Traffic light signal control system and method based on car-road cooperation

Abstract

The application belongs to the field of intelligent transportation, and provides a traffic light signal control system and method based on vehicle-road cooperation, which comprises a traffic information sensing and graph modeling module, a fuzzy game strategy generation module, a game optimization and traffic light regulation module, a vehicle intention analysis and path guidance module, and a traffic light instruction execution module. The traffic information sensing and graph modeling module is used for constructing a city road network graph model, collecting real-time traffic state data of each intersection and vehicle dynamic data uploaded by a vehicle terminal. The fuzzy game strategy generation module defines a fuzzy utility function for each candidate traffic light timing strategy, determines a vehicle intention according to the vehicle dynamic data, determines a fuzzy factor of the fuzzy utility function according to the vehicle intention, and weights the fuzzy membership degree of the traffic light timing strategy. The game optimization and traffic light regulation module is used for obtaining an optimal traffic light control strategy satisfying a fuzzy Nash equilibrium condition, and outputting an adjustment parameter of a traffic light length for a current period. The vehicle intention analysis and path guidance module is used for generating a detour suggestion for path guidance based on a game solution result. The traffic light instruction execution module is used for updating a traffic light state.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of intelligent transportation, and in particular relates to a traffic light signal control system and method based on vehicle-to-infrastructure cooperation. BACKGROUND

[0002] With the continuous growth of the number of motor vehicles in cities, the pressure on urban road traffic is increasing, and problems such as traffic congestion, low traffic efficiency, and high energy consumption are becoming increasingly prominent. In order to alleviate the above problems, the traditional traffic light signal control system has gradually developed towards intelligence and self-adaptation. Some existing systems make dynamic adjustments to traffic light timing to a certain extent by sensing traffic flow changes and combining historical data prediction, which has improved the traffic operation condition to a certain extent. However, most of these systems still take single intersection as the control object, lack regional coordination optimization capability, and cannot effectively deal with the chain congestion problem caused by high-density and strong-coupling road network structure.

[0003] In recent years, the development of vehicle-to-infrastructure (V2I) technology has provided a new breakthrough for urban traffic intelligence. By interconnecting road infrastructure, edge computing units, and vehicle terminals, more accurate and real-time dynamic perception of traffic status can be achieved, providing high-precision input conditions for traffic control systems. At the same time, the acquisition of vehicle path intention and other driving behavior information also provides new data support for predicting traffic evolution trends and optimizing control strategies.

[0004] Although some existing research has attempted to introduce vehicle-to-infrastructure information into the signal timing control model, there are still the following deficiencies: (1) Most control methods rely on rule libraries or heuristic strategies, making it difficult to handle timing linkage and strategy conflicts between multiple intersections; (2) There is a lack of modeling of traffic state uncertainty, especially in fast-changing and high-noise environments, with poor control stability; (3) The path intention information uploaded by the vehicle terminal is not fully utilized to guide the timing strategy, and the overall traffic efficiency and individual vehicle experience cannot be considered; (4) The timing strategy lacks a systematic modeling and optimization mechanism for regional synchronization characteristics such as green wave coordination and cycle offset.

[0005] Therefore, there is an urgent need to design an intelligent traffic light signal control system that can integrate vehicle-to-infrastructure data, consider traffic state uncertainty, and adapt to the structure of multiple intersections in cities, in order to achieve global coordination, adaptive adjustment, and path-level guidance, and comprehensively improve the traffic efficiency and traffic safety of urban road networks. SUMMARY

[0006] In order to solve the problems in the prior art, the present application provides a traffic light signal control system based on vehicle-to-infrastructure cooperation, comprising:

[0007] The traffic information perception and graph modeling module is configured to deploy edge control units at multiple intersections, construct a city road network graph model taking intersections as nodes and roads as edges, collect real-time traffic state data of each intersection and vehicle dynamic data uploaded by a vehicle terminal, and the like.

[0008] The fuzzy game strategy generation module is configured to construct a traffic light timing strategy set based on the real-time traffic state data of the intersection, define a fuzzy utility function for each candidate traffic light timing strategy, determine a vehicle intention according to the vehicle dynamic data, determine a fuzzy factor of the fuzzy utility function according to the vehicle intention, and weight fuzzy membership degrees of the traffic light timing strategies.

