Collaborative optimization method for intersection signal lamp and intelligent network connection vehicle track

Abstract

The invention relates to the technical field of intelligent network connection vehicles, in particular to an intersection signal lamp and intelligent network connection vehicle trajectory collaborative optimization method, which comprises the following steps: acquiring state information of all vehicles within a certain range of an intersection; obtaining mathematical expressions of signal lamp parameters of all lanes based on the entrance lanes of the intersection and a pre-established adaptive traffic signal model; generalized delays and vehicle tracks corresponding to different signal lamp parameters are obtained based on a pre-constructed vehicle delay estimation model; and solving an optimal signal lamp parameter and a corresponding vehicle track based on a heuristic search algorithm. Therefore, the problems that in the related technology, the real-time state of the traffic flow is not considered, the mixed traffic condition of mixed driving of the intelligent network connection vehicle and the human driving vehicle and the interaction relation between the signal lamp and the vehicle are not considered, so that the signal lamp parameter and the track of the intelligent network connection vehicle cannot be collaboratively optimized, the intersection traffic efficiency is low, and the traffic efficiency is poor are solved. And actual requirements cannot be met.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent connected vehicles, and in particular to a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories. Background Art

[0002] Urban intersections are the main bottleneck in the current traffic system. Due to the conflict of vehicle trajectories, there are safety risks between different vehicles. The intersection area is the intersection of multiple roads with different driving directions. The relatively limited road resources greatly affect the traffic efficiency near the intersection. In addition, the dynamic characteristics of traffic flow also bring challenges to traffic control at urban intersections.

[0003] In the related technology, the vehicle arrival information set corresponding to each intelligent connected vehicle can be determined based on the vehicle motion status information and traffic signal status information of multiple intelligent connected vehicles, and then the vehicle control strategy can be determined to control the driving of each intelligent connected vehicle; it is also possible to obtain the information of intelligent connected vehicles within the communication range of the intersection, establish an arrival time optimization model for the vehicle to arrive at the intersection, solve the optimal arrival time for the vehicle to arrive at the intersection, establish a traffic light optimization model and a vehicle speed trajectory optimization model to solve the optimal vehicle speed trajectory, and then realize the coordinated control of traffic lights and vehicles.

[0004] However, the relevant technologies lack consideration of the real-time status of traffic flow, nor do they take into account the mixed traffic conditions of intelligent connected vehicles and human-driven vehicles in future traffic systems. They also do not fully consider the interaction between traffic lights and vehicles, and fail to simultaneously optimize traffic light parameters and the trajectory of intelligent connected vehicles. The traffic efficiency at intersections is low and cannot meet the requirements of actual edge cloud deployment, which urgently needs to be improved. Summary of the invention

[0005] The present application provides a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories to address the problems in related technologies, such as the lack of consideration of the real-time status of traffic flow, the failure to consider the mixed traffic conditions of intelligent connected vehicles and human-driven vehicles in future traffic systems, the failure to fully consider the interaction between traffic lights and vehicles, the failure to simultaneously optimize traffic light parameters and the trajectories of intelligent connected vehicles, the low efficiency of intersection traffic, and the inability to meet the requirements for actual deployment of edge clouds.

[0006] The first aspect of the present application provides a method for collaborative optimization of intersection traffic lights and intelligent connected vehicle trajectories, comprising the following steps: collecting first state information of all intelligent connected vehicles and second state information of human-driven vehicles within a preset range of the intersection; determining an entrance lane of the intersection, and obtaining a mathematical expression of all lane traffic light parameters in the entrance lane based on the entrance lane and a pre-established adaptive traffic signal model; respectively inputting the mathematical expression of the traffic light parameters, the first state information, and the second state information into a pre-constructed vehicle delay estimation model to obtain the generalized delay of the traffic light parameters, the first trajectory of the intelligent connected vehicle, and the second trajectory of the human-driven vehicle; determining the traffic light parameters that meet the preset generalized delay conditions among the traffic light parameters based on the generalized delay, and generating the final trajectory of the intelligent connected vehicle using the traffic light parameters that meet the preset generalized delay conditions, the first trajectory, and the second trajectory.

[0007] Optionally, in one embodiment of the present application, the mathematical expression of the signal light parameters of all lanes in the import lane based on the import lane and a pre-established adaptive traffic signal model is obtained, including: obtaining the green light duration of the signal light in each import lane, and obtaining a parameterized expression of the green light duration based on the green light duration; determining the phase information of all import lane signal lights based on the import lane, and obtaining a parameterized expression of the phase information based on the phase information; based on the green light duration constraints, phase sequence blocking constraints and signal cycle constraints in the pre-established adaptive traffic signal model, the mathematical expression of the signal light parameters is obtained using the parameterized expression of the green light duration and the parameterized expression of the phase information.

