Traffic information real-time path planning system and method for vehicle-road cooperation

Abstract

The invention discloses a traffic information real-time path planning system and method for vehicle-road cooperation. Comprising a vehicle-road cooperative perception and information fusion module, a vehicle-road cooperative communication and information interaction module, a dynamic traffic state analysis and prediction module, a multi-target real-time path planning module, a path dynamic correction and obstacle avoidance decision module and a vehicle control and execution feedback module. According to the method, vehicle-mounted and roadside data are fused through vehicle-road collaborative sensing, a non-line-of-sight blind area is covered, the limitation of single vehicle sensing is solved, and the integrity of environment information is improved; based on a short-term traffic prediction and real-time correction mechanism, path planning can be dynamically adjusted along with traffic state changes, and the problem that a planned path is disjointed from the reality is avoided.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation and vehicle-road cooperation technology, specifically to a real-time traffic information path planning system and method for vehicle-road cooperation. Background Technology

[0002] With the development of intelligent transportation technology, autonomous vehicle navigation places higher demands on the real-time performance, dynamic adaptability, and multi-objective optimization of path planning. Existing path planning technologies largely rely on onboard sensors to independently acquire environmental information, which presents the following problems:

[0003] The limited perception range of a single vehicle makes it impossible to obtain real-time traffic conditions in non-line-of-sight areas, which can lead to the planned route becoming disconnected from actual traffic.

[0004] The V2X communication capability was not fully utilized, and key information such as traffic signal timing and road segment capacity collected by roadside equipment was not deeply involved in route decision-making, resulting in a lack of global optimality in the planning results.

[0005] Traditional path planning relies solely on individual vehicle obstacle avoidance adjustments after path generation, without incorporating real-time traffic changes from the roadside for global path replanning, which can easily lead to path planning failure.

[0006] Existing solutions often prioritize the shortest possible time, failing to consider multiple dimensions such as vehicle energy consumption and driving safety, making them unsuitable for the personalized needs of different scenarios.

[0007] Therefore, there is an urgent need for a real-time traffic information path planning system and method oriented towards vehicle-road cooperation to solve the problems mentioned above. Summary of the Invention

[0008] The purpose of this invention is to provide a real-time route planning system and method for vehicle-road cooperative traffic information to solve the problems existing in the prior art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a real-time traffic information path planning system for vehicle-road cooperation, comprising a vehicle-road cooperative perception and information fusion module, a vehicle-road cooperative communication and information interaction module, a dynamic traffic state analysis and prediction module, a multi-objective real-time path planning module, a path dynamic correction and obstacle avoidance decision-making module, and a vehicle control and execution feedback module.

[0010] The vehicle-road cooperative perception and information fusion module is used to acquire multi-source perception data from vehicles and roadsides. Through multi-source heterogeneous data fusion technology, it outputs unified three-dimensional environmental information on dynamic obstacles, traffic signals and road conditions, providing a basis for subsequent decision-making.

[0011] The vehicle-road cooperative communication and information interaction module is based on the V2X communication protocol to realize two-way information interaction between the on-board unit (OBU), the roadside unit (RSU), and the cloud platform, and transmit perception data, traffic status, route instructions and execution feedback to ensure the real-time and reliability of information.

[0012] The dynamic traffic state analysis and prediction module calculates the traffic flow state index of the current road segment based on the fused traffic information, and outputs the traffic flow change trend for the next 3-5 minutes through a short-term prediction model, providing a dynamic basis for route planning.

[0013] The multi-objective real-time path planning module takes time, energy consumption, and safety as objectives, constructs a comprehensive path cost model, and uses an improved dynamic weighted A* algorithm to generate the globally optimal path, satisfying vehicle kinematic constraints and user personalized needs.

[0014] The path dynamic correction and obstacle avoidance decision module monitors changes in traffic conditions and sudden obstacles in real time, triggers the path replanning mechanism, and generates obstacle avoidance decisions in combination with vehicle-road cooperative information to ensure that the path remains effective.

[0015] The vehicle control and execution feedback module converts the planned path into vehicle control commands (accelerator, brake, steering), adjusts the control parameters through closed-loop feedback, and transmits the execution status back to the roadside and the cloud, forming a vehicle-road cooperative closed loop.

[0016] Preferably, the vehicle-road cooperative perception and information fusion module includes an on-board perception unit, a roadside perception unit, and a multi-source information fusion unit;

[0017] Vehicle-mounted perception unit: Composed of vehicle-mounted camera, millimeter-wave radar and lidar, which respectively collect environmental images (identify pedestrians, non-motorized vehicles and traffic markings), obstacle distance and speed (identify dynamic vehicles and static obstacles), and high-precision 3D point cloud (construct local environmental contours) to achieve 360° environmental perception around the vehicle.

[0018] Roadside sensing unit: Composed of roadside high-definition camera, roadside radar, traffic signal controller, and road surface condition monitor, it collects traffic flow at intersections (number of vehicles, direction of travel), traffic signal phase and remaining time, road surface smoothness / water accumulation / icing status, and covers the non-line-of-sight blind spots of vehicle-mounted sensing;

[0019] Multi-source information fusion unit: Employs a three-level process of preprocessing, spatiotemporal calibration, and feature layer fusion.

[0020] Preprocessing: Denoising the vehicle-mounted and roadside data (such as removing outliers from radar point clouds and Gaussian filtering of images) and unifying the data format (converting radar data to Cartesian coordinates and image data to pixel coordinates).

[0021] Spatiotemporal calibration: Data time synchronization is achieved based on GPS timestamps, and spatial calibration is achieved based on a preset coordinate system transformation matrix;

[0022] Feature layer fusion: The DS evidence theory is used to fuse the obstacle recognition results of multiple sensors (such as the "pedestrian" identified by the camera and the "dynamic target" identified by the radar for feature matching to determine the target type and attributes). The output includes a dynamic obstacle list containing "target ID-location-speed-type", a traffic signal list containing "intersection ID-signal phase-remaining time", and a road information list containing "segment ID-road condition-traffic restriction".

[0023] Preferably, the vehicle-road cooperative communication and information interaction module includes a V2X communication unit, an information interaction unit, and a communication support unit;

[0024] V2X communication unit: Supports LTE-V2X / 5G-V2X dual-mode communication, enabling three types of interaction:

[0025] V2I (Vehicle-to-Infrastructure): Transmission of vehicle-mounted sensing data, roadside sensing data, and traffic signal commands between the vehicle-mounted OBU and the roadside RSU;

[0026] V2V vehicle-to-vehicle: Transmission of vehicle status and driving intentions between adjacent vehicle OBUs;

[0027] V2C Vehicle-to-Cloud: The vehicle-mounted OBU transmits global traffic status, historical route data, and user preference settings between itself and the cloud platform, using cellular networks as backup to ensure long-distance information transmission;

[0028] Information exchange unit: Defines a standardized data exchange format, including:

[0029] Sensing data frame: contains "sensor ID - acquisition timestamp - target list";

[0030] Traffic status frame: contains "segment ID - timestamp - traffic flow parameters";

[0031] Path instruction frame: contains "path ID - list of waypoints - control parameters";

[0032] Communication support unit: Employs data verification, retransmission mechanisms, and fault tolerance processing to ensure information reliability.

