unimpeded.

[0097] Considering the road section with reduced traffic capacity, in the scenario of bottleneck, such as Figure 3 As shown, the traffic flow density is greater than the critical density, i.e., ρ i >ρ m ′, the flow rate of the bottleneck head traffic evacuation flow will decrease linearly with a fixed slope w′, which is different from the traditional CTM, where the congested head traffic evacuation flow will be equal to the maximum traffic flow of the cell segment when the critical density is reached. The cell size division process is shown in Figure 4 .

[0098] Based on the basic traffic flow diagram showing the decrease in capacity, it can be obtained that the maximum number of vehicles that can be accepted in the bottleneck area in the modified cellular model is:

[0099]

[0100] The meaning of the parameters in the formula can be found in the above and Figure 3 .

[0101] See the cell position settings in the confluence section. Figure 5 Based on the CTM model, sub-cells are set for the inner, middle and outer lanes of the main line in the merging area and its upstream of the expressway, and the inner lane cell is indicated as a relatively independent cell that prohibits lane change in the merging area. Considering that the traffic flow level of the outer lane in the merging area, which is greatly affected by the merging flow, will be affected, by implementing early lane change information prompts upstream of the merging area, the traffic flow level of the inner lane cell is adjusted to the level of the outer lane cell in the merging area before merging, that is, the main line traffic flow is relatively concentrated to the inner lane before merging, so as to alleviate the traffic flow condition of the outer lane cell affected by the merging flow. At the same time, by setting the entrance ramp cell, the traffic flow status, queue length and delay of the entrance ramp can be clearly understood, which is convenient for the evaluation and analysis of the overall merging section.

[0102] The collaborative control process of combining multiple control methods in this embodiment is as follows: Figure 6 shown.

[0103] According to the traffic flow information (including but not limited to traffic flow, speed distribution, density distribution, microscopic vehicle control information and system evaluation of main line traffic conditions, etc.) received by the RIU at the entrance ramp via the communication system from the main line bottleneck area and its upstream RIU and the ramp queue status information of the queue detector, the main line traffic flow density, vehicle speed and queue situation are judged, and the corresponding vehicle merging optimization control strategy is adopted. The PI-ALINEA algorithm is improved to improve the adaptability and response speed of ramp control. At the same time, the different effects of the inner and outer lanes on the ramp control scheme are considered, and the lane conversion is added. The calculation formula of the entrance ramp rate is:

[0104]

[0105] Where: R(k) is the entrance ramp rate in the kth cycle after improvement. The entrance ramp rate refers to the difference between the cross-sectional occupancy rate downstream of the ramp entrance and its expected value; r(k-1) is the entrance ramp rate in the k-1th cycle; (α u ) u represents the reduction factor of lane u; o obj represents the set occupancy rate of the main traffic flow section; o(k-1) represents the occupancy rate of the main traffic flow section in the k-1th cycle; N v With N v+1 Respectively represent the number of lanes in the vth and v+1th intervals; K R is the adjustment parameter. If the value of α is 0.9, the reduction coefficient of the first lane close to the shoulder is (α1) 1 =0.9, and so on to 0.81, and so on. The ramp control method in the multi-lane traffic scenario is optimized by changing the reduction coefficient.

[0106] The variable speed limit of the main line is taken as the center of the main line traffic flow control, and the real-time update of the target density and the vehicle speed difference rate are taken into consideration to optimize the variable speed limit control:

[0107] Dynamically update critical density ρ based on least squares method cr , as follows:

[0108] Assume that the traffic flow and density satisfy:

[0109] q(k)=w(k)ρ(k)+b(k)

[0110] Where: w(k) is the slope, b(k) is the constant term, q(k) and ρ(k) are the vehicle flow and density in the merging bottleneck area respectively;

[0111] By using the least squares method, w(k) is obtained as:

[0112]

[0113] By setting the threshold β + With β - And introduce the parameter α to the critical density ρ cr To update:

[0114]

[0115] In the formula, ρ cr (k) is the critical density to be updated in the merging bottleneck area in the kth cycle; ρ(k) is the vehicle density in the merging bottleneck area in the kth cycle;

[0116] As the speed difference rate of vehicles in the overall system increases, a higher final speed limit value can be obtained, thereby reducing the impact on the main line capacity. Considering the actual speed difference rate, mainly based on the relationship between the difference rate of the merging bottleneck area and the speed limit display value, the expression is as follows:

[0117] V sl (k) = (1 + α c )·V c (k) = (1 + α c )·b·V f

[0118]

[0119] Where: V sl (k) is the variable speed limit value of the merging bottleneck area in the kth cycle; V c (k) is the optimal variable speed limit value output by the control strategy in the kth cycle; b is the variable speed limit; v f For smooth travel speed; is the average speed of the merging bottleneck section in the k-1th cycle; α c is the difference rate.

