Fleet dynamic scheduling method based on real-time vehicle location and task priority

By acquiring the real-time location of vehicles and task priorities, the availability of alternative vehicles for candidate vehicles and the degree to which response time is affected are determined. This solves the scheduling imbalance problem caused by multi-task competition in fleet scheduling and achieves dynamic balance between vehicles and tasks and scheduling stability.

CN123117670APending Publication Date: 2026-11-17HUNAN MOUNTAIN ELECTRIC VEHICLE CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611351720.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-02
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing fleet scheduling methods cannot effectively handle scheduling imbalances when the same vehicle is simultaneously assigned to multiple tasks, especially when task requests have uncertain arrival times, rigid response times, significant differences in service capabilities, and dynamic changes in task priorities. Furthermore, they fail to take into account the real-time location of vehicles, task priorities, service capabilities, and the availability of alternative vehicles.

Method used

By acquiring the real-time location of vehicles, current task status, and service capabilities, and combining the location, type, response time limit, and priority of the tasks to be executed, the availability of alternative vehicles for candidate vehicles and the degree to which response time limits are affected are determined. Based on the task priority and the availability of alternative vehicles, the scheduling relationship between vehicles and tasks is readjusted, and succession constraint records are generated to ensure the continuity and stability of scheduling.

Benefits of technology

It improves the accuracy of vehicle-task matching, reduces the risk of invalid and duplicate scheduling, maintains the real-time and continuous nature of the scheduling process, and improves fleet resource utilization and task response reliability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117670A_ABST
    Figure CN123117670A_ABST
Patent Text Reader

Abstract

This invention relates to a dynamic fleet scheduling method based on real-time vehicle location and task priority, belonging to the field of fleet scheduling technology. The method includes: acquiring the real-time location, current task status, estimated idle time, and service capacity of fleet vehicles; acquiring the task location, task type, response time limit, and task priority of tasks to be executed; extracting candidate vehicles based on the above information; when multiple task requests correspond to the same candidate vehicle, determining the availability of alternative vehicles and the impact on response time limits for each task after losing the candidate vehicle; determining the affiliation of the same candidate vehicle based on task priority; re-incorporating tasks that have not obtained the candidate vehicle into the scheduling sequence; updating the vehicle-task scheduling relationship; thereby forming a dynamic balance between task priority and task response time limits, reducing scheduling conflicts caused by multiple tasks competing for a single vehicle, and improving the continuity and stability of dynamic fleet scheduling.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of fleet scheduling technology, specifically a dynamic fleet scheduling method based on real-time vehicle location and task priority. Background Technology

[0002] In existing technologies, fleet dispatching for scenarios such as logistics transportation, autonomous driving operations, emergency response, and park operations typically requires comprehensive matching among vehicle real-time location, vehicle current status, task location, task type, response time limit, and task priority. Several dynamic dispatching schemes have been proposed in recently published Chinese patent documents. For example, patent publication number CN114331220A discloses a method and device for dispatching passenger vehicle transportation vehicles based on dynamic order priority. This method calculates the urgency of orders based on the current time, the latest delivery time of the order, the order joining time, and the direct transportation time, thereby determining the order priority. It then combines the current vehicle location to generate an initial dispatching solution, perform clustering, and optimization to obtain the vehicle dispatching solution. Another example is patent publication number CN116612654A, which discloses an autonomous driving fleet dispatching method, device, and electronic equipment. This method proposes global path planning for the autonomous driving fleet in the target area, determining local task paths based on the driving information of the task vehicles, and then optimizing the local task paths. These schemes can improve vehicle dispatching efficiency, path planning quality, and transportation resource utilization to a certain extent, providing a technical foundation for dynamic dispatching in multi-vehicle, multi-task environments.

[0003] However, in actual fleet scheduling, task requests are characterized by uncertain arrival times, rigid response times, significant differences in service capabilities, and dynamic changes in task priorities. Simply allocating tasks based on task priority, vehicle location, or path cost can easily lead to vehicle competition when the same vehicle is assigned to multiple tasks simultaneously. If only high-priority tasks are fixedly assigned to the optimal candidate vehicle, lower-priority tasks lacking alternatives may experience response timeouts, resulting in an imbalance in the overall scheduling outcome. Conversely, using proximity, lowest cost, or optimal path as the allocation criteria may ignore the vehicle's current task status, estimated idle time, available follow-up tasks, and the cascading impact of losing a candidate vehicle on response times. Existing publicly available solutions primarily focus on order urgency calculation, global or local path optimization, initial vehicle scheduling solution construction, and objective function optimization. However, when the same candidate vehicle corresponds to multiple task requests, there is a lack of more targeted mechanisms for simultaneously evaluating the availability of alternative vehicles for tasks not assigned to that vehicle and the degree of impact on response times, and using this information to determine candidate vehicle affiliation, reintegrate tasks not assigned to vehicles into the scheduling sequence, and update the vehicle-task scheduling relationship. Therefore, it is necessary to propose a dynamic fleet scheduling method that can take into account vehicle real-time location, task priority, service capacity, response time, and availability of alternative vehicles. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic fleet scheduling method based on real-time vehicle location and task priority, thereby addressing some of the drawbacks and shortcomings pointed out in the background art.

[0005] The present invention adopts the following technical solution to solve the above-mentioned technical problems:

[0006] The system obtains the real-time location, current task status, estimated idle time and service capacity of the fleet vehicles, obtains the task location, task type, response time limit and task priority of the task to be executed, and extracts candidate vehicles based on the real-time location, the current task status, the estimated idle time and the service capacity.