[0008] Optionally, in one embodiment of the present application, the respectively inputting the mathematical expression of the signal light parameters, the first state information and the second state information into a pre-constructed vehicle delay estimation model to obtain the generalized delay of the signal light parameters, the first trajectory of the intelligent connected vehicle and the second trajectory of the human-driven vehicle includes: obtaining the position information, speed information, speed upper and lower bounds information, acceleration upper and lower bounds information and final state requirement information of the intelligent connected vehicle; obtaining the intelligent connected vehicle trajectory planning model in the pre-constructed vehicle delay estimation model based on the position information, the speed information, the speed upper and lower bounds information, the acceleration upper and lower bounds information and the final state requirement information; and generating the generalized delay and the first trajectory based on the intelligent connected vehicle trajectory planning model and the first state information.

[0009] Optionally, in one embodiment of the present application, the respectively inputting the mathematical expression of the signal light parameters, the first state information and the second state information into a pre-constructed vehicle delay estimation model to obtain the generalized delay of the signal light parameters, the first trajectory of the intelligent connected vehicle and the second trajectory of the human-driven vehicle includes: in the human-driven vehicle trajectory generation model in the pre-constructed vehicle delay estimation model, judging whether the human-driven vehicle is within a preset range of the signal light; if the human-driven vehicle is not within the preset range, judging whether the driving state of the human-driven vehicle is a free driving state; if the driving state is the free driving state, obtaining the second trajectory according to the uniform speed trajectory of the human-driven vehicle; if the driving state is not the free driving state, generating the second trajectory using a target following model; if the human-driven vehicle is within the preset range, obtaining the second trajectory based on the signal light parameters.

[0010] Optionally, in one embodiment of the present application, determining the signal light parameters in the signal light parameters that meet the preset generalized delay conditions based on the generalized delay, and generating the final trajectory of the intelligent connected vehicle using the signal light parameters that meet the preset generalized delay conditions, the first trajectory and the second trajectory, includes: constructing a parallel search subtree of the signal light parameters based on the green light duration in the signal light parameters; determining the delay cost of each node of different subtrees in the parallel search subtree based on the green light duration of each node of different subtrees and the generalized delay; obtaining the signal light parameters that meet the preset generalized delay conditions based on the delay cost; and combining the first trajectory and the second trajectory based on the signal light parameters of the preset generalized delay conditions to obtain the final trajectory.

[0011] The embodiment of the present application can obtain the generalized delay of the signal light parameters, the trajectories of the intelligent networked vehicles and the human-driven vehicles based on the status information of all intelligent networked vehicles and human-driven vehicles within a certain range of the intersection, the import lanes of the intersection, the pre-established adaptive traffic signal model and the pre-built vehicle delay estimation model, and generate the final trajectory of the intelligent networked vehicle according to the signal light parameters that meet certain generalized delay conditions, solve the traffic optimization problem of urban intersections, consider the interactive characteristics of signal lights and vehicle time and mixed traffic conditions, make full use of the advantages of the cloud control system's wide-area perception and collaborative decision-making capabilities, and achieve the improvement of intersection traffic efficiency. The running time of the algorithm can meet the requirements of actual deployment in the edge cloud. Therefore, the problems in the related technology that lack consideration of the real-time state of traffic flow, the mixed traffic conditions of intelligent networked vehicles and human-driven vehicles in the future traffic system, the interaction between signal lights and vehicles, the failure to simultaneously optimize the signal light parameters and the trajectory of the intelligent networked vehicles, the low efficiency of intersection traffic, and the inability to meet the requirements of actual deployment of the edge cloud are solved.

[0012] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0014] Figure 1 A flowchart of a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories provided according to an embodiment of the present application;

[0015] Figure 2 A schematic block diagram of a city intersection scene provided according to an embodiment of the present application;

[0016] Figure 3 A block diagram of an adaptive signal model based on a dual-loop control framework provided according to an embodiment of the present application;

[0017] Figure 4 A block diagram of an optimal trajectory of an intelligent connected vehicle provided according to an embodiment of the present application;

[0018] Figure 5 A schematic block diagram of a human-driven vehicle trajectory considering the influence of traffic lights according to an embodiment of the present application;

[0019] Figure 6 A flow chart of a vehicle delay estimation model provided according to an embodiment of the present application;

[0020] Figure 7 A block diagram of a parameterized adaptive signal model provided according to an embodiment of the present application;

[0021] FIG8( a ) is a block diagram of a portion of a parallel search subtree provided according to an embodiment of the present application;