[0120] The lane management of the merging area in this embodiment first divides the lane layout: the main line lane is divided into an inner lane and an outer lane according to the distance from the ramp, and the inner lane away from the ramp is set as the main line flow lane in the merging section where the ramp merges, and lane changing is not allowed in this section. This can reduce unnecessary interweaving of the main line flow and the ramp flow. At the same time, at an appropriate position upstream of the merging, the intelligent network system issues advance lane change control so that the main line flow is relatively concentrated on the inner lane. Secondly, the ratio of the traffic volume to the capacity of the inner and outer lanes is calculated by the ratio of the expected flow and the capacity. If the operating status level of the inner lane of the main line is greater than that of the outer lane, the operating status level of the inner lane is adjusted to the same as that of any of the two outer lanes, or even one level lower, to avoid inducing more vehicles upstream of the main line to choose to drive in the two outer lanes, thereby reducing the operating efficiency of the merging area. The above results are input into the main line advance lane change control. Due to the change in the traffic flow service level restriction, the service level of the inner lane is lowered to accommodate more traffic flow to reduce the pressure of merging the outer lane and the ramp. The traffic flow growth space of the inner lane can use the intelligent network environment to publish the traffic operation status of each lane in the merging area ahead to remind the driver to choose the appropriate lane in advance. The process is as follows Figure 7 shown.

[0121] like Figure 8 As shown, this embodiment combines ramp control, variable speed limit control and lane management strategy under the MPC control strategy to achieve collaborative control combining multiple control methods.

[0122] The control framework mainly includes the following contents:

[0123] (1) Prediction model: improved CTM traffic flow model;

[0124] (2) Cooperative control system: including ramp control, variable speed limit control and lane management control system;

[0125] (3) Optimization objective: Select appropriate benefit indicators from the perspective of traffic efficiency to construct the objective function, including total travel time (TTT, represented by the running time of all vehicles in the road network during the entire simulation cycle), total queuing time (TWT, represented by the waiting time of all queued vehicles in the road network during the entire simulation cycle), and total time spent (TTS, represented by the running time of all vehicles in the road network and the waiting time of all queued vehicles in the entire simulation cycle). The total time is the sum of the total travel time and the total queue time, and the expression is as follows:

[0126]

[0127] Among them: J TTT is the total travel time; T is the sampling period; I all is the sequence number of the expressway merging area and its upstream and downstream sections; ρh (k) is the density of cell h in the kth sampling period; L u is the length of road section u; u is the number of lanes included in road section u; δ r , δ v With r o (l) are the penalty factors in the entry ramp rate and variable speed limit penalty items, which take positive values; the other parameters are as mentioned above.

[0128]

[0129] Among them: J TWT is the total queue time; θ(k) is the ramp queue length in the kth sampling period; α queue is the queue weight coefficient of the entrance ramp. From the relationship between TTS, travel and waiting time, we can get:

[0130] J TTS =J TTT +J TWT

[0131] In the formula, J TTs For total time.

[0132] (4) Constraints: Considering the traffic flow state and the actual situation of the bottleneck section, model constraints are added. Their respective meanings are as follows: the upstream flow of the merge should not be greater than the actual traffic capacity of the merge bottleneck section; the difference between the variable speed limit control in adjacent sections or adjacent time periods should not exceed 10 km / h; the ramp queue length should not exceed the maximum queue length; the difference rate of the entrance ramp rate output by the ramp control in adjacent cycles should not be higher than 15%. Specifically:

[0133] λ u ρ h-1 (k)v h-1 (k)≤q m

[0134] |v h-1 (k)-v h (k)|≤10

[0135] |v h (k+1)-v h (k)|≤10

[0136] θ(k)<θ max

[0137]

[0138] where q m is the maximum traffic flow in the bottleneck section; θ(k) is the ramp queue length; θ maxis the maximum queue length of the ramp.

[0139] This embodiment sets the optimization solution process of multiple control measures and their coordination strategies based on the NSGA-II algorithm.

[0140] The comparison diagram of the objective function values ​​under various control scenarios in this embodiment is as follows: Fig. 9 As shown, compared with the uncontrolled scenario, the TWT value optimization degree of the ramp control scenario is 46.5%, and the overall TTS optimization degree is 7.4%; the TTT value optimization degree of the main line control is 23.5%, the TWT value optimization degree is 35.6%, and the overall TTS optimization degree is 26.0%; the TTT value optimization degree of the coordinated control of ramps, variable speed limits and lane management is 40.5%, and the overall TTS optimization degree is 36.8%. In summary, the comprehensive optimization control method for ramp merging areas based on fine division of lane cells described in the present invention has the best effect in the above control groups.

