Multi-intersection cooperative scheduling method and system for networked autonomous vehicles

Abstract

The application provides a multi-intersection cooperative scheduling method and system for networked autonomous vehicles, and the method comprises the following steps: constructing a multi-intersection environment model, a lane conflict matrix and a time window occupation matrix; if a target vehicle has not reserved a time window, reserving a time window at a current intersection, if the target vehicle has reserved a time window, reserving a time window at a next intersection, and calculating an average target vehicle speed before and after the target vehicle passes through the current intersection according to the time window; if the average target vehicle speed is lower than a preset threshold, starting path re-planning to obtain a planned path; reserving time windows at the current and next intersections according to the planned path, and performing speed coordination calculation to obtain speed and acceleration instructions before and after the target vehicle passes through the current intersection; and making the target vehicle run according to the generated speed and acceleration. The application also provides a system corresponding to the method, and the application improves the traffic efficiency and path stability of the networked autonomous vehicles in a multi-intersection environment.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation technology, and specifically relates to a multi-intersection collaborative scheduling method and system for networked autonomous vehicles. Background Technology

[0002] With the continuous increase in the number of motor vehicles in cities, road traffic flow density has increased significantly. As the areas where traffic conflicts are most concentrated, the traffic efficiency of urban intersections directly affects the overall operational quality and safety level of the transportation system. At present, intersection control generally adopts fixed-phase signal control or inductive control methods. Although these methods can achieve traffic regulation to a certain extent, they are mostly based on historical average flow or local inductive feedback, lacking dynamic perception and forward-looking optimization capabilities for the overall traffic situation. Especially in scenarios where connected and autonomous vehicles are widely involved, their real-time performance and adaptability are clearly insufficient, making it difficult to meet the needs of future intelligent transportation development.

[0003] In recent years, with the rapid development of intelligent connected vehicles and vehicle-infrastructure cooperation (VIC) technologies, traffic light-less intersection scheduling methods based on V2I communication have gradually become a research hotspot. These methods achieve dynamic planning and control of vehicle paths, speeds, and passage timing through information interaction between onboard terminals and roadside infrastructure. Research shows that in single-intersection scenarios, reasonable vehicle passage order and speed control strategies can effectively reduce delays, improve traffic efficiency, and enhance the safety and robustness of intersection operation. In existing research, intersection vehicle scheduling often employs reservation mechanisms, with first-in, first-out (FIFO) strategies being simple in structure and easy to implement, but difficult to achieve globally optimal passage order. To improve scheduling performance, scholars have proposed optimization-based methods, such as partially observable Markov decision processes, particle swarm optimization, adaptive search trees, and queuing models, to solve the optimal passage sequence problem. Some studies also model vehicle scheduling as a discrete optimization problem in graph theory, such as job shop scheduling and minimum clique coverage, and design heuristic algorithms to improve solution efficiency. These methods have achieved good simulation results in a single-intersection, uncontrolled signal control scenario.

[0004] However, most existing publicly available methods focus on local optimization control of isolated intersections, while in typical urban road networks, vehicle routes often cross multiple consecutive intersections. If the scheduling of each intersection is independent and lacks information sharing and path coordination mechanisms, vehicles will only be scheduled when they approach the intersection, which can easily lead to problems such as drastic speed fluctuations, path conflicts, and congestion accumulation, reducing the overall operating efficiency and traffic balance of the system. Summary of the Invention

[0005] To address the aforementioned technical challenges, this invention proposes a multi-intersection collaborative scheduling method and system for connected autonomous vehicles. By constructing a unified time window model and lane conflict matrix mechanism, dynamic coordination and conflict detection of traffic resources across multiple intersections are achieved. Simultaneously, by combining a rolling speed adjustment strategy and a path replanning strategy, efficient, smooth, and conflict-free passage of vehicles through multiple intersections is ensured, thereby improving overall road network scheduling efficiency and traffic flow stability.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A multi-intersection cooperative scheduling method for connected autonomous vehicles includes the following steps:

[0008] A multi-intersection environment model and a lane conflict matrix are constructed; a time window occupancy matrix with fixed time granularity is constructed based on the multi-intersection environment model; and the safety of the time window occupancy matrix is ​​determined using the lane conflict matrix.

[0009] If the target vehicle has not reserved a time window, then reserve a time window at the current intersection; if the target vehicle has already reserved a time window, then reserve a time window at the next intersection. Combine the time windows at the current intersection and the next intersection to calculate the average target vehicle speed of the target vehicle on the road segment before and after the current intersection. If the average target vehicle speed is lower than a preset threshold, initiate a route replanning strategy to avoid congestion and obtain a replanned route.

[0010] According to the planned route, the time window for the current intersection and the time window for the next intersection are re-booked. The time windows for the current intersection and the time window for the next intersection are combined to perform speed coordination calculations to obtain the speed and acceleration instructions for the target vehicle before and after the current intersection. The vehicle runs according to the generated speed and acceleration, and the speed and acceleration adjustment of the road segments before and after the current intersection is completed.

[0011] This invention also proposes a multi-intersection collaborative scheduling system for connected autonomous vehicles, including a vehicle terminal, an intersection management unit, and a scheduling center;

[0012] The vehicle terminal is used to acquire target vehicle status data and send it to the intersection management unit, as well as to receive the planned path and the calculated speed and acceleration commands to execute operation control and complete the speed and acceleration adjustment of the road segments before and after the current intersection.