[0022] FIG8( b ) is a block diagram of another part of parallel search subtrees provided according to an embodiment of the present application;

[0023] Fig. 9 The present invention is a flowchart of the working principle of a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0025] The following describes a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories in an embodiment of the present application with reference to the accompanying drawings. In response to the lack of consideration of the real-time status of traffic flow mentioned in the above background technology, and the mixed traffic conditions of intelligent connected vehicles and human-driven vehicles in future traffic systems, and the interaction between traffic lights and vehicles, the traffic light parameters and the trajectory of intelligent connected vehicles cannot be optimized simultaneously, the traffic efficiency at the intersection is low, and the requirements of actual deployment of edge clouds cannot be met. The present application provides a method for collaborative optimization of intersection traffic lights and intelligent connected vehicle trajectories. In this method, the generalized delay of traffic light parameters, the trajectories of intelligent connected vehicles and human-driven vehicles can be obtained based on the status information of all intelligent connected vehicles and human-driven vehicles within a certain range of the intersection, the entrance lanes of the intersection, the pre-established adaptive traffic signal model and the pre-built vehicle delay estimation model, and the final trajectory of the intelligent connected vehicle is generated based on the traffic light parameters that meet certain generalized delay conditions, which solves the traffic optimization problem of urban intersections, takes into account the interactive characteristics of traffic lights and vehicle time and mixed traffic conditions, and makes full use of the advantages of the cloud control system's wide-area perception and collaborative decision-making capabilities to achieve improved intersection traffic efficiency. The running time of the algorithm can meet the requirements of actual deployment in the edge cloud. This solves the problems in related technologies, such as the lack of consideration of the real-time status of traffic flow, the failure to consider the mixed traffic conditions of intelligent connected vehicles and human-driven vehicles in future traffic systems, the failure to fully consider the interaction between traffic lights and vehicles, the failure to simultaneously optimize traffic light parameters and the trajectory of intelligent connected vehicles, the low efficiency of intersection traffic, and the inability to meet the requirements of actual edge cloud deployment.

[0026] Specifically, Figure 1 The present invention is a flowchart of a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories according to an embodiment of the present application.

[0027] like Figure 1 As shown, the method for collaboratively optimizing intersection signal lights and intelligent connected vehicle trajectories includes the following steps:

[0028] In step S101, first state information of all intelligent networked vehicles and second state information of human-driven vehicles within a preset range of the intersection are collected.

[0029] It can be understood that in the embodiments of the present application, intersections may include but are not limited to cross intersections, T-intersections, Y-intersections, X-intersections, etc., which can be specifically set by technicians in this field according to actual conditions, and this application does not impose specific limitations.

[0030] As a possible implementation method, the embodiment of the present application can collect the first state information of intelligent connected vehicles within a certain range of the intersection and the second state information of human-driven vehicles in real time through the roadside unit, where the state information may include but is not limited to the vehicle's location information and speed information, etc., and the present application does not impose specific restrictions.

[0031] In addition, a certain range can be understood as a control area, which can be specifically set by technicians in this field according to actual conditions, and this application does not impose any specific restrictions.

[0032] For example, Figure 2 As shown, a schematic diagram of an intersection scenario provided by an embodiment of the present application is shown, wherein a roadside unit collects position information and speed information of human-driven vehicles in an intersection control area in real time and reports it to an edge cloud, and an intelligent connected vehicle uploads its own position information and speed information to the edge cloud and receives a reference trajectory from the edge cloud decision.

[0033] In step S102, an entrance lane of the intersection is determined, and a mathematical expression of all lane signal light parameters in the entrance lane is obtained based on the entrance lane and a pre-established adaptive traffic signal model.

[0034] As a possible implementation method, the embodiment of the present application can first determine the entrance lane of the intersection, and then obtain the mathematical expression of the signal light parameters of all lanes according to the entrance lane and the pre-established adaptive traffic signal model.

[0035] It should be noted that, in the embodiment of the present application, the eight entrance lanes of the intersection can be numbered according to the NEMA (National Electrical Manufacturers Association) rules, which can be represented as 1 to 8.

[0036] Optionally, in one embodiment of the present application, a mathematical expression of the signal light parameters of all lanes in the import lane is obtained based on the import lane and a pre-established adaptive traffic signal model, including: obtaining the green light duration of the signal light in each import lane, and obtaining a parameterized expression of the green light duration based on the green light duration; determining the phase information of all import lane signal lights based on the import lane, and obtaining a parameterized expression of the phase information based on the phase information; based on the green light duration constraints, phase sequence blocking constraints and signal cycle constraints in the pre-established adaptive traffic signal model, the mathematical expression of the signal light parameters is obtained using the parameterized expression of the green light duration and the parameterized expression of the phase information.

