A multi-vehicle cooperative intelligent parking scheduling method fusing v2x communication

Abstract

This invention relates to the field of traffic control technology, and particularly to a multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication. The method acquires a real-time parking space occupancy grid map through roadside units and receives parking intention information sent by multiple vehicles waiting to park via V2X. It predicts the travel paths of each vehicle and their arrival times at the intersection of paths. When the estimated arrival time difference of at least three vehicles is less than a preset threshold, they are incorporated into a dynamic scheduling alliance. A virtual queue is generated for the dynamic scheduling alliance, and the spatiotemporal resources of the intersection area are sliced ​​to allocate exclusive passage time windows and speed curves to each vehicle. During the journey, the state is shared in real-time via V2V, and the vehicle speed is adjusted based on dynamic priority to achieve cooperative passage through the intersection area without stopping. After passing through the intersection area, the alliance is disbanded, and each vehicle completes parking. This invention effectively eliminates conflicts at path intersection points and improves parking lot traffic efficiency through multi-vehicle cooperative scheduling.

Description

Technical Field

[0001] This invention relates to the field of traffic control technology, and in particular to a multi-vehicle collaborative intelligent parking scheduling method integrating V2X communication. Background Technology

[0002] With the acceleration of urbanization and the continuous growth of car ownership, traffic congestion in large parking lots (such as underground parking lots in shopping malls, airports, and convention centers) is becoming increasingly prominent. Especially during peak hours, multiple vehicles enter the parking lot at the same time, easily creating conflict points at bottleneck areas such as entrances and intersections, leading to frequent starts and stops, queuing, and even serious collisions.

[0003] In existing technologies, parking lot scheduling mainly relies on the following methods: First, based on the entrance guidance screen displaying the total number of remaining parking spaces, drivers find parking spaces themselves; this method cannot resolve internal path conflicts. Second, static guidance systems based on parking space detectors assign fixed parking spaces to vehicles but do not intervene in the driving process; multiple vehicles may still conflict at path intersections. Third, the reservation parking system that has emerged in recent years only solves the parking space allocation problem and does not provide coordinated control over the driving process of multiple vehicles. All these methods treat each vehicle as an independent entity, lacking a coordination mechanism between vehicles, causing path intersections to become traffic bottlenecks, increasing the average vehicle delay and energy consumption.

[0004] Therefore, there is an urgent need for a method that enables multiple vehicles to travel collaboratively at intersections within parking lots, in order to improve the overall traffic efficiency of parking lots and reduce energy consumption and emissions caused by vehicle start-stop. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a multi-vehicle collaborative intelligent parking scheduling method that integrates V2X communication, which can effectively improve the overall traffic efficiency of parking lots.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication includes the following steps: S1. Obtain a real-time parking space occupancy grid map in the parking lot through the roadside unit, and receive parking intention information sent by multiple vehicles waiting to park via V2X communication; The real-time parking space occupancy grid map is a dynamic grid map generated by sensors sensing the global parking space occupancy status. S2. Based on the parking intention information, predict the driving path of each vehicle and the time to reach the intersection of the paths. When it is determined that the expected arrival time difference of at least three vehicles is less than a preset threshold, the at least three vehicles are incorporated into a dynamic scheduling alliance. S3. Generate a virtual queue for the dynamic scheduling alliance, slice the spatiotemporal resources of the path intersection area, and allocate an exclusive passage time window and corresponding speed curve to each vehicle in the dynamic scheduling alliance. S4. During the driving process of vehicles waiting to park in the dynamic scheduling alliance, the vehicle status is shared in real time through V2V communication, and the speed of vehicles waiting to park in the dynamic scheduling alliance is adjusted based on dynamic priority, so that they can achieve non-stop coordinated passing through the path intersection area. S5. After passing through the path intersection area, the dynamic scheduling alliance is disbanded, and each waiting vehicle completes parking based on the final path.

[0007] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A multi-vehicle cooperative intelligent parking scheduling device integrating V2X communication includes a processor, a memory, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps in the multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication described above.

[0008] The beneficial effects of this invention are as follows: This invention provides a multi-vehicle collaborative intelligent parking scheduling method integrating V2X communication. It establishes a global information foundation by using roadside units to perceive the global parking space status in real time and receive parking intentions from multiple vehicles. By predicting path conflicts and forming a dynamic scheduling alliance, it transforms multiple independently driving vehicles into a collaboratively schedulable group. Through refined slicing and exclusive allocation of spatiotemporal resources at intersections, it eliminates conflicts caused by multiple vehicles arriving at the same physical location simultaneously. Through real-time V2V communication and dynamic priority adjustment, it achieves adaptive correction of prediction deviations at the execution level, ensuring that vehicles can smoothly pass through intersections according to the planned time window. The entire process allows multiple vehicles to pass through bottleneck areas sequentially without stopping, significantly reducing start-stop delays caused by intersection conflicts, improving the traffic capacity of roads within parking lots, and reducing energy consumption and emissions from frequent vehicle acceleration and deceleration. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating a multi-vehicle collaborative intelligent parking scheduling method integrating V2X communication according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a multi-vehicle collaborative intelligent parking scheduling device integrating V2X communication according to an embodiment of the present invention; Label Explanation: 1. A multi-vehicle collaborative intelligent parking scheduling device integrating V2X communication; 2. Processor; 3. Memory. Detailed Implementation