[0009] The game optimization and traffic light regulation module is configured to call a fuzzy game solving mechanism according to an adjacency relationship between intersections in the city road network graph model and a change range of traffic light timing parameters, obtain an optimal traffic light control strategy satisfying a fuzzy Nash equilibrium condition, and output an adjustment parameter of a traffic light duration for a current period.

[0010] The vehicle intention analysis and path guidance module is configured to receive the vehicle dynamic data uploaded by the vehicle terminal, real-time traffic state data and game solving results, identify target vehicles that can still reach a destination as needed after changing a path intention, generate a rerouting suggestion for path guidance based on the game solving results, and push the suggestion to corresponding vehicles.

[0011] The traffic light instruction execution module is configured to receive the traffic light adjustment parameter output by the game optimization and traffic light regulation module, and deliver the adjustment parameter to a traffic light controller of each intersection to update a traffic light state.

[0012] Further, the traffic information perception and graph modeling module is configured to fuse the collected intersection traffic state data and dynamic information of vehicle positions, speeds and path intentions uploaded by the vehicle terminal, and construct a city road network graph model having node traffic attributes and edge path attributes, each node in the model corresponding to an intersection, each edge corresponding to a road connection relationship, and the nodes and edges respectively carrying traffic flow density, average travel time and flow direction weight attribute information.

[0013] Further, the fuzzy game strategy generation module is configured to generate multiple candidate traffic light timing strategies based on a traffic state vector of each intersection, historical data and control rules, the strategies including a signal period, a passing phase sequence, a green light duration of each phase and a yellow light warning duration, and construct a fuzzy utility function for each strategy, the fuzzy utility function scoring average waiting time, queue length and passing capacity indicators by using a preset membership function to reflect passing adaptability of the strategy.

[0014] Further, the fuzzy game strategy generation module is further configured to: according to path intention information carried in the vehicle dynamic data, count vehicle proportion of each passing direction in the current period of the intersection, and construct a direction intention fuzzy factor according to the same, and use the fuzzy factor to weight and correct the direction passing membership value in the fuzzy utility function, so as to enhance the utility response capability of the traffic light timing strategy in the main passing direction.

[0015] Further, the game optimization and traffic light regulation module is configured to: based on the adjacency relationship between intersection nodes in the city road network map, construct a fuzzy game model, the model taking the intersection as the participant and the candidate timing strategy as the strategy set, calculating the coordination penalty function between the local utility of the intersection and the adjacent strategy in each round of game, and iteratively solving the optimal control strategy combination meeting the fuzzy Nash equilibrium through the optimal response algorithm or the graph structure search algorithm.

[0016] Further, the coordination penalty function is based on the signal period difference between adjacent intersections, the phase green light start and end time offset, and the road segment green wave band continuity, to construct a coordination degree index for evaluating the coordination of the strategy combination in the regional range, so as to improve the signal synchronization and passing continuity in the whole region.

[0017] Further, the vehicle intention analysis and path guidance module is configured to: after identifying the vehicle with the possibility of changing the path intention, estimating the passing cost of the candidate path of the vehicle, the estimation index including path length, traffic light passing release probability, average delay and path congestion index, and evaluating the matching degree of each path in combination with the signal game result, so as to select the optimal path for guidance.

[0018] Further, the path guidance suggestion includes alternative path sequence, expected time saving, recommended entering intersection and corresponding lane change prompt, and is sent to the vehicle terminal through the V2I communication interface in the form of structured prompt information, the terminal presents the suggestion to the driver in the form of navigation prompt, screen pop-up window or voice broadcast, for assisting the vehicle dynamic decision.

[0019] Further, the traffic light instruction execution module is configured to: format convert the traffic light time length adjustment parameters output by the game optimization and traffic light regulation module, generate a structured control instruction including signal period total time length, green light time length of each traffic phase, yellow light warning time length and phase sequence,

[0020] and perform protocol adaptation based on the deployed traffic light controller communication protocol, and issue to the traffic light controller of the corresponding intersection in the form of serial communication, Ethernet communication or cellular network communication;

[0021] The traffic light control device has a state feedback mechanism for returning instruction execution state, current phase operation state or fault information, and the instruction execution module triggers emergency processing or retransmission process according to the feedback result, so as to guarantee the effective landing and stable operation of the signal control strategy.

[0022] The application also provides a traffic light signal control method based on vehicle-road cooperation, which uses the traffic light signal control system based on vehicle-road cooperation.