[0037] It is understandable that the embodiments of the present application can establish an adaptive traffic signal model under the dual-loop control framework of traffic signals.

[0038] During the actual implementation process, the embodiments of the present application can obtain the mathematical expression of the traffic light parameters based on the green light duration constraint, phase sequence barrier constraint and signal cycle constraint in the pre-established adaptive traffic signal model, and according to the parameterized expression of the green light duration of each import lane signal light and the parameterized expression of the phase information.

[0039] Exemplarily, the embodiments of the present application can use parameters to represent the green light duration and phase information corresponding to each import lane, thereby obtaining a mathematical expression of the signal light parameters, and use the mathematical expression of these parameters to express the start and end times of the green light for each import lane in the next signal cycle.

[0040] For example, Figure 3 As shown, the embodiment of the present application provides an adaptive traffic signal model based on a dual-loop control framework. By using the adaptive traffic signal model, the embodiment of the present application can parameterize the green light duration corresponding to each import lane. and phase information θ 12 ,θ 34 ,θ 56 ,θ 78 ,θ M ∈{0,1}, and use the mathematical expression of these parameters to express the starting time of the green light for each lane in the next signal cycle and end time

[0041] Furthermore, the embodiment of the present application takes into account that the green light duration and phase information are adjustable. Among them, the decision quantity is the green light duration corresponding to each import lane and five binary variables θ representing the order between them 12 ,θ 34 ,θ 56 ,θ 78 ,θ M ∈{0,1}. Let the start time of the signal cycle to be optimized be t0. Based on these decision quantities, the start time of the green light corresponding to each import lane can be expressed. For example, the calculation formula for the green light start time of lane 4 can be shown as formula (1), which can be but is not limited to being expressed as:

[0042] t 4,begin =t0+θ M (t 1,green +t 2,green )+θ 34 t 3,green , (1)

[0043] Furthermore, in the embodiment of the present application, these decision quantities need to satisfy the green light duration constraint, the phase sequence blocking constraint and the signal period constraint, and their expressions are shown in equations (2)-(4), which can be expressed as but not limited to:

[0044] g min ≤t i,green ≤g max , (2)

[0045]

[0046]

[0047] Among them, g min and g max are the predefined upper and lower limits of the green light length, C max It is the predefined maximum length of the signal cycle.

[0048] In step S103, the mathematical expression of the traffic light parameters, the first state information and the second state information are respectively input into a pre-built vehicle delay estimation model to obtain the generalized delay of the traffic light parameters, the first trajectory of the intelligent connected vehicle and the second trajectory of the human-driven vehicle.

[0049] It can be understood by those skilled in the art that the embodiments of the present application can input the mathematical expression of the traffic light parameters, the first state information and the second state information into a pre-built vehicle delay estimation model, thereby obtaining the generalized delay of the traffic light parameters, the first trajectory of the intelligent connected vehicle and the second trajectory of the human-driven vehicle.

[0050] Among them, the pre-built vehicle delay estimation model may include but is not limited to an intelligent connected vehicle trajectory planning model, a human-driven vehicle trajectory generation model, etc., and this application does not make specific restrictions.

[0051] For example, the embodiments of the present application can obtain a first trajectory of an intelligent connected vehicle, a second trajectory of a human-driven vehicle, and a generalized delay of traffic light parameters through a pre-built vehicle delay estimation model.

[0052] Optionally, in one embodiment of the present application, the mathematical expression of the signal light parameters, the first state information, and the second state information are respectively input into a pre-constructed vehicle delay estimation model to obtain the generalized delay of the signal light parameters, the first trajectory of the intelligent networked vehicle, and the second trajectory of the human-driven vehicle, including: obtaining the position information, speed information, speed upper and lower bounds information, acceleration upper and lower bounds information, and final state requirement information of the intelligent networked vehicle; obtaining the intelligent networked vehicle trajectory planning model in the pre-constructed vehicle delay estimation model based on the position information, speed information, speed upper and lower bounds information, acceleration upper and lower bounds information, and final state requirement information; generating the generalized delay and the first trajectory based on the intelligent networked vehicle trajectory planning model and the first state information. Among them, the calculation formula of the generalized delay can be, but is not limited to,:

[0053]

[0054] Among them, t ij,departure is the travel time of vehicle (i, j) calculated by the vehicle delay estimation model, t ij,arrival is the time when the vehicle enters the control area, t ij,expectation is the expected driving time of the vehicle in the control area, q ij is a binary variable that represents whether the vehicle can pass in the next signal cycle, ω is the weight factor, and N is the total number of vehicles.