[0033] Data verification: CRC32 cyclic redundancy check is used to detect data transmission errors;

[0034] Retransmission mechanism: When the receiving end detects data errors or loss, it triggers the sending end to retransmit via ACK / NACK feedback;

[0035] Fault tolerance: When V2I communication is interrupted, it automatically switches to V2V communication to obtain perception data from surrounding vehicles as a temporary supplement to environmental information.

[0036] Preferably, the dynamic traffic state analysis and prediction module includes a traffic state analysis unit and a short-term prediction unit;

[0037] Traffic State Analysis Unit: Based on the fused traffic flow data, a Traffic Flow State Index (SI) is constructed to assess the current road segment status. The formula for calculating the Traffic Flow State Index is:

[0038] ;

[0039] Where SI: Traffic Flow State Index, with a value range of [0,1]. The closer SI is to 1, the more congested the traffic is; the closer SI is to 0, the smoother the traffic is. , , Weighting coefficients, satisfying Adjustments will be made dynamically based on the type of road segment. The actual traffic flow density of a road segment is obtained by dividing the number of vehicles in the segment as counted by roadside radar by the length of the segment. : Maximum traffic flow density of the road segment, which is a design parameter for the road segment; The actual average vehicle speed on the road segment is obtained by averaging the speeds of all vehicles collected by roadside equipment. Free-flow speed on the road segment is the speed limit for that segment. The actual traffic flow of a road segment is obtained by counting the number of vehicles passing through the road segment section per unit time, as recorded by roadside cameras. Traffic capacity of a road segment is a design parameter for that segment. Traffic conditions are divided into four levels based on the SI value: Smooth traffic (SI < 0.2), Basically smooth traffic (0.2 ≤ SI < 0.4), Lightly congested traffic (0.4 ≤ SI < 0.6), and Severely congested traffic (SI ≥ 0.6).

[0040] Short-term prediction unit: Using a GRU gated recurrent unit neural network model, based on historical traffic data, real-time traffic data and influencing factors, it predicts the SI value change trend of each road segment in the next 3-5 minutes and outputs the prediction result of "road segment ID - prediction time - prediction SI value", providing dynamic traffic basis for route planning.

[0041] Preferably, the multi-objective real-time path planning module includes a multi-objective weight determination unit, a path comprehensive cost calculation unit, and a dynamically weighted A* path search unit;

[0042] Multi-objective weight determination unit: Dynamically adjusts the weights of time, energy consumption, and safety objectives based on user needs and scenarios.

[0043] Time weight Commuting scenarios Take 0.5, for leisure scenarios. Take 0.3;

[0044] Energy consumption weight Electric vehicle scenario Take 0.4, for gasoline vehicle scenarios Take 0.2;

[0045] Safety weight Rainy or snowy weather Take 0.4, sunny day Take 0.2; and satisfy ;

[0046] Route integrated cost calculation unit: Constructs a route integrated cost model; the route integrated cost calculation formula is as follows:

[0047] ;

[0048] in, : Overall path cost; the smaller the value, the better the path. : Path time cost, which is the length of each segment of the path. Divide by the predicted average speed of the road segment The sum, that is , This represents the number of road segments contained in the path. : Path energy consumption cost, calculated based on the vehicle energy consumption model, i.e. , , , The vehicle energy consumption coefficient is determined by the vehicle model. The road section slope is provided by roadside equipment; Path safety cost is the sum of the reciprocals of the safe distances between each segment of the path and obstacles, i.e. , The minimum safe distance between vehicles and obstacles within a road segment is provided by the vehicle-road cooperative perception and information fusion module. ≥5m;

[0049] Dynamically Weighted A* Path Search Unit: This unit introduces a dynamic heuristic function into the traditional A* algorithm. Dynamic adjustments based on predicted traffic conditions:

[0050] ;

[0051] in, For nodes The straight-line distance to the target point. For nodes Estimated energy consumption to reach the target point For maximum allowable energy consumption, For nodes The minimum safe distance to the target point; during the algorithm search process, priority is given to selecting the comprehensive cost. The path with the minimum value that satisfies the vehicle's kinematic constraints.

[0052] Preferably, the path dynamic correction and obstacle avoidance decision module includes a correction triggering unit, a path replanning unit, and an obstacle avoidance decision unit;

[0053] Correction trigger unit: Real-time monitoring of three types of trigger conditions; correction is initiated when any one condition is met.

[0054] Sudden changes in traffic conditions: The deviation between the predicted SI value and the actual SI value of a road segment is ≥0.2;

[0055] Sudden Obstacles: The perception module detects unpredictable static or dynamic obstacles within the path range, and the safe distance is [not specified]. <3m;

[0056] Traffic signal change: The traffic signal phase fed back by the roadside equipment deviates from the expected phase during route planning by ≥1 cycle (e.g., the planned green light is actually a red light when arriving at the intersection).

[0057] Route replanning unit: After the correction is initiated, based on the current traffic conditions and prediction results, the process of "multi-objective weight determination - comprehensive cost calculation - dynamic weighted A* search" is re-executed to generate a new route; during the replanning process, the target points and key waypoints of the original route are retained, and only local road segments are adjusted to reduce the amount of calculation;

[0058] Obstacle Avoidance Decision Unit: Generates obstacle avoidance behavior decisions by combining vehicle-road cooperative information, using a behavior decision tree model.

[0059] Step 1: Determine the type (static / dynamic) and location (front / side) of the obstacle;

[0060] Step 2: Obtain the driving intentions of surrounding vehicles based on V2V information;

[0061] Step 3: Generate decision options (lane change to avoid obstacle, slow down and wait, detour to avoid obstacle), and calculate the safety cost and time cost of each option;

[0062] Step 4: Select the decision with the lowest overall cost. For example, when there is a static obstacle ahead and no intention of changing lanes from vehicles in adjacent lanes, execute "lane change to avoid obstacle"; when there is a dynamic obstacle ahead and congestion in adjacent lanes, execute "decelerate and wait".

[0063] Preferably, the vehicle control and execution feedback module includes a control algorithm unit, an execution unit, and a feedback unit;

[0064] Control Algorithm Unit: Employs Model Predictive Control (MPC) algorithm to convert the planned path into control commands.

[0065] Path tracking: Using the desired trajectory of the path as a reference, which includes position, speed, and heading angle, calculate the deviation between the vehicle's actual trajectory and the desired trajectory, including the position deviation. Heading deviation ;

[0066] Control quantity calculation: Based on MPC-based rolling time-domain optimization, output throttle opening 0-100%, brake pedal travel 0-100%, and steering angle -30° to 30° to ensure deviation ≤0.5m, ≤5°;

[0067] Speed ​​adaptation: Combine roadside traffic signals and road section speed limits to dynamically adjust the desired speed (e.g., reduce to 30km / h when approaching an intersection, and increase to 60km / h on unobstructed sections).