[0023] The application provides a traffic light signal control system and method based on vehicle-road cooperation, which can fully integrate roadside traffic perception information and vehicle terminal dynamic information, establish a city road network graph model reflecting traffic evolution characteristics, and realize signal control linkage and global optimization between multiple intersections by combining fuzzy game theory, and has the following beneficial effects:

[0024] During the generation of the signal control strategy, a fuzzy utility function of the timing scheme is established by using a fuzzy membership function, so as to effectively deal with uncertain factors such as traffic data fluctuations and prediction errors, and improve the adaptability and stability of the strategy. The vehicle path intention ratio is introduced as a fuzzy factor to participate in game modeling, which effectively improves the score and priority of the main traffic direction strategy and enhances the response ability of the system to the current real traffic demand.

[0025] A game graph structure with intersections as participants is constructed, and an adjacent node coordination penalty function is introduced in the game, so that the signal control strategies of the intersections can evolve according to the regional traffic situation, avoiding problems such as signal island and green wave breakage, and significantly improving the regional traffic efficiency.

[0026] Through vehicle intention recognition and path accessibility analysis, the system can recommend a detour for some vehicles with adjustable intentions, relieve the pressure on the main channel without affecting the traffic efficiency, and realize the integrated scheduling of the control strategy and individual vehicle behavior.

[0027] In summary, the application constructs a multi-source perception, fuzzy modeling, game solving and path guidance integrated control architecture in a vehicle-road cooperation environment, significantly improves the intelligent level of urban road network traffic signal regulation, and provides effective technical support for relieving traffic congestion, improving traffic efficiency and supporting green travel. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0029] Figure 1 is a system block diagram of the present application. DETAILED DESCRIPTION

[0030] The application is described below in conjunction with the drawings and the detailed description.

[0031] The embodiment solves the above problems by the following steps:

[0032] In one embodiment, with reference to Figure 1 The application provides a traffic light signal control system based on vehicle-road cooperation, which uses a real-time communication mechanism between road infrastructure and vehicle-mounted equipment to realize a multi-source traffic state perception, intelligent traffic strategy game decision and dynamic signal timing control intelligent traffic control system. Specifically, the system includes the following modules:

[0033] A traffic information perception and graph modeling module is used to deploy edge control units at multiple intersections, construct a city road network graph model with intersections as nodes and roads as edges, and collect real-time traffic state data at each intersection and vehicle dynamic data uploaded by vehicle-mounted terminals.

[0034] The traffic information perception and graph modeling module collects real-time traffic state data at the intersection through various traffic perception devices deployed at the intersection, such as traffic cameras, geomagnetic detectors, radar sensors, microwave vehicle detectors, etc. The traffic state data includes but is not limited to:

[0035] The instantaneous number of vehicles passing through each direction lane;

[0036] The average length and waiting time of the queued vehicles on each lane;

[0037] The current phase, cycle and historical timing record of the traffic light in the intersection;

[0038] The road environment state (such as whether there is construction, whether there is traffic restriction, current weather, etc.);

[0039] The module also acquires real-time vehicle dynamic information through a V2I communication interface with the vehicle-mounted terminal, and the vehicle dynamic data includes:

[0040] The current position coordinates (GPS, Beidou) of the vehicle, instantaneous speed;

[0041] The set navigation destination and expected path intention (straight, left turn, right turn, etc.) of the vehicle;

[0042] The predicted arrival time and historical driving trajectory segment of the vehicle;

[0043] If it is a public vehicle or a logistics vehicle, it can also include its task priority information.

[0044] Based on the above multi-source data, the module models the urban traffic system as a heterogeneous graph structure, wherein:

[0045] The nodes represent various intersections, and each node is associated with its current traffic state vector;

[0046] The edges represent the road connection relationship between adjacent intersections, and the edge attributes can include road length, traffic capacity, average travel time, historical congestion coefficient, etc.;

[0047] If necessary, vehicle intention path prediction results can also be introduced to construct traffic flow prediction channel weights on the graph structure to assist path-level signal scheduling modeling.

[0048] By establishing the urban road network graph model, the system can perform traffic state propagation analysis based on spatial topology, strategy game modeling and signal timing strategy generation on the graph structure, significantly improving the regional coordination and overall traffic efficiency of traffic light control strategies.

