A tunnel bottleneck control system and method for giving priority to vehicles with high passenger load

Abstract

The present invention relates to a tunnel bottleneck control system and method for high-load vehicle priority, the method comprising: a data acquisition module collects the operation status information of CAV and CHV in the upstream control area and the tunnel bottleneck area; a decision generation module predicts the CHV trajectory plan, and then determines whether the CHV is on the conventional lane adjacent to the DHL and whether the CHV can obtain speed gain on the DHL, so as to determine whether the CHV needs to apply for CAV to change lanes to the DHL; if there is an application for changing lanes to the DHL, it is determined whether there is a CAV that can change lanes in front and whether the HOV lane can provide the corresponding right of way. If both are judged to be yes, the lane change trajectory optimization decision is made for the CAV, otherwise the vehicle cruise trajectory optimization decision is made for the CAV; finally, the control execution module converts the optimization decision information into a CAV trajectory control instruction, and sends it to the CAV in the road network for execution. Compared with the prior art, the present invention can reduce the traffic congestion in the bottleneck of the expressway tunnel while ensuring the fairness of benefits for traffic travelers.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent traffic control, and in particular to a tunnel bottleneck control system and method thereof for giving priority to vehicles with high passenger load rates. Background Art

[0002] Expressways play an important role in long-distance travel and transportation. With the gradual improvement of urban motorization, traffic congestion and supply-demand mismatch on expressways are becoming increasingly serious. Due to factors such as lighting conditions and speed limits, tunnels on expressways are often prone to sudden drops in capacity, becoming a bottleneck for traffic flow on expressways and causing serious traffic congestion. Expressway tunnel bottleneck management is to adjust the traffic demand upstream of the bottleneck to match the capacity of the expressway tunnel bottleneck, thereby achieving the goal of alleviating tunnel bottleneck congestion.

[0003] The existing methods for managing bottlenecks in expressway tunnels mainly include two strategies: variable speed limit and setting up high-occupancy vehicle priority lanes. Through corresponding control of traffic demand, the congestion of tunnel bottlenecks caused by disordered traffic demand can be reduced. Among them, the variable speed limit is to adjust the speed of upstream vehicles to match the traffic capacity of downstream tunnel bottlenecks as much as possible; the method of high-occupancy vehicle priority lanes is to open a dedicated priority lane on the expressway to provide priority right of way for high-occupancy vehicle passing. However, in the actual application process, there are still the following problems: on the one hand, the traditional variable speed limit control strategy only focuses on the benefit improvement of vehicles as the main body, and it is difficult to take into account the benefits of travelers, which is easy to cause poor fairness in traffic travel; on the other hand, although the existing high-occupancy vehicle priority lane strategy takes into account the benefits of travelers, the setting of priority lanes is difficult to accurately match the dynamic and random changes in traffic demand, resulting in inefficient use of road space resources and poor tunnel bottleneck control effect. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a tunnel bottleneck control system and method that gives priority to high-passenger-load vehicles. Based on the networked autonomous driving technology, it can dynamically and accurately regulate the upstream traffic demand for the traffic capacity of the bottleneck section of the expressway tunnel, and realize the scientific and reasonable control of the bottleneck of the expressway tunnel.

[0005] The object of the present invention can be achieved by the following technical solutions: a tunnel bottleneck control system for high passenger load vehicle priority, comprising a data acquisition module, a decision generation module and a control execution module connected in sequence, wherein the data acquisition module is used to collect the operation status information of the CAV (Connected-Automated Vehicle) and CHV (Connected Human-driven Vehicle) upstream of the bottleneck of the expressway tunnel, and output it to the decision generation module;

[0006] The decision generation module is used to make an optimization decision on the CAV trajectory and output corresponding decision information to the control execution module;

[0007] The control execution module is used to convert the decision information into vehicle control instructions and send them to the CAV in the road network for execution.

[0008] Furthermore, the data acquisition module includes a roadside detector and a vehicle-mounted detector.