[0141] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A comprehensive management and control method for merging bottleneck areas based on dynamic division of lane-level cells, characterized in that: The comprehensive management and control method for the confluence bottleneck area comprises the following steps: Step A: According to the high-precision map of the highway merging bottleneck area, real-time lane-level traffic information, and the control strategy of the inner and outer lanes, a dynamically adaptive-sized cell is set for each lane in the merging bottleneck area, including an entrance ramp cell and a main lane cell, wherein the main lane cell includes a main inner cell, a main middle cell, and a main outer cell; Step B, obtaining the initial distribution information of the traffic flow of each cell, and obtaining the improved CTM traffic flow model of the confluence bottleneck area; calculating the maximum inflow traffic flow, the maximum outflow traffic flow, and the transfer flow between adjacent cells of each cell of the CTM traffic flow model in the current time period, obtaining the traffic flow state information of each cell in the next time period, and updating the CTM traffic flow model; Step C, adopting the feedback collaborative method, constructing the collaborative control framework under the intelligent network environment, based on the model predictive control algorithm, combined with the improved CTM traffic flow model and the collaborative control framework under the intelligent network environment, comprehensively controls the traffic flow in the merging bottleneck area; specifically, it includes the following sub-steps: Step C1, obtaining the traffic flow status information of each cell of the main line outer lane in the merging bottleneck area in real time, and calculating the traffic flow of the main line outer lane; Step C2, determining whether the traffic flow of the outer lane of the main line reaches the outer lane flow threshold, if not, proceeding to step C1, if reached, proceeding to step C3; Step C3, in real time, obtain the traffic flow status information of each cell of the entrance ramp of the weaving area and each cell of the outer lane of the main line, calculate the weaving area traffic volume of the merging bottleneck area, and judge whether the weaving area traffic volume of the merging bottleneck area reaches the traffic volume threshold. If not, adopt the lane management strategy to reduce the level of the inner lane upstream of the main line in the merging bottleneck area, induce the upstream vehicles of the main line to concentrate on the inner lane, and then go to step C1; if it reaches, go to step C4; Step C4, according to the traffic flow state information of each cell of the entrance ramp of the merging bottleneck area, the queue length of the entrance ramp of the merging bottleneck area is obtained in real time. If the queue length of the entrance ramp does not reach the queue length threshold, the lane management strategy and the ramp control strategy are adopted at the same time until the traffic volume in the weaving area is less than the traffic volume threshold, and then the process goes to step C1; the ramp control strategy refers to: inputting the entrance ramp merging restriction control instruction, and reducing the traffic volume in the weaving area by reducing the entrance ramp merging rate; if the queue length of the entrance ramp reaches the queue length threshold, the lane management strategy, the ramp control strategy and the variable speed limit control strategy are adopted at the same time until the queue length of the entrance ramp is less than the queue length threshold, and then the process goes to step C1; the variable speed limit control strategy refers to: inputting the speed limit control instruction of the upstream main line section of the merging bottleneck area, and reducing the traffic volume in the weaving area and the queue length of the entrance ramp by reducing the vehicle speed of the upstream main line section of the merging bottleneck area; In step B, under the restriction of ramp control strategy, the traffic volume of ramp cell merging into main line cell is: s m (k)=min[δ j+1 (k),r m (k)] Among them, r m (k) is the traffic volume that can be merged into the main line by ramp cell m in the kth time interval under the restriction of ramp control strategy; σ m (k) is the maximum number of vehicles that flow from ramp cell m to the outer cell j+1 of the main line during the kth interval, which is recorded as the sending capacity of ramp cell m; δ j+1 (k) is the maximum number of vehicles that the mainline outer cell j+1 allows the ramp cell m to flow into during the kth interval, which is recorded as the acceptance capacity of the mainline outer cell j+1. The sending capacity and acceptance capacity of the cell are calculated through the basic traffic flow diagram; In step B, under the variable speed limit control strategy, the speed limit of the upstream main line lane in the kth time interval is set to v sl,h (k), the density of lanes when the speed limit is taken is ρ vsl,h (k), the maximum traffic flow of this lane is q vsl,h (k), the sending and receiving capacity of the main lane cell under variable speed limit control is obtained as: s h (k)=min{v h (k)·r h (k)·n h ,r vsl,h (k)·v sl,h (k)·n h }·Δt d h (k)=min{w·(ρ jam -r h (k))·n h ,r vsl，h (k)·v sl,h (k)·n h }·Δt in: Where: h (k), v h (k) are the traffic density and speed of the traffic flow of cell h