[0068] Execution unit: Transforms control commands into actions of the vehicle actuators.

[0069] Accelerator or brake: The accelerator opening and brake travel commands are sent to the engine controller ECU and brake controller EBCU via the CAN bus to control the vehicle's acceleration or deceleration.

[0070] Steering: Sends steering angle commands to the Electronic Power Steering (EPS) system to control vehicle steering;

[0071] Feedback Unit: Real-time acquisition of vehicle execution status, including actual speed, position, and steering angle, is transmitted back to roadside equipment and the cloud platform via V2I / V2C communication. This is used to: update the actual traffic flow data of the traffic state analysis module; and verify the effectiveness of route planning. If the execution deviation continues to exceed a threshold... If the path length is greater than 1m, a path correction will be triggered; historical path data will be optimized to provide empirical parameters for subsequent planning.

[0072] The planning method for a real-time traffic information route planning system oriented towards vehicle-road cooperation includes the following steps:

[0073] Step 1: System Initialization and Sensor Startup

[0074] 1.1 After the system is powered on, the on-board unit (OBU) and the roadside unit (RSU) complete hardware calibration: the on-board sensors perform internal parameter calibration, and the roadside unit completes GPS positioning and coordinate system initialization;

[0075] 1.2 The vehicle-road cooperative communication and information interaction module starts V2X communication, establishes the connection between the on-board OBU, the roadside RSU, and the cloud platform, and completes the communication link test;

[0076] 1.3 The vehicle-road cooperative perception and information fusion module starts the on-board and roadside sensors to collect initial environmental data, outputs the initial environmental information list through the multi-source information fusion unit, and the system enters the standby state;

[0077] Step 2: Vehicle-to-Infrastructure Communication Establishment and Information Exchange

[0078] 2.1 Receive user navigation commands and parse the starting coordinates Target point coordinates With user preferences;

[0079] 2.2 The vehicle-mounted OBU sends "origin-destination-user preference" information to the roadside RSU via V2I communication, and the roadside RSU feeds back "segment list-roadside sensing data-traffic signal timing" covering the origin to the destination.

[0080] 2.3 The vehicle-mounted OBU requests historical traffic data and influencing factors from the cloud platform via V2C communication, and the cloud platform returns the data in real time;

[0081] 2.4 The communication protection unit performs CRC32 verification on the received data. If a data error is detected, a retransmission mechanism is triggered to ensure information integrity.

[0082] Step 3: Dynamic Traffic Condition Analysis and Prediction

[0083] 3.1 The dynamic traffic state analysis and prediction module receives the fused real-time traffic data and, based on the formula... Calculate the current SI value for each road segment and classify the traffic status level;

[0084] 3.2 The short-term prediction unit loads the GRU neural network model, inputs "historical SI sequence + real-time SI sequence + influencing factors", trains the model parameters, and predicts the SI value of each road segment in the next 3-5 minutes;

[0085] 3.3 Output a traffic status report consisting of "segment ID - current SI - predicted SI - traffic status level" and send it to the multi-objective real-time path planning module;

[0086] Step 4: Multi-objective real-time path planning

[0087] 4.1 The multi-objective weight determination unit adjusts weights according to user preferences: If the user selects "priority time", then =0.5, =0.2, =0.3; if "Prioritize Energy Consumption" is selected, then =0.3, =0.4, =0.3;

[0088] 4.2 The route integrated cost calculation unit calculates the cost of each candidate route based on the predicted SI value from the traffic condition report. , , Through formula Calculate the overall cost of the path ;

[0089] 4.3 The dynamically weighted A* path search unit uses the starting point as the initial node and the target point as the ending node, based on the comprehensive cost C and a dynamic heuristic function. The system searches for the optimal path, generates a path plan containing "waypoint coordinates - desired speed - lane suggestions", and sends it to the path dynamic correction and obstacle avoidance decision module.

[0090] Step 5: Path Dynamic Correction and Obstacle Avoidance Decisions

[0091] 5.1 The correction trigger unit monitors traffic conditions, obstacles, and traffic signals in real time: compares the actual SI value of the road segment with the predicted SI value; if the deviation is ≥0.2, it is marked as "sudden change in traffic condition"; it receives obstacle information from the sensing module; if the safe distance is... <3m, marked as "sudden obstacle"; receive traffic signal change information from roadside equipment, if the phase deviation is ≥1 cycle, mark as "traffic signal change";

[0092] 5.2 If any marker exists, the route replanning unit initiates replanning, recalculates the overall cost based on the current traffic data, and generates a locally corrected new route;

[0093] 5.3 The obstacle avoidance decision unit generates obstacle avoidance decisions based on obstacle type and V2V information through a behavior decision tree, and sends the decision results to the vehicle control and execution feedback module;

[0094] Step 6: Vehicle control execution and feedback reset

[0095] 6.1 The control algorithm unit of the vehicle control and execution feedback module receives the path plan and obstacle avoidance decision, calculates the throttle, brake and steering commands using the MPC algorithm, and sends them to the execution unit via the CAN bus;

[0096] 6.2 The execution unit drives the vehicle to execute commands, collects the actual driving status in real time, and feeds it back to the control algorithm unit to adjust the control parameters to reduce trajectory deviation;

[0097] 6.3 The feedback unit transmits the execution status back to the roadside equipment and cloud platform via V2I / V2C to update traffic status data;

[0098] 6.4 When the vehicle reaches the target point, a navigation completion signal is output, all modules are reset to standby state, and the navigation data is recorded for subsequent model optimization.

[0099] Compared with the prior art, the beneficial effects of the present invention are:

[0100] This invention integrates vehicle-mounted and roadside data through vehicle-road cooperative perception, covering non-line-of-sight blind spots, overcoming the limitations of single-vehicle perception, and improving the completeness of environmental information. Based on short-term traffic prediction and real-time correction mechanisms, path planning can be dynamically adjusted according to changes in traffic conditions, avoiding the problem of planned paths being out of sync with reality. Combining the three objectives of time, energy consumption, and safety, and dynamically adjusting weights according to user preferences, it meets the personalized needs of different scenarios. Through V2X communication, it realizes a closed loop of vehicle-road cooperation in perception, planning, control, and feedback, improving the reliability and safety of path execution. Attached Figure Description

[0101] Figure 1 This is a system module diagram of the present invention;

[0102] Figure 2 This is a schematic diagram of the vehicle-road cooperative perception and information fusion module of the present invention;

[0103] Figure 3 This is a schematic diagram of the vehicle-road cooperative communication and information interaction module of the present invention;

[0104] Figure 4 This is a flowchart of the method of the present invention. Detailed Implementation

[0105] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0106] Please see Figure 1-4 This invention provides a real-time traffic information path planning system for vehicle-road cooperation, including a vehicle-road cooperative perception and information fusion module, a vehicle-road cooperative communication and information interaction module, a dynamic traffic state analysis and prediction module, a multi-objective real-time path planning module, a path dynamic correction and obstacle avoidance decision module, and a vehicle control and execution feedback module.

