Intersection and intelligent connected vehicle control method, system and storage medium

Abstract

The present invention discloses a method, system, and storage medium for controlling intersections and intelligent connected vehicles, including: obtaining the length of a virtual waiting area at each intersection; constructing a game model and an average vehicle delay model, determining intersection control information by finding a Nash equilibrium and minimizing the average delay time using a particle swarm optimization algorithm; determining a vehicle control plan; controlling the leading vehicle in the queue to travel according to the control plan; if the intersection has a green light in a direction associated with the direction of travel, controlling the appearance of a virtual waiting area consistent with the direction of travel, and obtaining the speed at the intersection; determining the duration of the corresponding virtual waiting area based on the speed and the length of the virtual waiting area; when the green light in the same direction of travel is on, controlling the leading vehicle in the queue to pass through the intersection; and controlling the virtual waiting area to disappear when the duration exceeds the duration. The present invention fully exploits the spatiotemporal resources of the intersection, improves the intersection's traffic capacity, and reduces average vehicle delay.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and more particularly to a method, system, and storage medium for controlling intersections and intelligent connected vehicles. Background Art

[0002] The rapid development of vehicle-to-everything (V2X) and connected and automated vehicle (CAV) technologies is providing more real-time vehicle information and better methods for optimizing vehicle trajectories at signalized intersections. Furthermore, in the future, heterogeneous traffic flows, where human-driven vehicles (HVs) and CAVs coexist, will become commonplace.

[0003] Prior art approaches to address the coexistence of HVs and CAVs have employed heterogeneous traffic flow collaborative control methods based on dedicated CAV lanes to improve intersection capacity while mitigating the uncertainty of HV driving behavior. However, existing technologies have not fully leveraged the advantages of CAV connectivity and exploited the spatiotemporal resources of intersections. Therefore, a heterogeneous traffic flow signal control method and vehicle control method that can fully utilize the spatiotemporal resources within intersections is urgently needed. Summary of the Invention

[0004] The present invention is proposed based on the above-mentioned requirements of the prior art. The technical problem to be solved by the present invention is to provide an intersection and intelligent connected vehicle control method, system and storage medium to fully utilize the time and space resources within the intersection, improve communication capabilities and reduce vehicle delays.

[0005] In order to solve the above problems, the present invention is implemented by adopting the following technical solutions:

[0006] Provided is a method for controlling an intersection and an intelligent connected vehicle, the method comprising: obtaining the length of a virtual waiting area at each intersection, the virtual waiting area comprising a straight virtual waiting area and a left-turn virtual waiting area, wherein the straight direction of the intersection is associated with the left-turn direction; constructing a game model and a vehicle average delay model, determining intersection control information by finding a Nash equilibrium and using a particle swarm optimization algorithm to find the minimum average delay time, the control information comprising a phase sequence and a phase duration of each cycle; based on the intersection control information, with the goal of the head vehicle of the team arriving at the intersection when the green light starts to coincide with the driving direction, minimizing the control error by solving a quadratic programming problem to obtain a vehicle control time. scheme; controlling the driving of the leading vehicle in the team according to the control scheme; the control error is the error between the expected state quantity and the actual state quantity; the state quantity includes speed, displacement and acceleration; if the intersection signal light shows a green light in a direction associated with the driving direction when the leading vehicle in the team arrives at the intersection, controlling the appearance of a virtual waiting area consistent with the driving direction of the leading vehicle in the team, and obtaining the speed of the leading vehicle in the team arriving at the intersection; determining the duration of the corresponding virtual waiting area according to the speed at the intersection and the length of the virtual waiting area; when the green light corresponding to the driving direction of the leading vehicle in the team is on, controlling the leading vehicle in the team to cross the intersection; when the duration exceeds the duration, controlling the corresponding virtual waiting area to disappear.

[0007] Optionally, obtaining the length of the virtual waiting area at each intersection includes: determining the length of the straight-ahead virtual waiting area by the distance from the stop line at the intersection entrance to the left-turn lane at the left-turn intersection exit; determining the length of the left-turn virtual waiting area by the distance from the stop line at the intersection entrance to the straight-ahead lane at the left-turn intersection exit.