[0010] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0011] Before detailing the embodiments of this application, some related concepts will first be explained: V2X: Vehicle-to-Everything, which includes vehicle-to-roadside unit (V2I) communication and vehicle-to-vehicle (V2V) communication.

[0012] Roadside Unit (RSU): A device with V2X communication and edge computing capabilities deployed in a parking lot.

[0013] Dynamic scheduling alliance: A temporary grouping of vehicles dynamically created by roadside units based on path conflict prediction, designed to coordinate the scheduling of multiple vehicles arriving at the same interchange point at the same time.

[0014] Spatiotemporal resource slicing: The physical space (such as lanes) and time axis of the intersection points of paths are discretized to form two-dimensional resource blocks that can be allocated.

[0015] In existing technologies, traffic management within large parking lots mainly relies on static guidance and independent parking location, lacking predictive handling of multi-vehicle path conflicts and collaborative passage mechanisms. When multiple vehicles approach an intersection simultaneously, they typically rely on driver observation and strategic maneuvering, which can easily lead to standoffs or collisions. Even with traffic signs, precise scheduling cannot be achieved, resulting in low internal traffic efficiency during peak hours and increased energy consumption due to frequent vehicle starts and stops.

[0016] To at least solve the above problems, please refer to Figure 1 This invention provides a multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication, comprising the following steps: S1. Obtain a real-time parking space occupancy grid map in the parking lot through the roadside unit, and receive parking intention information sent by multiple vehicles waiting to park via V2X communication; The real-time parking space occupancy grid map is a dynamic grid map generated by sensors sensing the global parking space occupancy status. S2. Based on the parking intention information, predict the driving path of each vehicle and the time to reach the intersection of the paths. When it is determined that the expected arrival time difference of at least three vehicles is less than a preset threshold, the at least three vehicles are incorporated into a dynamic scheduling alliance. S3. Generate a virtual queue for the dynamic scheduling alliance, slice the spatiotemporal resources of the path intersection area, and allocate an exclusive passage time window and corresponding speed curve to each vehicle in the dynamic scheduling alliance. S4. During the driving process of vehicles waiting to park in the dynamic scheduling alliance, the vehicle status is shared in real time through V2V communication, and the speed of vehicles waiting to park in the dynamic scheduling alliance is adjusted based on dynamic priority, so that they can achieve non-stop coordinated passing through the path intersection area. S5. After passing through the path intersection area, the dynamic scheduling alliance is disbanded, and each waiting vehicle completes parking based on the final path.

[0017] As described above, the beneficial effects of this invention are as follows: This invention establishes a global information foundation by real-time sensing of the global parking space status and receiving parking intentions from multiple vehicles through roadside units; it transforms multiple independently driving vehicles into a collaboratively dispatchable group by predicting path conflicts and forming a dynamic scheduling alliance; it eliminates conflicts caused by multiple vehicles arriving at the same physical location simultaneously by finely slicing and exclusively allocating spatiotemporal resources at intersections; and it achieves adaptive correction of prediction deviations at the execution level through real-time V2V communication and dynamic priority adjustment, ensuring that vehicles can smoothly pass through intersections according to the planned time window. The entire process allows multiple vehicles to pass through bottleneck areas sequentially without stopping, significantly reducing start-stop delays caused by intersection conflicts, improving the traffic capacity of roads within parking lots, and reducing energy consumption and emissions caused by frequent vehicle acceleration and deceleration.

[0018] Further, step S2 includes the following steps: Based on the real-time parking space occupancy grid map of the parking lot, the available parking spaces are matched according to the parking intention information, and the parking space closest to the parking intention information is determined to obtain the target parking space; Starting from the current position of the vehicle waiting to park and ending at the target parking space, an initial driving path is generated using a shortest path algorithm. Based on the initial driving paths of each waiting vehicle, identify the path intersection area, which is a path node or road segment that is shared by at least two initial driving paths. For each of the aforementioned path intersection zones, the arrival time difference of each waiting vehicle is predicted. If the time difference is less than a preset safe time threshold, a conflict is determined, and these vehicles are incorporated into the same dynamic scheduling alliance.

