Non-signalized intersection traffic control method and system based on multilayer game, and computer storage medium

Abstract

The invention discloses a non-signalized intersection traffic control method and system based on a multilayer game, and a computer storage medium, and belongs to the technical field of intelligent traffic and automatic driving, and the method comprises the following steps: 1, carrying out the grid division of an intersection; the method comprises the following steps: dividing a non-signalized intersection area into a plurality of equidistant grids to form a virtual logic grid, and establishing a logic coordinate system; 2, collecting state information; state information of each vehicle entering the intersection area is collected in real time; step 3, gaming in the grid; gaming optimization is carried out on the vehicles in each grid based on a revenue decision function, and the vehicles calculate an optimal strategy through the revenue decision function according to road conditions and states of surrounding vehicles; step 4, cross-grid gaming is carried out; the cross-grid game is realized through bidirectional information transmission between vehicles; 5, multi-layer game optimization and convergence are carried out; and the in-grid game and the cross-grid game are executed layer by layer, so that all vehicles reach global Nash equilibrium.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent transportation and automatic driving, and in particular relates to a traffic control method, system and computer storage medium for an unsignalized intersection based on multi-layer game. Background Art

[0002] At present, traffic management at unsignalized intersections has become an important research direction for intelligent connected transportation systems. Compared with traditional signal light control methods, unsignalized intersections can improve traffic efficiency, but they face problems such as unclear priority allocation and difficulty in vehicle coordination in complex dynamic traffic scenarios. In particular, the efficient coordinated scheduling of connected autonomous vehicles (CAVs) has not been effectively solved.

[0003] Most of the existing traffic strategies are based on centralized or distributed methods, but the centralized method has high computational complexity and poor adaptability; the distributed method is difficult to achieve global traffic flow optimization, and usually ignores the real-time interaction between vehicles, resulting in insufficient traffic efficiency and safety. In addition, how to make full use of the communication and decision-making capabilities of connected vehicles and combine multi-level traffic optimization strategies is still a key problem.

[0004] In response to the above problems, the present invention proposes a traffic control method and system for unsignalized intersections based on multi-layer game, which combines grid division, game optimization inside and outside the grid, and genetic algorithm to achieve efficient vehicle coordination and global optimal control, significantly improving the traffic efficiency and safety of intersections. Summary of the invention

[0005] The purpose of the present invention is to provide a method, system and computer storage medium for traffic control at an unsignalized intersection based on multi-layer game, so as to solve the problems of low efficiency, insufficient safety and difficulty in vehicle coordination in the existing traffic control technology for unsignalized intersections.

[0006] To achieve the above object, the present invention provides a traffic control method for an unsignalized intersection based on a multi-layer game, comprising the following steps:

[0007] Step 1: Intersection grid division: Divide the unsignalized intersection area into multiple equidistant grids to form a virtual logical grid and establish a logical coordinate system. The grid division is based on the geographical characteristics of the intersection and the complexity of the traffic flow. The size of each grid is determined by optimization to better adapt to the dynamic traffic environment. The calculation expression of the logical coordinate system is as follows:

[0008] x=i×l g ,y=j×l g ;

[0009] Among them, l gis the side length of the grid, i and j are the numbers of the grid in the horizontal and vertical directions respectively, and (x, y) is the coordinate of each grid;

[0010] Step 2: Status information collection: Real-time status information of each vehicle entering the intersection area is collected, including position, speed, acceleration, direction and path planning information. The status information is collected by the vehicle-mounted sensors and shared through the communication module to ensure real-time information exchange between all vehicles.

[0011] Step 3: Game within the grid: Vehicles within each grid perform game optimization based on the benefit decision function. The vehicle calculates the best strategy through the benefit decision function according to the road conditions and the status of surrounding vehicles. The expression of the benefit decision function is as follows:

[0012]

[0013] In the formula, J i represents the profit decision function, α, β and γ are used as weight coefficients, and the function f(·) represents the normalization operation with zero mean normalization. Indicates driving safety index, represents the driving efficiency index, represents the comfort index. Each vehicle dynamically adjusts its path and speed according to the game results, and finally makes the traffic flow in the grid reach dynamic equilibrium;

[0014] Step 4: Cross-grid game: Cross-grid game is achieved through two-way information transmission between vehicles. When a vehicle crosses a grid, it plays a collaborative game with vehicles in adjacent grids by sharing its latest status information. The cross-grid vehicle propagates its status information after entering the new grid back to the original grid, and updates the strategy based on the cross-grid information, adjusting the path and trajectory to ensure coordination and consistency between the new grid and the original grid.