[0007] When multiple task requests correspond to the same candidate vehicle, the availability of alternative vehicles and the degree of impact on response time limits are determined for each task after the loss of the same candidate vehicle. The availability of alternative vehicles refers to whether there are alternative vehicles that meet the service capacity and response time limit requirements. The degree of impact on response time limits refers to the expected response timeout duration or the result of not being able to meet the response time limit.

[0008] The affiliation of the same candidate vehicle is determined based on the task priority, the availability of the alternative vehicle, and the degree to which the response time limit is affected. Tasks that do not obtain the same candidate vehicle are re-included in the scheduling sequence, and the scheduling relationship between vehicles and tasks is updated.

[0009] Furthermore, when determining the availability of the alternative vehicle, the proposed alternative vehicle is verified together with the currently executing task and the subsequently assigned tasks; if, after calling the proposed alternative vehicle, there is no vehicle that meets the response time limit for the currently executing task or the subsequently assigned task, the proposed alternative vehicle is excluded from the alternative vehicles.

[0010] Furthermore, when determining the affiliation of the same candidate vehicle, if the priority of the first task is higher than that of the second task, the first task has an alternative vehicle that meets the service capability and response time requirements, and the second task cannot meet the response time requirement after losing the same candidate vehicle, the same candidate vehicle is assigned to the second task.

[0011] Furthermore, when a task that has not obtained the same candidate vehicle is re-included in the scheduling sequence, a succession constraint record with a constraint validity period is generated, and the succession constraint record specifies the candidate successor vehicle reserved for the task; before the task determines the successor vehicle, new tasks with the same task priority as the task are restricted from occupying the candidate successor vehicle.

[0012] Furthermore, when determining whether there is a vehicle that meets the response time limit for the currently executing task or the subsequent assigned task, a task succession chain is generated starting from the vehicle to be replaced, and subsequent succession vehicles are searched along the task succession chain; if there are tasks in the task succession chain for which subsequent succession vehicles cannot be determined, the vehicle to be replaced is excluded from the replacement vehicles.

[0013] Furthermore, when the currently executing task is in a state where it has reached the task location, started service, or the estimated remaining service time is less than a preset completion time threshold, the vehicle to be replaced is configured as a service protection vehicle, and the service protection vehicle is prohibited from being used as a replacement vehicle for tasks other than the currently executing task.

[0014] Furthermore, after excluding the proposed replacement vehicle, vehicles that meet the service capacity requirements are first selected from vehicles within a preset distance range to supplement the currently executing task or the subsequently assigned task with vehicles that meet the response time limit; before the vehicles are supplemented, tasks within a preset service area with a task priority lower than the currently executing task or the subsequently assigned task are restricted from calling vehicles within the preset distance range, where the preset service area is the vehicle dispatch area to which the proposed replacement vehicle belongs.

[0015] Furthermore, during the validity period of the constraint, if a successor vehicle has been determined for a task that has not obtained the same candidate vehicle, the occupancy restriction on the candidate successor vehicle corresponding to the succession constraint record is lifted; if a successor vehicle has not been determined for the task, the occupancy restriction is maintained.

[0016] Furthermore, during the period when the candidate successor vehicle corresponding to the succession constraint record is restricted from being occupied, when a new task with a task priority higher than the task priority of the task that has not obtained the same candidate vehicle requests to call the candidate successor vehicle, the occupation restriction of the candidate successor vehicle is lifted, and the task that has not obtained the same candidate vehicle is kept in the scheduling sequence.

[0017] Furthermore, after the restriction on the occupancy of the candidate vehicle is lifted, the task that has not obtained the same candidate vehicle is transferred to the reserved vehicle status; in the reserved vehicle status, the scheduling order of new tasks with the same task priority as the task is suspended from updating the task's scheduling order until the task determines a vehicle or the constraint expires.

[0018] This invention extracts candidate vehicles by acquiring the real-time location, current task status, estimated idle time and service capacity of fleet vehicles, and combining the task location, task type, response time limit and task priority of the task to be executed. This makes vehicle selection no longer dependent on the current location or distance factors, but is simultaneously limited by the vehicle's serviceability and task time limit constraints. This improves the matching accuracy between the candidate vehicle set and the actual execution needs, and reduces the risks of invalid scheduling, duplicate scheduling and failure to fulfill the task after scheduling.

[0019] When multiple task requests correspond to the same candidate vehicle, this invention does not mechanically determine vehicle allocation based on task priority. Instead, it further determines the availability of alternative vehicles after each task loses the candidate vehicle and the degree to which the response time is affected. This allows for the identification of tasks with lower priority but insufficient alternative resources, and those that may time out after losing a vehicle. Furthermore, the candidate vehicle allocation decision balances task urgency and resource scarcity, avoiding the problem of overall response failure due to local optima caused by a single priority rule.

[0020] This invention also re-includes tasks that have not obtained the same candidate vehicle into the scheduling sequence and updates the scheduling relationship between vehicles and tasks, enabling the scheduling process to have continuous correction capabilities. In scenarios where dynamic tasks continuously enter and vehicle states continuously change, it can maintain the real-time and continuity of scheduling relationships, improve fleet resource utilization, task response reliability, and scheduling stability under multi-task concurrency conditions. Attached Figure Description

[0021] Figure 1 This is a flowchart of the vehicle dynamic scheduling method of the present invention.

[0022] Figure 2 This is a diagram illustrating the determination of the competitive attribution of the same candidate vehicle in Embodiment 1 of the present invention.

[0023] Figure 3 This is a comparison chart of task response timeout duration and affected evaluation values ​​in Embodiment 1 of the present invention.

