Bulk logistics park vehicle intelligent queuing and scheduling system and method

Abstract

The invention provides an intelligent queuing and scheduling system and method for vehicles in a bulk logistics park, relates to the field of intelligent transportation, and solves the technical problem that high-time-efficiency goods are often delayed due to indifference queuing since the space-time conflict risk of multiple vehicles in an entrance area cannot be quantified in the prior art. The method comprises the following steps: acquiring state information of each vehicle to be dispatched in a park reservation range, and constructing a dynamic priority score of each vehicle to be dispatched; dividing the park entrance buffer area into space-time grid queues, and calculating space-time conflict potential energy values between every two to-be-scheduled vehicles based on the dynamic priority scores and the guide paths of the to-be-scheduled vehicles; and constructing a multi-target mixed linear programming model by taking minimization of the total waiting time and the total space-time conflict potential energy value as targets, and generating a rolling scheduling plan field of the to-be-scheduled vehicle in a preset time threshold based on the multi-target mixed linear programming model. The method and the device are used in the logistics park vehicle intelligent queuing and scheduling process.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation, and in particular to an intelligent queuing and scheduling system and method for vehicles in a bulk logistics park. Background Technology

[0002] Existing vehicle dispatching systems in bulk logistics parks generally employ queuing mechanisms based on first-come, first-served or fixed priority rules. These systems lack the dynamic perception and response capabilities to individual vehicle differences, such as contract delivery deadlines, loading and unloading resource requirements, and real-time traffic conditions. They also fail to quantify the potential conflict risks caused by multiple vehicles crossing paths and overlapping time and space at the entrance area. This results in delays for time-sensitive goods due to indiscriminate queuing, and densely packed vehicles are prone to close-range intersections and even congestion within narrow buffer zones, making it difficult to balance dispatching efficiency, service fairness, and operational safety. Therefore, this application provides an intelligent queuing and dispatching system and method for vehicles in bulk logistics parks. Summary of the Invention

[0003] This application provides an intelligent queuing and scheduling system and method for vehicles in bulk logistics parks, which solves the technical problem that the existing technology cannot quantify the risk of spatiotemporal conflicts among multiple vehicles in the entrance area, resulting in delays for high-time-efficiency goods due to indiscriminate queuing.

[0004] To achieve the above objectives, this application adopts the following technical solution: Firstly, a method for intelligent queuing and scheduling of vehicles in bulk logistics parks is provided, including: Real-time collection of status information of each vehicle waiting to be dispatched within the reservation area of ​​the park, including the type of loading and unloading resources required by the vehicle to be dispatched, the contract delivery deadline, and the average passage speed. Based on the status information, a dynamic priority score is constructed for each vehicle to be dispatched. The dynamic priority score is obtained by nonlinearly fusing the goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient and traffic disturbance sensitivity. Among them, the goods timeliness factor represents the urgency of goods delivery; the resource matching degree represents the degree of compatibility between the loading and unloading resources required by the vehicle to be dispatched and the currently available resources in the park; the queuing fairness adjustment coefficient represents the degree of inhibition exerted by the number of times priority was obtained in the past scheduling on the current priority; and the traffic disturbance sensitivity represents the degree to which the vehicle to be dispatched arrives at the park entrance on time due to the influence of the external traffic environment. The park entrance buffer zone is divided into a spatiotemporal grid queue. Based on the dynamic priority score and guidance path of each vehicle to be dispatched, the spatiotemporal conflict potential energy value between each pair of vehicles to be dispatched is calculated. The guidance path is one or more feasible driving trajectories from its current position to the designated entrance channel, which are pre-generated by the park's intelligent dispatch system for each vehicle to be dispatched. A multi-objective hybrid linear programming model is constructed with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy. Based on the multi-objective hybrid linear programming model, a rolling scheduling plan for vehicles to be scheduled within a preset time threshold is generated. The total spatiotemporal conflict potential energy is the sum of the spatiotemporal conflict potential energy values ​​between all pairs of vehicles to be scheduled.

[0005] Based on the above technical solutions, in the intelligent queuing and scheduling method for vehicles in a large-scale logistics park provided in this application, traditional scheduling methods struggle to balance efficiency, safety, and fairness in high-density vehicle entry scenarios within large logistics parks: rigid rules may lead to delays for urgent goods, or neglecting path interactions may cause congestion or even conflicts. To address this, this application constructs an intelligent collaborative scheduling mechanism: dynamically generating priority scores based on real-time vehicle status information to comprehensively reflect timeliness, resource matching, scheduling fairness, and the impact of external traffic disturbances; then, modeling the entrance buffer zone as a spatiotemporal grid queue, and accurately calculating the spatiotemporal conflict potential energy value between any two vehicles using the guidance path; based on this, establishing a multi-objective hybrid linear programming model with the objective of minimizing total waiting time and total spatiotemporal conflict potential energy value, and solving it under constraints such as resource capacity, time window, and maximum waiting time limit to generate a rolling scheduling plan. The advantages of this solution are: dynamic priority response to business changes, accurate risk assessment of spatial intersection and temporal overlap, spatiotemporal grid structure to support efficient optimization and instruction issuance, and rolling rescheduling mechanism to ensure the system has strong robustness to actual disturbances, thereby achieving safe, efficient and fair intelligent vehicle collaborative entry.

[0006] In conjunction with the first aspect above, in one possible implementation, the goods timeliness factor is calculated using a monotonically increasing nonlinear function based on the time difference between the current time and the contract delivery deadline of the goods carried by the vehicle to be dispatched; Through formula Calculate the product timeliness factor Where i is the index variable of the vehicle to be dispatched, i=1,2,…,n, and n is a positive integer. For the current moment, The contract delivery deadline for the goods carried by vehicle i to be dispatched. For time sensitivity coefficient, ; The resource matching degree is the similarity between the loading and unloading resource type vector required by the vehicle to be dispatched and the current available resource status vector in the park, which is calculated by cosine similarity. Through formula Calculate the resource matching degree ; in, Let be the vector of loading and unloading resource types required for vehicle i to be scheduled. This is the current state vector of available resources in the park. The queuing fairness adjustment coefficient is calculated using an exponential decay function based on the number of times the vehicle to be scheduled has been allowed to enter the venue early within a preset historical period. Through formula The queuing fairness adjustment coefficient was calculated. ;in, This represents the number of times vehicle i, to be dispatched, is allowed to enter the depot early within a preset historical period. This is a fairness penalty coefficient; The traffic disturbance sensitivity is calculated using a negative exponential function based on the ratio of the path distance from the current location of the vehicle to be dispatched to the park entrance to the current average speed of the path. Through formula Traffic disturbance sensitivity was calculated. ;in, Let i be the shortest path distance from the current location of the vehicle to be dispatched to the park entrance. This represents the current average travel speed along this route. Traffic sensitivity coefficient.