[0015] Step 5: Multi-layer game optimization and convergence: Execute intra-grid and inter-grid games layer by layer, and achieve global Nash equilibrium for all vehicles through local optimization within each grid and global coordination across grids based on genetic algorithms. The whole process is achieved through continuous iteration of games and state sharing until the vehicle strategy converges, minimizing traffic delays and avoiding vehicle conflicts.

[0016] Optimum driving safety index Driving efficiency index Comfort index The calculation expression is as follows:

[0017]

[0018] The following constraints limit the permissible speed and acceleration range and stipulate that the vehicle acceleration should not be interrupted:

[0019]

[0020] In the formula, represents the driving safety index, T i represents the passing time of vehicle i, T j represents the passing time of vehicle j, Indicates vehicle V i The speed at time t+1 is Indicates vehicle V i The maximum driving speed, ∈ represents a positive parameter, represents the current position of vehicle i, represents the position of vehicle i at the last moment, and Represents vehicle V i and V j At time t, the variable d represents the minimum safe distance between vehicles to maintain a safe distance, Indicates vehicle V i The minimum acceleration, Indicates vehicle V i The maximum acceleration.

[0021] Preferably, the process of cross-grid game in step 4 is as follows:

[0022] S41, cross-grid information sharing: When a cross-grid vehicle enters a new grid, it transmits its real-time status information to all vehicles in the new grid to ensure that the vehicles in the new grid understand the intention and status of the cross-grid vehicle; the specific expression is as follows:

[0023] I i (t) = {(x j (t),v j (t),p j (t))|j∈N i};

[0024] Among them, N i represents the set of all vehicles in the same grid as vehicle i, x j (t), v j (t), p j (t) represent the position, speed and path of vehicle j at time t, respectively. i (t) is the information sharing function, t represents time;

[0025] S42, cross-grid collaborative optimization: In the new grid, vehicles play games with existing vehicles based on shared cross-grid information. The game process is determined based on dynamic weight coefficients, and the calculation expression is as follows:

[0026]

[0027] In the formula, p ij represents the density ratio of grid i to grid j, ω ij Represents the basic weight factor between grids, indicating other factors besides traffic flow density (such as distance, priority, etc.). It can be a static value or a dynamically adjusted coefficient. i,j represents the dynamic weight coefficient, which is used to comprehensively describe the interaction priority between two grids, ρ i ,ρ j are the traffic flow densities of the two grids respectively;

[0028] S43. Achieve strategy optimization among grids through collaborative game.

[0029] Preferably, the process of adjusting the path and trajectory is as follows:

[0030] Path selection: The vehicle selects the optimal path based on the shared information. The path selection function P i The expression of (t) is as follows:

[0031]

[0032] Among them, i represents the set of all possible paths, C i (p) represents the congestion index of path p, E i (p) represents the travel time of path p, and γ and δ are weight coefficients; path selection provides a macro solution for vehicle traffic, optimizes the distribution of overall traffic flow, and ensures the best effect on congestion and time delay;

[0033] Trajectory adjustment: After the path is determined, the vehicle will further adjust the driving trajectory based on the path. The trajectory adjustment function T i The expression of (t) is as follows:

[0034]

[0035] Among them, θ i represents the set of all possible trajectories, D i (τ) represents the delay time corresponding to the trajectory τ, S i (τ) represents the safety index corresponding to the trajectory τ, and α and β are weight coefficients.

[0036] Preferably, the constraints of the global Nash equilibrium are as follows:

[0037] max U i =α·R i +(1-α)·J i ;

[0038]

[0039] Among them, δ is the tolerance of policy deviation, m is the number of participants, and U i Represents the participant V i The utility function, R i Denotes the deviation benefit of the ith participant, D i represents the strategy deviation of the ith participant, D i , U i and R i The calculation expression is as follows:

[0040]

[0041] U i =α·R i +(1-α)·J i ;

[0042] In the formula, S ij represents the j-th strategy of participant i, represents the optimal value of the jth strategy, and n is the number of strategies.