[0055] In some embodiments, the vehicle delay estimation model pre-built in the embodiments of the present application uses generalized delay as output.

[0056] Among them, the generalized delay t delay The definition of is shown in formula (5), which can be expressed as but not limited to:

[0057]

[0058] In some embodiments, for any given parameter of the signal cycle to be optimized, the embodiments of the present application, combined with the status information of the intelligent connected vehicles and human-driven vehicles in the control area, can output the generalized delay corresponding to the signal parameter group using a pre-built vehicle delay estimation model as the objective function of this optimization problem.

[0059] Furthermore, in the embodiment of the present application, the pre-built vehicle delay estimation model may include, but is not limited to, two sub-models: an intelligent networked vehicle trajectory planning model and a human-driven vehicle trajectory generation model, which are not specifically limited in the present application. The intelligent networked vehicle trajectory planning model uses the Pontryagin minimum principle to solve the optimal trajectory of the intelligent networked vehicle. Specifically, the position and speed of the intelligent networked vehicle at time t0 are p0 and v0 respectively, and the upper and lower bounds of the speed are [0,v max ], the upper and lower limits of acceleration are [a min ,a max ], the final state requirement is p f ,v f , the objective function can be expressed as but not limited to:

[0060]

[0061] Furthermore, the embodiment of the present application can derive the optimal speed curve by the minimum value principle: It is composed of no more than 3 straight lines, such as Figure 4 As shown, it can be expressed as but not limited to:

[0062]

[0063] Among them, t d1 ,t s2 is the dividing point, a1 and a2 are the acceleration of the first and third sections respectively, and the second section is a uniform speed, and their values ​​can be calculated through geometric analysis.

[0064] Optionally, in one embodiment of the present application, the mathematical expression of the signal light parameters, the first state information and the second state information are respectively input into a pre-constructed vehicle delay estimation model to obtain the generalized delay of the signal light parameters, the first trajectory of the intelligent connected vehicle and the second trajectory of the human-driven vehicle, including: in the human-driven vehicle trajectory generation model in the pre-constructed vehicle delay estimation model, judging whether the human-driven vehicle is within a preset range of the signal light; if the human-driven vehicle is not within the preset range, judging whether the driving state of the human-driven vehicle is a free driving state; if the driving state is a free driving state, obtaining the second trajectory according to the uniform speed trajectory of the human-driven vehicle; if the driving state is not a free driving state, generating the second trajectory using the target following model; if the human-driven vehicle is within the preset range, obtaining the second trajectory based on the signal light parameters.

[0065] As a possible implementation method, the human driving vehicle trajectory generation model in the embodiment of the present application takes into account the influence of the signal light state on the human driving vehicle trajectory. Specifically, a certain range of the signal light is obtained, such as the influence range p s, when the human driving the vehicle is outside the influence range, if it is in a free driving state, its trajectory is a uniform speed; if it is in a following state, its trajectory is described by the Newell following model, which can be expressed as but not limited to:

[0066] p ij (t) = p i,j-1 (t-τ ij )-d ij , (8)

[0067] Among them, τ ij is the expected time interval, d ij The static spacing.

[0068] Further, in the embodiment of the present application, when a human driving a vehicle is within the influence range, the trajectory will be different depending on the state of the signal light, such as Figure 5 When a human-driven vehicle just enters a certain range of an intersection, such as a control area, if the traffic light is red, it will slow down until the green light turns on and accelerate; if the traffic light is green, the speed of the human-driven vehicle will be adjusted to a speed that allows passing; if the traffic light is yellow, it will determine whether it can pass before the yellow light ends. If not, it will slow down and stop, otherwise it will be adjusted to a speed that allows passing.

[0069] In addition, it should be noted that Figure 6 The flowchart of the vehicle delay estimation model calculation provided in the embodiment of the present application, for all vehicles within a certain range, the trajectory of each vehicle is calculated by recursively pushing from downstream to upstream in different lanes, and its main contents are:

[0070] Step S601: i=1, j=1.

[0071] In this embodiment of the present application, (i, j) represents the j-th vehicle on the lane numbered i.

[0072] Step S602: Determine whether (i, j) is an intelligent connected vehicle.

[0073] In this embodiment of the present application, when (i, j) is an intelligent connected vehicle, step S603 is executed, otherwise, step S604 is executed.

[0074] Step S603: Intelligent connected vehicle trajectory planning.

[0075] In this embodiment of the present application, for lane i, if the leading vehicle (j=1) is an intelligent connected vehicle, its reference trajectory p is generated according to the trajectory planning model of the intelligent connected vehicle. ij (t), where the travel time is t i1,departure The starting time of the green light of the vehicle lane t i,begin and the earliest time the vehicle can pass.