[0009] A tunnel bottleneck control method for high-passenger-load vehicle priority includes the following steps:

[0010] S1, the data acquisition module collects the operating status information of CAV and CHV in the upstream control area and the bottleneck area of ​​the tunnel, and transmits it to the decision generation module;

[0011] S2, the decision generation module first predicts the time when the vehicle reaches the bottleneck of the tunnel and predicts the trajectory of the CHV, and then determines whether the CHV is in the regular lane adjacent to the DHL (Dynamic High-occupancy-vehicle Lane) and whether the CHV can obtain speed gain on the DHL, so as to determine whether the CHV needs to apply for the CAV to change lanes to the DHL;

[0012] S3. If there is an application for lane change to the DHL, it is determined whether there is a CAV that can change lanes ahead and whether the HOV (High-Occupancy Vehicle) lane can provide the corresponding right of way. If both are determined to be yes, the lane change trajectory optimization decision is made for the CAV. Otherwise, the CAV does not change lanes and the vehicle cruising trajectory optimization decision is made for the CAV.

[0013] S4. According to the lane change trajectory optimization decision or the vehicle cruising trajectory optimization decision, the control execution module converts the optimization decision information into corresponding CAV trajectory control instructions, and sends them to the CAV in the road network for execution, thereby controlling the driving behavior of the CAV.

[0014] Furthermore, the step S2 specifically includes the following steps:

[0015] S21, predicting the time when the vehicle arrives at the bottleneck of the tunnel, and performing CHV trajectory prediction based on the result of the predicted arrival time;

[0016] S22, determining whether the target CHV is in a regular lane adjacent to the DHL, if so, executing step S23, otherwise, not sending a lane change request to the DHL to the front CAV;

[0017] S23, determining whether the DHL lane has a speed gain compared to the current lane of the target CHV, if so, sending a lane change request to the DHL to the front CAV, otherwise, not sending a lane change request to the DHL to the front CAV.

[0018] Furthermore, in step S21, the time for all vehicles on the upstream section of the expressway tunnel bottleneck to arrive at the tunnel bottleneck is predicted according to the following formula:

[0019]

[0020]

[0021] Where n is the vehicle number index in the traffic flow, N is the vehicle number index set, k is the lane number index, and k is the lane index set (k = 1, 2, 3). is the earliest time that vehicle n in lane k can reach the bottleneck of the tunnel without considering the preceding vehicle, t 0 is the time corresponding to the current time step, L, v op and a max They represent the length of the upstream control area of ​​the expressway tunnel bottleneck, the optimal speed of the vehicle expected to pass through the tunnel bottleneck, and the maximum acceleration of the vehicle. is the initial speed of vehicle n in lane k, is the predicted time for vehicle n-1 in lane k to reach the bottleneck of the tunnel, t h is the expected vehicle safety headway;

[0022] The trajectory of the CHV reaching the bottleneck of the fast-speed tunnel is predicted according to the following formula:

[0023]

[0024]

[0025]

[0026]

[0027] Where m is the index of the number of connected human-driven vehicles in the traffic flow, Nm is the index set of all networked human-driven vehicles, t is the time step index, T is the time step index set, s 0 The safe distance for vehicles to travel. and They represent the position and speed of the online human-driven vehicle m in lane k at time step t, and They represent the acceleration of the online human-driven vehicle m in lane k at time step t when there is a car ahead and the acceleration of the online human-driven vehicle m in lane k at time step t when there is no car ahead. is the minimum expected distance between the connected human-driven vehicle m and the preceding vehicle in lane k, The actual distance between the connected human-driven vehicle m and the vehicle ahead in lane k, is the minimum expected interval of the connected human-driven vehicle m when there is no vehicle ahead, Represents the speed difference between the connected human-driven vehicle m and the vehicle in front, a d is the expected deceleration of the connected human-driven vehicle m.