[0043] Preferably, the genetic algorithm evaluates the effectiveness of the strategy through the following fitness function:

[0044] F=-(β1·T delay +β2·R collision );

[0045] Among them, T delay represents the average delay time of the vehicle, R collision represents the risk probability of vehicle collision, β1 and β2 are the weight coefficients of the fitness function, which are used to balance traffic efficiency and safety.

[0046] The present invention also provides a traffic control system for an unsignalized intersection based on multi-layer game, comprising: a grid division module; used to divide the intersection area into a plurality of equidistant grids, and establish a logical coordinate system for each grid to form a virtual logical network to adapt to the complex situation of dynamic entry and exit of vehicles;

[0047] Information collection module: used to collect vehicle status information in real time, including location, speed, acceleration and path planning information. Information collection is achieved through on-board equipment and intersection sensing equipment;

[0048] Game optimization module: used to perform game optimization within each grid and conduct cross-grid collaborative games when vehicles cross grids. Vehicles continuously adjust their own strategies based on the profit decision function to achieve dynamic equilibrium within and between grids. The game optimization module calculates the profit function between vehicles and executes multi-layer game processes within and across grids in sequence. Each layer of the game aims to maximize driving efficiency and safety, and ultimately all vehicles reach a dynamic equilibrium state.

[0049] Trajectory adjustment module: used to adjust the vehicle's driving trajectory and speed in real time according to the game optimization results to ensure the safety and efficiency of the vehicle when passing through the intersection. Trajectory adjustment includes path replanning and speed adjustment;

[0050] Global optimization module: used to globally optimize the preliminary results based on multi-layer games in combination with genetic algorithms. The optimization process includes fitness evaluation of traffic strategies, selection, crossover and mutation operations in genetic algorithms to generate the optimal traffic strategy. The global optimization module combines real-time traffic status information and optimizes the game results through genetic algorithms. The optimization process includes selection, crossover and mutation to adapt to complex situations in different traffic scenarios and reduce traffic delays and accident risks at intersections.

[0051] Communication module: used to realize information exchange between vehicles and systems, ensuring that vehicle status information can be shared in a timely manner within and between grids to support game optimization and global optimization processes. Communication is achieved through vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I);

[0052] The present invention also provides a computer storage medium on which computer-readable instructions are stored. When the instructions are executed by a computer, the above method is implemented.

[0053] Therefore, the present invention adopts the above-mentioned traffic control method, system and computer storage medium for signal-free intersections based on multi-layer game, and dynamically coordinates and optimizes the vehicles entering the intersection by gridding the intersection area and combining the multi-layer game mechanism; by utilizing the real-time status information and communication characteristics of the vehicle, the global optimization of path selection, speed adjustment and priority planning is achieved, and the traffic efficiency and safety of the intersection are effectively improved; compared with the traditional fixed priority control and first-come-first-served (FCFS) model, the method based on multi-layer game of the present invention can dynamically adjust the strategy and significantly reduce the delay time and collision risk of the vehicle. At the same time, the strategy of combining multi-layer game with genetic algorithm optimization overcomes the problems of high computational complexity and rigid strategy of traditional methods in high-traffic scenarios, and provides an efficient and reliable solution for the intelligent management of signal-free intersections. This technology not only improves the stability of the overall traffic flow, but also optimizes the passing efficiency of vehicles, laying a foundation for the application and promotion of intelligent networked transportation systems.

[0054] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flow chart of a traffic control method for an unsignalized intersection based on multi-layer game according to the present invention;

[0056] Figure 2 The present invention is an overall block diagram of a traffic control system for an unsignalized intersection based on multi-layer game. DETAILED DESCRIPTION

[0057] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] See also Figure 1 , a traffic control method for an unsignalized intersection based on multi-layer game, comprising the following steps:

[0059] Step 1: Intersection grid division: Divide the unsignalized intersection area into multiple equidistant grids to form a virtual logical grid and establish a logical coordinate system. The grid division is based on the geographical characteristics of the intersection and the complexity of the traffic flow. The size of each grid is determined by optimization to better adapt to the dynamic traffic environment. The calculation expression of the logical coordinate system is as follows:

[0060] x=i×l g ,y=j×l g ;

[0061] Among them, l g is the side length of the grid, i and j are the numbers of the grid in the horizontal and vertical directions respectively, and (x, y) is the coordinate of each grid;

[0062] Step 2: Status information collection: Real-time status information of each vehicle entering the intersection area is collected, including position, speed, acceleration, direction and path planning information. The status information is collected by the vehicle-mounted sensors and shared through the communication module to ensure real-time information exchange between all vehicles.