[0007] In conjunction with the first aspect above, in one possible implementation, the method for analyzing the dynamic priority score of each vehicle to be dispatched includes: Based on the current distribution of all vehicles awaiting dispatch across four dimensions—goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient, and traffic disturbance sensitivity—the information entropy method is used to dynamically determine the fusion weights for each dimension. ; ; ; Where k is the index variable for the four dimensions, k=1, 2, 3, 4; This is a dummy variable used in the summation process to iterate through the four dimensions; , They represent the first The information entropy values ​​of r dimensional indicators; This represents the total number of vehicles currently awaiting dispatch. Corresponding in sequence , , , ; The four dimensions are nonlinearly multiplied and fused according to their corresponding weights to obtain the dynamic priority score of the vehicle i to be scheduled. : ; in, For the first The non-linear weighted contribution of dimension to the priority of vehicle i.

[0008] In conjunction with the first aspect above, in one possible implementation, dividing the park entrance buffer into a spatiotemporal grid queue includes: The park entrance buffer zone is spatially divided into multiple continuous geographical grid units and temporally divided into equal-length scheduling time slots, forming a spatiotemporal grid queue defined by the spatial grid and the time slots. Each spatiotemporal grid represents a unique spatial-temporal resource that a vehicle to be scheduled can occupy within a specific time period.

[0009] In conjunction with the first aspect above, in one possible implementation, the calculation of the spatiotemporal conflict potential energy between each pair of vehicles to be scheduled includes: For any two vehicles i and j to be dispatched; where, ; Based on their respective planned guidance paths, calculate the minimum intersection angle of the guidance paths in space. and nearest neighbor spacing ;in, ; Based on the spatiotemporal grid sets of the two vehicles to be dispatched, which are either assigned or predicted to be occupied, determine the equivalent overlap duration of the two vehicles in the time dimension. ;in, ,when If it is , it means there is no time overlap; Obtain the dynamic priority scores of the two vehicles to be dispatched. and And calculate the priority difference coefficient between the two vehicles. ; The spatiotemporal conflict potential energy between any two vehicles to be dispatched is calculated using the following formula. : ; in, The preset safety distance threshold, , , , The positive adjustable weighting coefficients represent the basic conflict intensity, distance decay rate, path intersection sensitivity, and priority difference amplification factor, respectively.

[0010] In conjunction with the first aspect above, in one possible implementation, the calculation guide path has a minimum intersection angle in space. and nearest neighbor spacing ,include: Discretize the guidance paths for vehicles i and j into ordered line segment sequences; traverse all line segment pairs, calculate the angle between the direction vectors of each pair and the shortest Euclidean distance, and select the minimum of all angles as the minimum intersection angle. The minimum value among all distances is the nearest neighbor distance. ; where the angle between the direction vectors is an acute angle.

[0011] In conjunction with the first aspect mentioned above, one possible implementation involves analyzing the overlap duration of the two vehicles in the time dimension. ,include: Obtain the overlap size and location of the vehicles to be dispatched in the time dimension; Obtain the spatiotemporal grid occupancy sets associated with each vehicle i and vehicle j to be scheduled. Each set contains multiple spatial grids and their corresponding start and end times of occupancy. Iterate through all time interval pairs occupied by the spatiotemporal grids of the two vehicles to be dispatched, and calculate the overlap duration of each time interval pair; If any pair of spatiotemporal grids have the same spatial location, then the overlap duration of all corresponding time periods is summed to obtain the overlap duration of the two vehicles in the time dimension. ; Otherwise, if the two vehicles occupy different spatiotemporal grids but are located in adjacent or intersecting path regions, the overlap duration of all time periods is weighted and summed according to a preset spatial proximity weighting coefficient to obtain the overlap duration of the two vehicles in the time dimension. .

[0012] In conjunction with the first aspect above, in one possible implementation, the construction of the multi-objective hybrid linear programming model includes: Define decision variables, including integer variables. , indicating whether the vehicle i to be dispatched has been assigned to the spatiotemporal grid. Continuous variables This indicates vehicles awaiting dispatch. The actual waiting time for entry; Construct a multi-objective hybrid linear programming model Q: ; in, , All are adjustable weighting coefficients. For decision variables The potential energy value of spatiotemporal conflict; The constraints of the multi-objective mixed linear programming model include: Each vehicle is assigned only one spatiotemporal grid: ; Loading and unloading resource capacity constraints: ;in, For grid Corresponding physical capacity; Safe entry time window constraint: If ,but ;in, The earliest arrival time for vehicle i to be dispatched This refers to the latest permitted entry time for vehicle i to be dispatched. Maximum waiting time constraint: ;in, The actual waiting time for vehicle i to be dispatched. This represents the maximum waiting time for vehicles to be dispatched.

[0013] In conjunction with the first aspect above, in one possible implementation, the method for generating the rolling scheduling plan includes: Based on a multi-objective hybrid linear programming model, its time conflict potential energy value is analyzed. Perform a first-order Taylor expansion to construct a mixed-integer linear subproblem; The sequential convex approximation method is used to iteratively solve the mixed integer linear subproblem and update the linearization points until the change in the function value of the multi-objective mixed linear programming model between adjacent iterations is less than the preset convergence threshold or the maximum number of iterations is reached. Output the optimal allocation results of vehicles to be scheduled and the waiting time, and generate a rolling scheduling plan.

[0014] Secondly, this application provides an intelligent queuing and scheduling system for vehicles in a bulk logistics park, comprising: a data acquisition module, an analysis module, and a scheduling module; wherein, the data acquisition module is used to collect the status information of each vehicle to be scheduled within the reservation range of the park in real time; the analysis module is used to construct a dynamic priority score for each vehicle to be scheduled based on the status information; divide the park entrance buffer zone into a spatiotemporal grid queue, and calculate the spatiotemporal conflict potential energy value between each pair of vehicles to be scheduled based on the dynamic priority score and guidance path of each vehicle to be scheduled; the scheduling module is used to construct a multi-objective hybrid linear programming model with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy value, and generate a rolling scheduling plan for vehicles to be scheduled within a preset time threshold based on the multi-objective hybrid linear programming model.