[0013] The intersection management unit is used to receive vehicle status data, construct a multi-intersection environment model and a lane conflict matrix; construct a time window occupancy matrix with fixed time granularity based on the multi-intersection environment model; and use the lane conflict matrix to perform a safety judgment on the time window occupancy matrix; if the target vehicle has not reserved a time window, it reserves a time window at the current intersection; if the target vehicle has already reserved a time window, it reserves a time window at the next intersection; it calculates the average target vehicle speed of the target vehicle before and after passing the current intersection by combining the time windows of the current intersection and the next intersection; if the average target vehicle speed is lower than a preset threshold, it sends a route replanning request to the dispatch center; it receives the replanned route calculated by the dispatch center; it re-reserves the time windows of the current intersection and the next intersection according to the replanned route; it performs speed coordination calculation by combining the time windows of the current intersection and the next intersection to obtain the speed and acceleration commands of the target vehicle before and after passing the current intersection, and sends them to the vehicle terminal along with the replanned route;

[0014] The scheduling center is used to perform path replanning calculations to obtain the planned path when path replanning is required.

[0015] The effects described in the invention are merely those of the embodiments, and not all the effects of the invention. One of the above technical solutions has the following advantages or beneficial effects:

[0016] This invention proposes a multi-intersection cooperative scheduling method and system for connected autonomous vehicles. The method includes the following steps: constructing a multi-intersection environment model and a lane conflict matrix; constructing a time window occupancy matrix with fixed time granularity based on the multi-intersection environment model; and using the lane conflict matrix to perform a safety assessment on the time window occupancy matrix; if the target vehicle has not reserved a time window, then reserve a time window at the current intersection; if the target vehicle has already reserved a time window, then reserve a time window at the next intersection; calculating the average target vehicle speed of the target vehicle on the road segments before and after the current intersection by combining the time windows of the current intersection and the next intersection; if the average target vehicle speed is lower than a preset threshold, initiating a path replanning strategy to avoid congestion and obtain a planned path; re-reserving the time windows of the current intersection and the next intersection according to the planned path; performing speed coordination calculation by combining the time windows of the current intersection and the next intersection to obtain the speed and acceleration commands of the target vehicle on the road segments before and after the current intersection; and enabling the vehicle to run according to the generated speed and acceleration, completing the speed and acceleration adjustment of the road segments before and after the current intersection. This invention also proposes a system corresponding to the multi-intersection cooperative scheduling method for connected autonomous vehicles. This invention achieves dynamic coordination and conflict detection of traffic resources among multiple intersections by constructing a unified time window model and lane conflict matrix mechanism. At the same time, it combines a rolling speed adjustment strategy and a route replanning strategy to ensure efficient, smooth and conflict-free passage of vehicles in multi-intersection routes, thereby improving the overall road network scheduling efficiency and traffic flow stability.

[0017] This invention enables coordinated and unified scheduling of cross-intersection traffic resources, improving the traffic efficiency and path stability of connected autonomous vehicles in multi-intersection environments, and has good engineering feasibility and promotion value. Attached Figure Description

[0018] Figure 1 This is a flowchart of the multi-intersection cooperative scheduling method for networked autonomous vehicles proposed in Embodiment 1 of the present invention;

[0019] Figure 2 This is a typical four-way intersection model proposed in Embodiment 1 of the present invention;

[0020] Figure 3 This refers to the 2×3 multi-intersection urban road network constructed based on a typical four-way intersection model proposed in Embodiment 1 of the present invention;

[0021] Figure 4 This is a schematic diagram of the lane conflict matrix structure corresponding to a typical four-way intersection model proposed in Embodiment 1 of the present invention;

[0022] Figure 5 This is a schematic diagram of the time window model matrix structure proposed in Embodiment 1 of the present invention;

[0023] Figure 6 This is a schematic diagram of a multi-intersection collaborative scheduling system for networked autonomous vehicles proposed in Embodiment 1 of the present invention. Detailed Implementation

[0024] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. To simplify the disclosure of the invention, components and arrangements of specific examples are described below. Furthermore, reference numerals and / or letters may be repeated in different examples. This repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. It should be noted that the components illustrated in the drawings are not necessarily drawn to scale. Descriptions of well-known components, processing techniques, and processes are omitted in this invention to avoid unnecessarily limiting the invention.

[0025] Example 1

[0026] The multi-intersection cooperative scheduling method for connected autonomous vehicles proposed in Embodiment 1 of this invention aims to solve the efficiency bottlenecks and scheduling inconsistencies caused by the lack of linkage and coordination mechanisms in current urban traffic systems, difficulties in coordinating vehicle traffic rhythms, and one-sided reliance on local information in scheduling strategies. Especially in the context of the widespread deployment of connected autonomous vehicles, existing strategies generally focus on local optimization of single intersections, ignoring the reality that vehicle paths in urban road networks typically cross multiple intersections, making it difficult to achieve a systematic improvement in traffic efficiency and traffic balance.