[0063] Step 3: Game within the grid: Vehicles within each grid perform game optimization based on the benefit decision function. The vehicle calculates the best strategy through the benefit decision function according to the road conditions and the status of surrounding vehicles. The expression of the benefit decision function is as follows:

[0064]

[0065] In the formula, J i represents the profit decision function, α, β and γ are used as weight coefficients, and the function f(·) represents the normalization operation with zero mean normalization. Indicates driving safety index, represents the driving efficiency index, represents the comfort index. Each vehicle dynamically adjusts its path and speed according to the game results, and finally makes the traffic flow in the grid reach dynamic equilibrium. Among them, the driving safety index Driving efficiency index Comfort index The calculation expression is as follows:

[0066]

[0067] The following constraints limit the permissible speed and acceleration range and stipulate that the vehicle acceleration should not be interrupted:

[0068]

[0069] In the formula, represents the driving safety index, T i represents the passing time of vehicle i, T j represents the passing time of vehicle j, Indicates vehicle V i The speed at time t+1 is Indicates vehicle V i The maximum driving speed, ∈ represents a positive parameter, represents the current position of vehicle i, represents the position of vehicle i at the last moment, and Represents vehicle V i and V j At time t, the variable d represents the minimum safe distance between vehicles to maintain a safe distance, Indicates vehicle V i The minimum acceleration, Indicates vehicle V i The maximum acceleration.

[0070] Step 4: Cross-grid game: Cross-grid game is achieved through two-way information transmission between vehicles. When a vehicle crosses a grid, it plays a collaborative game with vehicles in adjacent grids by sharing its latest status information. The cross-grid vehicle propagates its status information after entering the new grid back to the original grid, and updates the strategy based on the cross-grid information, and adjusts the path and trajectory to ensure the coordination and consistency between the new grid and the original grid. The specific process is as follows:

[0071] S41, cross-grid information sharing: When a cross-grid vehicle enters a new grid, it transmits its real-time status information to all vehicles in the new grid to ensure that the vehicles in the new grid understand the intention and status of the cross-grid vehicle; the specific expression is as follows:

[0072] I i (t) = {(x j (t),v j (t),p j (t))|j∈N i};

[0073] Among them, N i represents the set of all vehicles in the same grid as vehicle i, x j (t), v j (t), p j (t) represent the position, speed and path of vehicle j at time t, respectively. i (t) is the information sharing function, t represents time;

[0074] S42, cross-grid collaborative optimization: In the new grid, vehicles play games with existing vehicles based on shared cross-grid information. The game process is determined based on dynamic weight coefficients, and the calculation expression is as follows:

[0075]

[0076] In the formula, p ij represents the density ratio of grid i to grid j, ω ij Represents the basic weight factor between grids, indicating other factors besides traffic flow density (such as distance, priority, etc.). It can be a static value or a dynamically adjusted coefficient. i,j represents the dynamic weight coefficient, which is used to comprehensively describe the interaction priority between two grids, ρ i ,ρ j are the traffic flow densities of the two grids respectively;

[0077] S43. Achieve strategy optimization among grids through collaborative game.

[0078] The process of adjusting the path and trajectory is as follows:

[0079] Path selection: The vehicle selects the optimal path based on the shared information. The path selection function P i The expression of (t) is as follows:

[0080]

[0081] Among them, i represents the set of all possible paths, C i(p) represents the congestion index of path p, E i (p) represents the travel time of path p, and γ and δ are weight coefficients; path selection provides a macro solution for vehicle traffic, optimizes the distribution of overall traffic flow, and ensures the best effect on congestion and time delay;

[0082] Trajectory adjustment: After the path is determined, the vehicle will further adjust the driving trajectory based on the path. The trajectory adjustment function T i The expression of (t) is as follows:

[0083]

[0084] Among them, θ i represents the set of all possible trajectories, D i (τ) represents the delay time corresponding to the trajectory τ, S i (τ) represents the safety index corresponding to the trajectory τ, and α and β are weight coefficients.