[0049] In a specific implementation example, taking a certain urban intersection "Intersection A" as an example, the intersection is deployed with an edge computing unit and a multi-modal perception device:

[0050] The traffic perception system detects that the eastbound lane queue exceeds 80 meters, with an average waiting time of 90 seconds;

[0051] The path intention information uploaded by the vehicle terminal indicates that about 60% of the vehicles plan to continue straight after crossing the intersection to go to Intersection B, and there is a short-term bottleneck on the road;

[0052] The system accordingly sets the edge between A and B in the graph model as a high-weight traffic path, and accordingly allocates higher priority to the green light timing for this direction in the game model;

[0053] Thus, intelligent timing control is realized under the support of path-level and graph structure.

[0054] The fuzzy game strategy generation module is configured to construct a set of traffic light timing strategies based on real-time traffic state data of the intersection, define a fuzzy utility function for each candidate traffic light timing strategy, determine vehicle intention based on vehicle dynamic data, determine the fuzzy factor of the fuzzy utility function based on the vehicle intention, and weight the fuzzy membership degree of the traffic light timing strategy.

[0055] The fuzzy game strategy generation module specifically includes:

[0056] The strategy candidate generation submodule:

[0057] This submodule automatically generates a set of feasible traffic light timing strategies based on the topological structure of the intersection (such as three-phase, four-phase, and five-phase intersections), traffic flow data, and historical timing parameters, denoted as:

[0058] A = {a1, a2, …, an} m}

[0059] where each strategy a m The following information elements are defined:

[0060] T i : total signal cycle length (unit: seconds);

[0061] g ij : green time of the jth traffic phase in strategy a i

[0062] r ij : red time;

[0063] y ij : yellow time;

[0064] s ij : adjacent phase interval time (e.g. safety transition time between right-turn green and opposite straight green);

[0065] φ i : execution order of phases in this strategy (e.g. "straight→left→right").

[0066] The generation of strategy set can be based on the following methods:

[0067] Manual setting: based on the experience of managers;

[0068] Rule base method: predefine experience timing strategies (e.g. "balanced scheme", "straight priority scheme", "high flow priority scheme", etc.);

[0069] Adaptive generation method: dynamically combine strategies according to the current cycle traffic state (e.g. extend the green time of the phase with long queue);

[0070] Genetic / reinforcement learning method: learn optimal strategy combination from historical timing results.

[0071] The specific strategy is not the focus of the present application, and those skilled in the art can use any of the above methods to implement the strategy set and update it appropriately.

[0072] All strategies must satisfy the road traffic safety constraints:

[0073] g ij ≥ g min (minimum green time should not be less than the relevant provisions);

[0074] g ij + y ij + s ij ≤ T​i

[0075] The conflict phase cannot be released at the same time.

[0076] Fuzzy utility index construction and membership function design submodule:

[0077] To evaluate the effect of candidate strategy a i , we need to define the fuzzy utility function μ i to describe the degree of adaptation of the strategy to the current traffic state.

[0078] Specifically, the fuzzy evaluation index includes:

[0079] Average waiting time

[0080] Maximum queue length Q i = max j (q j )

[0081] Traffic flow rate

[0082] Direction saturation

[0083] Where f ij is the phase flow, and c ij is the maximum release capacity of the channel.

[0084] Map each index through the fuzzy membership function to [0, 1] (traffic flow rate and direction saturation do not need to be mapped):

[0085] For the average waiting time

[0086]

[0087] For the maximum queue length

[0088]

[0089] Then combine into a fuzzy utility vector:

[0090]

[0091] Vehicle intention-driven fuzzy factor extraction submodule:

[0092] This submodule is responsible for statistical modeling and proportional modeling of path intention information uploaded by the vehicle terminal, and then generating direction weighted fuzzy factors.

[0093] The intersection has 3 types of directions: left turn (L), straight (S), and right turn (R). In this period:

[0094] n L ,n S ,n R : number of vehicles in three directions;

[0095] N = n L + n S + n R : total number of vehicles;

[0096] Then the intention ratio is:

[0097]

[0098] To improve the weight of the main traffic direction in fuzzy evaluation, introduce fuzzy factor θ k , defined as:

[0099] The basic weight is 1, and the maximum enhancement coefficient is κ, where 1.0 ≤ κ ≤ 2.0

[0100] A specific calculation example:

[0101] p L = 0.2, p S = 0.6, p R = 0.2

[0102] κ = 1.5

[0103] θ S = 1.5, θ L = θ R ≈ 1.133

[0104] Fuzzy utility function weighted integration and strategy score sub-module:

[0105] Used to combine the direction weighting factor and fuzzy index value, update the weighted fuzzy utility function of strategy a i :

[0106]

[0107] Where:

[0108] μ jk (a i ) represents the membership value of the jth fuzzy index in the kth direction;

[0109] w jk represents the weight in direction k;

[0110] θ k represents the direction fuzzy factor (vehicle intention decision).