[0008] Optionally, the game theory-based construction of an adjacent intersection model with the goal of minimizing average vehicle delay and determining intersection control information through Nash equilibrium includes: Step 1: constructing a game model M = {T, I, A, U}, where T represents a set of intersections, I represents information obtained for each intersection, A represents an action performed at each intersection, and U represents a utility function; Step 2: constructing a vehicle average delay time algorithm, wherein the average delay time algorithm is expressed as: Among them, T delay represents the delay time of all intersections; n represents one of all intersections; k represents an entrance road of intersection n; r n represents all entrances of intersection n; q k represents the road traffic flow at intersection entrance k, W represents the independent variable, λ represents the green signal ratio, d k (W,λ) represents the delay function, d w The expression is: Among them, d k Indicates vehicle delay time; represents the signal period; λ k represents the green-signal ratio of the entrance road k; w represents the traffic volume; η k Indicates saturation; S k Represents saturated traffic volume; Step 3: Use the particle swarm optimization algorithm to find the green-to-signal ratio strategy that minimizes the average delay time; Step 4: Under the current green-to-signal ratio strategy, use the game model analysis to obtain the first result, the first result includes the optimal strategy selection and the strategy to achieve Nash equilibrium; Step 5: According to the first result, determine the next search direction of the particle swarm optimization algorithm; Step 6: Update the green-to-signal ratio strategy according to the search direction; Repeat steps 4 to 6 more than once to make the green-to-signal ratio strategy reach a stable state, and obtain the corresponding optimal strategy selection as the intersection control information.

[0009] Optionally, the particle swarm optimization algorithm expression is: Among them, ω is the inertia weight, z is the current iteration number; is the particle velocity at iteration number z; is the particle velocity at iteration number z+1; is the particle position at iteration number z; is the particle position at iteration number z+1; is the individual extreme value; is the extreme value of the group; c1, r1, c2, r2 are hyperparameters, ω max ,ω min are the maximum and minimum values of inertia weight, N max is the maximum number of iterations.

[0010] Optionally, the expression for minimizing the control error is: minL diff , L diff =(X k -D k ) T Q(X k -D k ), X k =Ψx(k)+ΘU k , where L diff represents the control error; X k represents the actual state quantity in the future p control cycles at time k; Ψ represents a 3p*3 column vector, and its expression is Θ represents the 3p*p lower triangular matrix, which is expressed as in, T′=0.1s;x(k) represents the actual state quantity at time k, U k represents the expected state quantity in the future p control cycles at time k, Dk represents the expected value sequence of the state variables within p control cycles; Q represents the Jordan matrix, which is expressed as Among them p1, p2, p3 represent weights.

[0011] Optionally, determining the duration of the corresponding virtual waiting area based on the speed at the intersection and the length of the virtual waiting area includes: taking the speed at the intersection as the average speed of the head vehicle of the team traveling from the starting point to the end point of the virtual waiting area; and determining the duration of the corresponding virtual waiting area based on the average speed and the length of the virtual waiting area.

[0012] Optionally, it also includes: each intersection of the intersection includes an adjustment area and a dedicated lane area. In the adjustment area, vehicles are adjusted according to the driving direction to form a queue; in the dedicated lane area, team members in the queue drive according to the trajectory of the team's front vehicle.

[0013] Provided is an intersection and intelligent connected vehicle control system, the system comprising: an acquisition module for acquiring the length of a virtual waiting area at each intersection, the virtual waiting area comprising a straight virtual waiting area and a left-turn virtual waiting area, wherein the straight direction of the intersection is associated with the left-turn direction; a first processing module for constructing a game model and a vehicle average delay model, determining intersection control information by finding a Nash equilibrium and using a particle swarm optimization algorithm to find the minimum average delay time, the control information comprising a phase sequence and a phase duration of each cycle; a second processing module for minimizing the control error by solving a quadratic programming problem based on the intersection control information, with the goal of ensuring that the head vehicle of the team arrives at the intersection when the green light consistent with the driving direction starts. The vehicle control scheme is obtained by the difference; the control module controls the driving of the head vehicle of the team according to the control scheme; the control error is the error between the expected state quantity and the actual state quantity; the state quantity includes speed, displacement and acceleration; if the intersection signal light shows a green light in the direction associated with the driving direction when the head vehicle of the team arrives at the intersection, a virtual waiting area consistent with the driving direction of the head vehicle of the team is controlled to appear, and the speed of the head vehicle of the team arriving at the intersection is obtained; according to the speed at the intersection and the length of the virtual waiting area, the duration of the corresponding virtual waiting area is determined; when the green light corresponding to the driving direction of the head vehicle of the team is on, the head vehicle of the team is controlled to cross the intersection; when the duration exceeds the duration, the corresponding virtual waiting area is controlled to disappear.