[0028] Furthermore, the lane change trajectory optimization decision for the CAV and the vehicle cruising trajectory optimization decision for the CAV in step S3 are both performed based on set optimization decision objectives and constraints. The optimization decision objectives include travel time, fuel consumption, and optimal traffic state at the bottleneck of the tunnel. The constraints include kinematic constraints, vehicle collision avoidance constraints, lane management constraints, boundary solution constraints, and dynamic lane right of way allocation constraints.

[0029] Furthermore, the optimization decision objective is specifically:

[0030]

[0031]

[0032] Where c is the index of the number of the connected autonomous driving vehicle in the traffic flow, N c is the set of indexes of all connected human-driven vehicles, α, β and μ are weight coefficients, τ c is the travel time of the connected autonomous vehicle c from the starting position in the upstream control area to the bottleneck of the tunnel, is the acceleration of the connected autonomous driving vehicle c at time step t, and They are the position and speed of the connected autonomous driving vehicle c in lane k when it is predicted to reach the bottleneck of the tunnel.

[0033] Furthermore, the kinematic constraints are specifically:

[0034]

[0035]

[0036]

[0037]

[0038] in, and They represent the position, velocity and acceleration of the connected autonomous driving vehicle c in lane k at time step t, respectively. min 、v max 、a min and a max are the minimum speed, maximum speed, minimum acceleration and maximum acceleration of the vehicle respectively, and Δt is the unit time step interval.

[0039] Furthermore, the vehicle collision avoidance constraints are specifically:

[0040]

[0041]

[0042]

[0043]

[0044] Where h is the HOV number index, N h is a set of high-occupancy vehicle number indexes, is the position of the preceding vehicle of the connected autonomous driving vehicle c in lane k at time step t, is the position of the vehicle behind the connected autonomous driving vehicle c in lane k at time step t, is the position of high-occupancy vehicle h on lane k at time step t, is the position of the vehicle ahead of the high-occupancy vehicle h on lane k at time step t, is the position of the vehicle behind the high-occupancy vehicle h on lane k at time step t, l c and l h They are the length of connected autonomous driving vehicles and high-occupancy vehicles respectively.

[0045] Furthermore, the lane management constraint is specifically: the dynamic high passenger load lane is numbered k=3, and the other lanes are regular lanes (k=1, 2), then

[0046]

[0047]

[0048] in, is a binary auxiliary variable. If vehicle c is in lane k (k = 1, 2 or 3) at time step t, then on the contrary, and Binary auxiliary variables representing whether the connected autonomous driving vehicle c is located in the dynamic high-occupancy lane at time step t and time step t+1. If it is located, the auxiliary variable and is 1, otherwise it is 0.

[0049] Furthermore, the boundary constraint is used to narrow the solution scope and improve the efficiency of the solution of the vehicle trajectory control scheme, specifically:

[0050]

[0051] Wherein, M is a sufficiently large constant.

[0052] Furthermore, the dynamic lane right of way allocation constraints are specifically:

[0053]

[0054]

[0055] in, and They respectively represent the set of right-of-way segments allocated to the connected autonomous driving vehicle c and the high-occupancy vehicle h on the dynamic high-occupancy lane at time step t.

[0056] Compared with the prior art, the present invention has the following advantages:

[0057] 1. The present invention sets a data acquisition module, a decision generation module and a control execution module connected in sequence, uses the data acquisition module to sense and acquire vehicle operation status information (including physical information such as vehicle position, speed and acceleration, and passenger load information), and uses the above information as the input of the decision generation module downstream of the logic end, so as to dynamically optimize and make decisions on the trajectory plan of the networked autonomous driving vehicle according to the traffic conditions of the bottleneck in the tunnel ahead, and finally transmits the decision plan to the underlying control execution module to achieve trajectory control of the networked autonomous driving vehicle. In this way, a traffic flow control method for networked autonomous driving vehicles guided by passenger load factors is realized, based on networked autonomous driving technology, combined with vehicle passenger load factors, to control the demand of traffic flow upstream of the expressway bottleneck, which can reduce traffic congestion in the bottleneck of the expressway tunnel while also ensuring the fairness of benefits for traffic travelers.