[0111] Further, it can be simplified to matrix form for fast parallel calculation.

[0112] All strategy final scores Sequence of construction, for sorting screening.

[0113] In a specific calculation example:

[0114] Current cycle statistics: 50 left turns, 200 straight, 50 right turns;

[0115] The intention distribution ratio is: p L = 0.167, p S = 0.667, p R = 0.167;

[0116] Let the maximum fuzzy weight promotion coefficient κ = 1.5;

[0117] θ S = 1.5, θ L = θ R ≈ 1.125

[0118] Fuzzy scores are given to the three strategies, assuming the membership vectors are as follows:

[0119] Strategy number μ W ]]> μ Q ]]> μ T ]]> μ S ]]> μ final ]]> a1 0.6 0.5 0.7 0.6 0.65 a2 0.7 0.4 0.8 0.75 0.73 (highest) a3 0.6 0.6 0.6 0.5 0.62

[0120] The final strategy a2 is pushed to the fuzzy game solving module to participate in the Nash equilibrium iteration process.

[0121] Game optimization and traffic light control module, for calling fuzzy game solving mechanism according to the adjacency relationship between intersections in the city road network graph model and the change range of traffic light timing parameters, obtaining the optimal traffic light control strategy that meets the fuzzy Nash equilibrium condition, and outputting the adjustment parameters of traffic light duration for the current cycle.

[0122] This module receives the following inputs:

[0123] City traffic road network graph G = (V, E) from the traffic information perception and graph modeling module

[0124] The candidate traffic light control strategy set of each intersection v i

[0125] The fuzzy utility value μ i (a ij ) of each candidate strategy, including the direction weighted fuzzy factor correction;

[0126] Connection information between each node and adjacent nodes, i.e. adjacent set N(v i );

[0127] The traffic state vector s i of each node in the current cycle, including queue length, waiting time, intention distribution, etc.​

[0128] Further fuzzy game environment construction

[0129] Game modeling objects include:

[0130] Each intersection v i ∈V is a game player;

[0131] Each player v i selects a control strategy a i from the strategy set A i in the current control period;

[0132] The strategy combination of all nodes is denoted as a={a1,a2,...,a n};

[0133] The strategy influence relationship network between nodes is established through the graph structure E.

[0134] The joint fuzzy utility function of each node v i is denoted as:

[0135]

[0136] Wherein:

[0137] is the fuzzy utility of node v i for its own strategy a i ;

[0138] is the adjacent coordination penalty function between strategies;

[0139] λ is the coordination factor (0≤λ≤1);

[0140] a N(i) is the current strategy combination of adjacent nodes.

[0141] The adjacent coordination penalty function is determined in the following way according to the adjacent condition:

[0142] Method one, time offset conflict degree

[0143] Let T i (a i ) represent the total control period of strategy a i , if the period difference of two adjacent nodes v i ,v j is too large, it is easy to cause vehicle crosstalk, then:

[0144] Ψ1(a i ,a j )=|T i (a i )-Tj (a j )|

[0145] Method two: green wave band misalignment penalty

[0146] Let δ ij be the recommended green wave band propagation delay from v i to v j , if v j does not open the traffic phase on time, causing queue accumulation, then

[0147] Ψ2(a i ,a j ) = |green start time j - (green end time i + δ ij )|

[0148] After determining the above conditions, the fuzzy Nash equilibrium strategy mechanism is solved, and the purpose is to find the strategy combination

[0149]

[0150] such that for any node v i :

[0151]

[0152] For each node v i , the initial selection is the strategy with the highest fuzzy utility.

[0153] Optimal response updates are performed for all nodes in turn:

[0154]

[0155] All node strategies are stable, that is:

[0156] a (t+1) = a (t)

[0157] Converge to an approximate fuzzy Nash equilibrium when

[0158] Once the optimal strategy combination a * is determined, the system will:

[0159] Analyze the intersection strategy corresponding parameter set, including:

[0160] Each traffic phase green time;

[0161] Phase release order;

[0162] Total cycle length;

[0163] Green wave band time offset of adjacent intersections.