[0027] The overall method of the multi-intersection cooperative scheduling method for connected autonomous vehicles proposed in Embodiment 1 of this invention includes:

[0028] A multi-intersection environment model and a lane conflict matrix are constructed; a time window occupancy matrix with fixed time granularity is constructed based on the multi-intersection environment model; and the safety of the time window occupancy matrix is ​​determined using the lane conflict matrix.

[0029] If the target vehicle has not reserved a time window, then reserve a time window at the current intersection; if the target vehicle has already reserved a time window, then reserve a time window at the next intersection. Combine the time windows at the current intersection and the next intersection to calculate the average target vehicle speed of the target vehicle on the road segment before and after the current intersection. If the average target vehicle speed is lower than a preset threshold, initiate a route replanning strategy to avoid congestion and obtain a replanned route.

[0030] According to the planned route, the time window for the current intersection and the time window for the next intersection are re-booked. The time windows for the current intersection and the time window for the next intersection are combined to perform speed coordination calculations to obtain the speed and acceleration instructions for the target vehicle before and after the current intersection. The vehicle runs according to the generated speed and acceleration, and the speed and acceleration adjustment of the road segments before and after the current intersection is completed.

[0031] The method also includes:

[0032] If the average target vehicle speed is not lower than the preset threshold, the path replanning strategy will not be activated.

[0033] By combining the time windows of the current intersection and the next intersection, speed coordination calculations are performed to obtain the speed and acceleration commands for the target vehicle before and after the current intersection; the target vehicle then runs according to the generated speed and acceleration, completing the speed and acceleration adjustment of the road segments before and after the current intersection.

[0034] Before implementing the multi-intersection collaborative scheduling method, this application constructs a multi-intersection environment model and a lane conflict matrix; constructs a time window occupancy matrix with fixed time granularity based on the multi-intersection environment model; and uses the lane conflict matrix to perform a safety judgment on the time window occupancy matrix.

[0035] The process of constructing a multi-intersection environment model includes: constructing a grid-type multi-intersection urban road network, where each intersection includes an approach direction, and each approach direction includes multiple approach lanes, with each lane assigned a unique number.

[0036] This invention is based on Figure 2 Taking a typical four-way intersection (i.e., a crossroads) as an example, a 2×3 grid-type multi-intersection urban road network was constructed based on this. Figure 3 This invention presents a 2×3 multi-intersection urban road network constructed based on a typical four-way intersection model, as proposed in Embodiment 1 of this invention. In the multi-intersection environment model, each intersection contains four approach directions, and each direction has three approach lanes for left turns, straight ahead, and right turns (as shown by the arrows in the figure). Therefore, the intersection has a total of 12 approach lanes. Each lane is assigned a unique number. ( The numbering system starts from the northernmost right-turn lane and proceeds clockwise.

[0037] It should be noted that the scheduling method designed in this invention has good scalability in structure and is also applicable to other types of intersections such as T-shaped intersections and Y-shaped intersections, as well as other types of lane configuration scenarios.

[0038] To achieve traffic conflict detection, a lane conflict matrix containing all spatial conflict relationships between lanes is pre-established. The lane conflict matrix is ​​a two-dimensional Boolean matrix with dimension 1. ; wherein, the This represents the number of inbound lanes. In a two-dimensional Boolean matrix, a value of 1 in each cell indicates a conflict between two lanes within the intersection area, while a value of 0 in each cell indicates no conflict. Figure 4 This is a schematic diagram of the lane conflict matrix structure corresponding to a typical four-way intersection model proposed in Embodiment 1 of the present invention. The matrix is ​​a static structure, initialized and maintained by the intersection management unit, and provides a basis for the safety judgment of subsequent passage time windows.

[0039] To accurately depict the dynamic occupancy status of intersections over a future period, each intersection management unit maintains a dynamically updated time window occupancy matrix; the time window occupancy matrix is ​​a two-dimensional dynamic matrix, with rows representing lane numbers. The column represents the time window number to fix the time granularity. The time granularity is updated on a rolling basis; it represents the sum of the average travel time and safety margin of vehicles passing through the intersection; when the element value of the time window occupancy matrix is ​​0, it means that there are no vehicles occupying the lane in the corresponding time window; when the value is A, it means that vehicle A from that lane will occupy the intersection in the time window, where A is the vehicle number.

[0040] With this design, the matrix contains both the lane occupancy status within a future time window and the corresponding vehicle identity information. Figure 5 This is a schematic diagram of the time window model matrix structure proposed in Embodiment 1 of the present invention; in the time window No vehicles entered the intersection from any lane; during the time window From the lane Vehicle ID=5 and from lane Vehicle ID=2 will occupy the intersection; during the time window From the lane Vehicle ID=3, from lane Vehicle ID=1 and from lane Vehicle ID=6 will occupy the intersection. This is related to... Figure 4 The lane conflict matrix also matches, i.e., from lane and Vehicles or from the lane , and All vehicles can pass through the intersection simultaneously without conflict.

[0041] It should be noted that the time window model uses a sliding window mechanism for dynamic updates: after each time window length... The first column of the matrix will be removed, and a new future time window column will be appended to the end in real time. Because time advances continuously, the first column of the matrix will no longer correspond to a fixed interval after the update. Instead of mapping to the latest future time window, the other columns are mapped in the same way. Through this sliding update method, the time window occupancy matrix always maintains only a limited number of future time periods, which ensures the real-time performance of the model and avoids excessive storage space consumption.