[0085] Step 5, multi-layer game optimization and convergence; perform intra-grid and cross-grid games layer by layer, and achieve global Nash equilibrium for all vehicles through local optimization within each grid and global coordination across grids based on genetic algorithms. The whole process is achieved through continuous iteration of games and state sharing until the vehicle strategy converges, minimizing traffic delays and avoiding vehicle conflicts. Through the coordination of cross-grid vehicles, continuous information exchange and strategy adjustment are carried out between adjacent grids, realizing the organic combination of intra-grid and inter-grid game decisions. Since all grids form a closely connected virtual logical network, the global game dynamic equilibrium of the entire intersection can be finally achieved through multi-layer games within and between grids. On this basis, the genetic algorithm is used to solve the above-mentioned nonlinear optimization model with constraints to achieve strategy optimization, and finally the CAV traffic strategy optimization method for unsignaled intersections considering multi-layer games and genetic algorithms is obtained, realizing the global optimal game equilibrium of vehicle traffic strategy, thereby improving the traffic efficiency and safety of intersections. Among them. The constraints of the global Nash equilibrium are as follows:

[0086] max U i =α·R i +(1-α)·J i ;

[0087]

[0088] Among them, δ is the tolerance of strategy deviation, m is the number of participants, and U i Represents the participant V i The utility function, R i Denotes the deviation benefit of the ith participant, D i represents the strategy deviation of the ith participant, Di , U i and R i The calculation expression is as follows:

[0089]

[0090] U i =α·R i +(1-α)·J i ;

[0091] In the formula, S ij represents the j-th strategy of participant i, S ij * represents the optimal value of the jth strategy, and n is the number of strategies.

[0092] The genetic algorithm evaluates the effectiveness of the strategy through the following fitness function:

[0093] F=-(β1·T delay +β2·R collision );

[0094] Among them, T delay represents the average delay time of the vehicle, R collision represents the risk probability of vehicle collision, β1 and β2 are the weight coefficients of the fitness function, which are used to balance traffic efficiency and safety.

[0095] Among them, in the embodiment of the present invention, the intersection area grid division method and information collection process include the following steps:

[0096] According to the preset maximum speed of the test vehicle and the maximum speed limit of surrounding vehicles, an unsignalized intersection with four entrance lanes, all of which are two-lane, is established.

[0097] More specifically, in view of the fact that existing studies rarely consider the traffic scenario of connected automated vehicles at unsignalized intersections, SUMO is used in the embodiment of the present invention to establish the intersection to implement the method of the present invention. The simulated intersection range is a square decision area with a side length of 180m, and the vehicles that are used to play the game using the method of the present invention are vehicles that enter the decision area.

[0098] First, the unsignalized intersection area is divided into multiple equidistant virtual grids to form a logical grid structure. Each grid is an independent game unit. Assume that the intersection area is divided into m×n grids, and the logical coordinates of each grid are (x i ,y j ), where i and j represent the horizontal and vertical numbers of the grid respectively. The side length of the grid is l x ,l yDetermined through dynamic optimization to adapt to the dynamic characteristics of intersection traffic flow. Through grid division, the status of vehicles can be finely tracked and managed, thus providing a basis for subsequent game optimization.

[0099] The vehicle status information collection method comprises the following steps:

[0100] Each vehicle entering the intersection area needs to collect its own status information in real time, including position (x, y), speed v, acceleration a, and current path planning information. This information is collected by on-board sensors (such as lidar, camera) and shared with other vehicles in the same grid through vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication modules. Vehicles exchange information with other vehicles in the grid through communication modules to ensure that each vehicle can understand the dynamic information of other vehicles in the grid when making decisions, including each vehicle's current position, speed, acceleration, and path intention.