[0076] Step S604: Generate a trajectory of a human-driven vehicle.

[0077] In this embodiment of the present application, for lane i, if the leading vehicle is a human-driven vehicle, its reference trajectory p is generated according to the human-driven vehicle trajectory generation model. ij (t), and calculate the vehicle's passing time t based on the reference trajectory i1,departure .

[0078] Step S605: trajectory fusion.

[0079] In this embodiment of the present application, for the following vehicle (j>1), both the intelligent networked vehicle and the human-driven vehicle need to consider the influence of the preceding vehicle trajectory. Specifically, considering the calculated preceding vehicle trajectory p i,k-1 (t), a virtual following trajectory p can be obtained through the Newell following model ij,Newell (t). If the generated trajectory p ij (t) intersects with the following trajectory, then p ij (t) Smooth connection to p ij,Newell (t) To ensure safety.

[0080] Step S606: Determine whether j=N i or ij,departure ≥t i,end .

[0081] In this embodiment of the present application, j=N i Or the passing time t of a vehicle ij,departure If the green light time has expired, execute step S607; otherwise, execute step S608.

[0082] Step S607: j=j+1.

[0083] Among them, the embodiment of the present application will continue the above recursion until all vehicles on the vehicle lane have been traversed.

[0084] Step S608: Determine whether i=8.

[0085] In this embodiment of the present application, if i=8 is not true, step S609 is executed; if i=8 is true, step S610 is executed.

[0086] Step S609: i=i+1, j=1.

[0087] In this embodiment of the present application, step S610 is executed after all lanes have been recursively pushed.

[0088] Step S610: End.

[0089] Among them, the embodiment of the present application can obtain the generalized delay corresponding to the signal parameter.

[0090] In step S104, signal light parameters satisfying preset generalized delay conditions are determined among the signal light parameters based on the generalized delay, and the final trajectory of the intelligent connected vehicle is generated using the signal light parameters satisfying the preset generalized delay conditions, the first trajectory, and the second trajectory.

[0091] In the actual execution process, the embodiment of the present application can use a heuristic parallel tree search algorithm to find the signal light parameters that meet certain generalized delay conditions, such as the signal light parameters with the minimum generalized delay, as the signal scheme for the next signal cycle. The signal scheme is based on the signal light parameters that meet certain generalized delay conditions, combined with the first trajectory and the second trajectory generated by the pre-built vehicle delay estimation model, to obtain the final trajectory of the intelligent networked vehicle. Among them, the certain generalized delay conditions can be set by those skilled in the art according to actual conditions, and this application does not impose specific restrictions.

[0092] In addition, it should be noted that during the execution of this signal cycle, if a vehicle leaves the intersection or a new intelligent connected vehicle arrives at the control area, it will trigger the re-planning of the vehicle lane. The current signal parameters of the vehicle lane are used as input, and the trajectory of each vehicle is output through the pre-built vehicle delay estimation model, thereby updating the reference trajectory of the intelligent connected vehicles in the control area to adapt to the continuous traffic flow.

[0093] Optionally, in one embodiment of the present application, signal light parameters that meet preset generalized delay conditions in signal light parameters are determined based on generalized delays, and the final trajectory of the intelligent connected vehicle is generated using the signal light parameters that meet the preset generalized delay conditions, the first trajectory, and the second trajectory, including: constructing a parallel search subtree of signal light parameters based on the green light duration in the signal light parameters; determining the delay cost of each node of different subtrees in the parallel search subtree based on the green light duration and generalized delay of each node of different subtrees; obtaining signal light parameters that meet the preset generalized delay conditions based on the delay cost; combining the first trajectory and the second trajectory with the signal light parameters based on the preset generalized delay conditions to obtain the final trajectory.

[0094] In some embodiments, the embodiments of the present application may use a heuristic parallel tree search algorithm to find signal light parameters that meet certain generalized delay conditions, such as signal light parameters corresponding to the minimum generalized delay. Specifically, in the embodiments of the present application, the heuristic parallel tree search algorithm uses a depth-first and parallel search method to find the optimal signal solution and its corresponding intelligent connected vehicle reference trajectory, and decouples this intersection-level problem into several lane-level sub-problems by decoupling the green light phases on both sides of the isolation line.

[0095] For example, Figure 7 As shown, the embodiment of the present application provides a parameterized adaptive traffic signal model, wherein eight letters a to h represent eight green light phases in the adaptive signal model, and the total green light duration t of group 1 under each possible phase sequence is M With different values ​​of as the root node, two parallel search subtrees are constructed, as shown in Figure 8.