[0107] The vehicle-road cooperative perception and information fusion module is used to acquire multi-source perception data from vehicles and roadsides. Through multi-source heterogeneous data fusion technology, it outputs unified three-dimensional environmental information on dynamic obstacles, traffic signals and road conditions, providing a basis for subsequent decision-making.

[0108] The vehicle-road cooperative communication and information interaction module is based on the V2X communication protocol to realize two-way information interaction between the on-board unit (OBU), the roadside unit (RSU), and the cloud platform, and transmit perception data, traffic status, route instructions and execution feedback to ensure the real-time and reliability of information.

[0109] The dynamic traffic state analysis and prediction module calculates the traffic flow state index of the current road segment based on the fused traffic information, and outputs the traffic flow change trend for the next 3-5 minutes through a short-term prediction model, providing a dynamic basis for route planning.

[0110] The multi-objective real-time path planning module takes time, energy consumption, and safety as objectives, constructs a comprehensive path cost model, and uses an improved dynamic weighted A* algorithm to generate the globally optimal path, satisfying vehicle kinematic constraints and user-personalized needs.

[0111] The path dynamic correction and obstacle avoidance decision module monitors changes in traffic conditions and sudden obstacles in real time, triggers the path replanning mechanism, and generates obstacle avoidance decisions by combining vehicle-road cooperative information to ensure that the path remains effective.

[0112] The vehicle control and execution feedback module transforms the planned path into vehicle control commands (accelerator, brake, steering), adjusts control parameters through closed-loop feedback, and transmits the execution status back to the roadside and the cloud, forming a vehicle-road cooperative closed loop.

[0113] The vehicle-road cooperative perception and information fusion module includes an on-board perception unit, a roadside perception unit, and a multi-source information fusion unit.

[0114] Vehicle-mounted perception unit: Composed of vehicle-mounted camera, millimeter-wave radar and lidar, which respectively collect environmental images (identify pedestrians, non-motorized vehicles and traffic markings), obstacle distance and speed (identify dynamic vehicles and static obstacles), and high-precision 3D point cloud (construct local environmental contours) to achieve 360° environmental perception around the vehicle.

[0115] Roadside sensing unit: Composed of roadside high-definition camera, roadside radar, traffic signal controller, and road surface condition monitor, it collects traffic flow at intersections (number of vehicles, direction of travel), traffic signal phase and remaining time, road surface smoothness / water accumulation / icing status, and covers the non-line-of-sight blind spots of vehicle-mounted sensing;

[0116] Multi-source information fusion unit: Employs a three-level process of preprocessing, spatiotemporal calibration, and feature layer fusion.

[0117] Preprocessing: Denoising the vehicle-mounted and roadside data (such as removing outliers from radar point clouds and Gaussian filtering of images) and unifying the data format (converting radar data to Cartesian coordinates and image data to pixel coordinates).

[0118] Spatiotemporal calibration: Data time synchronization is achieved based on GPS timestamps, and spatial calibration is achieved based on a preset coordinate system transformation matrix;

[0119] Feature layer fusion: The DS evidence theory is used to fuse the obstacle recognition results of multiple sensors (such as the "pedestrian" identified by the camera and the "dynamic target" identified by the radar for feature matching to determine the target type and attributes). The output includes a dynamic obstacle list containing "target ID-location-speed-type", a traffic signal list containing "intersection ID-signal phase-remaining time", and a road information list containing "segment ID-road condition-traffic restriction".

[0120] The vehicle-road cooperative communication and information interaction module includes a V2X communication unit, an information interaction unit, and a communication support unit.

[0121] V2X communication unit: Supports LTE-V2X / 5G-V2X dual-mode communication, enabling three types of interaction:

[0122] V2I (Vehicle-to-Infrastructure): Transmission of vehicle-mounted sensing data, roadside sensing data, and traffic signal commands between the vehicle-mounted OBU and the roadside RSU;

[0123] V2V vehicle-to-vehicle: Transmission of vehicle status and driving intentions between adjacent vehicle OBUs;

[0124] V2C Vehicle-to-Cloud: The vehicle-mounted OBU transmits global traffic status, historical route data, and user preference settings between itself and the cloud platform, using cellular networks as backup to ensure long-distance information transmission;

[0125] Information exchange unit: Defines a standardized data exchange format, including:

[0126] Sensing data frame: contains "sensor ID - acquisition timestamp - target list";

[0127] Traffic status frame: contains "segment ID - timestamp - traffic flow parameters";

[0128] Path instruction frame: contains "path ID - list of waypoints - control parameters";

[0129] Communication support unit: Employs data verification, retransmission mechanisms, and fault tolerance processing to ensure information reliability.

[0130] Data verification: CRC32 cyclic redundancy check is used to detect data transmission errors;

[0131] Retransmission mechanism: When the receiving end detects data errors or loss, it triggers the sending end to retransmit via ACK / NACK feedback;

[0132] Fault tolerance: When V2I communication is interrupted, it automatically switches to V2V communication to obtain perception data from surrounding vehicles as a temporary supplement to environmental information.

[0133] The dynamic traffic condition analysis and prediction module includes a traffic condition analysis unit and a short-term prediction unit;

[0134] Traffic State Analysis Unit: Based on the fused traffic flow data, a Traffic Flow State Index (SI) is constructed to assess the current road segment status. The formula for calculating the Traffic Flow State Index is:

[0135] ;

[0136] Where SI: Traffic Flow State Index, with a value range of [0,1]. The closer SI is to 1, the more congested the traffic is; the closer SI is to 0, the smoother the traffic is. , , Weighting coefficients, satisfying Adjustments will be made dynamically based on the type of road segment. The actual traffic flow density of a road segment is obtained by dividing the number of vehicles in the segment as counted by roadside radar by the length of the segment. : Maximum traffic flow density of the road segment, which is a design parameter for the road segment; The actual average vehicle speed on the road segment is obtained by averaging the speeds of all vehicles collected by roadside equipment. Free-flow speed on the road segment is the speed limit for that segment. The actual traffic flow of a road segment is obtained by counting the number of vehicles passing through the road segment section per unit time, as recorded by roadside cameras. Traffic capacity of a road segment is a design parameter for that segment. Traffic conditions are divided into four levels based on the SI value: Smooth traffic (SI < 0.2), Basically smooth traffic (0.2 ≤ SI < 0.4), Lightly congested traffic (0.4 ≤ SI < 0.6), and Severely congested traffic (SI ≥ 0.6).