[0101] See also Figure 2 , a traffic control system for an unsignalized intersection based on multi-layer game, comprising: a grid division module; for dividing the intersection area into a plurality of equidistant grids, and establishing a logical coordinate system for each grid to form a virtual logical network, adapting to the complex situation of dynamic entry and exit of vehicles;

[0102] Information collection module: used to collect vehicle status information in real time, including location, speed, acceleration and path planning information. Information collection is achieved through on-board equipment and intersection sensing equipment;

[0103] Game optimization module: used to perform game optimization within each grid and conduct cross-grid collaborative games when vehicles cross grids. Vehicles continuously adjust their own strategies based on the profit decision function to achieve dynamic equilibrium within and between grids. The game optimization module calculates the profit function between vehicles and executes multi-layer game processes within and across grids in sequence. Each layer of the game aims to maximize driving efficiency and safety, and ultimately all vehicles reach a dynamic equilibrium state.

[0104] Trajectory adjustment module: used to adjust the vehicle's driving trajectory and speed in real time according to the game optimization results to ensure the safety and efficiency of the vehicle when passing through the intersection. Trajectory adjustment includes path replanning and speed adjustment;

[0105] Global optimization module: used to globally optimize the preliminary results based on multi-layer games in combination with genetic algorithms. The optimization process includes fitness evaluation of traffic strategies, selection, crossover and mutation operations in genetic algorithms to generate the optimal traffic strategy. The global optimization module combines real-time traffic status information and optimizes the game results through genetic algorithms. The optimization process includes selection, crossover and mutation to adapt to complex situations in different traffic scenarios and reduce traffic delays and accident risks at intersections.

[0106] Communication module: used to realize information exchange between vehicles and systems, ensuring that vehicle status information can be shared in a timely manner within and between grids to support game optimization and global optimization processes. Communication is achieved through vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I);

[0107] The present invention provides a computer storage medium on which computer-readable instructions are stored. When the instructions are executed by a computer, the above method is implemented.

[0108] Therefore, the present invention adopts the above-mentioned signal-free intersection traffic control method, system and computer storage medium based on multi-layer game, and dynamically coordinates and optimizes vehicles entering the intersection by gridding the intersection area and combining the multi-layer game mechanism. By utilizing the real-time status information and communication characteristics of the vehicle, the global optimization of path selection, speed adjustment and priority planning is achieved, effectively improving the traffic efficiency and safety of the intersection.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A traffic control method for unsignalized intersections based on multi-layer game, characterized in that: The following steps are involved: Step 1: Intersection grid division: Divide the unsignalized intersection area into multiple equidistant grids to form a virtual logical grid and establish a logical coordinate system. The grid division is based on the geographical characteristics of the intersection and the complexity of the traffic flow. The size of each grid is determined by optimization. Step 2: Status information collection; Real-time collection of status information of each vehicle entering the intersection area, including location, speed, acceleration, direction and path planning information. The status information is collected by the vehicle-mounted sensors and shared through the communication module; Step 3: Game within the grid: Vehicles within each grid perform game optimization based on the benefit decision function. The vehicle calculates the best strategy through the benefit decision function according to the road conditions and the status of surrounding vehicles. The expression of the benefit decision function is as follows: In the formula, J i represents the profit decision function, α, β and γ are used as weight coefficients, and the function f(·) represents the normalization operation with zero mean normalization. Indicates driving safety index, represents the driving efficiency index, represents the comfort index; Step 4: Cross-grid game: Cross-grid game is achieved through two-way information transmission between vehicles. When a vehicle crosses a grid, it plays a collaborative game with vehicles in adjacent grids by sharing its latest status information. The cross-grid vehicle propagates its status information after entering the new grid back to the original grid, and updates the strategy based on the cross-grid information to adjust the path and trajectory. Step 5: Multi-layer game optimization and convergence: Execute intra-grid and inter-grid games layer by layer, and achieve global Nash equilibrium for all vehicles through local optimization within each grid and global coordination across grids based on genetic algorithms.

2. The method for controlling traffic at an unsignalized intersection based on multi-layer game according to claim 1, characterized in that: Driving safety index Driving efficiency index Comfort index The calculation expression is as follows: In the formula, represents the driving safety index, T i represents the passing time of vehicle i, T j represents the passing time of vehicle j, Indicates vehicle V i The speed at time t+1 is Indicates vehicle V i The maximum driving speed, ∈ represents a positive parameter, represents the current position of vehicle i, represents the position of vehicle i at the previous moment.