[0096] Among them, Figure 8(a) is subtree 1, which represents the values ​​of the 4 green light durations of group 1. The first-level child nodes are all possible values ​​of the two green lights a and b, and the second-level child nodes are all possible values ​​of the two green lights e and f, satisfying:

[0097] t a,green +t b,green =t e,green +t f,grren =t M , (9)

[0098] FIG8(b) is subtree 2. Subtree 2 represents the values ​​of the 4 green light durations of group 2. The first-level child nodes are all possible values ​​of the green light c, the second-level child nodes are all possible values ​​of the green light d, and the third-level child nodes are all possible values ​​of the two green lights g and h, satisfying:

[0099] t c,green +t d,green =t g,green +t h,green , (10)

[0100] Furthermore, in the embodiment of the present application, the two subtrees are searched independently and in parallel. The generalized delay output after the green light duration of the subnode is input into the vehicle delay estimation model is the cost of the subnode. The path with the smallest total cost among all paths from the root node to the leaf node is the optimal path, and the union of the optimal paths of the two subtrees represents the phase sequence and t m The optimal solution that can be achieved by taking the value of , and the one with the minimum cost among all the optimal solutions found in this way is the global optimal solution.

[0101] Furthermore, in an embodiment of the present application, the pruning rule is that during a subtree search, if the total cost of the current path is no less than the calculated minimum cost, the corresponding subsequent nodes will be pruned and will not participate in the calculation.

[0102] The working principle of the method for collaborative optimization of intersection traffic lights and intelligent connected vehicle trajectories proposed in the embodiment of the present application is explained below in conjunction with a specific embodiment.

[0103] in, Fig. 9The present invention is a flowchart of the working principle of a method for collaboratively optimizing intersection traffic lights and intelligent connected vehicle trajectories according to an embodiment of the present application.

[0104] Step S901: The first state information of the intelligent connected vehicles and the second state information of the human-driven vehicles in the intersection control area are collected in real time by the roadside unit, and the traffic lights of all lanes in the import lane are parameterized according to the pre-established adaptive traffic signal model.

[0105] Among them, in the embodiment of the present application, the schematic diagram of the intersection scene is shown in the figure, and the eight import lanes of the intersection can be numbered according to the NEMA rule, which can be represented as 1 to 8.

[0106] Furthermore, the embodiments of the present application can parameterize the green light duration and phase information corresponding to each import lane of the signal cycle to be optimized based on a pre-established adaptive traffic signal model, and use these parameter combinations to express the start and end times of the green light for each import lane in the next signal cycle.

[0107] Step S902: For any given parameter of the signal cycle to be optimized, the generalized delay corresponding to the signal parameter group is output using a pre-built vehicle delay estimation model in combination with the status information of the vehicles in the control area.

[0108] In this embodiment of the present application, the pre-built vehicle delay estimation model execution process is as follows: Figure 6 shown.

[0109] Step S903: A heuristic parallel tree search algorithm is used to find the signal light parameters corresponding to the minimum generalized delay, which is executed as the signal scheme for the next signal cycle. The signal scheme is based on the first trajectory and the second trajectory generated by the pre-built vehicle delay estimation model to obtain the final trajectory of the intelligent connected vehicle.

[0110] Step S904: During the execution of the signal cycle, if a vehicle leaves the intersection or a new intelligent connected vehicle arrives at the control area, it will trigger the re-planning of the vehicle lane. The current signal parameters of the vehicle lane are used as input, and the trajectory of each vehicle is output through the pre-built vehicle delay estimation model, so as to update the reference trajectory of the intelligent connected vehicles in the control area to adapt to the continuous traffic flow.

[0111] According to the method for collaborative optimization of intersection signal lights and intelligent networked vehicle trajectories proposed in the embodiment of the present application, the generalized delay of signal light parameters, the trajectories of intelligent networked vehicles and human-driven vehicles can be obtained based on the status information of all intelligent networked vehicles and human-driven vehicles within a certain range of the intersection, the import lanes of the intersection, the pre-established adaptive traffic signal model and the pre-built vehicle delay estimation model, and the final trajectory of the intelligent networked vehicle is generated according to the signal light parameters that meet certain generalized delay conditions, which solves the traffic optimization problem at urban intersections, considers the interactive characteristics of signal lights and vehicle time and mixed traffic conditions, and fully utilizes the advantages of the cloud control system's wide-area perception and collaborative decision-making capabilities to achieve the improvement of intersection traffic efficiency, and the running time of the algorithm can meet the requirements of actual deployment in the edge cloud. Therefore, the problems in the related technology that lack consideration of the real-time state of traffic flow, the mixed traffic conditions of intelligent networked vehicles and human-driven vehicles in the future traffic system, the interaction between signal lights and vehicles, the failure to simultaneously optimize signal light parameters and the trajectory of intelligent networked vehicles, the low efficiency of intersection traffic, and the inability to meet the requirements of actual deployment of edge clouds are solved.