[0024] Figure 4 This is a vehicle configuration diagram for task succession chain verification and service protection in Embodiment 1 of the present invention.

[0025] Figure 5 This is a diagram showing the generation and occupancy restriction of continuation constraint records in Embodiment 2 of the present invention.

[0026] Figure 6 This is a diagram showing the remaining time of the constraint validity period in Embodiment 2 of the present invention.

[0027] Figure 7 This is a diagram showing the retention of succession status and scheduling order in Embodiment 2 of the present invention. Detailed Implementation

[0028] As attached Figure 1 As shown, this embodiment provides a dynamic fleet scheduling method based on real-time vehicle location and task priority. This method can be executed by a fleet scheduling platform, a scheduling server, or an on-board scheduling system. The scheduling system first obtains the real-time location, current task status, estimated idle time, and service capability of each vehicle in the fleet, and obtains the task location, task type, response time limit, and task priority of the task to be executed. Real-time location can be reported through vehicle positioning devices. The current task status is used to indicate whether the vehicle is idle, has traveled to the task location, is executing a task, or is waiting for a follow-up task. The estimated idle time is used to determine the time when the vehicle can participate in subsequent scheduling after completing the current task; when the vehicle is idle, the estimated idle time is the time when the vehicle can currently participate in scheduling. Service capability is used to determine whether the vehicle can execute the corresponding task type. Service capability includes at least one of the following: vehicle carrying capacity, vehicle operating equipment, driver qualifications, vehicle availability status, or task type matching relationship.

[0029] The dispatching system selects candidate vehicles from the fleet based on the distance between the real-time vehicle location and the task location, the vehicle's current task status, estimated idle time, and service capacity. The distance relationship includes the travel distance or estimated arrival time from the vehicle to the task location, and candidate vehicles are those that can reach the task location within the response time limit and meet the service capacity requirements. When the same candidate vehicle simultaneously meets the allocation conditions for multiple task requests within the same dispatching cycle, the dispatching system analyzes the scheduling impact of losing the same candidate vehicle for each task, determines whether there is a substitute vehicle that meets the service capacity and response time limit requirements, and calculates or judges the degree to which the response time limit is affected after losing the same candidate vehicle. The degree of impact on the response time limit can be represented by the estimated response timeout duration or by whether the response time limit cannot be met. The dispatching system then combines the task priority, the availability of substitute vehicles, and the degree of impact on the response time limit to determine which task the same candidate vehicle is ultimately assigned to. In cases where task priorities are the same, the same candidate vehicle is preferentially assigned to tasks where substitute vehicles are unavailable or where the response time limit is more significantly affected. For tasks that do not obtain the same candidate vehicle, the scheduling system re-includes it in the scheduling sequence and redetermines the scheduling relationship between the vehicle and the task based on the updated vehicle status and task status, thereby forming a dynamic balance between task priority and task response time limit and reducing scheduling conflicts caused by multiple tasks competing for a single vehicle.

[0030] When determining the availability of alternative vehicles, the scheduling system not only assesses whether the proposed alternative vehicle itself has the service capability to execute the task to be performed, but also simultaneously verifies the currently executing tasks and subsequent assigned tasks associated with the proposed alternative vehicle. This is to avoid disrupting the time constraints of the original task succession chain due to calling the proposed alternative vehicle. Specifically, starting with the proposed alternative vehicle, the scheduling system generates a task succession chain by combining the proposed alternative vehicle's current location, the task location and estimated completion time of the currently executing task, the task location and response time limit of the subsequent assigned tasks, and sequentially searches for vehicles that can take over the subsequent assigned tasks along the task succession chain. The task succession chain indicates that after the proposed alternative vehicle is called, the order in which the currently executing tasks and subsequent assigned tasks associated with that proposed alternative vehicle need to be rearranged. If a task in the task succession chain cannot determine a subsequent vehicle that meets the response time limit, it indicates that calling the proposed alternative vehicle would cause a risk of response time limit issues for the currently executing task or subsequent assigned tasks, and the scheduling system will exclude the proposed alternative vehicle from the set of alternative vehicles. If the currently executing task is in a state where it has reached its location, started service, or the estimated remaining service time is less than a preset completion time threshold, the dispatching system will configure the vehicle to be replaced as a service protection vehicle and prohibit the service protection vehicle from being used as a replacement vehicle for tasks other than the currently executing task, to ensure that tasks in progress or about to be completed are not interrupted. After the vehicle to be replaced is excluded, the dispatching system will prioritize vehicles within a preset distance range that meet the service capacity requirements to replenish the currently executing task or subsequent assigned tasks with vehicles that meet the response time limit. Before replenishing the vehicles, the dispatching system will restrict tasks within a preset service area with a task priority lower than the currently executing task or subsequent assigned tasks from calling vehicles within a preset distance range. The preset service area is the vehicle dispatching area to which the vehicle to be replaced belongs, thereby ensuring that the selection of replacement vehicles will not cause the existing task continuity chain to break and improving the continuity and stability of the fleet dynamic dispatching process.