[0014] A computer-readable storage medium stores a computer program thereon. The computer-readable storage medium stores an intersection and intelligent connected vehicle control program. When the intersection and intelligent connected vehicle control program is executed by a processor, it implements the intersection and intelligent connected vehicle control method.

[0015] Compared with the existing technology, the present invention fully exploits the time and space resources of the intersection by setting up a virtual waiting area, and can reduce the average vehicle delay on the basis of effectively improving the traffic capacity of the intersection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 This is a flow chart of an intersection and intelligent connected vehicle control method provided by a specific embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of intersection road division according to an intersection and an intelligent connected vehicle control method provided by a specific embodiment of the present invention;

[0019] Figure 3A 1 is a schematic diagram of a first intersection of an intersection and an intelligent connected vehicle control method provided by a specific embodiment of the present invention;

[0020] Figure 3B is a phase corresponding to a first intersection schematic diagram of an intersection and an intelligent connected vehicle control method provided in a specific embodiment of the present invention;

[0021] Figure 3C 1 is a schematic diagram of a second intersection of an intersection and an intelligent connected vehicle control method provided by a specific embodiment of the present invention;

[0022] Figure 3D is a phase corresponding to a second intersection schematic diagram of an intersection and an intelligent connected vehicle control method provided in a specific embodiment of the present invention;

[0023] Figure 4 It is a schematic diagram of adjacent intersections of an intersection and an intelligent connected vehicle control method provided in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0025] To facilitate understanding of the embodiments of the present invention, the following will be further explained with reference to specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the scope of protection of the present invention.

[0026] Example 1

[0027] This embodiment provides a method for controlling an intersection and an intelligent connected vehicle, such as Figure 1 Shown, including:

[0028] S1: Obtain the length of the virtual waiting area of each intersection.

[0029] The virtual waiting area includes a straight virtual waiting area and a left-turn virtual waiting area. In this embodiment, the straight direction of the intersection is associated with the left-turn direction.

[0030] The principle of setting the length of the virtual waiting area is to maximize the use of intersection space value and CAV network connection value without affecting intersection traffic.

[0031] This embodiment targets the scenario where HV and CAV vehicles coexist, and proposes to plan a variable virtual waiting area inside the intersection in front of the CAV dedicated lane. The length of the virtual waiting area is variable and affected by the actual geographical environment of the intersection. The setting principle is to maximize the spatial value of the intersection and the value of CAV network connection without affecting the traffic flow of the intersection. Figure 2 As shown, the layout of an isolated intersection with four arms is shown. Each arm has three different directions of travel for vehicles, namely left turn, straight and right turn, and each arm has a dedicated CAV lane for straight and left-turning CAVs. Among them, right-turning CAVs and HVs share the right-turn lane and are not controlled by signals; CAVs and HVs share the signal phase. Each arm is divided into an adjustment area, a dedicated lane area and a virtual waiting area. HVs do not change lanes in the adjustment area, and a sensor device is provided in the adjustment area. The sensor device can collect or predict the arrival information of CAVs at the intersection in the next signal cycle. CAVs can change lanes and complete formation in the adjustment area. After the CAVs form a queue, they wait to enter the dedicated lane area and the virtual waiting area. Then, the CAV queue crosses the intersection without stopping according to the planned trajectory in the dedicated lane composed of the dedicated lane area and the virtual waiting area.