[0058] 2. The present invention first predicts the time when a vehicle arrives at a bottleneck in a tunnel, predicts the trajectory of a human-driven vehicle based on the predicted arrival time, and then determines whether the human-driven vehicle is in a conventional lane adjacent to a dynamic high-load lane (DHL) and whether the vehicle can obtain a speed gain in the dynamic high-load lane to determine whether to issue a lane change application. Then, it determines whether there is a qualified connected autonomous driving vehicle in front that can change lanes, and then optimizes the lane change trajectory or cruise trajectory based on the optimization decision goal and constraints, thereby realizing a dynamic high-load rate priority lane method. The method reduces the dynamic mismatch between supply and demand caused by the traditional fixed lane setting, can dynamically identify the running status of traffic flow, and takes the current expressway tunnel bottleneck capacity as the target. The traffic flow upstream of the tunnel bottleneck is dynamically and accurately controlled by lane priority. While meeting the priority demand of high-load rate vehicles, the problem of low efficiency in the use of priority lane right-of-way resources is avoided, thereby realizing dynamic and accurate control of expressway tunnel bottlenecks.

[0059] 3. When making vehicle trajectory optimization decisions, the present invention takes the travel efficiency of travelers as the goal, including travel time, fuel consumption and the best traffic state at the bottleneck of the tunnel. It can improve the traffic operation state at the bottleneck of the expressway tunnel while ensuring the rationality and fairness of the allocation of road rights resources, and improve the stability of the overall traffic system; in addition, the corresponding kinematic constraints, vehicle collision avoidance constraints, lane management constraints, boundary constraints and dynamic lane road right allocation constraints are designed, thereby allocating priority road rights to upstream high-occupancy rate vehicles based on the real-time traffic capacity at the bottleneck of the expressway tunnel, dynamically opening and setting high-occupancy rate priority lanes, improving the efficiency of the use of road rights resources on the priority lanes, realizing active control of the bottleneck traffic of the expressway tunnel, and forming a trajectory plan for networked autonomous driving vehicles that prioritizes high-occupancy rate vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0061] Figure 2 It is a schematic diagram of the method flow of the present invention;

[0062] Figure 3 It is a schematic diagram of an application scenario of an embodiment;

[0063] Figure 4 It is a schematic diagram of the application process of the embodiment;

[0064] Description of the marks in the figure: 1. Data acquisition module, 2. Decision making module, 3. Control execution module. DETAILED DESCRIPTION

[0065] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0066] Example

[0067] like Figure 1 As shown, a tunnel bottleneck control system for high-passenger-load vehicle priority includes a data acquisition module 1, a decision-making module 2 and a control execution module 3 connected in sequence, wherein the data acquisition module 1 is used to collect the operation status information of the CAV and CHV upstream of the bottleneck of the expressway tunnel, and output it to the decision-making module 2;

[0068] The decision generation module 2 is used to make an optimization decision on the CAV trajectory and output the corresponding decision information to the control execution module 3;

[0069] The control execution module 3 is used to convert the decision information into vehicle control instructions and send them to the CAV in the road network for execution.

[0070] Among them, the data acquisition module 1 includes a roadside detector and a vehicle-mounted detector, which can collect vehicle operation status information in the current upstream control area and the tunnel bottleneck area as input to the decision generation module 2.

[0071] Based on the above system, a tunnel bottleneck control method is implemented to prioritize high-passenger-load vehicles. Figure 2 As shown, the following steps are included:

[0072] S1, the data acquisition module collects the operating status information of CAV and CHV in the upstream control area and the bottleneck area of ​​the tunnel, and transmits it to the decision generation module;

[0073] S2, the decision generation module first predicts the time when the vehicle reaches the bottleneck of the tunnel and predicts the trajectory of the CHV, and then determines whether the CHV is in the regular lane adjacent to the DHL and whether the CHV can obtain a speed gain on the DHL (compared to the current lane of the CHV), so as to determine whether the CHV needs to apply for the CAV to change lanes to the DHL;

[0074] S3. If there is an application for lane change to the DHL, it is determined whether there is a CAV ahead that can change lanes and whether the HOV lane can provide the corresponding right of way. If both are determined to be yes, the lane change trajectory optimization decision is made for the CAV. Otherwise, the CAV does not change lanes and the vehicle cruising trajectory optimization decision is made for the CAV.