[0015] This application provides an intelligent vehicle queuing and scheduling system and method for bulk logistics parks, which can effectively address the pain points commonly encountered in current parks during peak hours, such as disorderly vehicle queuing, intense resource competition, delays of time-sensitive goods, and safety hazards caused by cross-traffic. Traditional scheduling relies on manual experience or simple first-come-first-served rules, which cannot perceive the urgency of goods and the availability of resources, nor can it predict multi-vehicle path conflicts, resulting in low overall efficiency and difficulty in guaranteeing service levels. To address this, this solution innovatively integrates dynamic priority modeling and spatiotemporal conflict quantification mechanisms: on the one hand, it constructs a multi-dimensional nonlinear priority score based on real-time vehicle status (such as contract deadlines, loading and unloading requirements, and traffic speed) to achieve intelligent hierarchical classification of "fast when it should be fast, and wait when it should be waited for"; on the other hand, by discretizing the entrance area into a spatiotemporal grid queue and combining it with the guidance path to accurately calculate the spatiotemporal conflict potential energy value between vehicle pairs, the scheduling decision has spatial safety awareness. Based on this, with the objective of minimizing the total waiting time and the total spatiotemporal conflict potential energy value, a solvable mixed integer optimization model is constructed, and a rolling scheduling strategy is adopted to achieve continuous adaptation to the dynamic environment. This solution not only significantly improves the park's throughput efficiency and the fulfillment rate of high-quality vehicles, but also reduces operational risks from the source.

[0016] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0017] Figure 1 A system architecture diagram of an intelligent vehicle queuing and scheduling system for bulk logistics parks provided in this application embodiment; Figure 2 A flowchart illustrating an intelligent queuing and scheduling method for vehicles in a bulk logistics park, provided as an embodiment of this application; Figure 3 This is a schematic flowchart of a method for analyzing time conflict potential energy values ​​provided in an embodiment of this application. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] The intelligent queuing and scheduling method for vehicles in bulk logistics parks provided in this application embodiment can be applied to, for example... Figure 1 The system shown is an intelligent queuing and scheduling system for vehicles in a bulk logistics park. The system includes: a data acquisition module, an analysis module, and a scheduling module. The data acquisition module is used to collect real-time status information of each vehicle waiting to be dispatched within the reservation area of ​​the park; The analysis module is used to construct a dynamic priority score for each vehicle to be dispatched based on status information. The park entrance buffer zone is divided into a spatiotemporal grid queue. Based on the dynamic priority score and guidance path of each vehicle to be dispatched, the spatiotemporal conflict potential energy value between each pair of vehicles to be dispatched is calculated. The scheduling module is used to construct a multi-objective hybrid linear programming model with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy. Based on the multi-objective hybrid linear programming model, a rolling scheduling plan for vehicles to be scheduled within a preset time threshold is generated.

[0020] To address the technical problem that existing technologies cannot quantify the risk of spatiotemporal conflicts among multiple vehicles in the entrance area, leading to delays in high-speed cargo due to indiscriminate queuing, this application provides an intelligent queuing and scheduling method for vehicles in bulk logistics parks. This method includes: real-time collection of status information of each vehicle awaiting scheduling within the park's reservation area; constructing a dynamic priority score for each vehicle based on the status information; dividing the park's entrance buffer zone into a spatiotemporal grid queue; calculating the spatiotemporal conflict potential energy between each pair of vehicles based on their dynamic priority scores and guidance paths; constructing a multi-objective hybrid linear programming model with the objective of minimizing total waiting time and total spatiotemporal conflict potential energy; and generating a rolling scheduling plan for vehicles awaiting scheduling within a preset time threshold based on the multi-objective hybrid linear programming model. Based on this, the technical solution effectively solves the core problem of balancing efficiency and safety in high-density vehicle entry into logistics parks through a closed loop of "perception—evaluation—modeling—optimization—execution". Its advantages are as follows: First, a dynamic priority mechanism, based on real-time collected delivery deadlines, resource requirements, and traffic status, integrates timeliness, resource matching, fairness, and traffic disturbance to achieve intelligent sorting driven by business value; second, accurate conflict quantification, relying on spatiotemporal grid queues and guidance paths, abstracts complex vehicle interactions into calculable spatiotemporal conflict potential values, accurately capturing high-risk scenarios such as path intersections and close-range parallelism; third, optimized goal coordination, aiming to minimize both total waiting time and total conflict potential value, balancing traffic efficiency and operational safety; and fourth, robust scheduling execution, through a rolling planning mechanism, solving and dynamically updating within a limited time window, combined with local rescheduling to address actual disturbances, ensuring the system is efficient, stable, and implementable. Overall, it achieves a leapfrog upgrade from experience-based queuing to data-driven, safety-first, and adaptive collaborative scheduling.

[0021] like Figure 2 As shown in the embodiment of this application, a method for intelligent queuing and scheduling of vehicles in a bulk logistics park includes: S201. Real-time collection of status information of each vehicle waiting to be dispatched within the reservation area of ​​the park.

[0022] The status information includes the type of loading and unloading resources required for the vehicles to be dispatched, the contract delivery deadline, and the average traffic speed.

[0023] S202. Based on the status information, construct a dynamic priority score for each vehicle to be dispatched.

[0024] The dynamic priority score is obtained by nonlinearly integrating the goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient and traffic disturbance sensitivity.

[0025] The goods timeliness factor represents the urgency of goods delivery; the resource matching degree represents the degree of compatibility between the loading and unloading resources required by the vehicle to be dispatched and the currently available resources in the park; the queuing fairness adjustment coefficient represents the degree of inhibition exerted by the frequent acquisition of priority in historical dispatch on the current priority; and the traffic disturbance sensitivity represents the degree to which the vehicle to be dispatched arrives at the park entrance on time due to the influence of the external traffic environment.

[0026] S203. Divide the park entrance buffer zone into a spatiotemporal grid queue. Based on the dynamic priority score and guidance path of each vehicle to be dispatched, calculate the spatiotemporal conflict potential energy value between each pair of vehicles to be dispatched.

[0027] S204. Under the conditions of satisfying the constraints of loading and unloading resource capacity, safe entry time window and maximum waiting time limit, construct a multi-objective hybrid linear programming model with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy; based on the multi-objective hybrid linear programming model, generate a rolling scheduling plan for vehicles to be scheduled with a preset time threshold.

[0028] The total spatiotemporal conflict potential energy value is the sum of the spatiotemporal conflict potential energy values ​​between all pairs of vehicles to be dispatched.