[0019] As described above, the shortest path algorithm is used to obtain the travel path of each vehicle, and the intersection points of the paths are identified as potential conflict areas. The probability of conflict is judged by predicting the arrival time difference. This conflict detection method based on path prediction can identify problems in advance before the vehicles reach the conflict point, leaving sufficient adjustment time window for subsequent coordinated scheduling, and realizing the transformation from passive response to proactive prevention.

[0020] Furthermore, the predicted time difference for the arrival of each waiting vehicle includes: Based on the length of each segment in the initial driving path and the average driving speed of vehicles in the parking lot, calculate the time required for the vehicle to pass through each segment at a constant speed. The time required for each waiting vehicle to travel from the starting point to the intersection of the paths is accumulated to determine the arrival time of each waiting vehicle at the intersection of the paths, and the time difference is calculated.

[0021] As described above, the method of predicting arrival time by accumulating road segments is simple and efficient, and suitable for the edge computing capabilities of roadside units. Using the average speed of the parking lot as the prediction benchmark avoids prediction bias caused by differences in the historical speed of individual vehicles, and improves the fairness of the alliance formation.

[0022] Further, step S3 includes the following steps: S31. Generate the initial sorting of the virtual queue based on the estimated arrival time, vehicle type, and urgency of each waiting vehicle in the dynamic scheduling alliance. S32. Divide the path intersection area into spatial channels according to the lanes, and divide the time into time slots in a preset time slot unit to generate a two-dimensional time-space resource grid composed of the spatial channels and the time-space slots; S33. Based on the initial sorting of the virtual queue, with the goal of minimizing the total travel time of all vehicles in the dynamic scheduling alliance, a spatiotemporal resource allocation optimization model is constructed, and each waiting vehicle is allocated an exclusive spatiotemporal resource cell, wherein the spatiotemporal resource cell corresponds to a specific spatial channel and travel time window. S34. Based on the allocated spatiotemporal resource cells, and in conjunction with the vehicle's dynamic constraints and current driving status, plan the speed curve for each vehicle waiting to park.

[0023] As described above, abstracting physical spacetime into a quantifiable resource grid transforms the multi-vehicle conflict problem into a resource allocation problem. The optimization model, aiming to minimize total travel time, can achieve system optimality while ensuring no conflict. The planning of speed curves ensures the feasibility of allocation, allowing the theoretical time window to be supported by actual vehicle dynamics.

[0024] Further, step S4 includes the following steps: S41. Obtain the status information of each waiting vehicle in the dynamic scheduling alliance via V2V broadcast. The status information includes real-time position, speed, acceleration, heading angle, and remaining distance to the intersection area with the path. S42. Calculate the dynamic priority of each waiting vehicle in real time based on the status information. S43. For each vehicle waiting to park, perform distributed speed coordination based on a consensus algorithm, and calculate the expected acceleration according to the dynamic priority and the state information; S44. Adjust the speed of the vehicles waiting to be parked by the desired acceleration so that the deviation between the time when each vehicle arrives at the intersection of the paths and the time window is less than a preset threshold.

[0025] As described above, a closed-loop control mechanism at the execution level has been established. V2V broadcasting enables state awareness between vehicles, dynamic priority reflects real-time adjustment needs, and the consensus algorithm guarantees cooperative convergence. This distributed control structure does not rely on real-time commands from the central node. Even if roadside unit communication is briefly interrupted, vehicles can still maintain cooperative states by relying on neighbor information, thus improving the robustness of the system.

[0026] Furthermore, the calculation of the dynamic priority is expressed as follows: ; in, This indicates the dynamic priority of vehicle i waiting to park. This indicates the static priority of vehicle i waiting to park. This indicates the deviation between the time when the waiting vehicle i arrives at the intersection of the paths and the time window. Factors indicating urgency α , β , γ The preset weighting coefficients, This indicates the elimination of the zero constant.

[0027] As described above, multiple factors are integrated into a single scalar value: static priority ensures priority for special vehicles (such as ambulances); the reciprocal form of the time deviation term ensures that vehicles with greater deviations receive higher priority compensation, guiding the system to correct deviations as quickly as possible; the urgency factor can incorporate personalized needs such as remaining battery power. This multi-factor weighted design allows the priority to comprehensively reflect the adjustment needs of vehicles, providing a reasonable weighting basis for the consensus algorithm.