[0032] The virtual waiting area is divided into a straight virtual waiting area and a left-turn virtual waiting area. The direction of the virtual waiting area is determined by the control system. The straight direction of the intersection is associated with the left-turn direction. The maximum length of the straight virtual waiting area is the distance from the stop line at the intersection entrance to the left-turn lane at the left-turn intersection exit; the maximum length of the left-turn virtual waiting area is the distance from the stop line at the intersection entrance to the straight lane at the left-turn intersection exit. Figure 2 shown.

[0033] When the left-turn signal at the intersection turns green, the straight-moving vehicles in the CAV dedicated lane enter the intersection at the same time or after a period of delay as the left-turning vehicles, and run to the straight-moving virtual waiting area lane in the intersection. When the straight-moving green light turns on, the CAV passes through the straight-moving virtual waiting area without stopping and enters the intersection, and finally passes through the signalized intersection without stopping. The schematic diagram is shown as follows Figure 3A As shown, the corresponding phase is Figure 3B shown.

[0034] When the green light for going straight at the intersection is on, in the late green light period, the left-turning vehicle in the CAV dedicated lane enters the intersection along with the straight-moving vehicle in front, and runs to the left-turn waiting area lane in the intersection. When the left-turn green light is on, it passes through the left-turn virtual waiting area without stopping and enters the intersection, and finally passes through the signalized intersection without stopping. The schematic diagram is shown as follows. Figure 3C As shown, the corresponding phase is Figure 3D shown.

[0035] Throughout this process, there's no need to increase the effective green light duration during the through phase or the number of through entrance lanes. However, the virtual waiting area effectively lengthens the entrance lanes. This variable virtual waiting area fully utilizes the intersection area, reducing signal control time and shortening the entire intersection cycle.

[0036] S2: Construct a game model and an average vehicle delay model, and determine the intersection control information by finding the Nash equilibrium and using the particle swarm optimization algorithm to minimize the average delay time.

[0037] The control information includes a phase sequence and a phase duration of each cycle.

[0038] S200: Construct a game model M={T, I, A, U}, where T represents a set of intersections, I represents information obtained for each intersection, A represents an action performed at each intersection, and U represents a utility function.

[0039] The timing scheme of this embodiment is based on a fixed phase sequence at the intersection, ie, going straight first and then turning left, and uses a standard National Marine Electronics Association (NEMA) ring barrier structure to optimize the signal light cycle length and phase duration.

[0040] Due to the limited overall carrying capacity of the road network at any given time and space, optimizing only a single intersection often results in a local optimum with poor global performance. Therefore, this embodiment utilizes a multi-agent coordinated control algorithm based on game theory to construct a game tree between adjacent intersections, searching for Nash equilibria and achieving phase coordination between them.

[0041] like Figure 4 As shown in Figure 1, consider the game between the current intersection and the four adjacent intersections. The road network traffic resources in the game are represented by the current queue length of vehicles. The game goal is to minimize the queue length as much as possible, that is, to avoid vehicles being stranded in the road network. The model is constructed as follows:

[0042] M={T,I,A,U}

[0043] M is the entire game control model, T is the set of intersections, there are a total of 5 intersections, namely T = {T1, T2, T2, T4, T5], and I represents the information obtained at each intersection. This embodiment is based on full networking, that is, the information between all intersections is fully interconnected. A is the action that can be performed at each intersection, that is, which phase it is currently in. As mentioned above, we adopt NEMA's dual-loop eight-phase timing scheme, which corresponds to 8 phase actions, A = {a1, a2, a3, a4, a5, a6, a7, a8}, and U is the utility function, which is an evaluation indicator in the game. The highest utility is when there are no stranded vehicles in the road network, and U = 0 at this time. If there are vehicles, the current queue length is subtracted from 0.

[0044] Remember L i (t) is the queue length at the i-th intersection at time t, including the queue lengths in the four directions of east, south, west and north. Its corresponding utility function is U i .

[0045] L i (t) = {L ie (t),L is (t),L iw (t),L in (t)}

[0046] The process of the game is to seek Nash equilibrium, that is

[0047]

[0048] in, Indicates the optimal a i .