[0042] Once a connected autonomous vehicle (CAV) enters the communication range of the intersection management unit (ORU), it automatically reports key information to the ORU, including its vehicle number, current location, current speed, and travel route. Upon receiving the vehicle status information, the ORU initiates a scheduling process based on the current time window occupancy and conflict relationship model. It allocates the earliest reachable safe time window to the vehicle, calculates the target speed or acceleration parameters to be maintained, and sends real-time scheduling instructions to the target vehicle.

[0043] Figure 1 This is a flowchart of the multi-intersection cooperative scheduling method for networked autonomous vehicles proposed in Embodiment 1 of the present invention;

[0044] In step S100, the process begins.

[0045] In step S101, it is determined whether the target vehicle has entered the communication range of the intersection management unit. If it has entered the communication range of the intersection management unit, step S102 is executed; otherwise, step S101 is executed again.

[0046] Once a target vehicle enters the communication range of the intersection management unit, it will proactively report the following information to the intersection management unit: vehicle ID, lane number, current speed, current location, and subsequent route information. After receiving the information, the intersection management unit will initiate the dispatching process.

[0047] In step S102, it is determined whether the vehicle has a reserved time window; if no reserved time window is available, step S103 is executed; if a reserved time window is available, step S104 is executed.

[0048] The method for determining whether a vehicle has a reserved time window is as follows: The intersection management unit searches the time window occupancy matrix for the row corresponding to the lane to which the vehicle belongs. If the vehicle already has a reserved time window, that is, the vehicle ID number exists in that row. The presence or absence of the vehicle ID number determines whether a reserved time window exists.

[0049] In step S103, the time window for reserving the target vehicle at the current intersection is determined. Specifically:

[0050] Query the lane conflict matrix of the current intersection to find the set of conflicting lanes in the lane to which the target vehicle belongs. (Based on the lane number of the target vehicle...) For example, its conflict lane set is as follows: ;

[0051] Extract the rows corresponding to these conflicting lanes from the current intersection time window occupancy matrix, and find the first column with all zeros (representing the first safe time window in the future).

[0052] The candidate time window is then used for dynamic reachability assessment, specifically by calculating the constant acceleration required for the vehicle to reach the target intersection at the start of the candidate time window.

[0053] ;

[0054] in: The remaining driving distance; Current vehicle speed; The remaining available time is the candidate time window start time minus the current time. It is a constant acceleration value; , The maximum permissible acceleration and the maximum permissible deceleration of the vehicle are listed in order.

[0055] First step in determining: If or Then, without exceeding the vehicle acceleration constraint, the vehicle will remain in the remaining time. If the target intersection cannot be reached within the time window, it means that the time window is kinetically unreachable and can be directly determined as unreachable without proceeding to the next step.

[0056] Second step judgment: If If there are no speed limits for road sections or intersections, the time window can be directly determined to be kinetically reachable, meaning that the constant acceleration scheme exists and satisfies the kinetic constraints. Otherwise, it is necessary to further verify whether the trajectory speed exceeds the speed limit for road sections or intersections. Only when both are satisfied can the time window be determined to be reachable.

[0057] Check if the trajectory speed exceeds the road section speed limit. Speed ​​limits at intersections Specifically:

[0058] The target vehicle's speed at the intersection, calculated based on uniform acceleration or uniform deceleration. :

[0059] ;

[0060] If the constraint is satisfied: for any , ;and If the time window is found to be dynamically reachable, meaning the constant acceleration scheme exists and satisfies all hard constraints, then the shortest reachable time is calculated. :

[0061] When there is no cruise segment, the peak velocity is solved according to the triangular velocity profile. :

[0062] ;

[0063] The target vehicle's arrival speed at the intersection is calculated based on the triangular velocity profile.

[0064] in, (In practice, it can be set as follows) If the solution is found ,but

[0065] ;

[0066] When a cruise segment exists, i.e., according to the trapezoidal velocity profile, the top velocity is set... ,calculate:

[0067]

[0068] ;

[0069] The time taken to travel the distance from the current position to the maximum speed; The time it takes to decelerate from maximum speed to the intersection; The length of the road traveled from the current position to the maximum speed; This represents the length of the road section from the point of deceleration at maximum speed to the intersection.

[0070] like Then cruise time Therefore, the following calculations can be made:

[0071] ;

[0072] when If the time window is determined to be kinetically reachable, that is, the triangular velocity profile scheme or the trapezoidal velocity profile scheme exists and satisfies all hard constraints; otherwise, the time window is kinetically unreachable.

[0073] If the candidate time window is reachable after the above determination, then the time window is allocated and the vehicle number is written into the time window occupancy matrix of the current intersection; if it is not reachable, then try the next idle time window until a feasible solution is found.

[0074] In step S104, if the target vehicle has already reserved a time window at the current intersection, then it reserves a time window at the next intersection; specifically:

[0075] If the target vehicle has already booked a time window at the current intersection, and the current intersection is not the end of the route, calculate the fastest time the vehicle can reach the next intersection after passing through the current intersection, assuming it travels at the maximum permitted speed. ;

[0076] Based on the lane number reported by the vehicle after passing through the current intersection, query the lane conflict matrix of the next intersection to find the set of conflicting lanes;

[0077] Extract the rows corresponding to the conflicting lanes from the time window occupancy matrix of the next intersection, starting from the time when they arrive at the next intersection fastest. Search for the first safe time window with all zeros;

[0078] Assign the corresponding safe time window to the vehicle and record the vehicle ID in the time window occupancy matrix of the next intersection.