[0028] Furthermore, the calculation of the desired acceleration is expressed as follows: ; in, This represents the real-time expected acceleration of vehicle i waiting to park. This represents the set of vehicles waiting to park in the dynamic scheduling alliance, excluding vehicle i. w ij This represents the communication weight between waiting vehicle i and waiting vehicle j, which is inversely proportional to the distance between the vehicles. s i ( tThe real-time location of vehicle i waiting to park. v i ( t () represents the real-time speed of vehicle i waiting to park. This indicates the safe following distance, which is positively correlated with speed. This represents the target speed at the current moment, derived from the speed curve of the vehicle i waiting to park. This represents the average dynamic priority of vehicles waiting to park in the dynamic scheduling alliance. k 1. k 2 and k 3 indicates the preset control gain.

[0029] As described above, the desired acceleration formula consists of three parts: the first term enables coordinated following between vehicles to ensure a safe distance; the second term tracks the planned target speed to achieve precise alignment within the time window; and the third term adjusts the speed according to priority differences, accelerating higher-priority vehicles and decelerating lower-priority vehicles, ultimately converging the priorities of all vehicles to near the average value. This three-channel control structure ensures both safety and achieves the coordinated objective, and its simple calculation makes it suitable for real-time operation of the onboard controller.

[0030] The multi-vehicle collaborative intelligent parking scheduling method integrating V2X communication described above is applicable to traffic management within various large parking lots, especially in scenarios where multiple vehicles simultaneously enter during peak hours. The following describes specific implementation methods: Please refer to Figure 1 One embodiment of the present invention is as follows: A multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication includes the following steps: S1. Obtain a real-time parking space occupancy grid map in the parking lot through the roadside unit, and receive parking intention information sent by multiple vehicles waiting to park via V2X communication; The real-time parking space occupancy grid map is a dynamic grid map generated by sensors sensing the global parking space occupancy status.

[0031] In this embodiment, roadside units (RSUs) are deployed at key locations at the parking lot entrance and inside. They utilize integrated high-definition cameras, millimeter-wave radar, and lidar sensors to perceive the global parking space occupancy status in real time. The RSUs fuse multi-sensor data to generate a high-precision dynamic grid map, i.e., a real-time parking space occupancy grid map. This grid map represents the occupancy status (vacant / occupied) of each parking space using a fixed-size grid (e.g., 0.5m × 0.5m), and is updated every fixed time interval (e.g., 100 milliseconds) to ensure the real-time nature of parking space information.

[0032] Meanwhile, the RSU continuously monitors parking intention information sent by waiting vehicles via V2X communication interfaces (such as the C-V2X PC5 interface). Each waiting vehicle automatically sends this information upon entering the parking lot area, including the vehicle's unique identifier, vehicle type (such as sedan, SUV, emergency vehicle), vehicle size, target parking space number (such as reserved by the driver or assigned by the system), current GPS location, and estimated arrival time at the parking lot entrance. The RSU receives and caches this information, providing a data foundation for subsequent scheduling.

[0033] S2. Based on the parking intention information, predict the driving path of each vehicle and the time to reach the intersection of the paths. When it is determined that the expected arrival time difference of at least three vehicles is less than a preset threshold, the at least three vehicles are incorporated into a dynamic scheduling alliance. Step S2 includes the following steps: Based on the real-time parking space occupancy grid map of the parking lot, the available parking spaces are matched according to the parking intention information, and the parking space closest to the parking intention information is determined to obtain the target parking space; Starting from the current position of the vehicle waiting to park and ending at the target parking space, an initial driving path is generated using a shortest path algorithm. Based on the initial driving paths of each waiting vehicle, identify the path intersection area, which is a path node or road segment that is shared by at least two initial driving paths. For each of the aforementioned path intersection zones, the arrival time difference of each waiting vehicle is predicted. If the time difference is less than a preset safe time threshold, a conflict is determined, and these vehicles are incorporated into the same dynamic scheduling alliance.

[0034] The predicted time difference for the arrival of each waiting vehicle includes: Based on the length of each segment in the initial driving path and the average driving speed of vehicles in the parking lot, calculate the time required for the vehicle to pass through each segment at a constant speed. The time required for each waiting vehicle to travel from the starting point to the intersection of the paths is accumulated to determine the arrival time of each waiting vehicle at the intersection of the paths, and the time difference is calculated.

[0035] In this embodiment, the RSU first determines a target parking space for each vehicle waiting to park based on the received parking intention information and the real-time parking space occupancy grid. If the vehicle has already specified a target parking space and that space is available, it is used directly; otherwise, the RSU matches the nearest available parking space as the target parking space according to the vehicle type and current location.

[0036] Next, the RSU (Roadside Unit) uses the current location of the waiting vehicle as the starting point and the target parking space as the ending point, and utilizes the parking lot's electronic map (containing information such as road network, nodes, and intersections) to plan an initial driving path for each vehicle using a shortest path algorithm (such as the A* algorithm or Dijkstra's algorithm). Path planning must consider factors such as road direction and turning restrictions.