3. The method for controlling traffic at an unsignalized intersection based on multi-layer game according to claim 2, characterized in that: The process of cross-grid game in step 4 is as follows: S41, cross-grid information sharing: When a cross-grid vehicle enters a new grid, it transmits its real-time status information to all vehicles in the new grid; S42, cross-grid collaborative optimization: In the new grid, vehicles play games with existing vehicles based on shared cross-grid information. The game process is determined based on dynamic weight coefficients, and the calculation expression is as follows: In the formula, p ij Denotes the density ratio of grid i to grid j, ω ij Represents the basic weight factor between grids, W i,j represents the dynamic weight coefficient, ρ i ,ρ j are the traffic flow densities of the two grids respectively; S43. Achieve strategy optimization among grids through collaborative game.

4. The method for controlling traffic at an unsignalized intersection based on multi-layer game according to claim 3, characterized in that: The process of adjusting the path and trajectory is as follows: Path selection; the vehicle selects the optimal path based on shared information, and the path selection function P i The expression of (t) is as follows: Among them, i represents the set of all possible paths, C i (p) represents the congestion index of path p, E i (p) represents the travel time of path p, γ and δ are weight coefficients; Trajectory adjustment: After the path is determined, the vehicle will further adjust the driving trajectory based on the path. The trajectory adjustment function T i The expression of (t) is as follows: Among them, θ i represents the set of all possible trajectories, D i (τ) represents the delay time corresponding to the trajectory τ, S i (τ) represents the safety index corresponding to the trajectory τ, and α and β are weight coefficients.

5. The method for controlling traffic at an unsignalized intersection based on multi-layer game according to claim 4, characterized in that: The constraints of the global Nash equilibrium are as follows: max U i =α·R i +(1-a)·J i ; Among them, δ is the tolerance of strategy deviation, m is the number of participants, and U i Represents the participant V i The utility function, R i Denotes the deviation benefit of the ith participant, D i represents the strategy deviation of the i-th participant.

6. The method for controlling traffic at an unsignalized intersection based on multi-layer game according to claim 5, characterized in that: The genetic algorithm evaluates the effectiveness of the strategy through the following fitness function: F=-(β1·T delay +β2·R collision ); Among them, T delay represents the average delay time of the vehicle, R collision represents the risk probability of vehicle collision, β1 and β2 are the weight coefficients of the fitness function, which are used to balance traffic efficiency and safety.

7. A traffic control system for an unsignalized intersection based on multi-layer game, characterized in that: A method for controlling traffic at an unsignalized intersection based on multi-layer game according to any one of claims 1 to 6, comprising: Grid division module: used to divide the intersection area into multiple equidistant grids and establish a logical coordinate system for each grid to form a virtual logical network to adapt to the complex situation of dynamic entry and exit of vehicles; Information collection module: used to collect vehicle status information in real time, including location, speed, acceleration and path planning information. Information collection is achieved through on-board equipment and intersection sensing equipment; Game optimization module: used to perform game optimization in each grid and conduct cross-grid collaborative games when vehicles cross grids. Vehicles continuously adjust their strategies based on the profit decision function to achieve dynamic equilibrium within and between grids. Trajectory adjustment module: used to adjust the vehicle's driving trajectory and speed in real time according to the game optimization results to ensure the safety and efficiency of the vehicle when passing through the intersection. Trajectory adjustment includes path replanning and speed adjustment; Global optimization module: used to globally optimize the preliminary results based on multi-layer game by combining genetic algorithm. The optimization process includes fitness evaluation of traffic strategy, selection, crossover and mutation operations in genetic algorithm to generate the optimal traffic strategy. Communication module: used to realize information exchange between vehicles and systems, ensuring that vehicle status information can be shared timely within and between grids to support game optimization and global optimization processes. Communication is realized through vehicle-to-vehicle and vehicle-to-infrastructure.

8. A computer storage medium, characterized in that: Computer-readable instructions are stored thereon, and when the instructions are executed by a computer, the method according to any one of claims 1 to 6 is implemented.

Images