[0031] Building upon the above embodiments, when determining the availability of alternative vehicles, the scheduling system does not solely judge whether the proposed alternative vehicle can reach the location of the task to be executed. Instead, it includes the currently executing task and subsequent assigned tasks associated with the proposed alternative vehicle in the verification scope. The scheduling system obtains the real-time location of the proposed alternative vehicle, the task location and estimated completion time of the currently executing task, the task location and response time limit of the subsequent assigned tasks, and generates a task succession chain starting from the proposed alternative vehicle. The task succession chain represents the order of tasks that need to be re-accepted between the currently executing task and the subsequent assigned tasks after the proposed alternative vehicle is invoked. The scheduling system searches for subsequent succession vehicles that can meet the response time limit along the task succession chain. Subsequent succession vehicles are determined from vehicles that are idle, waiting for a succession task, or whose estimated idle time meets the corresponding task succession requirements. Furthermore, the subsequent succession vehicle must meet the service capability requirements of the corresponding task type. When no subsequent succession vehicle can be determined for any task in the task succession chain, it indicates that invoking the proposed alternative vehicle would pose a response time limit risk to the currently executing task or the subsequent assigned tasks. Therefore, the scheduling system excludes the proposed alternative vehicle from the set of alternative vehicles. For situations where the currently executing task has already reached its location, commenced service, or its estimated remaining service time is less than a preset completion time threshold, the dispatching system will configure the proposed replacement vehicle as a service protection vehicle and prohibit it from being used as a replacement vehicle for tasks other than the currently executing task, to prevent interruption of tasks that are currently executing or nearing completion. After the proposed replacement vehicle is excluded, the dispatching system prioritizes replenishing the currently executing task or subsequent assigned tasks with vehicles within a preset distance range that meet service capacity requirements, ensuring the response time is met. Before replenishment, the dispatching system restricts tasks within a preset service area with a priority lower than the currently executing task or subsequent assigned task from calling vehicles within a preset distance range. The preset service area is the dispatching area of ​​the proposed replacement vehicle. If no replacement vehicle is found within the preset distance range, the dispatching system maintains the original dispatching relationship of the currently executing task or subsequent assigned task and re-includes the task in the dispatching sequence for subsequent scheduling. This ensures that the replacement vehicle selection process does not break the existing task continuity chain and improves the continuity and stability of the fleet dynamic dispatching process.

[0032] Based on the above embodiments, when a task that has not obtained the same candidate vehicle is re-included in the scheduling sequence, the scheduling system generates a succession constraint record for the task and configures a constraint validity period for the succession constraint record. The succession constraint record records the task identifier, task location, task priority, response time limit, the same candidate vehicle that was not obtained, and the candidate successor vehicles reserved for the task. Candidate successor vehicles can be determined based on the vehicle's real-time location, estimated idle time, service capacity, and estimated arrival time at the task location. Candidate successor vehicles do not include vehicles that have been configured as service protection vehicles and cannot be called. Before a task determines a successor vehicle, the scheduling system sets an occupancy limit on the candidate successor vehicles, preventing newly entering tasks with the same task priority from occupying candidate successor vehicles, thereby avoiding the task being squeezed out by tasks of the same priority during rescheduling. During the constraint validity period, the scheduling system continuously determines whether a task that has not obtained the same candidate vehicle has already determined a successor vehicle. If a task has determined a successor vehicle, the occupancy limit on the candidate successor vehicle corresponding to the succession constraint record is released, and the scheduling relationship between the task and the successor vehicle is updated. If a task has not yet secured a replacement vehicle, the occupancy restriction remains in place to ensure that candidate replacement vehicles are still available for subsequent task continuation. While a candidate replacement vehicle is under occupancy restriction, if a new task with a higher priority than a task that has not secured the same candidate vehicle requests to use that vehicle, the scheduling system removes the occupancy restriction and allows the new task with the higher priority to participate in scheduling the candidate replacement vehicle. Simultaneously, the scheduling system keeps the task that has not secured the same candidate vehicle in the scheduling sequence and, after removing the occupancy restriction, moves the task to a reserved continuation state. In the reserved continuation state, the scheduling system suspends updating the scheduling order of new tasks with the same priority as the task until a replacement vehicle is secured or the constraint expires. When the constraint expires and the task still has not secured a replacement vehicle, the scheduling system removes the occupancy restriction of the candidate replacement vehicle and regenerates the continuation constraint record or re-determines the scheduling order of the task in the scheduling sequence based on the vehicle's real-time location, estimated idle time, service capacity, and task status at the time of expiration. By employing the above methods, we can ensure timely response for high-priority tasks while preventing tasks that have not yet obtained the same candidate vehicle from being repeatedly pushed back by newly entered tasks with the same priority during the successive scheduling process, thereby improving the stability and continuity of fleet dynamic scheduling.

[0033] Example 1:

[0034] This embodiment provides a dynamic fleet scheduling method based on real-time vehicle location and task priority, which can be executed by a fleet scheduling platform configured by an organization. The scheduling system obtains the real-time location, current task status, estimated idle time and service capacity of the fleet vehicles, and obtains the task location, task type, response time limit and task priority of the tasks to be executed. The current task status includes idle, driving to the task location, executing the task, and waiting for the next task. The service capacity includes at least one of the following: vehicle carrying capacity, vehicle operating equipment status, driver qualifications and task type matching relationship.

[0035] In a dispatching scenario, a fleet consists of vehicles V1, V2, V3, V4, and V5. Vehicle V1 is idle and capable of handling both regular transportation and emergency delivery tasks. Vehicle V2 is waiting for a follow-up task and is capable of handling regular transportation tasks. Vehicle V3 is waiting for a follow-up task and is capable of handling emergency delivery tasks. Vehicle V4 is idle and capable of handling regular transportation tasks. Vehicle V5 is currently executing task T0, with an estimated remaining service time of 4 minutes, and is capable of handling emergency delivery tasks. Tasks to be executed include tasks T1, T2, and T3. Task T1 is an emergency delivery task with a remaining response time of 15 minutes and a high priority. Task T2 is a regular transportation task with a remaining response time of 10 minutes and a medium priority. Task T3 is a regular transportation task with a remaining response time of 20 minutes and a low priority.