[0037] Then, the RSU performs cross-analysis on the initial travel paths of all vehicles to identify all possible path intersections. A path intersection is a node (such as an intersection) or overlapping section that is shared by at least two travel paths.

[0038] For each identified path junction, the RSU predicts the arrival time of each waiting vehicle at that junction. The prediction method involves dividing the path from the origin to the junction into several segments. Based on the length of each segment and the average speed of vehicles in the parking lot (which can be obtained from historical data or dynamically adjusted according to the current congestion level), the time required for a vehicle to pass through each segment at a constant speed is calculated. These times are then summed to obtain the estimated arrival time. To improve prediction accuracy, adjustments can be made based on the actual driving status of the vehicles (such as real-time speed obtained via V2X).

[0039] For each path junction, the RSU calculates the difference in arrival times between any two vehicles. If the maximum time difference between any two pairs of expected arrival times of at least three vehicles is less than a preset safe time threshold, then these vehicles are determined to have a conflict in that junction. In this case, the RSU incorporates these at least three vehicles into a dynamic scheduling alliance and assigns a unique identifier to the alliance. Member vehicles of the alliance will be treated as a coordinated whole in subsequent scheduling.

[0040] S3. Generate a virtual queue for the dynamic scheduling alliance, slice the spatiotemporal resources of the path intersection area, and allocate an exclusive passage time window and corresponding speed curve to each vehicle in the dynamic scheduling alliance.

[0041] In this embodiment, for each dynamic scheduling coalition, the RSU performs the following sub-steps for fine-grained scheduling.

[0042] Step S3 includes the following steps: S31. Generate the initial sorting of the virtual queue based on the estimated arrival time, vehicle type, and urgency level of each waiting vehicle in the dynamic scheduling alliance.

[0043] In this embodiment, an initial order of the virtual queue is generated based on the estimated arrival time, vehicle type, and urgency level (such as remaining battery power, special passenger needs, etc.) of each vehicle in the alliance. The estimated arrival time is typically the primary order, and if the times are the same, a static priority order is used (e.g., priority is given to emergency vehicles (ambulances, fire trucks, etc.), large vehicles, etc.), thus forming the basis for the order in which vehicles pass through the junction area.

[0044] S32. Divide the path intersection area into spatial channels according to the lanes, and divide the time into time slots in a preset time slot unit to generate a two-dimensional spatiotemporal resource grid composed of the spatial channels and the time slots.

[0045] In this embodiment, the path intersection area corresponding to the alliance is divided into spatial channels (such as straight lanes, left-turn lanes, and right-turn lanes) according to the actual lane layout. Simultaneously, the time axis is discretized in fixed time slots (e.g., 0.5 seconds) to form a two-dimensional spatiotemporal resource grid composed of spatial channels and time slots. Each grid cell represents a resource block of a specific spatial channel that can be exclusively used by one vehicle within a specific time slot.

[0046] S33. Based on the initial sorting of the virtual queue, with the goal of minimizing the total travel time of all vehicles in the dynamic scheduling alliance, a spatiotemporal resource allocation optimization model is constructed, and each waiting vehicle is allocated an exclusive spatiotemporal resource cell, which corresponds to a specific spatial channel and travel time window.

[0047] In this embodiment, based on the initial sorting of the virtual queue, a spatiotemporal resource allocation optimization model is constructed with the objective of minimizing the total travel time (or total energy consumption) of all vehicles within the alliance. This model is an integer programming model, with the decision variable being whether each vehicle occupies a specific spatiotemporal cell. Constraints include: each vehicle can only be allocated one spatiotemporal cell (i.e., exclusively occupying one channel and one initial time slot); each spatiotemporal cell can be allocated to at most one vehicle; and the allocation order must be consistent with the sorting of the virtual queue (i.e., an earlier-arriving vehicle cannot be allocated a later time slot). By solving this optimization model (using branch and bound or heuristic algorithms), each vehicle is allocated an exclusive spatiotemporal resource cell. This cell corresponds to a specific spatial channel and travel time window (initial time slot and the number of continuous time slots, the number of continuous time slots being determined by vehicle length and travel speed).

[0048] Specifically, we can define the following variables and parameters: The vehicle set of the dynamic dispatch alliance is: V ={1,2,..., N}

[0049] The spatiotemporal resource grid consists of a spatial channel set L={1,2,...,M} and a time discrete time slot set T={1,2,...,K}.

[0050] Let binary decision variables be used. x i,l,t ∈{0,1}, if and only if vehicle i is assigned to use the channel. l And starting from time slot t, when it is occupied (usually occupying consecutive slots) d i Each time slot d i (Depending on vehicle length and speed), this variable is 1.