[0075] S4. According to the lane change trajectory optimization decision or the vehicle cruising trajectory optimization decision, the control execution module converts the optimization decision information into corresponding CAV trajectory control instructions, and sends them to the CAV in the road network for execution, thereby controlling the driving behavior of the CAV.

[0076] The application scenarios of this embodiment are as follows Figure 3 As shown, the application process is as follows Figure 4 As shown, the main ones are:

[0077] The data acquisition module is used to collect the operating status information (including physical information such as vehicle position, speed and acceleration, trajectory information and passenger load factor information) of the connected autonomous driving vehicle CAV and the connected human-driven vehicle CHV upstream of the bottleneck of the expressway tunnel as input to the decision generation module;

[0078] The decision generation module takes the traffic capacity at the bottleneck of the tunnel as the target, optimizes the decision of the trajectory of the connected autonomous driving vehicle, and then controls the overall traffic flow to pass through the bottleneck of the tunnel efficiently and smoothly. The working process of the decision generation module is as follows:

[0079] S21, first predicting the time when the vehicle will arrive at the bottleneck of the tunnel and predicting the trajectory of the CHV, then judging whether the CHV is in the adjacent DHL lane and whether the DHL lane has a speed gain compared to the current lane of the CHV, so as to determine whether the CHV needs to apply for the CAV to change lanes to the DHL, if both of the above two conditions are met and it is judged to be yes, then sending a lane change application to the front CAV to the DHL;

[0080] S22. If there is a lane change application, continue to determine whether there is a CAV that meets the conditions and can change lanes ahead, and whether the HOV lane can provide the corresponding right of way. If the above two conditions are met, the CAV makes a trajectory optimization decision for lane change; otherwise, the CAV does not need to change lanes and makes a trajectory optimization decision under lane cruising;

[0081] Finally, the control execution module converts the decision plan generated in the decision generation module into vehicle control instructions and sends them to the CAV in the road network for execution, so that the CAV executes the corresponding trajectory control commands.

[0082] Among them, the decision generation module predicts the time for all vehicles on the upstream section of the expressway tunnel bottleneck to arrive at the tunnel bottleneck, and predicts the trajectory of the CHV arriving at the expressway tunnel bottleneck based on the predicted arrival time results, specifically:

[0083]

[0084]

[0085] Where n is the vehicle number index in the traffic flow, N is the vehicle number index set, k is the lane number index, and k is the lane index set (k = 1, 2, 3). is the earliest time that vehicle n in lane k can reach the bottleneck of the tunnel without considering the preceding vehicle, t 0 is the time corresponding to the current time step. L, v op and a maxThey represent the length of the upstream control area of ​​the expressway tunnel bottleneck, the optimal speed of the vehicle expected to pass through the tunnel bottleneck, and the maximum acceleration of the vehicle. is the initial speed of vehicle n in lane k, is the predicted time for vehicle n-1 in lane k to reach the bottleneck of the tunnel, t h is the expected safe headway between vehicles.

[0086]

[0087]

[0088]

[0089]

[0090] Where m is the index of the number of connected human-driven vehicles in the traffic flow, N m is the index set of all networked human-driven vehicles, t is the time step index, and T is the time step index set. 0 The safe distance for vehicles to travel. and They represent the position and speed of the online human-driven vehicle m in lane k at time step t respectively. and They respectively represent the acceleration of the connected human-driven vehicle m in lane k at time step t when there is a car ahead and the acceleration of the connected human-driven vehicle m in lane k at time step t when there is no car ahead. is the minimum expected distance between the connected human-driven vehicle m and the preceding vehicle in lane k, The actual distance between the connected human-driven vehicle m and the vehicle ahead in lane k. is the minimum expected interval of the connected human-driven vehicle m when there is no vehicle ahead, Represents the speed difference between the connected human-driven vehicle m and the vehicle in front. a d is the expected deceleration of the connected human-driven vehicle m.