[0036] The dispatching system extracts candidate vehicles based on the real-time location of each vehicle and the distance of its travel path to each task location, the current task status, estimated idle time, and service capacity. For task T1, both vehicles V1 and V3 have service capacity; for task T2, vehicles V1, V2, and V4 all have service capacity; for task T3, vehicles V2 and V4 all have service capacity. The dispatching system determines that vehicle V1 meets the allocation conditions for both task T1 and task T2, therefore vehicle V1 constitutes the same candidate vehicle. Figure 2 As shown, both tasks T1 and T2 point to vehicle V1 through a candidate relationship, and vehicle V1 is simultaneously subject to competition from both task requests; vehicles V3, V2, and V4 serve as alternative or supplementary verification objects, respectively, while vehicle V5 participates in constraint judgment as a service-protected vehicle state. Figure 2 The final home node in the diagram represents the vehicle and task scheduling relationship after the scheduling system assigns vehicle V1 to task T2.

[0037] When the same candidate vehicle V1 corresponds to two task requests, task T1 and task T2, the scheduling system determines the availability of alternative vehicles and the degree to which response time limits are affected after task T1 loses vehicle V1. The availability of alternative vehicles refers to whether there are alternative vehicles that meet the service capacity and response time limit requirements; the degree to which response time limits are affected indicates whether a predicted response timeout will occur after task T1 loses vehicle V1, or whether the response time limit cannot be met.

[0038] To facilitate the assessment of the impact on response time limits, the scheduling system uses the following formula to determine the expected response timeout duration for a task:

[0039]

[0040] in, Indicates the first The estimated response timeout duration for each task. The replacement vehicle is expected to arrive at the [number]th [location]. The time required for each task location Indicates the first The remaining response time for each task Indicates the relationship between 0 and Take the larger value between the two. The derivation logic of this formula is: when the estimated arrival time of the alternative vehicle is no later than the remaining response time limit, If the value is not greater than 0, the task has no expected response timeout, therefore the value is 0; when the expected arrival time of the replacement vehicle is later than the remaining response time limit, If the value is positive, it represents the expected response timeout duration.

[0041] In this embodiment, after task T1 loses vehicle V1, the scheduling system uses vehicle V3 as the proposed replacement vehicle. The estimated time for vehicle V3 to reach the location of task T1 is 12 minutes, and the remaining response time for task T1 is 15 minutes. Therefore:

[0042]

[0043] Therefore, vehicle V3 can meet the response time requirement of task T1. After task T2 loses vehicle V1, the scheduling system will use vehicle V2 as the replacement vehicle. The estimated time for vehicle V2 to reach the location of task T2 is 14 minutes, and the remaining response time of task T2 is 10 minutes. Therefore:

[0044]

[0045] Therefore, it can be seen that task T2 has an estimated response timeout of 4 minutes, indicating that the response time of task T2 is significantly affected after losing vehicle V1. Figure 3As shown, the expected response timeout for task T1 is 0 minutes, while the expected response timeout for task T2 is 4 minutes. Furthermore, the impact evaluation value for task T2 is significantly higher than that for task T1. Figure 3 The highly affected region is used to characterize the state where the task has insufficient replacement vehicles and an increased risk of response timeout after losing the same candidate vehicle.

[0046] When determining whether vehicle V3 can serve as a replacement for task T1, the scheduling system does not only check whether vehicle V3 can reach the location of task T1, but also includes all subsequent assigned tasks associated with vehicle V3 in the verification scope. Starting with vehicle V3, the scheduling system generates a task succession chain based on vehicle V3's current location, the task locations of subsequent assigned tasks, and their response time limits, and then searches along this task succession chain for a subsequent vehicle capable of taking over the subsequent assigned tasks. For example... Figure 4 As shown, after vehicle V3 is to replace task T1, it needs to undergo subsequent task continuity verification first, and then vehicle V4 will complete the succession task. Only if the succession path can be kept continuous can vehicle V3 be retained as the replacement vehicle for task T1.

[0047] If the scheduling system determines, along the task continuation chain, that vehicle V4 can continue the subsequent assigned task associated with vehicle V3 within a preset distance, and vehicle V4 meets the service capability requirements of that subsequent assigned task, then vehicle V3 has not caused a break in the existing task continuation chain and can be retained in the alternative vehicle set for task T1. If the scheduling system determines that any task in the task continuation chain cannot determine a subsequent assigned vehicle, it means that calling vehicle V3 would cause a response time limit risk for the subsequent assigned task, and the scheduling system will exclude vehicle V3 from the alternative vehicle set for task T1.

[0048] For vehicle V5 currently executing task T0, the scheduling system determines that the estimated remaining service time for task T0 is 4 minutes, which is less than the preset completion time threshold of 5 minutes. Therefore, vehicle V5 is configured as a service protection vehicle and is prohibited from being used as a substitute for any task other than the currently executing task T0. This process prevents ongoing or near-completion tasks from being interrupted and avoids breaking the existing task continuation chain due to calling a service protection vehicle. Figure 4 The protection path in the diagram shows that vehicle V5 enters the service protection vehicle state from the current task execution state, and further forms a scheduling restriction that prohibits the substitution of tasks other than T0. This restriction, together with the task continuation chain verification, works to determine the availability of substitute vehicles.