in the main lane in time period k; ρ jam is the blocking density; n h is the number of lanes in the main lane cell h; w is the wave speed; the interval time Δt is the time it takes for a vehicle to pass through the main lane cell at free flow speed; In the next time interval, the speed of each main lane cell in the variable speed limit control strategy is determined by the traffic flow density of the cell and the speed limit value in the previous time interval: Among them, v f is the free-flow speed, that is, the average speed when the traffic density approaches zero and vehicles travel unimpeded; In step B, in the case of a bottleneck, the traffic flow density ρ i Greater than the critical density ρ m ′, the traffic evacuation flow at the bottleneck head will decrease linearly with a fixed slope w′; the maximum number of vehicles that the bottleneck area can accept in the modified cellular model is calculated as: In step C4, considering the different impacts of inner and outer lanes on the ramp control scheme, the improved ALINEA algorithm is used to convert the lane flow; the calculation formula for the entrance ramp rate is: Where: R(k) is the entrance ramp rate in the kth cycle after improvement. The entrance ramp rate refers to the difference between the cross-sectional occupancy rate downstream of the ramp entrance and its expected value; r(k-1) is the entrance ramp rate in the k-1th cycle; (α u ) u represents the reduction factor of lane u; o obj represents the set occupancy rate of the main traffic flow section; o(k-1) represents the occupancy rate of the main traffic flow section in the k-1th cycle; N v With N v+1 Respectively represent the number of lanes in the vth and v+1th intervals; K R To adjust the parameters; In step C4, the process of simultaneously adopting the lane management strategy, the ramp control strategy and the variable speed limit control strategy includes the following sub-steps: The variable speed limit control strategy is used as the center of mainline traffic flow control, and the MTFC-VSL algorithm is improved. The critical density in the algorithm is updated in real time and the vehicle speed difference rate is considered to optimize the variable speed limit control: Assume that the traffic flow and density satisfy: q(k)=w(k)ρ(k)+b(k) Where: w(k) is the slope, b(k) is the constant term, q(k) and ρ(k) are the vehicle flow and density in the merging bottleneck area respectively; By using the least squares method, w(k) is obtained as: By setting the threshold β + With β - And introduce the parameter α to the critical density ρ cr To update: In the formula, ρc r (k) is the critical density to be updated in the merging bottleneck area in the kth cycle; ρ(k) is the vehicle density in the merging bottleneck area in the kth cycle; The relationship between the vehicle speed difference rate and the speed limit display value in the merging bottleneck area is as follows: V sl (k)=(1+α c )·V c (k)=(1+α c )·b·V f Where: V sl (k) is the variable speed limit value of the merging bottleneck area in the kth cycle; V c (k) is the optimal variable speed limit value output by the control strategy in the kth cycle; b is the variable speed limit; v f For smooth travel speed; is the average speed of the merging bottleneck section in the k-1th cycle; α c is the difference rate; In step C, the benefit indicator total travel time J is selected from the aspect of traffic efficiency. TTT , total queue time J TWT and the total time spent J TTs Construct the collaborative control objective function of the highway merging bottleneck area control scenario: I TTS =J TTT +J TWT In the formula, the total travel time J TTT It is expressed as the running time of all vehicles in the road network during the entire simulation cycle, and the total queue time J TWT It is expressed as the waiting time of all vehicles in the queue in the road network during the entire simulation cycle, and the total time spent is J TTS It is represented by the running time of all vehicles in the road network and the waiting time of all queued vehicles in the whole simulation period; T is the sampling period; H all is the sequence number of the expressway merging area and its upstream and downstream sections; ρ h (k) is the density of cell h in the kth sampling period; L u is the length of road section u; u is the number of lanes included in road section u; δ r , δ v With r o (l) are the penalty factors in the entry ramp rate and variable speed limit penalty items, which are positive values; θ(k) is the ramp queue length in the kth sampling period; α queue is the entrance ramp queue weight coefficient; The constraints are: l u r h-1 (k)v h-1 (k)≤q m |v h-1 (k)-v h (k)|≤10 |v h (k+1)-v h (k)|≤10 θ(k)<θ max where q m is the maximum traffic flow on the bottleneck section; θ(k) is the ramp queue length; θ max is the maximum queue length of the ramp.

2. The method for comprehensive management and control of merging bottleneck areas based on dynamic division of lane-level cells according to claim 1 is characterized in that: In step A, the real-time lane-level traffic information includes real-time acquired vehicle flow, average vehicle speed, and microscopic vehicle motion information.

3. The method for comprehensive management and control of merging bottleneck areas based on dynamic division of lane-level cells according to claim 1 is characterized in that: Genetic algorithm is introduced to optimize the objective function of highway merging bottleneck area control scenario.

Images