[0079] In step S105, the average target vehicle speed of the target vehicle before and after the current intersection is calculated by combining the time window of the current intersection and the time window of the next intersection.

[0080] Calculate the average target speed of the target vehicle on the road segment before the current intersection. and the average target speed of the road section after the exit :

[0081] ;

[0082] in: This represents the distance from the current location to the current intersection. This represents the distance from the current intersection to the next intersection. Subtract the current time from the start time of the current intersection reservation time window; Subtract the start time of the current intersection's reservation time window from the start time of the next intersection's reservation time window.

[0083] In step S106, it is determined whether path replanning is triggered. If path replanning is triggered, step S107 is executed; otherwise, step S110 is executed.

[0084] If the above calculations indicate that the average target speed that vehicles need to maintain on the road segment before the current intersection... The average target speed that the vehicle needs to maintain after passing through the road segment is below the preset threshold. If the value is still lower than expected, it is determined that there is congestion ahead of the vehicle's original route, and route replanning is required.

[0085] In step S107, the road network weights are adjusted and the shortest time path is recalculated.

[0086] In the road network graph model constructed based on real-time average travel time data of road segments, the dispatch center locates the road segment that a vehicle will enter after passing the current intersection, and updates the edge weight of the corresponding road segment in the road network graph model to a larger value or set it to infinity. This reduces or completely excludes the selection priority of the road segment in the next path calculation, preventing congested or inefficient road segments from being included in the new path scheme.

[0087] Using the current intersection as the starting point and the destination as the target node, the Dijkstra algorithm is used to calculate the shortest time path from the current location to the destination in the road network model.

[0088] In step S108, a new route plan is sent to the intersection management unit. The dispatch center sends the calculated new route plan to the intersection management unit, triggering the intersection management unit to reallocate the time windows of the current intersection and the next intersection in the new route, and to perform speed coordination calculations.

[0089] In step S109, the time window for the vehicle at the current intersection and the next intersection in the new route is rescheduled. After the new route is determined, the safe time window for the vehicle at the current and next intersections needs to be rescheduled according to the methods in steps S103 and S104.

[0090] In step S110, the speed and acceleration commands that the target vehicle needs to maintain before and after the current intersection are calculated. By combining the safety time window of the current intersection with the reserved time window of the next intersection, speed coordination calculations are performed to obtain the segmented "acceleration / constant speed - duration" commands for the target vehicle before and after the current intersection, specifically:

[0091] The system sequentially calculates the speed and acceleration commands that the vehicle must maintain on the road segment before and after the current intersection. The inputs for this calculation include the distance from the current location to the current intersection. Distance from the current intersection to the next intersection The start time of the current intersection reservation window The start time of the next intersection reservation window Current moment Current vehicle speed Vehicle acceleration upper and lower limits For scenarios with speed limits for road sections and intersections, the input also includes: the speed limits for the preceding / following road sections. Speed ​​limit at intersections .

[0092] The remaining travel time for the vehicle before and after the current intersection is:

[0093] ;

[0094] The calculation process involves, given the remaining travel time and distance, reverse-engineering an executable speed profile that satisfies the acceleration boundary (and meets the speed limit in scenarios with road segment speed limits and intersection speed limits). This profile can be constant acceleration, triangular, or trapezoidal. The calculation process for the vehicle on the road segment before the current intersection is illustrated below. The calculation process for the subsequent road segment is similar; simply use the final speed of the previous segment as the current speed for the subsequent segment. First, we attempt the constant acceleration scheme to calculate the required constant acceleration value:

[0095] ;

[0096] For scenarios without speed limits on road sections or intersections, the constant acceleration scheme is feasible, i.e., it satisfies the constraints, since the time window satisfies the dynamic reachability criterion. Initial velocity at the end:

[0097] ;

[0098] For scenarios with speed limits on road sections and at intersections, if the following constraints are met: ① For any Just check the peak value. ;② If a constant acceleration scheme is adopted, the acceleration value is maintained. Otherwise, try a triangular velocity profile or a trapezoidal velocity profile, using the core equation:

[0099] ;

[0100] And require:

[0101] ;

[0102] First choice Solving for the peak velocity .like Then, a trapezoidal velocity profile scheme is adopted, and the durations of acceleration, cruising, and deceleration are calculated to obtain the corresponding acceleration / constant speed-duration commands; if Then, a triangular velocity profile scheme is adopted and the acceleration and deceleration durations are calculated to obtain the corresponding acceleration / constant velocity-duration command.

[0103] In step S111, the scheduling results are fed back to the target vehicle and the scheduling ends. The intersection management unit feeds back the scheduling results to the target vehicle all at once, including the "acceleration / constant speed-duration" instructions to be maintained before the current intersection, the "acceleration / constant speed-duration" instructions to be maintained after passing the current intersection, and the planned route. The target vehicle executes the instructions, thereby driving smoothly at a near-optimal speed through the continuous intersections.