[0049] After the proposed replacement vehicle is ruled out, the dispatching system prioritizes replenishing the current task or subsequent assigned tasks with vehicles within a preset distance range that meet service capacity requirements, ensuring that a vehicle can meet the response time limit. Before a vehicle is replenished, the dispatching system restricts tasks within a preset service area with a lower priority than the current task or subsequent assigned tasks from using vehicles within that preset distance range. The preset service area is the vehicle dispatching area to which the proposed replacement vehicle belongs, thereby ensuring that the selection of a replacement vehicle does not break the existing task continuity chain. Figure 4 The supplementary path in the system means that the scheduling system first locks vehicles within a preset distance range, then executes vehicle supplementation for the current task or subsequent tasks, and restricts low-priority tasks from calling related vehicles before the supplementation is completed, thereby maintaining the continuity and stability of the connection chain.

[0050] The scheduling system further considers task priority, availability of alternative vehicles, and the degree to which response time is affected to determine the allocation of the same candidate vehicle V1. To ensure a consistent comparison of the impact on different tasks, the scheduling system uses the following formula to determine the task impact evaluation value:

[0051]

[0052] in, Indicates the first The impact on the evaluation value when a task loses the same candidate vehicle. This indicates the impact parameters corresponding to task priorities. This indicates the impact parameters corresponding to the availability of alternative vehicles. This indicates the aforementioned estimated response timeout duration. When no alternative vehicle meets the service capacity and response time requirements, Take the larger value; when there is an alternative vehicle that meets the service capacity and response time requirements for the task, Take the smaller value. The derivation logic of this formula is as follows: task priority, lack of alternative vehicles, and expected response timeout will all increase the scheduling impact after a task loses the same candidate vehicle. Therefore, the three types of impact parameters are combined to form an impact evaluation value, which is used to help determine the ownership of the same candidate vehicle.

[0053] In this embodiment, task T1 has a higher task priority, and the scheduling system uses its priority influence parameters. Set to 3; if there is an alternative vehicle V3 for task T1 that meets the service capacity and response time requirements, the dispatch system will... Set to 1; the task corresponding to T1 If it is 0, then:

[0054]

[0055] Task T2 has a medium priority, and the scheduling system sets its priority-affecting parameters accordingly. Set to 2; After task T2 loses vehicle V1, the estimated arrival time of vehicle V2 exceeds the response time limit, and the estimated arrival time of vehicle V4 also exceeds the response time limit. Therefore, there is no alternative vehicle that meets the service capacity and response time limit requirements, and the dispatch system will... Set to 5; corresponding to task T2 If the value is 4, then:

[0056]

[0057] This indicates that the impact of losing vehicle V1 on the evaluation value of task T2 is higher than that of task T1, suggesting that task T2 is more dependent on the same candidate vehicle V1. Figure 3 By comparing the expected response timeout duration with the affected evaluation value, it is further reflected that task T2 not only has an actual timeout, but also has a higher scheduling impact due to the unavailability of alternative vehicles. Therefore, it can be used as one of the main bases for determining the ownership of the same candidate vehicle V1.

[0058] Based on the above judgment, the scheduling system assigns the same candidate vehicle V1 to task T2 and reinstates task T1 into the scheduling sequence. Although task T1 has a higher task priority than task T2, task T1 has an alternative vehicle V3 that meets the service capacity and response time requirements, while task T2 cannot meet the response time requirements after losing vehicle V1. Therefore, assigning vehicle V1 to task T2 can avoid the risk of task T2 directly encountering the response time limit and create a dynamic balance between task priority and task response time. Figure 2 The assignment result of vehicle V1 to task T2 corresponds to the above allocation result. This reflects that the scheduling system does not allocate based solely on task priority, but determines the final assignment of the same candidate vehicle under the joint constraints of the availability of alternative vehicles and the degree to which response time is affected.

[0059] For task T1 that does not obtain a candidate vehicle V1, the scheduling system redetermines the scheduling relationship between vehicles and tasks based on the updated vehicle and task states. The scheduling system continues to acquire the real-time locations, current task states, estimated idle times, and service capabilities of vehicles V2, V3, V4, and V5. When vehicle V3 meets the task succession chain verification requirements, it is designated as the successor vehicle for task T1. This process reduces scheduling conflicts caused by multiple tasks competing for a single vehicle and improves the continuity and stability of the fleet dynamic scheduling process.

[0060] Example 2:

[0061] This embodiment provides a dynamic fleet scheduling method based on real-time vehicle location and task priority. This method can be executed by a fleet scheduling platform configured by an organization. After the same candidate vehicle is assigned to one of the tasks, the scheduling system will re-include the tasks that have not been assigned to the same candidate vehicle into the scheduling sequence, and continue to determine the scheduling relationship between vehicles and tasks based on the updated vehicle status and task status.

[0062] In a scheduling scenario, task T4 is re-included in the scheduling sequence after failing to secure a replacement vehicle V5. Task T4 is a regular transportation task with a medium priority and a remaining response time of 18 minutes. Based on the vehicle's real-time location, estimated idle time, service capacity, and estimated arrival time at task T4's location, the scheduling system identifies candidate vehicles V6 and V7 for task T4. Vehicle V6 is in a waiting state for a replacement task and is expected to arrive at task T4's location in 8 minutes, while vehicle V7 is in an idle state and is expected to arrive at task T4's location in 11 minutes. Both vehicles meet the service capacity requirements of task T4.

[0063] The scheduling system generates a succession constraint record C4 for task T4 and configures the constraint validity period for record C4. The succession constraint record C4 specifies the task identifier, task location, task priority, response time limit, unobtained candidate vehicle V5, and candidate successor vehicles V6 and V7 reserved for task T4. Before determining the successor vehicle for task T4, the scheduling system sets occupancy restrictions on candidate successor vehicles V6 and V7, preventing new tasks with the same task priority as task T4 from occupying candidate successor vehicles V6 and V7. Figure 5 The processing relationship of the succession constraint record C4 generated after task T4 rejoins the queue is shown. The succession constraint record C4 is associated with candidate succession vehicles V6 and V7 respectively, and prevents the same-level task T5 from calling vehicle V6 through occupancy restriction, while keeping task T4 in a state of waiting for succession vehicle. Figure 5 The record fields below further indicate that the succession constraint record C4 not only records the task number, but also records data related to subsequent succession judgment, such as task identifier, response time limit, task priority, and candidate succession vehicles.