[0091] The decision generation module uses CHV trajectory information as input conditions to optimize and make decisions on the CAV trajectory. The optimization decision objectives include travel time, fuel consumption, and the best traffic state at the bottleneck of the tunnel. Specifically:

[0092]

[0093]

[0094] Where c is the index of the number of the connected autonomous driving vehicle in the traffic flow, N cis the set of indexes of all connected human-driven vehicles. α, β and μ are weight coefficients, τ c is the travel time of the connected autonomous vehicle c from the starting position in the upstream control area to the bottleneck of the tunnel, is the acceleration of the connected autonomous driving vehicle c at time step t, and They are the position and speed of the connected autonomous driving vehicle c in lane k when it is predicted to reach the bottleneck of the tunnel.

[0095] The corresponding kinematic constraints need to be met during the CAV decision-making and control process, specifically:

[0096]

[0097]

[0098]

[0099]

[0100] in, and They represent the position, velocity and acceleration of the connected autonomous driving vehicle c in lane k at time step t. min , v max , a min and a max are the minimum speed, maximum speed, minimum acceleration and maximum acceleration of the vehicle respectively, and Δt is the unit time step interval (1s).

[0101] CAVs need to meet safety requirements during movement, so they need to comply with collision avoidance constraints, specifically:

[0102]

[0103]

[0104]

[0105]

[0106] Where h is the high occupancy vehicle (HOV) number index, N h It is a collection of high-occupancy vehicle number indexes. is the position of the preceding vehicle of the connected autonomous driving vehicle c in lane k at time step t, is the position of the vehicle behind the connected autonomous driving vehicle c in lane k at time step t, is the position of high-occupancy vehicle h on lane k at time step t, is the position of the vehicle ahead of the high-occupancy vehicle h on lane k at time step t, is the position of the vehicle behind the high-occupancy vehicle h on lane k at time step t. c and l h They are the length of connected autonomous driving vehicles and high-occupancy vehicles respectively.

[0107] The CAV needs to meet the corresponding lane usage criteria during the movement. In this embodiment, the dynamic high passenger load lane is numbered k=3, and the other lanes are regular lanes (k=1, 2), specifically:

[0108]

[0109]

[0110] in, is a binary auxiliary variable. If vehicle c is in lane k (k = 1, 2 or 3) at time step t, then on the contrary, and are binary auxiliary variables that represent whether the connected autonomous driving vehicle c is in the dynamic high-occupancy lane at time step t and time step t+1. If it is, the auxiliary variable and is 1, otherwise it is 0.

[0111] In addition, the boundary constraints of the design model are used to narrow the solution scope and improve the efficiency of solving the vehicle trajectory control scheme, specifically:

[0112]

[0113] Here, M is a constant that is always large enough.

[0114] After a CAV that meets the lane-changing conditions to a dynamic high-occupancy lane obtains a lane-changing application, the CAV's right of way in the dynamic high-occupancy lane will be dynamically optimized, specifically:

[0115]

[0116]

[0117] in, and They respectively represent the set of right-of-way segments allocated to the connected autonomous driving vehicle c and the high-occupancy vehicle h on the dynamic high-occupancy lane at time step t.

[0118] In summary, this solution is different from the traditional approach that ignores the traffic benefits of passenger-oriented travelers. It uses connected autonomous driving technology to provide priority right of way for vehicles with high passenger loads. It also uses connected autonomous driving technology to achieve micro-level decision-making and control of vehicles. By controlling connected autonomous driving vehicles and then controlling the overall traffic flow operation mode, it provides a solution to the traffic congestion problem that is prone to occur in bottlenecks on expressway tunnels.