[0029] Based on the above technical solutions, this application provides an intelligent queuing and scheduling method for vehicles in large-scale logistics parks. In scenarios with high-concurrency vehicle entry in large logistics parks, traditional scheduling methods struggle to coordinate and optimize efficiency, safety, and fairness: fixed priorities easily lead to delays for time-sensitive goods, while first-come-first-served ignores resource matching and path conflicts, causing congestion and even safety hazards. To address this, this application proposes an integrated intelligent scheduling solution: It constructs a dynamic priority score by collecting vehicle status data in real time, integrating timeliness urgency, resource adaptability, fairness mitigation, and traffic disturbance to achieve business value perception; it discretizes the entry buffer into a spatiotemporal grid queue and accurately quantifies the spatiotemporal conflict potential energy value between vehicles based on the guidance path; and then constructs a multi-objective hybrid linear programming model with the weighted sum of total waiting time and total time conflict potential energy value as the objective, generating a rolling scheduling plan under strict constraints such as resource capacity and time windows. The advantages of this solution are: dynamic priority adapts to business changes, conflict potential energy accurately characterizes collaborative risks, the spatiotemporal grid supports efficient optimization and execution, and the rolling mechanism ensures robust response to disturbances. Overall, it has achieved a leap from "passive queuing" to "proactive collaboration, safety and efficiency" in intelligent scheduling, significantly improving the park's throughput capacity and service quality.

[0030] In one possible implementation of this application embodiment, the above-mentioned S202 can be specifically implemented by the following S301, S302 and S303, which are described in detail below: S301. The cargo timeliness factor is calculated based on the time difference between the current time and the contract delivery deadline of the cargo carried by the vehicle to be dispatched, using a monotonically increasing S-shaped nonlinear function, namely the Sigmoid function. Through formula Calculate the product timeliness factor Where i is the index variable of the vehicle to be dispatched, i=1,2,…,n, and n is a positive integer. For the current moment, The contract delivery deadline for the goods carried by vehicle i to be dispatched. For time sensitivity coefficient, ; It should be pointed out that, If the value is negative (e.g., -2 hours): it means there are still 2 hours until the deadline → not urgent. If the value is positive (e.g., +0.5 hours): it means the deadline has passed by 30 minutes → extremely urgent.

[0031] The smaller (earlier) the input to the Sigmoid function, the closer the output is to 1; the larger the input (later / overdue), the closer the output is to 0; the most drastic changes occur near the critical point (such as 1 hour before the deadline).

[0032] The larger the value, the more lenient the timeframe (priority can be slightly lower); the smaller the value, the more urgent the timeframe (scheduling priority should be increased).

[0033] The larger the value → when the deadline is approaching The steeper the descent, the more sensitive the system is to "approaching expiration"; The smaller the value, the more gradual the change, and the gentler the priority adjustment.

[0034] Resource matching degree is the directional similarity between the vector of loading and unloading resource types required by the vehicle to be dispatched and the vector of the current available resource status in the park, which is calculated by cosine similarity. Through formula Calculate the resource matching degree ; in, Let be the vector of loading and unloading resource types required for vehicle i to be scheduled. This is the current state vector of available resources in the park. The queuing fairness adjustment coefficient is calculated using an exponential decay function based on the number of times the vehicle to be dispatched was allowed to enter the field early within a preset historical period. Through formula The queuing fairness adjustment coefficient was calculated. ;in, This represents the number of times vehicle i, to be dispatched, is allowed to enter the depot early within a preset historical period. This is a fairness penalty coefficient; Traffic disturbance sensitivity is calculated using a negative exponential function based on the ratio of the path distance from the current location of the vehicle to be dispatched to the park entrance to the current average speed of the path. Through formula Traffic disturbance sensitivity was calculated. ;in, Let i be the shortest path distance from the current location of the vehicle to be dispatched to the park entrance. This represents the current average travel speed along this route. Traffic sensitivity coefficient.

[0035] S302. Based on the distribution of indicators of all vehicles to be dispatched in four dimensions—goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient, and traffic disturbance sensitivity—the information entropy method is used to dynamically determine the fusion weight of each dimension. ; ; ; Where k is the index variable for the four dimensions, k=1, 2, 3, 4; r is a dummy variable in the summation process, used to traverse the four dimensions; , These represent the information entropy values ​​of the k-th and r-th dimension indicators, respectively. This represents the total number of vehicles currently awaiting dispatch. Corresponding in sequence , , , ; S303. Perform nonlinear multiplication and fusion of the four factors according to their corresponding weights to obtain the dynamic priority score of the vehicle i to be scheduled. Assume k=1 represents the product's timeliness factor. k=2 indicates the resource matching degree k=3 represents the queuing fairness adjustment coefficient, and k=4 represents the traffic disturbance sensitivity. ; ; in, For the first Dimensions of vehicles Non-linear weighted contribution terms for priority.

[0036] It should be noted that information entropy It is an indicator that measures the "disorder" of a variable's distribution; if all vehicle values ​​in a certain dimension are roughly the same → uniform distribution → →Small weights; if there are large differences in a certain dimension (e.g., some vehicles are severely overdue, while others are very lenient) →uneven distribution→ →The weight is large. Therefore, the weight... In other words, the greater the difference between vehicles in a given dimension, the better it can distinguish vehicle priorities, and therefore it should be given higher weight.

[0037] For example, suppose there are 3 vehicles currently waiting to be dispatched, as shown in Table 1 below; Table 1 Current Scheduling Information

[0038] Step 1: Calculate the information entropy for each dimension : by For example: Normalization: Similarly, we get 0.333 and 0.292; Calculate entropy: More uniform distribution → higher entropy → lower weight; Let's look again. Values ​​of 0.7, 0.8, and 0.6 indicate greater differences and lower entropy (approximately 0.95). Larger → Higher weight; Step 2: Calculate the weights: Assume the four-dimensional entropies are as follows: , , , ,but , , , The sum is 0.29; Weighting coefficients: , , , ; Conclusion: Fairness dimension showed the greatest difference and had the highest weight; timeliness dimension showed the smallest difference and had the lowest weight.

[0039] Continuing with the previous example, let's take vehicle A: , , , Weights: , , , ; calculate: ; Similarly, we can obtain , ; Therefore, the order is: C>B>A. Even though A has the best timeliness, it is still ranked last because of its low fairness factor (it has been waiting for a long time).