[0064] To determine whether a continuation constraint record is still valid, the scheduling system uses the following formula to determine the remaining validity time of the continuation constraint record:

[0065]

[0066] in, Indicates the first The remaining validity period of the continuation constraint record. Indicates the first The validity period of the continuation constraint record. Indicates the current judgment time. Indicates the first The generation time of a continuation constraint record. The derivation logic of this formula is as follows: The scheduling system first determines the duration of the continuation constraint record based on the difference between the current judgment time and the generation time, and then subtracts the duration from the constraint validity period to obtain the remaining time for the continuation constraint record to maintain the occupancy limit.

[0067] In this embodiment, the generation time of continuation constraint record C4 is taken as the 0-minute baseline. The constraint validity period of continuation constraint record C4 is 12 minutes. The current judgment time is 7 minutes after generation. Therefore, the remaining validity time of continuation constraint record C4 is:

[0068]

[0069] Therefore, the succession constraint record C4 still has 5 minutes of remaining valid time, and the scheduling system continues to maintain the restriction on the occupancy of candidate successor vehicles V6 and V7. Within this remaining valid time, if a new task T5 with the same task priority as task T4 requests to call vehicle V6, the scheduling system will reject task T5's request for vehicle V6 and keep vehicle V6 as a candidate successor vehicle for task T4. Figure 6 The validity period determination process of the continuation constraint record C4 is shown in the form of a timeline. The time from 0 minutes to 7 minutes corresponds to the duration, and the time from 7 minutes to 12 minutes corresponds to the remaining validity period. The current determination time is at 7 minutes, and the expiration time is at 12 minutes. This intuitively shows that the continuation constraint record C4 has not yet expired at the current determination time and can still maintain the occupancy restriction.

[0070] During the constraint validity period, the scheduling system continuously determines whether a successor vehicle has been identified for task T4. If vehicle V6 meets the estimated arrival time requirement, service capacity requirement, and current vehicle status requirement in subsequent scheduling decisions, the scheduling system identifies vehicle V6 as the successor vehicle for task T4, removes the candidate successor vehicle occupancy restriction corresponding to successor constraint record C4, and updates the scheduling relationship between task T4 and vehicle V6. If a successor vehicle has not yet been identified for task T4, the scheduling system continues to maintain the aforementioned occupancy restriction.

[0071] When candidate vehicles V6 and V7 are under occupancy restrictions, if a new task T6 with a higher priority than task T4 requests to use candidate vehicle V6, the scheduling system removes the occupancy restriction on vehicle V6 and allows the new task T6 with the higher priority to participate in the scheduling of vehicle V6. At this time, the scheduling system does not remove task T4 from the scheduling sequence, but keeps task T4 in the scheduling sequence, and after removing the occupancy restriction on vehicle V6, it puts task T4 into a reserved succession state. Figure 7 The above shows the path of a newly entered task T6 with high task priority requesting to call vehicle V6, removing the occupancy restriction of vehicle V6, and allowing task T6 to participate in scheduling. Figure 7 The path shown below is that task T4 remains in the scheduling sequence, enters the hold-and-continue state, suspends the update of peer tasks, and continues to retain candidate vehicle V7. The linkage constraint between the two paths indicates that high-priority tasks can be inserted into the schedule, but this will not cause task T4 to be removed from the scheduling sequence.

[0072] While maintaining the continuity status, the scheduling system suspends the updating of the scheduling priority of new tasks with the same priority as task T4 until task T4 determines its successor vehicle or the validity period of the continuity constraint record C4 expires. In other words, new tasks with the same priority can enter the scheduling sequence, but their entry does not lower the relative priority of task T4 in the scheduling sequence, thus preventing task T4 from being repeatedly moved back due to consecutive entries of tasks of the same priority.

[0073] To determine the scheduling priority retention status of tasks in the reserved continuation state, the scheduling system uses the following formula to determine the scheduling priority retention value:

[0074]

[0075] in, Indicates the first For tasks that do not obtain the same candidate vehicle, the scheduling priority value is maintained in the retained succession state. This indicates the scheduling priority baseline value before the task enters the reserved continuation state. This represents the number of newly entered tasks of the same level that are suspended from being counted while in the hold-for-continuation state. The derivation logic of this formula is as follows: the scheduling system uses the scheduling priority before the task enters the hold-for-continuation state as a benchmark, and deducts the number of newly entered tasks of the same level that are suspended from being counted from this benchmark to reflect the scheduling position maintained by the task due to the suspension of the counting of newly entered tasks of the same level; in this embodiment, the smaller the scheduling priority value, the earlier the task is in the scheduling sequence.

[0076] In this embodiment, the scheduling priority baseline value of task T4 before entering the hold-and-continue state is 3. If two newly entered tasks of the same priority are paused during the hold-and-continue state, then the scheduling priority value of task T4 is:

[0077]

[0078] Therefore, task T4 will not be moved further down the scheduling sequence due to two new tasks of the same level entering the sequence. Based on this scheduling priority retention value, the scheduling system maintains task T4 in a position where it can continue to be scheduled, and as long as vehicle V7 still meets the service capacity and response time requirements, vehicle V7 will be retained as a candidate vehicle to succeed task T4. Figure 7 The scheduling priority data on the right side is represented by bar charts corresponding to G4, W4, and H4, which show the baseline priority, the number of newly entered tasks at the same level that are paused, and the priority result after retention. This makes the priority retention logic in the retention continuation state correspond to the calculation process of the scheduling system.