[0164] At the same time, the parameters are verified for boundary legality:

[0165] The green light must not be shorter than the minimum safe travel time;

[0166] Adjacent phases are not in conflict;

[0167] The cycle must not exceed the system maximum cycle limit (such as 180s);

[0168] Finally, the specific adjustment parameters are obtained.

[0169] A vehicle intention analysis and path guidance module is configured to receive vehicle dynamic data uploaded by a vehicle terminal, real-time traffic state data and game solving results, identify target vehicles that can still reach the destination as needed after changing the path intention, generate a detour suggestion for path guidance based on the game solving results, and push the suggestion to the corresponding vehicle.

[0170] The unit is configured to receive vehicle dynamic data uploaded by a vehicle terminal in real time through a V2I communication link, including but not limited to the following fields:

[0171] p v (t): current position coordinates of the vehicle (GPS);

[0172] v v (t): instantaneous speed of the vehicle;

[0173] θ v (t): heading angle of the vehicle;

[0174] Destination coordinates (obtained through a navigation system);

[0175] π v (t): vehicle path planning trajectory (which can be represented as a sequence of intersections);

[0176] Expected arrival time at each intersection;

[0177] κ v : vehicle category (such as ordinary car, bus, rescue vehicle, etc.).

[0178] On the basis of the above data, the unit combines with the city road network graph G=(V, E) to identify the vehicle path intention:

[0179] Extract the set of vehicles V i that plan to pass through the intersection v i in the current cycle;

[0180] For each vehicle v, according to its path trajectory v (t) derive its passing direction intention at intersection v i :

[0181] σ v (v i ) ∈ {left turn, straight, right turn}

[0182] The above results are used to support fuzzy game strategy weighting and subsequent guidance judgment.

[0183] To avoid pushing path adjustment suggestions to all vehicles, the system only identifies vehicles that can be guided to change the passing path but can still successfully reach the original destination, forming a target vehicle set V target The specific method is as follows:

[0184] For each vehicle v, if its original path trajectory is:

[0185]

[0186] The system searches all feasible alternative path sets from the current position p v (t):

[0187]

[0188] And calculates the estimated cost (such as predicted time, traffic load, etc.) of each path:

[0189]

[0190] If there is satisfying:

[0191]

[0192] And the traffic light strategy of the intersection contained in the current period has better passing conditions (such as high green light release probability), the vehicle is included in the target vehicle set.

[0193] Based on the game optimization and the optimal traffic light timing strategy combination output by the traffic light control module in the current period

[0194]

[0195] Select the path scheme with the maximum passing guarantee degree from the alternative paths.

[0196] The calculation method is as follows:

[0197] For each path , count the control strategy coordination degree of the intersection it passes through:

[0198]

[0199] The ranking index is constructed by comprehensively considering the path cost and the coordination degree:

[0200]

[0201] The path with the highest index value is selected as the recommended path

[0202] The system generates a structured path adjustment suggestion for each target vehicle, including:

[0203] Recommended path trajectory (intersection sequence or navigation route segment);

[0204] Estimated time saving;

[0205] Comparison and analysis of the current traffic direction and the recommended direction;

[0206] Whether there is a green wave passing interval;

[0207] Behavioral suggestions (such as "maintain current speed, change lane into right turn lane after 3 seconds").

[0208] Specific examples are:

[0209] During the morning rush hour in a city, intersections A, B, and C form a congestion bottleneck on a main road. The original path of the vehicle is A→B→C→D, and the estimated time is 17 minutes.

[0210] Traffic game results show that the traffic light release time at C is insufficient, which is prone to queue accumulation;

[0211] The alternative path A→E→F→D can pass through multiple green wave release intersections in the current period;

[0212] The system judges that C(π alt ) = 16.5 min ≤ 17 + ε;

[0213] Guide the vehicle to turn right into E road in advance to avoid the bottleneck at C intersection.

[0214] Finally, the system pushes the path adjustment suggestion to the vehicle and prompts it: "It is recommended to turn right in advance to avoid the high congestion intersection C on E road, and it is estimated to save 1.2 minutes."

[0215] Traffic light instruction execution module, for receiving the traffic light adjustment parameters output by the game optimization and traffic light control module, and issuing the adjustment parameters to the traffic light controllers of each intersection to realize the update of the traffic light state.