[0040] Based on the aforementioned technical solutions, in high-density bulk logistics parks, vehicle types are diverse, delivery time requirements vary greatly, and resources are limited. Using fixed priorities or single indicators (such as first-come, first-served) for scheduling can easily lead to delays in time-sensitive goods, low resource matching efficiency, and long-term waiting vehicles experiencing "starvation," making it difficult to balance efficiency, fairness, and service quality. Therefore, this application proposes a dynamic priority calculation mechanism: quantifying four key dimensions—goods timeliness, resource matching degree, queuing fairness, and traffic disturbance sensitivity; then, using the information entropy method to adaptively determine the weights of each dimension, allowing the most distinctive indicator to dominate the ranking; and using a weighted geometric average for nonlinear fusion to strengthen the "weakest link effect," preventing a severe deterioration in one aspect from being masked by others. The advantages of this solution are: dynamically adaptive weights, eliminating the need for manual parameter tuning; multi-dimensional collaborative decision-making, balancing business urgency and system fairness; and nonlinear fusion that aligns with scheduling intuition, significantly improving overall traffic efficiency and providing high-quality input for subsequent conflict potential modeling and rolling optimization, forming the core foundation for intelligent, robust, and interpretable scheduling.

[0041] In one possible implementation of the embodiments of this application, combined with Figure 2 ,like Figure 3 As shown, the above S203 can be specifically implemented through the following S401 and S402, which are explained in detail below: S401. The park entrance buffer zone is spatially divided into multiple continuous geographical grid units and temporally divided into equal-length scheduling time slots, forming a spatiotemporal grid queue defined by the spatial grid and the time slots. Each spatiotemporal grid represents a unique spatial-temporal resource that a vehicle to be scheduled can occupy within a specific time period.

[0042] For example, the core idea of ​​this step is to abstract the "entry buffer" in the physical world into a discrete resource pool composed of "spatial units × time units", namely, a "spatiotemporal grid queue".

[0043] Spatial division: geographic grid unit; Divide the buffer area in front of the park entrance (such as queuing lanes and waiting areas) into several continuous, non-overlapping geographical grids (e.g., each 5m×5m). Each cell is called a spatial grid cell and is identified by a unique number, such as G1, G2, ..., Gm; A vehicle can only occupy one or a few consecutive adjacent grids at any given time.

[0044] Time allocation: Equal-length scheduling time slots; Divide a future period of time (such as 10 minutes within a rolling scheduling window) into fixed-length time segments, such as 30 seconds per time slot; each time slot is represented by a time index, such as t=0 (current moment), t=1 (30 seconds later), t=2 (60 seconds later), and so on; all vehicle entry, movement, and stop must be aligned to these time slots.

[0045] Spatiotemporal grid = spatial grid + time slot; A spatiotemporal grid is represented as This means: "Within time slot t, spatial grid g is occupied by a certain vehicle"; It is a unique spatiotemporal resource unit that cannot be occupied by multiple vehicles at the same time (hard constraint). Multiple consecutive When combined, they form a projection of a vehicle's guiding path in space and time.

[0046] For example, in a coal logistics park, there is currently a 100-meter-long buffer queuing lane at the east entrance of the park. The average length of the vehicles is about 15 meters, with a safety distance of 5 meters → one vehicle can be accommodated every 20 meters. Therefore, the area is divided into 5 spatial grids: G1 (closest to the gate), G2, G3, G4, and G5 (farthest). The scheduling time slot is set to 30 seconds per slot; the current time is 10:00:00, and the rolling scheduling window is for the next 5 minutes (a total of 10 time slots: t=0 to t=9).

[0047] Dispatch plan for vehicle A: The system assigns the following spatiotemporal grid sequence to it: t=2 (10:01:00) → occupy G5, t=3 (10:01:30) → occupy G4, t=4 (10:02:00) → occupy G3, t=5 (10:02:30) → occupy G2, t=6 (10:03:00) → occupy G1 (enter the gate); that is, vehicle A will slowly drive in from the farthest end G5 starting at 10:01 and arrive at the gate at 10:03.

[0048] Dispatch plan for vehicle B: If the system simultaneously allocates B to occupy G5 at t=3, it will not conflict with A occupying G5 at t=2 (the times are staggered). However, if allocation B also occupies G5 at t=2, it violates the spatiotemporal grid exclusive constraint and will be automatically excluded by the model.

[0049] Conflict detection example: If path A is G5→G4→G3 and path B is G3→G4→G5 (muting in opposite directions), and both occupy G4 at t=4, the spatiotemporal grid (G4, t=4) is repeatedly allocated, resulting in high conflict potential. The scheduling model will avoid this approach.

[0050] S402. The analysis process of the spatiotemporal conflict potential energy values ​​between any two vehicles to be dispatched includes: S4021. For any two vehicles i and j to be dispatched; where... ; For example, assuming there are currently 5 cars in the park, 10 [unclear] need to be calculated. .

[0051] S4022. Based on the respective planned guidance paths, calculate the minimum intersection angle of the guidance paths in space. and nearest neighbor spacing ;in, The details are as follows: S40221, Vehicles awaiting dispatch Vehicles to be dispatched The guiding paths are discretized into ordered line segment sequences. and ; Among them, each line segment Defined by two geographic coordinate points, As the starting point of the guiding path, This is the endpoint of the guiding path.

[0052] For example: if a car drives from outside the park into the gate and the route passes through 3 turning points, it can be divided into 4 line segments.

[0053] S40222, Traverse all line segment pairs ;in, , ; S40223, Calculate line segments and The direction vector of the line and ; S40224, Calculation and The angle between and take , making ; S40225, Calculate line segments and The shortest Euclidean distance between The endpoint effect is handled using a point-to-line segment distance algorithm; From all The minimum value is selected as the minimum cross angle: ; From all The minimum value is selected as the nearest neighbor distance: .

[0054] It should be noted that the guidance path is one or more feasible driving trajectories pre-generated by the park's intelligent dispatch system for each vehicle waiting to enter the park, from its current location (or reserved entry point) to the designated entrance channel (or buffer zone target grid), used to guide the vehicles waiting to be dispatched to pass safely, orderly and without conflict in the park's perimeter and entrance areas.

[0055] Minimum Cross Angle The path with the smallest included angle (acute angle) among all intersections of the two paths. Parallel and in the same direction (low risk); Vertical intersection (high risk).

[0056] Nearest neighbor distance The minimum Euclidean distance between any two points on two paths (unit: meters).

[0057] For example, suppose vehicle A travels straight from east to west → the path is a single line segment. (Unit: meters, Y=50 is the main east-west road); Vehicle B: Turn right from south to north onto the main road → The route is split into two segments: (North and South Sections) (Curve section).

[0058] Line segment pair combinations (1×2=2 pairs in total): Pair1: Pair2: ; Pair 1 Analysis (Straight vs. North / South Segment): Direction vector: , included angle: ; Shortest distance: A is at Y=50, B1 is at X=60, Y∈[0,30] → the closest point is from (60,30) to (60,50) → ; Pair2 Analysis (Straight vs. Curving Segment): ; ; ; .