[0104] The multi-intersection cooperative scheduling method for connected autonomous vehicles proposed in Embodiment 1 of this invention organically combines speed adjustment strategies with path replanning strategies to achieve safe and efficient passage at local intersections and rolling optimization and cooperative control for global operation across intersections. The scheduling algorithm adopts an "event-triggered + rolling optimization" strategy. The scheduling process is automatically triggered when a vehicle enters the IM communication range of any intersection, executing in a closed-loop logic of "time window allocation → speed judgment → replanning and reallocating time windows when necessary → speed coordination calculation → unified feedback". The entire process exhibits good real-time performance and scalability, making it suitable for high-density scheduling scenarios of connected autonomous vehicles in typical grid-like urban road network environments.

[0105] The multi-intersection collaborative scheduling method for connected autonomous vehicles proposed in Embodiment 1 of this invention realizes the coordinated and unified scheduling of cross-intersection traffic resources, improves the traffic efficiency and path stability of connected autonomous vehicles in multi-intersection environments, and has good engineering feasibility and promotion value.

[0106] Example 2

[0107] Embodiment 2 of the present invention also proposes a multi-intersection collaborative scheduling system for connected autonomous vehicles, which is used to execute the multi-intersection collaborative scheduling method for connected autonomous vehicles proposed in Embodiment 1 of the present invention. Figure 6 This is a schematic diagram of a multi-intersection collaborative scheduling system for networked autonomous vehicles proposed in Embodiment 2 of the present invention: the system includes a vehicle terminal, an intersection management unit, and a scheduling center;

[0108] The vehicle terminal is used to acquire target vehicle status data and send it to the intersection management unit, as well as to receive the planned path and the calculated speed and acceleration commands to execute operation control and complete the speed and acceleration adjustment of the road segments before and after the current intersection.

[0109] The intersection management unit is used to receive vehicle status data, construct a multi-intersection environment model and a lane conflict matrix; construct a time window occupancy matrix with fixed time granularity based on the multi-intersection environment model; and use the lane conflict matrix to perform a safety judgment on the time window occupancy matrix; if the target vehicle has not reserved a time window, it reserves a time window at the current intersection; if the target vehicle has already reserved a time window, it reserves a time window at the next intersection; it calculates the average target vehicle speed of the target vehicle before and after passing the current intersection by combining the time windows of the current intersection and the next intersection; if the average target vehicle speed is lower than a preset threshold, it sends a route replanning request to the dispatch center; it receives the replanned route calculated by the dispatch center; it re-reserves the time windows of the current intersection and the next intersection according to the replanned route; it performs speed coordination calculation by combining the time windows of the current intersection and the next intersection to obtain the speed and acceleration commands of the target vehicle before and after passing the current intersection, and sends them to the vehicle terminal along with the replanned route;

[0110] The scheduling center is used to perform path replanning calculations to obtain the planned path when path replanning is required.

[0111] In this application, a dedicated direct communication link for vehicle-road cooperation, such as DSRC or C-V2X PC5 mode, is used between the vehicle terminal and the intersection management unit; a wireless communication network, such as cellular network (4G / 5G), dedicated optical fiber, or government and enterprise private network, is used between the intersection management unit and the dispatch center.

[0112] Adjacent intersection management units communicate directly without going through a dispatch center. After allocating a time window at the current intersection to the target vehicle, the upstream intersection management unit C sends a reservation request to the downstream intersection management unit N through this link, negotiates and locks the time window for the target vehicle at the next intersection, and returns the result to the upstream intersection management unit C.

[0113] Vehicle terminals and dispatch centers generally do not communicate directly. Instead, route replanning results are relayed through the intersection management unit. This reduces the communication burden between vehicles and remote locations and allows the local intersection management unit to uniformly manage dispatch instructions.

[0114] Through this communication system, the multi-intersection collaborative scheduling system for connected autonomous vehicles proposed in this application can ensure low-latency scheduling response while taking into account the aggregation of global information across the entire network and path optimization, thereby improving the traffic efficiency and safety of connected autonomous vehicles in multi-intersection scenarios.

[0115] In this application, the vehicle terminal is responsible for collecting and reporting real-time status data of the target vehicle, including location data, speed data, and path data. It stores the vehicle's latest planned route and receives scheduling results from the intersection management unit, namely acceleration / constant speed - duration and the new planned route, adjusting its own behavior accordingly. In this application, the vehicle terminal itself does not participate in complex scheduling calculations, thereby reducing the requirements for onboard computing resources.

[0116] As the core of local scheduling, the intersection management unit possesses the basic functions of sensing real-time status information of vehicles within its communication range and dynamically maintaining lane conflict matrices and time window occupancy matrices. As the execution unit of the rolling speed adjustment strategy, the intersection management unit undertakes the main scheduling calculation tasks, including allocating safe time windows for target vehicles at the current intersection, coordinating with downstream intersection management units to reserve safe time windows for target vehicles at the next intersection, performing speed coordination calculations, and determining path replanning trigger conditions. It then feeds back the scheduling results (acceleration / constant speed - duration, new planned route scheme) to the vehicle terminal in one go.