[0051] The optimization objective is typically chosen to minimize the total communication time, i.e., minimize the time it takes for the last vehicle to complete the intersection. The objective function can be expressed as: ; To achieve exclusive allocation, the following exclusive constraint must be satisfied (ensuring that the same spatiotemporal resource block cannot be allocated to two vehicles simultaneously): ; In addition, accessibility constraints need to be considered: based on the vehicle i The estimated arrival time and current speed, and the allocated time window must be within a physically feasible range ( ).

[0052] By solving the above 0-1 integer programming model, a unique spatiotemporal resource can be allocated to each vehicle.

[0053] S34. Based on the allocated spatiotemporal resource cells, and in conjunction with the vehicle's dynamic constraints and current driving status, plan the speed curve for each vehicle waiting to park.

[0054] In this embodiment, based on the spatiotemporal resource cell allocated to each vehicle, combined with the vehicle's dynamic constraints (such as maximum acceleration, maximum deceleration, and comfort limits) and current driving state, a smooth speed curve is planned for each vehicle. This speed curve should ensure that the vehicle arrives at the intersection entrance at a preset speed (such as the parking lot speed limit) within a specified time window and passes through the intersection at a constant speed. Speed ​​planning can employ polynomial interpolation or model predictive control methods to generate the desired speed-time trajectory from the current location to the intersection.

[0055] Specifically, first, a simplified longitudinal dynamics model of the vehicle is established (ignoring lateral control and assuming the vehicle travels along the centerline of the lane): ; ; in, s i ( k ) is a vehicle i At any moment k Location, v i ( k ) is speed. a i ( k ) is acceleration (control input).

[0056] To achieve zero-stop passage, it is necessary to ensure that vehicles arrive at the intersection entrance (location) s junction At that time, it was exactly at the start of its allocated time window. And the speed is equal to the preset traffic speed. This constitutes a terminal constraint: ; ; During driving, the controller at each time step k Solving an optimization problem in a finite-time domain typically involves the objective of tracking a reference trajectory (an ideal position-time curve derived from the results of S33) while minimizing the rate of change of acceleration (i.e., the impact, to ensure comfort). The objective function can be expressed as: ; in, H To predict the time domain, For reference position, w 1. w 2 and w 3 represents the preset weighting coefficient.

[0057] S4. During the driving process of vehicles waiting to park in the dynamic scheduling alliance, the vehicle status is shared in real time through V2V communication, and the speed of vehicles waiting to park in the dynamic scheduling alliance is adjusted based on dynamic priority, so that they can achieve non-stop coordinated passing through the path intersection area. Step S4 includes the following steps: S41. Obtain the status information of each waiting vehicle in the dynamic scheduling alliance via V2V broadcast. The status information includes real-time position, speed, acceleration, heading angle, and remaining distance to the intersection area with the path.

[0058] In this embodiment, each vehicle in the alliance broadcasts its own status information via V2V at a fixed interval (e.g., 50ms), forming a closed loop of "cooperative perception." The broadcast state vector includes: the vehicle's unique identifier, current latitude and longitude coordinates, instantaneous speed, instantaneous acceleration, heading angle, and remaining distance to the path intersection point. In this way, each vehicle can monitor the motion status of other members in the alliance in real time, providing a data foundation for cooperative control.

[0059] S42. Calculate the dynamic priority of each waiting vehicle in real time based on the status information. The calculation of the dynamic priority is expressed as follows: ; in, This indicates the dynamic priority of vehicle i waiting to park. This indicates the static priority of vehicle i waiting to park. This indicates the deviation between the time when the waiting vehicle i arrives at the intersection of the paths and the time window. Factors indicating urgency α , β , γ The preset weighting coefficients, This indicates the elimination of the zero constant.

[0060] In this embodiment, a dynamic priority mechanism is needed to address disturbances during driving (such as deceleration of the vehicle in front or temporary obstacles). The calculation of dynamic priority is shown above, and it is recalculated every control cycle as the basis for speed adjustment.

[0061] S43. For each vehicle waiting to park, perform distributed speed coordination based on a consensus algorithm, and calculate the expected acceleration according to the dynamic priority and the state information; The calculation of the desired acceleration is expressed as follows: ; in, This represents the real-time expected acceleration of vehicle i waiting to park. This represents the set of vehicles waiting to park in the dynamic scheduling alliance, excluding vehicle i. w ij This represents the communication weight between waiting vehicle i and waiting vehicle j, which is inversely proportional to the distance between the vehicles. s i ( t The real-time location of vehicle i waiting to park. v i ( t () represents the real-time speed of vehicle i waiting to park. This indicates the safe following distance, which is positively correlated with speed. This represents the target speed at the current moment, derived from the speed curve of the vehicle i waiting to park. This represents the average dynamic priority of vehicles waiting to park in the dynamic scheduling alliance. k 1. k 2 and k 3 indicates the preset control gain.