[0137] Short-term prediction unit: Using a GRU gated recurrent unit neural network model, based on historical traffic data, real-time traffic data and influencing factors, it predicts the SI value change trend of each road segment in the next 3-5 minutes and outputs the prediction result of "road segment ID - prediction time - prediction SI value", providing dynamic traffic basis for route planning.

[0138] The multi-objective real-time path planning module includes a multi-objective weight determination unit, a path comprehensive cost calculation unit, and a dynamic weighted A* path search unit;

[0139] Multi-objective weight determination unit: Dynamically adjusts the weights of time, energy consumption, and safety objectives based on user needs and scenarios.

[0140] Time weight Commuting scenarios Take 0.5, for leisure scenarios. Take 0.3;

[0141] Energy consumption weight Electric vehicle scenario Take 0.4, for gasoline vehicle scenarios Take 0.2;

[0142] Safety weight Rainy or snowy weather Take 0.4, sunny day Take 0.2; and satisfy ;

[0143] Route integrated cost calculation unit: Constructs a route integrated cost model; the route integrated cost calculation formula is as follows:

[0144] ;

[0145] in, : Overall path cost; the smaller the value, the better the path. : Path time cost, which is the length of each segment of the path. Divide by the predicted average speed of the road segment The sum, that is , This represents the number of road segments contained in the path. : Path energy consumption cost, calculated based on the vehicle energy consumption model, i.e. , , , The vehicle energy consumption coefficient is determined by the vehicle model. The road section slope is provided by roadside equipment; Path safety cost is the sum of the reciprocals of the safe distances between each segment of the path and obstacles, i.e. , The minimum safe distance between vehicles and obstacles within a road segment is provided by the vehicle-road cooperative perception and information fusion module. ≥5m;

[0146] Dynamically Weighted A* Path Search Unit: This unit introduces a dynamic heuristic function into the traditional A* algorithm. Dynamic adjustments based on predicted traffic conditions:

[0147] ;

[0148] in, For nodes The straight-line distance to the target point. For nodes Estimated energy consumption to reach the target point For maximum allowable energy consumption, For nodes The minimum safe distance to the target point; during the algorithm search process, priority is given to selecting the comprehensive cost. The path with the minimum value that satisfies the vehicle's kinematic constraints.

[0149] The path dynamic correction and obstacle avoidance decision-making module includes a correction triggering unit, a path replanning unit, and an obstacle avoidance decision-making unit.

[0150] Correction trigger unit: Real-time monitoring of three types of trigger conditions; correction is initiated when any one condition is met.

[0151] Sudden changes in traffic conditions: The deviation between the predicted SI value and the actual SI value of a road segment is ≥0.2;

[0152] Sudden Obstacles: The perception module detects unpredictable static or dynamic obstacles within the path range, and the safe distance is [not specified]. <3m;

[0153] Traffic signal change: The traffic signal phase fed back by the roadside equipment deviates from the expected phase during route planning by ≥1 cycle (e.g., the planned green light is actually a red light when arriving at the intersection).

[0154] Route replanning unit: After the correction is initiated, based on the current traffic conditions and prediction results, the process of "multi-objective weight determination - comprehensive cost calculation - dynamic weighted A* search" is re-executed to generate a new route; during the replanning process, the target points and key waypoints of the original route are retained, and only local road segments are adjusted to reduce the amount of calculation;

[0155] Obstacle Avoidance Decision Unit: Generates obstacle avoidance behavior decisions by combining vehicle-road cooperative information, using a behavior decision tree model.

[0156] Step 1: Determine the type (static / dynamic) and location (front / side) of the obstacle;

[0157] Step 2: Obtain the driving intentions of surrounding vehicles based on V2V information;

[0158] Step 3: Generate decision options (lane change to avoid obstacle, slow down and wait, detour to avoid obstacle), and calculate the safety cost and time cost of each option;

[0159] Step 4: Select the decision with the lowest overall cost. For example, when there is a static obstacle ahead and no intention of changing lanes from vehicles in adjacent lanes, execute "lane change to avoid obstacle"; when there is a dynamic obstacle ahead and congestion in adjacent lanes, execute "decelerate and wait".

[0160] The vehicle control and execution feedback module includes a control algorithm unit, an execution unit, and a feedback unit;

[0161] Control Algorithm Unit: Employs Model Predictive Control (MPC) algorithm to convert the planned path into control commands.

[0162] Path tracking: Using the desired trajectory of the path as a reference, which includes position, speed, and heading angle, calculate the deviation between the vehicle's actual trajectory and the desired trajectory, including the position deviation. Heading deviation ;

[0163] Control quantity calculation: Based on MPC-based rolling time-domain optimization, output throttle opening 0-100%, brake pedal travel 0-100%, and steering angle -30° to 30° to ensure deviation ≤0.5m, ≤5°;

[0164] Speed ​​adaptation: Combine roadside traffic signals and road section speed limits to dynamically adjust the desired speed (e.g., reduce to 30km / h when approaching an intersection, and increase to 60km / h on unobstructed sections).

[0165] Execution unit: Transforms control commands into actions of the vehicle actuators.

[0166] Accelerator or brake: The accelerator opening and brake travel commands are sent to the engine controller ECU and brake controller EBCU via the CAN bus to control the vehicle's acceleration or deceleration.

[0167] Steering: Sends steering angle commands to the Electronic Power Steering (EPS) system to control vehicle steering;

[0168] Feedback Unit: Real-time acquisition of vehicle execution status, including actual speed, position, and steering angle, is transmitted back to roadside equipment and the cloud platform via V2I / V2C communication. This is used to: update the actual traffic flow data of the traffic state analysis module; and verify the effectiveness of route planning. If the execution deviation continues to exceed a threshold... If the path length is greater than 1m, a path correction will be triggered; historical path data will be optimized to provide empirical parameters for subsequent planning.

[0169] The planning method for a real-time traffic information route planning system oriented towards vehicle-road cooperation includes the following steps:

[0170] Step 1: System Initialization and Sensor Startup

[0171] 1.1 After the system is powered on, the on-board unit (OBU) and the roadside unit (RSU) complete hardware calibration: the on-board sensors perform internal parameter calibration, and the roadside unit completes GPS positioning and coordinate system initialization;

[0172] 1.2 The vehicle-road cooperative communication and information interaction module starts V2X communication, establishes the connection between the on-board OBU, the roadside RSU, and the cloud platform, and completes the communication link test;

[0173] 1.3 The vehicle-road cooperative perception and information fusion module starts the on-board and roadside sensors to collect initial environmental data, outputs the initial environmental information list through the multi-source information fusion unit, and the system enters the standby state;

[0174] Step 2: Vehicle-to-Infrastructure Communication Establishment and Information Exchange

[0175] 2.1 Receive user navigation commands and parse the starting coordinates Target point coordinates With user preferences;

[0176] 2.2 The vehicle-mounted OBU sends "origin-destination-user preference" information to the roadside RSU via V2I communication, and the roadside RSU feeds back "segment list-roadside sensing data-traffic signal timing" covering the origin to the destination.