[0059] Shortest distance: Calculate the shortest distance between line segments (100,50)-(20,50) and (60,30)-(40,50) → It is found that the two line segments are close to each other near (50,50), which is actually... ; Global minimum: , ; Conclusion: The two vehicles have a medium intersection angle (45°) and are very close (2.2m) at the turning section, which poses a high risk of conflict.

[0060] S4023. Based on the spatiotemporal grid sets of the two vehicles to be dispatched that are allocated or predicted to be occupied, determine the equivalent overlap duration of the two vehicles in the time dimension. ;in, ,when When the time interval is 1, it indicates that there is no time overlap; specifically as follows: S40231. Obtain the overlap size and position of the vehicles to be dispatched in the time dimension; S40232, Obtain vehicles to be dispatched Vehicles to be dispatched Each associated spatiotemporal grid occupancy set contains multiple spatial grids and their corresponding occupancy start and end times; S40233, Traverse all time period pairs occupied by spatiotemporal grids for the two vehicles to be dispatched, and calculate the overlap duration of each time period pair; S40234. If any pair of spatiotemporal grids have the same spatial location, then sum up the overlap duration of all corresponding time periods to obtain the overlap duration of the two vehicles in the time dimension. ; Otherwise, if the two vehicles occupy different spatiotemporal grids but are located in adjacent or intersecting path regions, the overlap duration of all time periods is weighted and summed according to a preset spatial proximity weighting coefficient to obtain the overlap duration of the two vehicles in the time dimension. .

[0061] For example, assume the spatiotemporal grid layout: G1–G5: east-west main lanes (G1 is close to the gate), H1–H3: north-south branch roads, H2 intersects G3 perpendicularly; Dispatch plan: Vehicle A (straight on the main road): (G5, 10:01:00–10:01:30), (G4, 10:01:30–10:02:00), (G3, 10:02:00–10:02:30); Vehicle B (merging from a side road): (H2,10:01:45–10:02:15), (G3,10:02:15–10:02:45).

[0062] Conflict Analysis: Overlapping of the same type: G3; A occupies G3 from 10:02:00 to 10:02:30, and B occupies G3 from 10:02:15 to 10:02:45. Overlapping time period: 10:02:15–10:02:30 → Second; Adjacent / intersecting overlap: H2 and G3; H2 and G3 intersect perpendicularly at the intersection → defined as spatial proximity, weighted. ; A is in G3: 10:02:00–10:02:30, B is in H2: 10:01:45–10:02:15; Overlapping time period: 10:02:00–10:02:15 → Seconds; Weighted contribution: Second; Total equivalent overlap duration: Second; Even though B has not yet entered G3, its presence at the intersection with A already creates an additional risk of conflict.

[0063] S4024. Obtain the dynamic priority scores of the two vehicles to be dispatched. and And calculate the priority difference coefficient between the two vehicles. ; S4025, the potential energy value of the spatiotemporal conflict between two vehicles awaiting dispatch. The calculation is performed using the following formula: ; in, The preset safety distance threshold, , , , The positive adjustable weighting coefficients represent the basic conflict intensity, distance decay rate, path intersection sensitivity, and priority difference amplification factor, respectively.

[0064] It should be noted that, assuming vehicle A: hazardous materials transport vehicle → dynamic priority Vehicle B: Ordinary coal car → Dynamic priority ; Geometric and time parameters: , , Seconds, safety threshold: =10m → Meets the "otherwise" condition; parameter settings: , , , ; Priority differences: ; Distance term: ; Angle item: ; Priority amplification item: ; Total potential energy: ; Comparison: If the two cars have the same priority (e.g.) ), Priority item = 1 That is, when a high-priority vehicle and a low-priority vehicle travel in similar spatiotemporal regions, the calculated spatiotemporal conflict potential energy is 2.35 times higher than the conflict potential energy generated by two vehicles of the same priority under the same geometric and temporal conditions.

[0065] Based on the above technical solutions, in high-density vehicle entry scenarios in logistics parks, traditional scheduling methods struggle to accurately quantify the spatiotemporal interaction risks between vehicles, easily leading to path conflicts, congestion, and even safety accidents. This is especially true when multiple vehicles converge at narrow entrances from different directions; relying solely on time or space is insufficient to guarantee safe coordinated passage. To address this, this application proposes a spatiotemporal conflict potential energy modeling mechanism: by dividing the spatiotemporal grid queue, continuous vehicle movement is abstracted into discrete resource occupancy, providing a unified coordinate system for conflict analysis; furthermore, it integrates four key elements—nearest neighbor distance, minimum intersection angle, time overlap duration, and dynamic priority difference—to calculate the temporal conflict potential energy value. The advantages of this scheme are: clear physical meaning, accurately capturing high-risk scenarios such as close-range intersections and perpendicular merging; the introduction of a priority difference amplification factor to proactively avoid the risk of high-priority vehicles mixing with ordinary vehicles; support for efficient computation and optimized embedding, enabling the solution of large-scale problems through sequence convex approximation; and integrated with rolling scheduling closed-loop linkage to achieve "prediction-evaluation-avoidance" as a unified whole. This technology fundamentally solves the safety and smoothness challenges of multi-vehicle collaborative entry under complex road networks, significantly improving the park's throughput capacity and scheduling intelligence.

[0066] In one possible implementation of this application embodiment, the above-mentioned S204 can be specifically described as follows: Define decision variables, including integer variables. , indicating whether the vehicle i to be dispatched has been assigned to the spatiotemporal grid. Continuous variables This indicates vehicles awaiting dispatch. The actual waiting time for entry; Construct a multi-objective hybrid linear programming model Q: ; in, , All are adjustable weighting coefficients. The spatiotemporal conflict potential energy value depends on the decision variable x; This represents the total potential energy of spatiotemporal conflict.

[0067] The constraints of the multi-objective mixed linear programming model include: Each vehicle is assigned only one spatiotemporal grid: ; It should be noted that each vehicle is only allocated one entry opportunity during the entire scheduling cycle (usually referring to the first grid in the entry buffer).

[0068] Loading and unloading resource capacity constraints: ;in, For grid The corresponding physical capacity, such as the number of lanes, parking spaces, and number of guides; It should be noted that in each spatiotemporal grid Above, the total number of vehicles occupying a grid does not exceed its capacity. .