[0216] This module mainly includes the following functions:

[0217] I. Control parameter receiving function

[0218] The module receives traffic light timing parameters output by upstream modules of the system, which include:

[0219] total duration of the current control cycle;

[0220] green light time, yellow light time and red light time of each traffic phase;

[0221] the order of each phase;

[0222] optional offset time, coordination control information, etc.

[0223] These parameters are generated by the system according to traffic flow state, vehicle intention information and multi-intersection coordination strategy, and represent the optimal traffic light release scheme for the current period.

[0224] II. Instruction conversion and formatting function

[0225] In order to ensure that the timing strategy can be correctly recognized and executed by the actual traffic light controller, the module analyzes and formats the received control parameters to generate standard control instructions. The instruction format can be adapted according to the brand and protocol requirements of the traffic controller, and common formats include structured data files (such as JSON or XML) or control code instruction sets.

[0226] The module also checks the legality of the parameters, such as whether the green light time meets the minimum release requirement, whether the total cycle time exceeds the upper limit of the controller, etc.

[0227] III. Communication delivery function

[0228] The module sends the formatted control instructions to the traffic light controller deployed at the intersection through the communication interface. The communication method can be serial communication, Ethernet communication, 4G / 5G cellular communication or Internet of Things communication, etc., which is determined according to the controller model and deployment environment.

[0229] The module supports batch delivery, timed delivery, multicast delivery, etc. to meet the control needs of different scale road networks.

[0230] In another embodiment, the present application also provides a traffic light signal control method based on car-road cooperation, which uses the aforementioned embodiment to control the traffic light signal.

[0231] It should be noted that the explanation and description of the aforementioned traffic light signal control system based on car-road cooperation embodiment also applies to the method of the present application, which will not be repeated here.

[0232] Those skilled in the art can clearly understand that the units and algorithm steps described in the embodiments disclosed herein can be realized by electronic hardware, computer software and a combination of the two. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0233] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0234] In several embodiments provided in the present application, any function realized in the form of a software function unit and sold or used as an independent product can be stored in a computer-readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory; hereinafter referred to as: ROM), a random access memory (Random Access Memory; hereinafter referred to as: RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0235] The above is only a specific implementation of the present application. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. The parts of the present application not specifically mentioned should be subject to the content recorded in the prior art. The prior art mentioned in the foregoing background section and the specific embodiment section of the present application can be used as a part of the present application to understand the meaning of some technical features or parameters.