Claims

1. A method for collaborative optimization of intersection signal lights and intelligent connected vehicle trajectories, characterized in that: The following steps are involved: Collecting first state information of all intelligent networked vehicles and second state information of human-driven vehicles within a preset range of the intersection; Determine an entrance lane of the intersection, and obtain mathematical expressions of all lane signal light parameters in the entrance lane based on the entrance lane and a pre-established adaptive traffic signal model; Inputting the mathematical expression of the signal light parameter, the first state information, and the second state information into a pre-built vehicle delay estimation model respectively to obtain the generalized delay of the signal light parameter, the first trajectory of the intelligent connected vehicle, and the second trajectory of the human-driven vehicle; Based on the generalized delay, signal light parameters among the signal light parameters that meet preset generalized delay conditions are determined, and a final trajectory of the intelligent connected vehicle is generated using the signal light parameters that meet the preset generalized delay conditions, the first trajectory, and the second trajectory.

2. The method according to claim 1, characterized in that The mathematical expression of all lane signal light parameters in the import lane is obtained based on the import lane and the pre-established adaptive traffic signal model, including: Obtaining the green light duration of the signal light in each entrance lane, and obtaining a parameterized expression of the green light duration based on the green light duration; Determining phase information of all traffic lights of the import lane based on the import lane, and obtaining a parameterized expression of the phase information based on the phase information; Based on the green light duration constraint, phase sequence barrier constraint and signal cycle constraint in the pre-established adaptive traffic signal model, the mathematical expression of the traffic light parameters is obtained by using the parameterized expression of the green light duration and the parameterized expression of the phase information.

3. The method according to claim 1, characterized in that The step of inputting the mathematical expression of the signal light parameter, the first state information, and the second state information into a pre-built vehicle delay estimation model to obtain the generalized delay of the signal light parameter, the first trajectory of the intelligent connected vehicle, and the second trajectory of the human-driven vehicle comprises: Obtaining the position information, speed information, speed upper and lower bound information, acceleration upper and lower bound information, and final state requirement information of the intelligent connected vehicle; Obtaining a trajectory planning model for an intelligent connected vehicle in the pre-constructed vehicle delay estimation model based on the position information, the speed information, the speed upper and lower bounds information, the acceleration upper and lower bounds information, and the final state requirement information; The generalized delay and the first trajectory are generated based on the intelligent connected vehicle trajectory planning model and the first state information.

4. The method according to claim 1, characterized in that: The step of inputting the mathematical expression of the signal light parameter, the first state information, and the second state information into a pre-built vehicle delay estimation model to obtain the generalized delay of the signal light parameter, the first trajectory of the intelligent connected vehicle, and the second trajectory of the human-driven vehicle comprises: In the human-driven vehicle trajectory generation model in the pre-built vehicle delay estimation model, determining whether the human-driven vehicle is within a preset range of a traffic light; If the human-driven vehicle is not within the preset range, determining whether the driving state of the human-driven vehicle is a free driving state; If the driving state is the free driving state, obtaining the second trajectory according to the uniform speed trajectory of the human-driven vehicle; If the driving state is not the free driving state, generating the second trajectory by using a target following model; If the human-driven vehicle is within the preset range, the second trajectory is obtained based on the signal light parameters.

5. The method according to claim 2, characterized in that: The determining, based on the generalized delay, signal light parameters among the signal light parameters that meet the preset generalized delay condition, and generating the final trajectory of the intelligent connected vehicle by using the signal light parameters that meet the preset generalized delay condition, the first trajectory, and the second trajectory, includes: Constructing a parallel search subtree of the signal light parameters based on the green light duration in the signal light parameters; Determine the delay cost of each node of the different subtrees in the parallel search subtree based on the green light duration of each node of the different subtrees and the generalized delay; Obtaining the signal light parameters satisfying the preset generalized delay condition based on the delay cost; Based on the signal light parameters of the preset generalized delay condition, the first trajectory and the second trajectory are combined to obtain the final trajectory.

6. The method according to claim 1, characterized in that The calculation formula for the generalized delay is: Among them, t ij，departure is the travel time of vehicle (i, j) calculated by the vehicle delay estimation model, t ij，arrival is the time when the vehicle enters the control area, t ij，expectation is the expected driving time of the vehicle in the control area, q ij is a binary variable that represents whether the vehicle can pass in the next signal cycle, ω is the weight factor, and N is the total number of vehicles.

Images