[0069] Safe entry time window constraint: If ,but ;in, The earliest arrival time for vehicle i to be dispatched This refers to the latest permitted entry time for vehicle i to be dispatched. It should be noted that vehicles may only enter the site during the time window permitted by their contract and when traffic is feasible.

[0070] Maximum waiting time constraint: ;in, The actual waiting time for vehicle i to be dispatched. This represents the maximum waiting time for vehicles to be dispatched.

[0071] It should be noted that to prevent vehicles from waiting indefinitely, the Service Scope Agreement (SLA) must be guaranteed.

[0072] The method for generating a rolling schedule is as follows: Based on a multi-objective hybrid linear programming model, the potential energy value of time conflict is calculated. Perform a first-order Taylor expansion to construct a mixed-integer linear subproblem; The sequential convex approximation method is used to iteratively solve the mixed integer linear subproblem and update the linearization points until the change in the function value of the multi-objective mixed linear programming model between adjacent iterations is less than the preset convergence threshold or the maximum number of iterations is reached. The sequential convex approximation method stops iteration when any of the following termination conditions are met: (1) The absolute value of the difference between the objective function values ​​of two adjacent iterations is less than the preset convergence threshold; (2) The number of iterations reaches the preset maximum value.

[0073] Output the optimal allocation results of vehicles to be scheduled and the waiting time, and generate a rolling scheduling plan.

[0074] According to the rolling dispatch plan, the park's intelligent dispatch system pushes the following instructions to the vehicle terminals or mobile applications of each vehicle to be dispatched: Specify the entry time period (corresponding to time slot t); Assign the target grid (spatial grid g) to the ingress buffer; Issue a guiding path (composed of an ordered spatiotemporal grid sequence) from the current location to the target grid. Within a preset time window (e.g., the next 5 minutes): Real-time acquisition of vehicle location, speed, and environmental disturbance information; If a major deviation is detected (such as severe vehicle delays, sudden congestion, or equipment failure), a local rescheduling is triggered: only the affected vehicles and their neighboring vehicles are re-solved in their subproblems, and their assignments and paths are updated. The remaining vehicles retain their original plans, in order to balance response speed and system stability.

[0075] It should be noted that the rolling scheduling plan optimizes the entry behavior of vehicles to be scheduled within a limited time window (such as 5–15 minutes) in each decision cycle and generates specific scheduling instructions. As the time window progresses, the system re-optimizes the plan for the next window based on the latest status of the vehicles to be scheduled and environmental information. This process is repeated to achieve continuous adaptation to the dynamic environment.

[0076] Based on the above technical solutions, in scenarios with high-concurrency vehicle entry in large logistics parks, traditional scheduling methods struggle to simultaneously balance traffic efficiency, conflict safety, and service fairness: minimizing waiting time alone can easily lead to multi-vehicle path intersection conflicts; avoiding conflicts alone may cause severe delays for time-sensitive vehicles. To address this, this invention constructs a mixed-integer optimization model with the weighted sum of total waiting time and total spatiotemporal conflict potential energy as the objective, and transforms continuous motion into a computable resource allocation problem through spatiotemporal grid discretization. The core advantages of this solution are: the objective function integrates both efficiency and safety requirements, with adjustable weights to adapt to different operational strategies; the constraint system strictly ensures time window compliance, resource capacity not exceeding limits, and SLA fulfillment; the use of sequential convex approximation to locally linearize the nonlinear conflict potential energy enables efficient iterative solutions to complex problems; and rolling scheduling and local rescheduling mechanisms ensure the system can still respond in real-time and operate stably under dynamic disturbances. Overall, this technology achieves a leap from "empirical rules" to intelligent scheduling based on "data-driven, safety-first, and adaptive optimization," significantly improving park throughput and service quality.

[0077] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

Claims

1. A method for intelligent queuing and scheduling of vehicles in a bulk logistics park, characterized in that, include: Real-time collection of status information of each vehicle waiting to be dispatched within the reservation area of ​​the park, including the type of loading and unloading resources required by the vehicle to be dispatched, the contract delivery deadline, and the average passage speed. Based on the status information, a dynamic priority score is constructed for each vehicle to be dispatched. The dynamic priority score is obtained by nonlinearly fusing the goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient and traffic disturbance sensitivity. Among them, the goods timeliness factor represents the urgency of goods delivery; the resource matching degree represents the degree of compatibility between the loading and unloading resources required by the vehicle to be dispatched and the currently available resources in the park; the queuing fairness adjustment coefficient represents the degree of inhibition exerted by the number of times priority was obtained in the past scheduling on the current priority; and the traffic disturbance sensitivity represents the degree to which the vehicle to be dispatched arrives at the park entrance on time due to the influence of the external traffic environment. The park entrance buffer zone is divided into a spatiotemporal grid queue. Based on the dynamic priority score and guidance path of each vehicle to be dispatched, the spatiotemporal conflict potential energy value between each pair of vehicles to be dispatched is calculated. The guidance path is one or more feasible driving trajectories from its current position to the designated entrance channel, which are pre-generated by the park's intelligent dispatch system for each vehicle to be dispatched. A multi-objective hybrid linear programming model is constructed with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy. Based on the multi-objective hybrid linear programming model, a rolling scheduling plan for vehicles to be scheduled within a preset time threshold is generated. The total spatiotemporal conflict potential energy is the sum of the spatiotemporal conflict potential energy values ​​between all pairs of vehicles to be scheduled.

2. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 1, characterized in that, The cargo timeliness factor is calculated using a monotonically increasing nonlinear function based on the time difference between the current moment and the contract delivery deadline of the cargo carried by the vehicle to be dispatched. Through formula Calculate the product timeliness factor Where i is the index variable of the vehicle to be dispatched, i=1,2,…,n, and n is a positive integer. For the current moment, The contract delivery deadline for the goods carried by vehicle i to be dispatched. For time sensitivity coefficient, ; The resource matching degree is the similarity between the loading and unloading resource type vector required by the vehicle to be dispatched and the current available resource status vector in the park, which is calculated by cosine similarity. Through formula Calculate the resource matching degree ; in, Let be the vector of loading and unloading resource types required for vehicle i to be scheduled. This is the state vector of available resources in the park at the current time t; The queuing fairness adjustment coefficient is calculated using an exponential decay function based on the number of times the vehicle to be scheduled has been allowed to enter the venue early within a preset historical period. Through formula The queuing fairness adjustment coefficient was calculated. ;in, This represents the number of times vehicle i, to be dispatched, is allowed to enter the depot early within a preset historical period. This is a fairness penalty coefficient; The traffic disturbance sensitivity is calculated using a negative exponential function based on the ratio of the path distance from the current location of the vehicle to be dispatched to the park entrance to the current average speed of the path. Through formula Traffic disturbance sensitivity was calculated. ;in, Let i be the shortest path distance from the current location of the vehicle to be dispatched to the park entrance. This represents the current average travel speed along this route. Traffic sensitivity coefficient.

3. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 2, characterized in that, The method for analyzing the dynamic priority score of each vehicle to be dispatched includes: Based on the current distribution of all vehicles awaiting dispatch across four dimensions—goods timeliness factor, resource matching degree, queuing fairness adjustment coefficient, and traffic disturbance sensitivity—the information entropy method is used to determine the fusion weights for each dimension. ; ； ； Where k is the index variable of the four dimensions, k=1, 2, 3, 4; r is a dummy variable in the summation process, used to traverse the four dimensions; , These represent the information entropy values ​​of the k-th and r-th dimension indicators, respectively. This represents the total number of vehicles currently awaiting dispatch. Corresponding in sequence , , , ; The four dimensions are nonlinearly multiplied and fused according to their corresponding weights to obtain the dynamic priority score of the vehicle i to be scheduled. : ； in, For the first The non-linear weighted contribution of dimension to the priority of vehicle i.

4. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 1, characterized in that, The process of dividing the park entrance buffer zone into a spatiotemporal grid queue includes: The park entrance buffer zone is spatially divided into multiple continuous geographic grid units and temporally divided into equal-length scheduling time slots, forming a spatiotemporal grid queue defined by the spatial grid and the time slots. Each spatiotemporal grid represents a unique spatial-temporal resource that a vehicle to be scheduled can occupy within a specific time period.

5. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 2, characterized in that, The calculation of the spatiotemporal conflict potential energy between each pair of vehicles to be dispatched includes: For any two vehicles i and j to be dispatched; where, ; Based on their respective planned guidance paths, calculate the minimum intersection angle of the guidance paths in space. and nearest neighbor spacing ;in, ; Based on the spatiotemporal grid sets of the two vehicles to be dispatched, which are either assigned or predicted to be occupied, determine the equivalent overlap duration of the two vehicles in the time dimension. ;in, ,when If it is , it means there is no time overlap; Obtain the dynamic priority scores of the two vehicles to be dispatched. and And calculate the priority difference coefficient between the two vehicles. ; The spatiotemporal conflict potential energy between any two vehicles to be dispatched is calculated using the following formula. : ； in, The preset safety distance threshold, , , , The positive adjustable weighting coefficients represent the basic conflict intensity, distance decay rate, path intersection sensitivity, and priority difference amplification factor, respectively.

6. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 5, characterized in that, The calculation guide path has the minimum intersection angle in space. and nearest neighbor spacing ,include: Discretize the guidance paths for vehicles i and j into ordered line segment sequences; traverse all line segment pairs, calculate the angle between the direction vectors of each pair and the shortest Euclidean distance, and select the minimum of all angles as the minimum intersection angle. The minimum value among all distances is the nearest neighbor distance. ; where the angle between the direction vectors is an acute angle.

7. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 5, characterized in that, Analyze the overlap duration of the two vehicles in the time dimension. ,include: Obtain the overlap size and location of the vehicles to be dispatched in the time dimension; Obtain vehicles to be dispatched Each spatiotemporal grid occupancy set is associated with a vehicle j to be dispatched, and each set contains multiple spatial grids and their corresponding occupancy start and end times; Iterate through all time interval pairs occupied by the spatiotemporal grids of the two vehicles to be dispatched, and calculate the overlap duration of each time interval pair; If any pair of spatiotemporal grids have the same spatial location, then the overlap duration of all corresponding time periods is summed to obtain the overlap duration of the two vehicles in the time dimension. ; Otherwise, if the two vehicles occupy different spatiotemporal grids but are located in adjacent or intersecting path regions, the overlap duration of all time periods is weighted and summed according to a preset spatial proximity weighting coefficient to obtain the overlap duration of the two vehicles in the time dimension. .

8. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 2, characterized in that, The construction of the multi-objective hybrid linear programming model includes: Define decision variables, including integer variables. This indicates vehicles awaiting dispatch. Whether it is assigned to the spatiotemporal grid Continuous variables This indicates vehicles awaiting dispatch. The actual waiting time for entry; Construct a multi-objective hybrid linear programming model Q: ； in, , All are adjustable weighting coefficients. The spatiotemporal conflict potential energy value depends on the decision variable x; The constraints of the multi-objective mixed linear programming model include: Each vehicle is assigned only one spatiotemporal grid: ; Loading and unloading resource capacity constraints: ;in, For grid Corresponding physical capacity; Safe entry time window constraint: If ,but ;in, The earliest arrival time for vehicle i to be dispatched This refers to the latest permitted entry time for vehicle i to be dispatched. Maximum waiting time constraint: ;in, The actual waiting time for vehicle i to be dispatched. This represents the maximum waiting time for vehicles to be dispatched.

9. The intelligent queuing and scheduling method for vehicles in a bulk logistics park according to claim 1, characterized in that, The method for generating the rolling scheduling plan includes: Based on a multi-objective hybrid linear programming model, the potential energy value of spatiotemporal conflict is... Perform a first-order Taylor expansion to construct a mixed-integer linear subproblem; The sequential convex approximation method is used to iteratively solve the mixed integer linear subproblem and update the linearization points until the change in the function value of the multi-objective mixed linear programming model between adjacent iterations is less than the preset convergence threshold or the maximum number of iterations is reached. The optimal vehicle allocation result and waiting time are output to generate a rolling scheduling plan.

10. A vehicle intelligent queuing and scheduling system for bulk logistics parks, operating based on the vehicle intelligent queuing and scheduling method for bulk logistics parks as described in any one of claims 1-9, characterized in that, It includes a data acquisition module, an analysis module, and a scheduling module; The data acquisition module is used to collect the status information of each vehicle waiting to be dispatched within the reservation area of ​​the park in real time. The analysis module is used to construct a dynamic priority score for each vehicle to be dispatched based on the status information. The park entrance buffer zone is divided into a spatiotemporal grid queue. Based on the dynamic priority score and guidance path of each vehicle to be dispatched, the spatiotemporal conflict potential energy value between each pair of vehicles to be dispatched is calculated. The scheduling module is used to construct a multi-objective hybrid linear programming model with the goal of minimizing the total waiting time and the total spatiotemporal conflict potential energy. Based on a multi-objective hybrid linear programming model, a rolling scheduling plan for vehicles to be scheduled within a preset time threshold is generated.

Images