[0049] By solving the above model, based on the NEMA timing scheme, a better global and more optimal phase sequence can be obtained, thereby calculating the optimal green-to-signal ratio.

[0050] S210: Constructing a vehicle average delay time algorithm.

[0051] The average delay time algorithm is expressed as: Among them, T delayrepresents the delay time of all intersections; n represents one of all intersections; k represents an entrance road of intersection n; r n represents all entrances of intersection n; q k represents the road traffic flow at intersection entrance k, W represents the independent variable, λ represents the green signal ratio, d k (W,λ) represents the delay function, d w The expression is: Among them, d w Indicates vehicle delay time; represents the signal period; λ k represents the green-signal ratio of the entrance road k; w represents the traffic volume; η k Indicates saturation; S k Indicates saturated traffic volume.

[0052] S220: Find a green-to-signal ratio strategy that minimizes the average delay time through a particle swarm optimization algorithm.

[0053] By optimizing the green-to-signal ratio to minimize the average vehicle delay, this embodiment uses an improved particle swarm optimization algorithm for solution, specifically including: step 1: initializing the parameter values; step 2: calculating the particle fitness; step 3: updating the particle extreme value; step 4: updating the inertia weight; step 5: finally determining whether convergence has occurred. If converged, the algorithm ends; otherwise, steps 2 to 5 are repeated.

[0054] The corresponding formula is as follows:

[0055]

[0056]

[0057]

[0058] Among them, ω is the inertia weight, z is the current iteration number; is the particle velocity at iteration number z; is the particle velocity at iteration number z+1; is the particle position at iteration number z; is the particle position at iteration number z+1; is the individual extreme value; is the extreme value of the group; c1, r1, c2, r2 are hyperparameters, ω max ,ω min are the maximum and minimum values of inertia weight, N max is the maximum number of iterations.

[0059] S230: Under the current green-credit ratio strategy, a first result is obtained by using a game model analysis, where the first result includes an optimal strategy selection and a strategy for achieving a Nash equilibrium.

[0060] S240: Determine the next search direction of the particle swarm optimization algorithm according to the first result.

[0061] S250: Update the green-to-signal ratio strategy according to the search direction.

[0062] S260: Repeat S230 to S250 for more than one time to make the green-to-signal ratio strategy reach a stable state, and obtain the corresponding optimal strategy selection as intersection control information.

[0063] S3: Based on the intersection control information, with the goal of the head vehicle of the team arriving at the intersection when the green light consistent with the driving direction starts, a vehicle control plan is obtained by solving the minimization control error.

[0064] In this embodiment, the derivative of vehicle acceleration is used as the control variable, and its expression is: Where a represents acceleration and u represents the derivative of acceleration.

[0065] The acceleration, velocity and displacement of the vehicle are combined as the state quantity, and the expression is X = [avx] T , where v is velocity, x is displacement, and X is state quantity.

[0066] According to the system dynamics equation Derivation of X and Relationship, written as In the form of, A and B are parameters, because So observing the above formula, we can get:

[0067] In actual control, since it is impossible to directly take the differential, the differential is used instead of the differential to derive the relationship between the state quantities at two adjacent moments. The expression is:

[0068] Among them, x(k+1) represents the state quantity at time k+1, x(k) represents the state quantity at time k, T ′ represents the change time, in this embodiment, T=0.1s; u(k) represents the relationship between the derivative of the acceleration at the previous moment and the derivative of the acceleration at the next moment.

[0069] By shifting the above expression, we can get: x(k+1)=(I+T ′ A)x(k)+T ′ Bu(k)

[0070] In order to simplify the calculation, this embodiment defines the parameters in, Right now

[0071] This embodiment considers the actual state quantity X in the next p control cycles k and the expected state quantity U in the next p control cycles k , X k The expression is:

[0072] X k =[x(k+1|k) T x(k+2|k) T …x(k+p|k) T ] T

[0073] U k The expression is: U k =[u=k|k) T u(k+1|k) T …u(k+p-1|k) T ] T

[0074] Specifically,

[0075]

[0076]

[0077]

[0078] By analogy, we get the general formula:

[0079]

[0080] X k Expressed in matrix form, the expression is: X k =Ψx(k)+ΘU k , where Ψ is a 3p*3 column vector. Specifically, Θ is a 3p*p lower triangular matrix, specifically,

[0081] Furthermore, the control error L is derived diff In the application scenario of this embodiment, the control error is the error between the expected state quantity and the actual state quantity. The expected state quantity includes the expected displacement x d , expected speed v d and the expected acceleration a dThe actual state variables include the actual displacement x, the actual velocity v, and the actual acceleration a. Take the 2-norm of their deviations and record this difference as L diff , where p1, p2, and p3 are the weights of the corresponding quantities. The derived control error expression is:

[0082] L diff =p3(xx d ) 2 +p2(vv d ) 2 +p1(aa d ) 3

[0083] Define the expected value sequence D of the state variables in the next p periods k , whose expression is:

[0084] D k =[d(k+1) T d(k+2) T …d(k+p) T ] T

[0085] Among them, d(k+1) T represents the expected state at time k+1, d(k+2) T and d(k+p) T Same thing.

[0086] Rewrite L in terms of state variables and expected value sequences diff We can get: L diff =(X k -D k ) T Q(X k -D k ), where Q is a 3p×3p Jordan matrix with all diagonal elements being q. Specifically

[0087] Through quadratic programming, the objective function minL diff By solving the problem, we can get the corresponding CAV control scheme.

[0088] S4: According to the control scheme, control the leading vehicle of the team to move.

[0089] S5: If the intersection signal light shows a green light in the direction associated with the driving direction when the first vehicle in the queue arrives at the intersection, a virtual waiting area consistent with the driving direction of the first vehicle in the queue is controlled to appear, and the speed at which the first vehicle in the queue arrives at the intersection is obtained; according to the speed at the intersection and the length of the virtual waiting area, the duration of the corresponding virtual waiting area is determined.

[0090] Since the phase sequence is fixed, that is, going straight first and then turning left, the actual speed of vehicles at different intersections v CAV are not the same, and the maximum length S of the waiting area at different intersections is different, so the virtual waiting area holding time t waiting The maximum length S of the waiting area and the average speed of CAV in this area Impact, namely:

[0091]

[0092] Where S is the maximum length of the virtual waiting area at the intersection, which is determined by the actual intersection spatial location and vehicle travel direction; It is the average value of the actual operating speed of the CAV vehicle in the virtual waiting area of the intersection, which is affected by the CAV trajectory planning and its value does not exceed the speed limit of the virtual waiting area.

[0093] In this embodiment, the CAV passes through the virtual waiting area at a constant speed, and the speed at which the CAV reaches the intersection can be used as the average speed of the head vehicle of the team traveling from the starting point to the end point of the virtual waiting area.

[0094] Setting the time of the virtual waiting area can avoid the existence of useless virtual waiting area and save computer storage resources.

[0095] S6: When the green light corresponding to the driving direction of the first vehicle in the team is on, the first vehicle in the team is controlled to cross the intersection; when the duration exceeds the duration, the corresponding virtual waiting area is controlled to disappear.

[0096] Compared with the prior art, this embodiment fully exploits the spatial and temporal resources of the intersection by setting up a virtual waiting area, and can reduce the average vehicle delay on the basis of effectively improving the traffic capacity of the intersection.

[0097] Example 2

[0098] Provided is an intersection and intelligent connected vehicle control system, including:

[0099] The acquisition module acquires the length of the virtual waiting area of each intersection, wherein the virtual waiting area includes a straight virtual waiting area and a left-turn virtual waiting area.

[0100] The first processing module constructs a game model and an average vehicle delay model, and determines intersection control information by finding a Nash equilibrium and using a particle swarm optimization algorithm to minimize the average delay time. The control information includes the phase sequence and phase duration of each cycle, where the straight direction and left turn direction of the intersection are associated.

[0101] The second processing module obtains a vehicle control plan by minimizing the control error based on the intersection control information and taking the arrival of the first vehicle in the team at the intersection when the green light consistent with the driving direction starts as the goal.