[0117] In addition to the aforementioned computational and communication tasks, the intersection management unit also periodically collects data on vehicle travel time in the approach lanes within a past time window. Based on this, it calculates the average travel time for each approach road segment within the current window and uploads this data to the dispatch center. The dispatch center then aggregates the data from the entire network and constructs a road network map model. When the route replanning trigger condition is met, the intersection management unit sends a route replanning calculation request to the dispatch center and receives the new planned route scheme issued by the dispatch center. Because the intersection management unit's computation is centralized and the communication link is short, it can effectively reduce response latency and improve dispatching accuracy.

[0118] The dispatch center is responsible for aggregating the road segment traffic status data reported by each intersection management unit and periodically updating the average travel time of each road segment across the entire network to construct a real-time weighted road network map. Route replanning calculations are uniformly executed by the dispatch center, and new planned route schemes are issued to vehicle terminals through the intersection management units, thereby ensuring the global optimality of route selection.

[0119] The multi-intersection collaborative scheduling system for connected autonomous vehicles proposed in Embodiment 2 of this invention achieves dynamic coordination and conflict detection of traffic resources among multiple intersections by constructing a unified time window model and lane conflict matrix mechanism. At the same time, it combines a rolling speed adjustment strategy and a path replanning strategy to ensure efficient, smooth and conflict-free passage of vehicles in multi-intersection paths, thereby improving the overall road network scheduling efficiency and traffic flow stability.

[0120] The multi-intersection collaborative scheduling system for connected autonomous vehicles proposed in Embodiment 2 of this invention realizes the coordinated and unified scheduling of cross-intersection traffic resources, improves the traffic efficiency and path stability of connected autonomous vehicles in multi-intersection environments, and has good engineering feasibility and promotion value.

[0121] The description of the relevant parts of the multi-intersection cooperative scheduling system for networked autonomous vehicles provided in Embodiment 2 of this application can be found in the detailed description of the corresponding parts of the multi-intersection cooperative scheduling method for networked autonomous vehicles provided in Embodiment 1 of this application, and will not be repeated here.

[0122] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that the elements inherent in a process, method, article, or apparatus that includes a list of elements are included. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. Additionally, portions of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of corresponding technical solutions in the prior art have not been described in detail to avoid excessive elaboration.

[0123] While specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art can make other modifications or variations based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A multi-intersection cooperative scheduling method for connected autonomous vehicles, characterized in that, Includes the following steps: Construct a multi-intersection environment model and lane conflict matrix; Construct a time window occupancy matrix with fixed time granularity based on a multi-intersection environment model; And a lane conflict matrix is ​​used to determine the safety of the time window occupancy matrix; If the target vehicle has not reserved a time window, then reserve a time window at the current intersection; if the target vehicle has already reserved a time window, then reserve a time window at the next intersection. Combine the time windows at the current intersection and the next intersection to calculate the average target vehicle speed of the target vehicle on the road segment before and after the current intersection. If the average target vehicle speed is lower than a preset threshold, initiate a route replanning strategy to avoid congestion and obtain a replanned route. According to the planned route, the time window for the current intersection and the time window for the next intersection are re-booked. The time windows for the current intersection and the time window for the next intersection are combined to perform speed coordination calculations to obtain the speed and acceleration instructions for the target vehicle before and after the current intersection. The vehicle runs according to the generated speed and acceleration, and the speed and acceleration adjustment of the road segments before and after the current intersection is completed. If the target vehicle has not booked a time window, the process of booking a time window at the current intersection is as follows: If the target vehicle has already booked a time window, the process of booking a time window at the next intersection includes: If the target vehicle has not booked a time window, extract the row corresponding to the conflict lane from the time window occupancy matrix, find the first all-zero column as the first candidate time window, and verify the dynamic reachability of the candidate time window. If it is reachable, allocate the time window and write the vehicle number into the time window occupancy matrix; if it is not reachable, try the next available time window until a feasible solution is found. If the target vehicle has already booked a time window and the current intersection is not the end of the path, calculate the fastest time to reach the next intersection; query the set of conflicting lanes at the next intersection based on the lane number after passing the current intersection; extract the row corresponding to the conflicting lane from the time window occupancy matrix of the next intersection, and search for the first time window with all zero columns starting from the fastest time to reach the next intersection. The process of verifying the dynamic reachability of candidate time windows includes: Calculate the constant acceleration required for the vehicle to reach the target intersection at the start of the candidate time window: ； in: The remaining driving distance; Current vehicle speed; The remaining available time is the candidate time window start time minus the current time. It is a constant acceleration value; , The maximum permissible acceleration and the maximum permissible deceleration of the vehicle are listed in order. if or It was determined to be unreachable; if Next, verify whether the trajectory speed exceeds the speed limit of the road section. Speed ​​limits at intersections Specifically: The target vehicle's speed at the intersection, calculated based on uniform acceleration or uniform deceleration. : ； If the constraint is satisfied: for any , ;and If the time window is found to be dynamically reachable, then it is determined that the time window is dynamically reachable; otherwise, the shortest reachable time is further calculated. : When there is no cruise segment, the peak velocity is solved according to the triangular velocity profile. : ； The target vehicle's arrival speed at the intersection is calculated based on the triangular velocity profile. in, If the solution is ,but ； When a cruise segment exists, i.e., according to the trapezoidal velocity profile, the top velocity is set... ,calculate: ； The time taken to travel the distance from the current position to the maximum speed; The time it takes to decelerate from maximum speed to the intersection; The length of the road traveled from the current position to the maximum speed; This represents the length of the road section from the point of deceleration at maximum speed to the intersection. like Then cruise time Therefore, the following calculations can be made: ； when If the time window is reached, it is determined that the time window is dynamically reachable; otherwise, the time window is dynamically unreachable.

2. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 1, characterized in that, The process of constructing a multi-intersection environment model includes: constructing a grid-type multi-intersection urban road network, where each intersection includes an approach direction, and each approach direction includes multiple approach lanes, with each lane assigned a unique number.

3. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 1, characterized in that, The lane conflict matrix is ​​a two-dimensional Boolean matrix with dimension 1. ; Among them, the This represents the number of inbound lanes. In a two-dimensional Boolean matrix, a value of 1 in each cell indicates a conflict between two lanes within the intersection area, while a value of 0 in each cell indicates no conflict.

4. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 1, characterized in that, The time window occupancy matrix is ​​a two-dimensional dynamic matrix, where rows represent lane numbers and columns represent time window numbers; and it is updated on a rolling basis with a fixed time granularity. The fixed time granularity represents the sum of the average travel time of vehicles passing through the intersection and the safety margin; when the element value of the time window occupancy matrix is ​​0, it means that there are no vehicles occupying the lane in the corresponding time window; when the value is A, it means that vehicle A from that lane will occupy the intersection in the time window, where A is the vehicle number. Furthermore, the time window occupancy matrix is ​​dynamically updated using a sliding window mechanism. Every fixed time interval, the first column of the matrix is ​​removed and a future time window column is added.

5. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 1, characterized in that, The triggering condition for path replanning is: Calculate the average target vehicle speed of the target vehicle on the road segment before and after the current intersection by combining the time window of the current intersection and the time window of the next intersection; ； in: This represents the distance from the current location to the current intersection. This represents the distance from the current intersection to the next intersection. Subtract the current time from the start time of the current intersection reservation time window; Subtract the start time of the current intersection's reservation time window from the start time of the next intersection's reservation time window; If the vehicle needs to maintain the average target speed in the road section before the current intersection The average target speed that the vehicle needs to maintain after passing through the road segment is below the preset threshold. If the value is still lower than expected, it is determined that there is congestion ahead of the vehicle's original route, and route replanning is required.

6. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 5, characterized in that, The process of performing speed coordination calculations by combining the time window of the current intersection and the time window of the next intersection to obtain the speed and acceleration commands of the target vehicle before and after the current intersection includes: Based on the current location of the target vehicle and the reservation time window between the current intersection and the next intersection, calculate the target speed profile of the road segment before and after the current intersection. Under acceleration constraints, the velocity profile is solved in reverse to obtain the segmented acceleration duration or constant velocity duration. When it is necessary to meet the speed limits of road sections and intersections, the peak speed and intersection speed are adjusted to ensure that the speed of each section does not exceed the corresponding speed threshold. The target vehicle accelerates according to the calculated segmented speed to ensure that it arrives at each intersection on time at the start of the scheduled time window.

7. The multi-intersection cooperative scheduling method for networked autonomous vehicles according to claim 1, characterized in that, The process of path replanning strategy includes: Based on real-time average travel time data of road segments, the road network map model is constructed to locate the road segment that the target vehicle will enter after passing the current intersection; the weight of the congested road segment that the vehicle is about to enter in the road network map model is set to the maximum value; The Dijkstra algorithm is used to calculate the shortest time path from the current intersection to the target node in the road network model as the planned path.

8. A multi-intersection cooperative scheduling system for connected autonomous vehicles, used to execute the multi-intersection cooperative scheduling method for connected autonomous vehicles as described in any one of claims 1 to 7, characterized in that, This includes vehicle terminals, intersection management units, and dispatch centers; The vehicle terminal is used to acquire target vehicle status data and send it to the intersection management unit, as well as to receive the planned path and the calculated speed and acceleration commands to execute operation control and complete the speed and acceleration adjustment of the road segments before and after the current intersection. The intersection management unit is used to receive vehicle status data and construct a multi-intersection environment model and lane conflict matrix. Construct a time window occupancy matrix with fixed time granularity based on a multi-intersection environment model; And a lane conflict matrix is ​​used to determine the safety of the time window occupancy matrix; If the target vehicle has not reserved a time window, then reserve a time window at the current intersection. If the target vehicle has already reserved a time window, then reserve a time window at the next intersection. Calculate the average target vehicle speed of the target vehicle on the road segment before and after the current intersection by combining the time windows at the current intersection and the next intersection. If the average target vehicle speed is lower than a preset threshold, send a route replanning request to the dispatch center. Receive the planned path calculated by the dispatch center; According to the planned route, the time window for the current intersection and the time window for the next intersection are re-booked. The time windows for the current intersection and the time windows for the next intersection are combined to perform speed coordination calculations to obtain the speed and acceleration instructions of the target vehicle before and after passing the current intersection. These instructions, along with the planned route, are then sent to the vehicle terminal. The scheduling center is used to perform path replanning calculations to obtain the planned path when path replanning is required.

Images