[0079] If a successor vehicle has been identified for task T4 before the constraint expires, the scheduling system will remove the occupancy restriction of the candidate successor vehicle corresponding to successor constraint record C4 and update the scheduling relationship between task T4 and the successor vehicle. If a successor vehicle has not been identified for task T4 when the constraint expires, the scheduling system will remove the occupancy restriction of the original successor constraint record C4 and, based on the real-time location of the vehicle, the estimated idle time, service capacity, and task status at the time of expiration, regenerate the successor constraint record or re-determine the scheduling order of task T4 in the scheduling sequence.

[0080] In this way, the scheduling system ensures that high-priority tasks can promptly call upon candidate vehicles for succession, while preventing tasks that have not obtained the same candidate vehicle from being repeatedly shifted back during the succession scheduling process due to new tasks of the same priority, thereby improving the stability and continuity of fleet dynamic scheduling.

Claims

1. A method for dynamic fleet scheduling based on real-time vehicle location and task priority, characterized in that, include: The system obtains the real-time location, current task status, estimated idle time and service capacity of the fleet vehicles, obtains the task location, task type, response time limit and task priority of the task to be executed, and extracts candidate vehicles based on the real-time location, the current task status, the estimated idle time and the service capacity. When multiple task requests correspond to the same candidate vehicle, the availability of alternative vehicles and the degree of impact on response time limits are determined for each task after the loss of the same candidate vehicle. The availability of alternative vehicles refers to whether there are alternative vehicles that meet the service capacity and response time limit requirements. The degree of impact on response time limits refers to the expected response timeout duration or the result of not being able to meet the response time limit. The affiliation of the same candidate vehicle is determined based on the task priority, the availability of the alternative vehicle, and the degree to which the response time limit is affected. Tasks that do not obtain the same candidate vehicle are re-included in the scheduling sequence, and the scheduling relationship between vehicles and tasks is updated.

2. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 1, characterized in that, When determining the availability of the alternative vehicle, the proposed alternative vehicle is verified together with the currently executing task and the subsequently assigned tasks; if, after calling the proposed alternative vehicle, there is no vehicle that meets the response time limit for the currently executing task or the subsequently assigned task, the proposed alternative vehicle is excluded from the alternative vehicles.

3. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 1, characterized in that, When determining the affiliation of the same candidate vehicle, if the priority of the first task is higher than that of the second task, the first task has an alternative vehicle that meets the service capability and response time requirements, and the second task cannot meet the response time requirement after losing the same candidate vehicle, the same candidate vehicle shall be assigned to the second task.

4. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 1, characterized in that, When a task that has not obtained the same candidate vehicle is re-included in the scheduling sequence, a succession constraint record with a constraint validity period is generated, and the succession constraint record specifies the candidate successor vehicle reserved for the task; before the task determines the successor vehicle, new tasks with the same task priority as the task are restricted from occupying the candidate successor vehicle.

5. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 2, characterized in that, When it is determined whether there is a vehicle that meets the response time limit for the currently executing task or the subsequent assigned task, a task succession chain is generated starting from the vehicle to be replaced, and subsequent succession vehicles are searched along the task succession chain. If there is a task in the task succession chain where the next successor vehicle cannot be determined, then the proposed replacement vehicle is excluded from the list of replacement vehicles.

6. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 2, characterized in that, When the currently executing task is in a state where it has reached the task location, started service, or the estimated remaining service time is less than a preset completion time threshold, the vehicle to be replaced is configured as a service protection vehicle, and the service protection vehicle is prohibited from being used as a replacement vehicle for tasks other than the currently executing task.

7. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 2, characterized in that, After excluding the proposed replacement vehicles, vehicles that meet the service capacity requirements within a preset distance range are selected to supplement the current task or the subsequent assigned task with vehicles that meet the response time limit. Before the vehicle is replenished, tasks within a preset service area with a task priority lower than the currently executing task or the subsequently assigned task are restricted from calling vehicles within the preset distance range. The preset service area is the vehicle dispatch area to which the vehicle to be replaced belongs.

8. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 4, characterized in that, During the validity period of the constraint, if a successor vehicle has been determined for a task that has not obtained the same candidate vehicle, the occupancy restriction on the candidate successor vehicle corresponding to the succession constraint record is lifted; if a successor vehicle has not been determined for the task, the occupancy restriction is maintained.

9. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 4, characterized in that, During the period when the candidate successor vehicle corresponding to the succession constraint record is restricted from being occupied, when a new task with a task priority higher than the task priority of the task that has not obtained the same candidate vehicle requests to call the candidate successor vehicle, the occupation restriction of the candidate successor vehicle is lifted, and the task that has not obtained the same candidate vehicle is kept in the scheduling sequence.

10. The fleet dynamic scheduling method based on real-time vehicle location and task priority according to claim 9, characterized in that, After the occupancy restriction of the candidate vehicle is lifted, the task that has not obtained the same candidate vehicle is transferred to the reserved vehicle status. In the reserved vehicle status, the scheduling order of new tasks with the same task priority as the task is suspended from updating the task until the task determines a vehicle or the constraint expires.

Citation Information

Patent Citations

  • Passenger vehicle transport vehicle scheduling method and device based on order dynamic priority

    CN114331220A

  • Unmanned motorcade scheduling method and device and electronic equipment

    CN116612654A