Claims

1. A traffic light signal control system based on vehicle infrastructure integration, characterized by, The system comprises the following modules: A traffic information perception and graph modeling module is configured to deploy edge control units at multiple intersections, construct a city road network graph model taking intersections as nodes and roads as edges, collect real-time traffic state data of each intersection and vehicle dynamic data uploaded by a vehicle terminal, and the like. A fuzzy game strategy generation module is configured to construct a traffic light timing strategy set based on the real-time traffic state data of the intersection, define a fuzzy utility function for each candidate traffic light timing strategy, determine a vehicle intention based on the vehicle dynamic data, determine a fuzzy factor of the fuzzy utility function based on the vehicle intention, and weight fuzzy membership degrees of the traffic light timing strategies. A game optimization and traffic light regulation module is configured to call a fuzzy game solving mechanism based on an adjacency relationship between intersections in the city road network graph model and a change range of traffic light timing parameters, obtain an optimal traffic light control strategy satisfying a fuzzy Nash equilibrium condition, and output an adjustment parameter of a traffic light duration for a current period. A vehicle intention analysis and path guidance module is configured to receive vehicle dynamic data uploaded by the vehicle terminal, real-time traffic state data, and game solving results, identify target vehicles that can still reach a destination as needed after changing a path intention, generate a rerouting suggestion for path guidance based on the game solving results, and push the suggestion to corresponding vehicles. A traffic light instruction execution module is configured to receive the traffic light adjustment parameter output by the game optimization and traffic light regulation module, and distribute the adjustment parameter to traffic light controllers of each intersection to update a traffic light state. The determination of the fuzzy factor of the fuzzy utility function based on the vehicle intention and the weighting of the fuzzy membership degrees of the traffic light timing strategies specifically comprises the following steps: An intersection has three types of directions: left turn, straight, and right turn, and in the current period: : number of three-way vehicles; : total number of vehicles; The intention proportion is: ， Introducing a fuzz factor defined as: The base weight is 1 and the maximum enhancement coefficient is wherein, , The direction weighting factor is combined with the fuzzy index value, and the strategy is updated The weighted fuzzy utility function of the strategy is: ， Wherein: representing the first fuzzy index in the direction of membership value of the first fuzzy index in the direction of representing a weight in the direction on the direction denotes the direction ambiguity factor; Further, it can be simplified into a matrix form for fast parallel calculation. All strategy final scores Constitute sequence, for ordering screening. 2.The vehicle infrastructure integration based traffic light signal control system according to claim 1, wherein, The traffic information perception and graph modeling module is configured to fuse the collected intersection traffic state data and dynamic information of vehicle position, speed, and path intention uploaded by the vehicle terminal, construct a city road network graph model having node traffic attributes and edge path attributes, and the like. 3.The vehicle infrastructure integration based traffic light signal control system according to claim 1, wherein, The fuzzy game strategy generation module is configured to generate multiple candidate traffic light timing strategies based on a traffic state vector of each intersection, historical data, and control rules, the strategies including a signal period, a passing phase sequence, a green light duration of each phase, and a yellow light warning duration, and construct a fuzzy utility function for each strategy, the fuzzy utility function scoring average waiting time, queue length, and passing capacity indicators through a preset membership function to reflect strategy passing adaptability. 4.The vehicle infrastructure integration based traffic light signal control system according to claim 3, wherein, The fuzzy game strategy generation module is further configured to: according to path intention information carried in the vehicle dynamic data, count vehicle proportion of each passing direction in the current period of the intersection, and construct a direction intention fuzzy factor according to the same, and use the fuzzy factor to weight and correct a direction passing membership value in the fuzzy utility function. 5.The vehicle infrastructure integration based traffic light signal control system according to claim 1, wherein, The game optimization and traffic light regulation module is configured to: based on an adjacency relationship between intersection nodes in a city road network map, construct a fuzzy game model, the model taking the intersection as a participant and taking a candidate timing strategy as a strategy set, calculating a coordination penalty function between a local utility of the intersection and an adjacent strategy in each round of game, and solving an optimal control strategy combination satisfying a fuzzy Nash equilibrium through an optimal response algorithm. 6.The vehicle infrastructure integration based traffic light signal control system according to claim 5, wherein, The coordination penalty function is based on signal period difference between adjacent intersections, phase green light start and end time offset, and road segment green wave band continuity, and constructs a coordination degree index for evaluating the coordination of the strategy combination in a regional range. 7.The vehicle infrastructure integration based traffic light signal control system according to claim 1, wherein, The vehicle intention analysis and path guidance module is configured to: after identifying a vehicle having a possibility of changing path intention, estimating a passing cost of a candidate path of the vehicle, the estimation index including path length, traffic light passing release probability, average delay, and path congestion index, and combining a signal game result to evaluate matching degrees of each path, so as to select an optimal path for guidance. 8.The vehicle infrastructure integration based traffic light signal control system as claimed in claim 1, wherein, The path guidance suggestion includes a sequence of alternative paths, a predicted time saving, a recommended entering intersection and a corresponding lane change prompt, and is sent to a vehicle terminal through a V2I communication interface in a structured prompt information, the terminal presenting the suggestion to a driver in a navigation prompt, a screen pop-up window or a voice broadcast mode, for assisting vehicle dynamic decision-making. 9.The vehicle infrastructure integration based traffic light signal control system according to claim 1, wherein, The traffic light instruction execution module is configured to: perform format conversion on traffic light time length adjustment parameters output by the game optimization and traffic light regulation module, generate a structured control instruction including a total time length of a signal period, a green light time length of each traffic phase, a yellow light warning time length, and a phase sequence, and perform protocol adaptation based on a deployed traffic light controller communication protocol, and issue the same to a traffic light controller of a corresponding intersection in a serial communication, Ethernet communication or cellular network communication mode; The traffic light controller has a state feedback mechanism for returning instruction execution state, current phase running state or fault information, and the instruction execution module triggers an emergency processing or reissue process according to the feedback result. 10.A traffic light signal control method based on vehicle infrastructure integration, characterized in that, Using the traffic light signal control system based on vehicle-road cooperation according to any one of claims 1-9 for traffic light signal control. Using the traffic light signal control system based on vehicle-road cooperation according to any one of claims 1-9 for traffic light signal control.

Images