[0119] On the one hand, with the travel efficiency of travelers as the goal, while ensuring the rationality and fairness of the allocation of road rights resources, the traffic operation status at the bottleneck of the expressway tunnel is improved, and the stability of the overall transportation system is improved; on the other hand, based on the real-time traffic capacity at the bottleneck of the expressway tunnel, priority road rights are allocated to upstream high-occupancy vehicles, and high-occupancy priority lanes are dynamically opened and set up to improve the efficiency of the use of road rights resources on the priority lanes. Using networked autonomous driving vehicles as a starting point, active control of the bottleneck traffic in the expressway tunnel is achieved.

Claims

1. A tunnel bottleneck control method, applied to a tunnel bottleneck control system with priority for high passenger load vehicles, characterized in that: The tunnel bottleneck control system comprises a data acquisition module (1), a decision generation module (2) and a control execution module (3) which are connected in sequence, wherein the data acquisition module (1) is used to collect the operation status information of the CAV and CHV upstream of the bottleneck of the expressway tunnel, and output it to the decision generation module (2); The decision generation module (2) is used to make an optimization decision on the CAV trajectory and output corresponding decision information to the control execution module (3); The control execution module (3) is used to convert the decision information into vehicle control instructions and send them to the CAV in the road network for execution; The tunnel bottleneck control method includes the following steps: S1, the data acquisition module collects the operating status information of CAV and CHV in the upstream control area and the bottleneck area of ​​the tunnel, and transmits it to the decision generation module; S2, the decision generation module first predicts the time when the vehicle reaches the bottleneck of the tunnel and predicts the trajectory of the CHV, and then determines whether the CHV is in the regular lane adjacent to the DHL and whether the CHV can obtain a speed gain on the DHL, so as to determine whether the CHV needs to apply for the CAV to change lanes to the DHL; S3. If there is an application for lane change to the DHL, it is determined whether there is a CAV ahead that can change lanes and whether the HOV lane can provide the corresponding right of way. If both are determined to be yes, the lane change trajectory optimization decision is made for the CAV. Otherwise, the CAV does not change lanes and the vehicle cruising trajectory optimization decision is made for the CAV. S4. According to the lane change trajectory optimization decision or the vehicle cruising trajectory optimization decision, the control execution module converts the optimization decision information into a corresponding CAV trajectory control instruction, and sends it to the CAV in the road network for execution, thereby controlling the driving behavior of the CAV; The step S2 specifically includes the following steps: S21, predicting the time when the vehicle arrives at the bottleneck of the tunnel, and performing CHV trajectory prediction based on the result of the predicted arrival time; S22, determining whether the target CHV is in a regular lane adjacent to the DHL, if so, executing step S23, otherwise, not sending a lane change request to the DHL to the front CAV; S23, determining whether the DHL lane has a speed gain compared to the current lane of the target CHV, if so, sending a lane change request to the DHL to the front CAV, otherwise, not sending a lane change request to the DHL to the front CAV; In step S21, the time for all vehicles on the upstream section of the expressway tunnel bottleneck to arrive at the tunnel bottleneck is predicted according to the following formula: Where n is the vehicle number index in the traffic flow, N is the vehicle number index set, k is the lane number index, k is the lane index set, k = 1, 2, 3, is the earliest time that vehicle n in lane k can reach the bottleneck of the tunnel without considering the preceding vehicle, t 0 is the time corresponding to the current time step, L, v op and a max They represent the length of the upstream control area of ​​the expressway tunnel bottleneck, the optimal speed of the vehicle expected to pass through the tunnel bottleneck, and the maximum acceleration of the vehicle. is the initial speed of vehicle n in lane k, is the predicted time for vehicle n-1 in lane k to reach the bottleneck of the tunnel, t h is the expected vehicle safety headway; The trajectory of the CHV reaching the bottleneck of the fast-speed tunnel is predicted according to the following formula: Where m is the index of the number of connected human-driven vehicles in the traffic flow, N m is the index set of all networked human-driven vehicles, t is the time step index, T is the time step index set, s0 is the safe distance for vehicles to travel, and They represent the position and speed of the online human-driven vehicle m in lane k at time step t, and They represent the acceleration of the online human-driven vehicle m in lane k at time step t when there is a car ahead and the acceleration of the online human-driven vehicle m in lane k at time step t when there is no car ahead. is the minimum expected distance between the connected human-driven vehicle m and the preceding vehicle in lane k, The actual distance between the connected human-driven vehicle m and the vehicle ahead in lane k, is the minimum expected interval of the connected human-driven vehicle m when there is no vehicle ahead, Represents the speed difference between the connected human-driven vehicle m and the vehicle in front, is the expected deceleration of the connected human-driven vehicle m.