[0102] a control module for controlling the leading vehicle in the queue to travel according to the control scheme; if the light at the intersection is green when the leading vehicle in the queue arrives at the intersection and is in a direction associated with the travel direction, controlling the appearance of a virtual waiting area in the same direction as the leading vehicle in the queue, and obtaining the speed at which the leading vehicle in the queue arrives at the intersection; determining the duration of the corresponding virtual waiting area based on the speed at the intersection and the length of the virtual waiting area; controlling the leading vehicle in the queue to cross the intersection when the green light in the same direction as the travel direction of the leading vehicle in the queue is on; and controlling the disappearance of the corresponding virtual waiting area when the duration exceeds the duration.

[0103] Example 3

[0104] A computer-readable storage medium having a computer program stored thereon, wherein the computer-readable storage medium stores an intersection and intelligent connected vehicle control program. When the intersection and intelligent connected vehicle control program is executed by a processor, it implements the intersection and intelligent connected vehicle control method corresponding to the embodiment.

[0105] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for controlling an intersection and an intelligent connected vehicle, characterized in that: include: Obtaining the length of a virtual waiting area at each intersection, wherein the virtual waiting area includes a straight virtual waiting area and a left-turn virtual waiting area, wherein the straight direction of the intersection is associated with the left-turn direction; Constructing a game model and an average vehicle delay model, determining intersection control information by finding a Nash equilibrium and minimizing the average delay time using a particle swarm optimization algorithm, and determining the phase sequence and phase duration of each cycle; Based on the intersection control information, with the goal of ensuring that the leading vehicle in the group arrives at the intersection when the green light starts to appear in the direction of travel, a vehicle control solution is obtained by minimizing the control error through quadratic programming. The expression for minimizing the control error is: in, represents the control error; Represents the actual state quantity in the future p control cycles at time k; express The column vector of , express The lower triangular matrix of ,in, , , ; represents the actual state quantity at time k, represents the expected state quantity in the future p control cycles at time k, represents the expected value sequence of the state variables within p control cycles; represents the Jordan matrix, which is expressed as ,in , , , represents weight; Controlling the leading vehicle of the team to travel according to the control scheme; the control error is the error between the expected state quantity and the actual state quantity; the state quantity includes speed, displacement and acceleration; If the intersection signal light shows green in the direction associated with the travel direction when the first vehicle in the queue arrives at the intersection, a virtual waiting area consistent with the travel direction of the first vehicle in the queue is controlled to appear, and the speed of the first vehicle in the queue arriving at the intersection is obtained; the duration of the corresponding virtual waiting area is determined based on the speed at the intersection and the length of the virtual waiting area; When the green light corresponding to the driving direction of the first vehicle in the team is on, the first vehicle in the team is controlled to cross the intersection; when the duration exceeds the duration, the corresponding virtual waiting area is controlled to disappear.

2. The method for controlling an intersection and an intelligent connected vehicle according to claim 1, characterized in that: The step of obtaining the length of the virtual waiting area at each intersection includes: The length of the straight virtual waiting area is determined by the distance from the stop line at the intersection entrance to the left-turn lane at the left-turn intersection exit; The length of the left-turn virtual waiting area is determined by the distance from the stop line at the intersection entrance to the straight lane at the left-turn intersection exit.

3. The method for controlling an intersection and an intelligent connected vehicle according to claim 1, characterized in that: The game model and the average vehicle delay model are constructed, and intersection control information is determined by finding the Nash equilibrium and minimizing the average delay time using the particle swarm optimization algorithm, including: Step 1: Build a game model , T represents the set of intersections, I represents the information obtained for each intersection, A represents the action performed at each intersection, and U represents the utility function; Step 2: Construct a vehicle average delay time algorithm, which is expressed as: ,in, represents the delay time of all intersections; Represents one of all intersections; k represents the intersection An entrance road; Indicates an intersection All entrances; represents the road traffic flow at intersection entrance k, represents the independent variable, Indicates the green letter ratio, represents the delay function, The expression is: , ;in, Indicates vehicle delay time; Indicates the signal period; represents the green-to-signal ratio of import channel k; Indicates traffic volume; Indicates saturation; Indicates saturated traffic volume; Step 3: Use particle swarm optimization algorithm to find the green-to-signal ratio strategy that minimizes the average delay time; Step 4: Under the current green credit ratio strategy, use the game model to analyze and obtain a first result, which includes the best strategy selection and the strategy that reaches the Nash equilibrium; Step 5: Determine the next search direction of the particle swarm optimization algorithm based on the first result; Step 6: Update the green-to-signal ratio strategy according to the search direction; Repeat steps 4 to 6 more than once to make the green-to-signal ratio strategy reach a stable state and obtain the corresponding optimal strategy selection as the intersection control information.