[0062] In this embodiment, to achieve "no-stop passage," a consensus algorithm is used to converge the states of all vehicles to a cooperative state. Each vehicle adjusts its desired acceleration as shown above based on the received states of its neighbors and its own priority. Vehicles with higher priority (such as those with greater delays) receive positive acceleration compensation, while vehicles with lower priority appropriately decelerate to give way. Ultimately, all vehicles converge their errors to within an acceptable range when they reach the intersection point.

[0063] S44. Adjust the speed of the vehicles waiting to be parked by the desired acceleration so that the deviation between the time when each vehicle arrives at the intersection of the paths and the time window is less than a preset threshold.

[0064] In this embodiment, the vehicle applies the calculated desired acceleration to the throttle / brake control system to achieve real-time speed adjustment. Through continuous control cycles, the speeds of all vehicles are coordinated, gradually reducing the deviation between the arrival time of each vehicle at the junction and the allocated time window, eventually falling below a preset error threshold (e.g., 0.1 seconds). When each vehicle arrives at the junction sequentially, due to the exclusivity of the time window, they can pass through without stopping, achieving a seamless coordinated junction.

[0065] In this embodiment, during the driving process, if it is detected that a vehicle cannot arrive within the allowable error range due to severe disturbances (such as sudden malfunctions or road obstacles), and adjusting the speeds of other vehicles cannot avoid the conflict, then a local rescheduling is triggered. The triggering condition can be set as follows: satisfy: ; Furthermore, its speed regulation capability is saturated.

[0066] Once triggered, the current alliance enters "emergency coordination mode," and vehicles slow down significantly (or stop). The RSU or the alliance leader then resolves the S33 optimization problem to allocate new time windows for the remaining vehicles.

[0067] S5. After passing through the path intersection area, the dynamic scheduling alliance is disbanded, and each waiting vehicle completes parking based on the final path.

[0068] In this embodiment, when all vehicles in the dynamic scheduling alliance successfully pass through the path intersection area, the roadside unit or the lead vehicle in the alliance (such as the first vehicle to pass) initiates a dynamic scheduling alliance disbandment command. Upon receiving the command, each vehicle exits the alliance state and resumes independent driving mode. Subsequently, each vehicle, based on its target parking space, performs local path planning from its current location (such as using the dynamic window method or artificial potential field method) to complete the final stage of parking. After successful parking, the vehicle sends parking completion information to the roadside unit via V2X communication. The roadside unit then updates the real-time parking space occupancy grid map accordingly, marking the corresponding parking space as occupied for reference by subsequent vehicles.

[0069] According to another aspect of the invention, Figure 2 This is a schematic diagram illustrating a multi-vehicle collaborative intelligent parking scheduling device that integrates V2X communication according to an embodiment of the present invention.

[0070] A multi-vehicle cooperative intelligent parking scheduling device 1 integrating V2X communication includes a processor 2, a memory 3, and a computer program stored in the memory 3 and executable on the processor 2. When the processor 2 executes the computer program, it implements the various steps of the multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication as described above.

[0071] This invention presents a multi-vehicle collaborative intelligent parking scheduling method integrating V2X communication. It establishes a global information foundation by using roadside units to perceive the global parking space status in real time and receive parking intentions from multiple vehicles. By predicting path conflicts and forming a dynamic scheduling alliance, it transforms multiple independently moving vehicles into a collaboratively scheduled group. Through refined slicing and exclusive allocation of spatiotemporal resources at intersections, it eliminates conflicts caused by multiple vehicles arriving at the same physical location simultaneously. Through real-time V2V communication and dynamic priority adjustment, it achieves adaptive correction of prediction deviations at the execution level, ensuring that vehicles can smoothly pass through intersections according to the planned time window. The entire process allows multiple vehicles to pass through bottleneck areas sequentially without stopping, significantly reducing start-stop delays caused by intersection conflicts, improving the traffic capacity of roads within parking lots, and reducing energy consumption and emissions from frequent vehicle acceleration and deceleration.

[0072] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication, characterized in that, Including the following steps: S1. Obtain a real-time parking space occupancy grid map in the parking lot through the roadside unit, and receive parking intention information sent by multiple vehicles waiting to park via V2X communication; The real-time parking space occupancy grid map is a dynamic grid map generated by sensors sensing the global parking space occupancy status. S2. Based on the parking intention information, predict the driving path of each vehicle and the time to reach the intersection of the paths. When it is determined that the expected arrival time difference of at least three vehicles is less than a preset threshold, the at least three vehicles are incorporated into a dynamic scheduling alliance. S3. Generate a virtual queue for the dynamic scheduling alliance, slice the spatiotemporal resources of the path intersection area, and allocate an exclusive passage time window and corresponding speed curve to each vehicle in the dynamic scheduling alliance. S4. During the driving process of vehicles waiting to park in the dynamic scheduling alliance, the vehicle status is shared in real time through V2V communication, and the speed of vehicles waiting to park in the dynamic scheduling alliance is adjusted based on dynamic priority, so that they can achieve non-stop coordinated passing through the path intersection area. S5. After passing through the path intersection area, the dynamic scheduling alliance is disbanded, and each waiting vehicle completes parking based on the final path.