2. A tunnel bottleneck control method according to claim 1, characterized in that: The lane change trajectory optimization decision for the CAV and the vehicle cruising trajectory optimization decision for the CAV in step S3 are both performed based on set optimization decision objectives and constraints. The optimization decision objectives include travel time, fuel consumption, and optimal traffic state at the bottleneck of the tunnel. The constraints include kinematic constraints, vehicle collision avoidance constraints, lane management constraints, boundary solution constraints, and dynamic lane right of way allocation constraints.

3. A tunnel bottleneck control method according to claim 2, characterized in that: The optimization decision objectives are specifically: Where c is the index of the number of the connected autonomous driving vehicle in the traffic flow, N c is the set of indexes of all connected human-driven vehicles, α, β and μ are weight coefficients, τ c is the travel time of the connected autonomous vehicle c from the starting position in the upstream control area to the bottleneck of the tunnel, is the acceleration of the connected autonomous driving vehicle c at time step t, and They are the position and speed of the connected autonomous driving vehicle c in lane k when it is predicted to reach the bottleneck of the tunnel.

4. A tunnel bottleneck control method according to claim 3, characterized in that: The kinematic constraints are specifically: in, and They represent the position, velocity and acceleration of the connected autonomous driving vehicle c in lane k at time step t, respectively. min 、v max 、a min and a ,ax are the minimum speed, maximum speed, minimum acceleration and maximum acceleration of the vehicle respectively, and Δt is the unit time step interval.

5. A tunnel bottleneck control method according to claim 4, characterized in that: The vehicle collision avoidance constraints are specifically: Where h is the HOV number index, N h is a set of high-occupancy vehicle number indexes, is the position of the preceding vehicle of the connected autonomous driving vehicle c in lane K at time step t, is the position of the vehicle behind the connected autonomous driving vehicle c in lane k at time step t, is the position of high-occupancy vehicle h on lane k at time step t, is the position of the vehicle ahead of the high-occupancy vehicle h on lane k at time step t, is the position of the vehicle behind the high-occupancy vehicle h on lane k at time step t, l c and l h They are the length of connected autonomous driving vehicles and high-occupancy vehicles respectively.

6. A tunnel bottleneck control method according to claim 5, characterized in that: The lane management constraint is as follows: set the dynamic high passenger load lane number to k=3, and the other lanes to be regular lanes, i.e. k=1, 2, then in, is a binary auxiliary variable. If vehicle c is in lane k at time step t, i.e., k = 1, 2, or 3, then on the contrary, and Binary auxiliary variables representing whether the connected autonomous driving vehicle c is located in the dynamic high-occupancy lane at time step t and time step t+1. If it is located, the auxiliary variable and is 1, otherwise it is 0.

7. A tunnel bottleneck control method according to claim 6, characterized in that: The solution boundary constraint is used to narrow the solution scope and improve the solution efficiency of the vehicle trajectory control solution, specifically: Where M is a sufficiently large constant; The dynamic lane right of way allocation constraints are specifically: in, and They respectively represent the set of right-of-way segments allocated to the connected autonomous driving vehicle c and the high-occupancy vehicle h on the dynamic high-occupancy lane at time step t.

Images