4. The method for controlling an intersection and an intelligent connected vehicle according to claim 3, wherein: The particle swarm optimization algorithm expression is: in, is the inertia weight, z is the current iteration number; is the particle velocity at iteration number z; is the particle velocity at iteration number z+1; is the particle position at iteration number z; is the particle position at iteration number z+1; is the individual extreme value; is the extreme value of the group; , , , is a hyperparameter, , are the maximum and minimum values of inertia weight, is the maximum number of iterations.

5. The method for controlling an intersection and an intelligent connected vehicle according to claim 1, wherein: Determining the duration of the corresponding virtual waiting area according to the arrival speed at the intersection and the length of the virtual waiting area includes: The speed at which the vehicle reaches the intersection is taken as the average speed of the first vehicle in the queue from the beginning to the end of the virtual waiting area. The duration of the corresponding virtual waiting area is determined according to the average speed and the length of the virtual waiting area.

6. The method for controlling an intersection and an intelligent connected vehicle according to claim 1, characterized in that: Also includes: Each intersection includes an adjustment area and a dedicated lane area. In the adjustment area, vehicles are adjusted according to the direction of travel to form a queue; in the dedicated lane area, the team members in the queue travel according to the trajectory of the team's front vehicle.

7. An intersection and intelligent connected vehicle control system, characterized in that: include: An acquisition module is configured to acquire the length of a virtual waiting area at each intersection, wherein the virtual waiting area includes a straight virtual waiting area and a left-turn virtual waiting area, wherein the straight direction of the intersection is associated with the left-turn direction; The first processing module constructs a game model and a vehicle average delay model, and determines intersection control information by finding a Nash equilibrium and using a particle swarm optimization algorithm to minimize the average delay time. The control information includes a phase sequence and a phase duration of each cycle. The second processing module, based on the intersection control information, takes the goal of ensuring that the leading vehicle in the group arrives at the intersection when the green light starts to turn green in the direction of travel as the goal, and obtains a vehicle control plan by minimizing the control error through quadratic programming. The expression for minimizing the control error is: in, represents the control error; Represents the actual state quantity in the future p control cycles at time k; express The column vector of , express The lower triangular matrix of ,in, , , ; represents the actual state quantity at time k, represents the expected state quantity in the future p control cycles at time k, represents the expected value sequence of the state variables within p control cycles; represents the Jordan matrix, which is expressed as ,in , , , represents weight; A control module controls the driving of the leading vehicle in the team according to the control scheme; the control error is the error between the expected state quantity and the actual state quantity; the state quantity includes speed, displacement and acceleration; if the intersection signal light shows a green light in a direction associated with the driving direction when the leading vehicle in the team arrives at the intersection, a virtual waiting area consistent with the driving direction of the leading vehicle in the team is controlled to appear, and the speed of the leading vehicle in the team arriving at the intersection is obtained; according to the speed at the intersection and the length of the virtual waiting area, the duration of the corresponding virtual waiting area is determined; when the green light corresponding to the driving direction of the leading vehicle in the team is on, the leading vehicle in the team is controlled to cross the intersection; when the duration exceeds the duration, the corresponding virtual waiting area is controlled to disappear.

8. A computer-readable storage medium having a computer program stored thereon, wherein the computer-readable storage medium stores an intersection and intelligent connected vehicle control program, and when the intersection and intelligent connected vehicle control program is executed by a processor, it implements the intersection and intelligent connected vehicle control method described in any one of claims 1-6.

Images