2. The multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 1, characterized in that, Step S2 includes the following steps: Based on the real-time parking space occupancy grid map of the parking lot, the available parking spaces are matched according to the parking intention information, and the parking space closest to the parking intention information is determined to obtain the target parking space; Starting from the current position of the vehicle waiting to park and ending at the target parking space, an initial driving path is generated using a shortest path algorithm. Based on the initial driving paths of each waiting vehicle, identify the path intersection area, which is a path node or road segment that is shared by at least two initial driving paths. For each of the aforementioned path intersection zones, the arrival time difference of each waiting vehicle is predicted. If the time difference is less than a preset safe time threshold, a conflict is determined, and these vehicles are incorporated into the same dynamic scheduling alliance.

3. The multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 2, characterized in that, The predicted time difference for the arrival of each waiting vehicle includes: Based on the length of each segment in the initial driving path and the average driving speed of vehicles in the parking lot, calculate the time required for the vehicle to pass through each segment at a constant speed. The time required for each waiting vehicle to travel from the starting point to the intersection of the paths is accumulated to determine the arrival time of each waiting vehicle at the intersection of the paths, and the time difference is calculated.

4. The multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 1, characterized in that, Step S3 includes the following steps: S31. Generate the initial sorting of the virtual queue based on the estimated arrival time, vehicle type, and urgency of each waiting vehicle in the dynamic scheduling alliance. S32. Divide the path intersection area into spatial channels according to the lanes, and divide the time into time slots in a preset time slot unit to generate a two-dimensional time-space resource grid composed of the spatial channels and the time-space slots; S33. Based on the initial sorting of the virtual queue, with the goal of minimizing the total travel time of all vehicles in the dynamic scheduling alliance, a spatiotemporal resource allocation optimization model is constructed, and each waiting vehicle is allocated an exclusive spatiotemporal resource cell, wherein the spatiotemporal resource cell corresponds to a specific spatial channel and travel time window. S34. Based on the allocated spatiotemporal resource cells, and in conjunction with the vehicle's dynamic constraints and current driving status, plan the speed curve for each vehicle waiting to park.

5. The multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 1, characterized in that, Step S4 includes the following steps: S41. Obtain the status information of each waiting vehicle in the dynamic scheduling alliance via V2V broadcast. The status information includes real-time position, speed, acceleration, heading angle, and remaining distance to the intersection area with the path. S42. Calculate the dynamic priority of each waiting vehicle in real time based on the status information. S43. For each vehicle waiting to park, perform distributed speed coordination based on a consensus algorithm, and calculate the expected acceleration according to the dynamic priority and the state information; S44. Adjust the speed of the vehicles waiting to be parked by the desired acceleration so that the deviation between the time when each vehicle arrives at the intersection of the paths and the time window is less than a preset threshold.

6. The multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 5, characterized in that, The calculation of the dynamic priority is expressed as follows: ； in, This indicates the dynamic priority of vehicle i waiting to park. This indicates the static priority of vehicle i waiting to park. This indicates the deviation between the time when the waiting vehicle i arrives at the intersection of the paths and the time window. Factors indicating urgency α , β , γ The preset weighting coefficients, This indicates the elimination of the zero constant.

7. A multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication according to claim 5, characterized in that, The calculation of the desired acceleration is expressed as follows: ； in, This represents the real-time expected acceleration of vehicle i waiting to park. This represents the set of vehicles waiting to park in the dynamic scheduling alliance, excluding vehicle i. w ij This represents the communication weight between waiting vehicle i and waiting vehicle j, which is inversely proportional to the distance between the vehicles. s i ( t The real-time location of vehicle i waiting to park. v i ( t () represents the real-time speed of vehicle i waiting to park. This indicates the safe following distance, which is positively correlated with speed. This represents the target speed at the current moment, derived from the speed curve of the vehicle i waiting to park. This represents the average dynamic priority of vehicles waiting to park in the dynamic scheduling alliance. k 1. k 2 and k 3 indicates the preset control gain.

8. A multi-vehicle cooperative intelligent parking scheduling device integrating V2X communication, comprising a processor, a memory, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the multi-vehicle cooperative intelligent parking scheduling method integrating V2X communication as described in any one of claims 1-7.

Images