[0177] 2.3 The vehicle-mounted OBU requests historical traffic data and influencing factors from the cloud platform via V2C communication, and the cloud platform returns the data in real time;

[0178] 2.4 The communication protection unit performs CRC32 verification on the received data. If a data error is detected, a retransmission mechanism is triggered to ensure information integrity.

[0179] Step 3: Dynamic Traffic Condition Analysis and Prediction

[0180] 3.1 The dynamic traffic state analysis and prediction module receives the fused real-time traffic data and, based on the formula... Calculate the current SI value for each road segment and classify the traffic status level;

[0181] 3.2 The short-term prediction unit loads the GRU neural network model, inputs "historical SI sequence + real-time SI sequence + influencing factors", trains the model parameters, and predicts the SI value of each road segment in the next 3-5 minutes;

[0182] 3.3 Output a traffic status report consisting of "segment ID - current SI - predicted SI - traffic status level" and send it to the multi-objective real-time path planning module;

[0183] Step 4: Multi-objective real-time path planning

[0184] 4.1 The multi-objective weight determination unit adjusts weights according to user preferences: If the user selects "priority time", then =0.5, =0.2, =0.3; if "Prioritize Energy Consumption" is selected, then =0.3, =0.4, =0.3;

[0185] 4.2 The route integrated cost calculation unit calculates the cost of each candidate route based on the predicted SI value from the traffic condition report. , , Through formula Calculate the overall cost of the path ;

[0186] 4.3 The dynamically weighted A* path search unit uses the starting point as the initial node and the target point as the ending node, based on the comprehensive cost C and a dynamic heuristic function. The system searches for the optimal path, generates a path plan containing "waypoint coordinates - desired speed - lane suggestions", and sends it to the path dynamic correction and obstacle avoidance decision module.

[0187] Step 5: Path Dynamic Correction and Obstacle Avoidance Decisions

[0188] 5.1 The correction trigger unit monitors traffic conditions, obstacles, and traffic signals in real time: compares the actual SI value of the road segment with the predicted SI value; if the deviation is ≥0.2, it is marked as "sudden change in traffic condition"; it receives obstacle information from the sensing module; if the safe distance is... <3m, marked as "sudden obstacle"; receive traffic signal change information from roadside equipment, if the phase deviation is ≥1 cycle, mark as "traffic signal change";

[0189] 5.2 If any marker exists, the route replanning unit initiates replanning, recalculates the overall cost based on the current traffic data, and generates a locally corrected new route;

[0190] 5.3 The obstacle avoidance decision unit generates obstacle avoidance decisions based on obstacle type and V2V information through a behavior decision tree, and sends the decision results to the vehicle control and execution feedback module;

[0191] Step 6: Vehicle control execution and feedback reset

[0192] 6.1 The control algorithm unit of the vehicle control and execution feedback module receives the path plan and obstacle avoidance decision, calculates the throttle, brake and steering commands using the MPC algorithm, and sends them to the execution unit via the CAN bus;

[0193] 6.2 The execution unit drives the vehicle to execute commands, collects the actual driving status in real time, and feeds it back to the control algorithm unit to adjust the control parameters to reduce trajectory deviation;

[0194] 6.3 The feedback unit transmits the execution status back to the roadside equipment and cloud platform via V2I / V2C to update traffic status data;

[0195] 6.4 When the vehicle reaches the target point, a navigation completion signal is output, all modules are reset to standby state, and the navigation data is recorded for subsequent model optimization.

[0196] Example:

[0197] Hardware configuration:

[0198] Vehicle-mounted units: vehicle-mounted camera, millimeter-wave radar, lidar, OBU, MPC controller;

[0199] Roadside units: roadside HD cameras, roadside radar, RSUs, traffic signal controllers;

[0200] Cloud platform: Servers that deploy GRU prediction models and databases that store historical traffic data.

[0201] Implementation process:

[0202] Step 1: The vehicle-mounted sensors complete calibration, the OBU and RSU establish a V2I connection, the perception module collects data around cell A (3 private vehicles, 1 pedestrian), and outputs an initial environment list after fusion;

[0203] Step 2: The user inputs "A Community → B Hospital, priority time". The OBU requests traffic data for the route (5 road segments) from the RSU and requests the SI sequence of the morning rush hour (8:00-8:30) from the cloud. After the data is verified to be correct, it is received.

[0204] Step 3: Calculate the SI value of current road segments 1-5 (segment 3 is 0.5, light congestion). The GRU model predicts that the SI value of segment 3 will rise to 0.65 (severe congestion) in the next 3 minutes, and outputs a traffic status report.

[0205] Step 4: Set the weights =0.5, =0.2, =0.3, calculate candidate path 1 (via road segments 1-2-3-4-5) =0.8, candidate route 2 (via road segments 1-2-6-4-5) =0.6, dynamically weighted A* selects path 2;

[0206] Step 5: When driving to section 2, the perception module detects sudden construction at section 6 (safe distance 2m), triggers correction, and replans the route as "section 1-2-7-4-5". The obstacle avoidance decision is "decelerate to 30km / h and change lanes to section 7".

[0207] Step 6: The MPC algorithm outputs a throttle opening of 30% and a steering angle of 15°, executes a lane change, and the actual position deviation is 0.3m. This is fed back to the RSU, and the system finally arrives at Hospital B with a position deviation of 0.4m. The system then resets.

[0208] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time route planning system for vehicle-road cooperative traffic information, characterized in that: It includes a vehicle-road cooperative perception and information fusion module, a vehicle-road cooperative communication and information interaction module, a dynamic traffic state analysis and prediction module, a multi-objective real-time path planning module, a path dynamic correction and obstacle avoidance decision-making module, and a vehicle control and execution feedback module. The vehicle-road cooperative perception and information fusion module is used to acquire multi-source perception data from vehicles and roadsides, and outputs unified three-dimensional environmental information on dynamic obstacles, traffic signals and road conditions through multi-source heterogeneous data fusion technology. The vehicle-road cooperative communication and information interaction module is based on the V2X communication protocol to realize two-way information interaction between the on-board unit (OBU), the roadside unit (RSU), and the cloud platform, and to transmit perception data, traffic status, route instructions, and execution feedback. The dynamic traffic state analysis and prediction module calculates the traffic flow state index of the current road segment based on the fused traffic information, and outputs the traffic flow change trend for the next 3-5 minutes through a short-term prediction model. The multi-objective real-time path planning module takes time, energy consumption, and safety as objectives, constructs a comprehensive path cost model, and uses an improved dynamic weighted A* algorithm to generate the globally optimal path, satisfying vehicle kinematic constraints and user personalized needs. The path dynamic correction and obstacle avoidance decision module monitors changes in traffic conditions and sudden obstacles in real time, triggers the path replanning mechanism, and generates obstacle avoidance decisions in combination with vehicle-road cooperative information to ensure that the path remains effective. The vehicle control and execution feedback module converts the planned path into vehicle control commands, adjusts control parameters through closed-loop feedback, and transmits the execution status back to the roadside and the cloud, forming a vehicle-road cooperative closed loop.

2. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The vehicle-road cooperative perception and information fusion module includes an on-board perception unit, a roadside perception unit, and a multi-source information fusion unit. Vehicle-mounted perception unit: Composed of vehicle-mounted camera, millimeter-wave radar and lidar, which respectively collect environmental images, obstacle distance and speed and high-precision 3D point cloud, to realize 360° environmental perception around the vehicle; Roadside sensing unit: Composed of roadside high-definition camera, roadside radar, traffic signal controller, and road surface condition monitor, it collects traffic flow at intersections, traffic signal phase and remaining time, road surface smoothness / water accumulation / icing status, and covers the non-line-of-sight blind spots of vehicle-mounted sensing; Multi-source information fusion unit: Employs a three-level process of preprocessing, spatiotemporal calibration, and feature layer fusion. Preprocessing: Denoising and standardizing the data format of vehicle-mounted and roadside data; Spatiotemporal calibration: Data time synchronization is achieved based on GPS timestamps, and spatial calibration is achieved based on a preset coordinate system transformation matrix; Feature layer fusion: The obstacle recognition results of multiple sensors are fused using DS evidence theory to output a dynamic obstacle list containing "target ID-location-speed-type", a traffic signal list containing "intersection ID-signal phase-remaining time", and a road information list containing "segment ID-road condition-traffic restriction".

3. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The vehicle-road cooperative communication and information interaction module includes a V2X communication unit, an information interaction unit, and a communication support unit. V2X communication unit: Supports LTE-V2X / 5G-V2X dual-mode communication, enabling three types of interaction: V2I (Vehicle-to-Infrastructure): Transmission of vehicle-mounted sensing data, roadside sensing data, and traffic signal commands between the vehicle-mounted OBU and the roadside RSU; V2V vehicle-to-vehicle: Transmission of vehicle status and driving intentions between adjacent vehicle OBUs; V2C Vehicle-to-Cloud: The vehicle-mounted OBU transmits global traffic status, historical route data, and user preference settings between itself and the cloud platform, using cellular networks as backup to ensure long-distance information transmission; Information exchange unit: Defines a standardized data exchange format, including: Sensing data frame: contains "sensor ID - acquisition timestamp - target list"; Traffic status frame: contains "segment ID - timestamp - traffic flow parameters"; Path instruction frame: contains "path ID - list of waypoints - control parameters"; Communication support unit: Employs data verification, retransmission mechanisms, and fault tolerance processing to ensure information reliability. Data verification: CRC32 cyclic redundancy check is used to detect data transmission errors; Retransmission mechanism: When the receiving end detects data errors or loss, it triggers the sending end to retransmit via ACK / NACK feedback; Fault tolerance: When V2I communication is interrupted, it automatically switches to V2V communication to obtain perception data from surrounding vehicles as a temporary supplement to environmental information.

4. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The dynamic traffic state analysis and prediction module includes a traffic state analysis unit and a short-term prediction unit. Traffic State Analysis Unit: Based on the fused traffic flow data, a Traffic Flow State Index (SI) is constructed to assess the current road segment status. The formula for calculating the Traffic Flow State Index is: ； Where SI: Traffic Flow State Index, with a value range of [0,1]. The closer SI is to 1, the more congested the traffic is; the closer SI is to 0, the smoother the traffic is. , , Weighting coefficients, satisfying Adjustments will be made dynamically based on the type of road segment. The actual traffic flow density of a road segment is obtained by dividing the number of vehicles in the segment as counted by roadside radar by the length of the segment. : Maximum traffic flow density of the road segment, which is a design parameter for the road segment; The actual average vehicle speed on the road segment is obtained by averaging the speeds of all vehicles collected by roadside equipment. Free-flow speed on the road segment is the speed limit for that segment. The actual traffic flow of a road segment is obtained by counting the number of vehicles passing through the road segment section per unit time, as recorded by roadside cameras. Traffic capacity of a road segment is a design parameter for that segment. Traffic conditions are divided into four levels based on the SI value: Smooth traffic (SI < 0.2), Basically smooth traffic (0.2 ≤ SI < 0.4), Lightly congested traffic (0.4 ≤ SI < 0.6), and Severely congested traffic (SI ≥ 0.6). Short-term prediction unit: Using a GRU gated recurrent unit neural network model, based on historical traffic data, real-time traffic data and influencing factors, it predicts the SI value change trend of each road segment in the next 3-5 minutes and outputs the prediction result of "road segment ID - prediction time - prediction SI value".

5. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The multi-objective real-time path planning module includes a multi-objective weight determination unit, a path comprehensive cost calculation unit, and a dynamic weighted A* path search unit; Multi-objective weight determination unit: Dynamically adjusts the weights of time, energy consumption, and safety objectives based on user needs and scenarios. Time weight Commuting scenarios Take 0.5, for leisure scenarios. Take 0.3; Energy consumption weight Electric vehicle scenario Take 0.4, for gasoline vehicle scenarios Take 0.2; Safety weight Rainy or snowy weather Take 0.4, sunny day Take 0.2; and satisfy ; Route integrated cost calculation unit: Constructs a route integrated cost model; the route integrated cost calculation formula is as follows: ； in, : Overall path cost; the smaller the value, the better the path. : Path time cost, which is the length of each segment of the path. Divide by the predicted average speed of the road segment The sum, that is , This represents the number of road segments contained in the path. : Path energy consumption cost, calculated based on the vehicle energy consumption model, i.e. , , , The vehicle energy consumption coefficient is determined by the vehicle model. The road section gradient is provided by roadside equipment; Path safety cost is the sum of the reciprocals of the safe distances between each segment of the path and obstacles, i.e. , The minimum safe distance between vehicles and obstacles within a road segment is provided by the vehicle-road cooperative perception and information fusion module. ≥5m; Dynamically Weighted A* Path Search Unit: This unit introduces a dynamic heuristic function into the traditional A* algorithm. Dynamic adjustments based on predicted traffic conditions: ； in, For nodes The straight-line distance to the target point. For nodes Estimated energy consumption to reach the target point For maximum allowable energy consumption, For nodes The minimum safe distance to the target point; during the algorithm search process, priority is given to selecting the comprehensive cost. The path with the minimum value that satisfies the vehicle's kinematic constraints.

6. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The path dynamic correction and obstacle avoidance decision-making module includes a correction triggering unit, a path replanning unit, and an obstacle avoidance decision-making unit. Correction trigger unit: Real-time monitoring of three types of trigger conditions; correction is initiated when any one condition is met. Sudden changes in traffic conditions: The deviation between the predicted SI value and the actual SI value of a road segment is ≥0.2; Sudden Obstacles: The perception module detects unpredictable static or dynamic obstacles within the path range, and the safe distance is [not specified]. <3m; Traffic signal change: The traffic signal phase fed back by the roadside equipment deviates from the expected phase during route planning by ≥1 cycle; Route replanning unit: After the correction is initiated, based on the current traffic conditions and prediction results, the "multi-objective weight determination - comprehensive cost calculation - dynamic weighted A* search" process is re-executed to generate a new route; During the replanning process, the target points and key waypoints of the original route are retained, and only local road sections are adjusted to reduce the amount of calculation. Obstacle Avoidance Decision Unit: Generates obstacle avoidance behavior decisions by combining vehicle-road cooperative information, using a behavior decision tree model. Step 1: Determine the type and location of the obstacle; Step 2: Obtain the driving intentions of surrounding vehicles based on V2V information; Step 3: Generate decision options and calculate the security and time costs of each option; Step 4: Choose the decision with the lowest overall cost.

7. The real-time route planning system for vehicle-road cooperative traffic information according to claim 1, characterized in that: The vehicle control and execution feedback module includes a control algorithm unit, an execution unit, and a feedback unit; Control Algorithm Unit: Employs Model Predictive Control (MPC) algorithm to convert the planned path into control commands. Path tracking: Using the desired trajectory of the path as a reference, which includes position, speed, and heading angle, calculate the deviation between the vehicle's actual trajectory and the desired trajectory, including the position deviation. Heading deviation ; Control quantity calculation: Based on MPC-based rolling time-domain optimization, output throttle opening 0-100%, brake pedal travel 0-100%, and steering angle -30° to 30° to ensure deviation ≤0.5m, ≤5°; Speed ​​adaptation: Dynamically adjust the desired speed by combining roadside traffic signals and road segment speed limits; Execution unit: Transforms control commands into actions of the vehicle actuators. Accelerator or brake: The accelerator opening and brake travel commands are sent to the engine controller ECU and brake controller EBCU via the CAN bus to control the vehicle's acceleration or deceleration. Steering: Sends steering angle commands to the Electronic Power Steering (EPS) system to control vehicle steering; Feedback Unit: Real-time acquisition of vehicle execution status, including actual speed, position, and steering angle, is transmitted back to roadside equipment and the cloud platform via V2I / V2C communication. This is used to: update the actual traffic flow data of the traffic state analysis module; and verify the effectiveness of route planning. If the execution deviation continues to exceed a threshold... If the path length is greater than 1m, a path correction will be triggered; historical path data will be optimized to provide empirical parameters for subsequent planning.

8. The planning method of the real-time route planning system for vehicle-road cooperative traffic information according to any one of claims 1-7, characterized in that: Includes the following steps: Step 1: System Initialization and Sensor Startup (1.1) After the system is powered on, the on-board unit (OBU) and the roadside unit (RSU) complete hardware calibration: the on-board sensors perform internal parameter calibration, and the roadside equipment completes GPS positioning and coordinate system initialization; (1.2) The vehicle-road cooperative communication and information interaction module starts V2X communication, establishes the connection between the on-board OBU, the roadside RSU, and the cloud platform, and completes the communication link test; (1.3) The vehicle-road cooperative perception and information fusion module starts the on-board and roadside sensors, collects the initial environmental data, outputs the initial environmental information list through the multi-source information fusion unit, and the system enters the standby state; Step 2: Vehicle-to-Infrastructure Communication Establishment and Information Exchange (2.1) Receive user navigation instructions and parse the starting coordinates Target point coordinates With user preferences; (2.2) The vehicle-mounted OBU sends "origin-destination-user preference" information to the roadside RSU via V2I communication, and the roadside RSU feeds back "segment list-roadside sensing data-traffic signal timing" covering the origin to the destination; (2.3) The vehicle-mounted OBU requests historical traffic data and influencing factors from the cloud platform via V2C communication, and the cloud platform returns the data in real time; (2.4) The communication protection unit performs CRC32 verification on the received data. If a data error is detected, a retransmission mechanism is triggered to ensure information integrity. Step 3: Dynamic Traffic Condition Analysis and Prediction (3.1) The dynamic traffic state analysis and prediction module receives the fused real-time traffic data and, based on the formula... Calculate the current SI value for each road segment and classify the traffic status level; (3.2) The short-term prediction unit loads the GRU neural network model, inputs "historical SI sequence + real-time SI sequence + influencing factors", trains the model parameters, and predicts the SI value of each road segment in the next 3-5 minutes; (3.3) Output a traffic status report with "segment ID - current SI - predicted SI - traffic status level" and send it to the multi-objective real-time path planning module; Step 4: Multi-objective real-time path planning (4.1) The multi-objective weight determination unit adjusts the weights according to user preferences: if the user selects "priority time", then =0.5, =0.2, =0.3; if "prioritize energy consumption" is selected, then =0.3, =0.4, =0.3; (4.2) The route integrated cost calculation unit calculates the cost of each candidate route based on the predicted SI value from the traffic state report. , , Through formula Calculate the overall cost of the path ; (4.3) The dynamically weighted A* path search unit uses the starting point as the initial node and the target point as the ending node, based on the comprehensive cost C and the dynamic heuristic function. The system searches for the optimal path, generates a path plan containing "waypoint coordinates - desired speed - lane suggestions", and sends it to the path dynamic correction and obstacle avoidance decision module. Step 5: Path Dynamic Correction and Obstacle Avoidance Decisions (5.1) The correction trigger unit monitors traffic conditions, obstacles, and traffic signals in real time: compares the actual SI value of the road segment with the predicted SI value, and if the deviation is ≥0.2, it is marked as "sudden change in traffic condition"; it receives obstacle information from the sensing module, and if the safe distance is within range... <3m, marked as "Sudden Obstacle"; Receive traffic signal change information from roadside equipment. If the phase deviation is ≥1 cycle, mark it as "traffic signal change". (5.2) If any marker exists, the route replanning unit initiates replanning, recalculates the comprehensive cost based on the current traffic data, and generates a new route with local correction; (5.3) The obstacle avoidance decision unit generates obstacle avoidance decisions based on obstacle type and V2V information through behavior decision tree, and sends the decision results to vehicle control and execution feedback module; Step 6: Vehicle control execution and feedback reset (6.1) The control algorithm unit of the vehicle control and execution feedback module receives the path scheme and obstacle avoidance decision, calculates the throttle, brake and steering commands using the MPC algorithm, and sends them to the execution unit via the CAN bus; (6.2) The execution unit drives the vehicle to execute commands, collects the actual driving status in real time, feeds it back to the control algorithm unit, and adjusts the control parameters to reduce trajectory deviation; (6.3) The feedback unit transmits the execution status back to the roadside equipment and the cloud platform via V2I / V2C to update the traffic status data; (6.4) When the vehicle reaches the target point, it outputs a navigation completion signal, and each module is reset to standby state. At the same time, the navigation data is recorded for subsequent model optimization.

Images