Real-time dynamic distributed task flow distribution method based on iteration strategy

Abstract

The invention discloses a real-time dynamic distributed task flow distribution method based on an iteration strategy, and belongs to the technical field of task distribution. The invention discloses a real-time dynamic distributed task flow distribution method based on an iteration strategy. The method comprises the following steps: issuing a sub-task information set; the unmanned aerial vehicle group bids each sub-task and sends a bidding quotation to the task distribution system; the task distribution system selects the bid-winning unmanned aerial vehicle according to the bidding quotation, introduces a judgment mechanism to check a task distribution result and adjusts a task distribution scheme; the unmanned aerial vehicle sends a task allocation request at a fixed frequency, and the sub-tasks of the new task and the sub-tasks which do not meet the allocation requirement in the last round form a new round of task set; and repeating the steps S1 to S4. According to the real-time dynamic distributed task flow allocation method based on the iteration strategy, a dynamic allocation mechanism is adopted, potential errors in the task flow allocation process are reduced under the condition that the dynamic iteration strategy is considered, and the real-time and dynamic requirements of system task allocation are met.

Description

Technical Field

[0001] The present invention relates to the technical field of task allocation, and in particular to a real-time dynamic distributed task flow allocation method based on an iteration strategy. Background Art

[0002] Drones have high mobility, self-organization ability and flexible scalability. These advantages have attracted widespread attention in both civil and military fields. Task allocation in drone swarms is related to collaborative decision-making and control. Drones can perform multiple continuous tasks on multiple targets. The main goal of real-time dynamic task allocation is to allocate multiple tasks to multiple drones according to the task environment and task target status under multiple constraints such as battlefield environment, task requirements, and drone performance, so that the drone formation can maximize combat effectiveness and minimize costs under constraints such as time, space, and resources.

[0003] Task flow allocation can be formulated as a graph theory problem, where UAVs need to share information and allocate tasks at the appropriate time. Appropriate communication and information exchange between UAV swarms can improve the capabilities of UAVs to meet the performance requirements of fast and reliable execution of related tasks. With the widespread application of multi-UAV task planning systems, the real-time dynamic task allocation process of UAVs has become an emerging problem. The essence of task flow allocation is a complex optimization problem with constraints, which has the characteristics of multiple complexities, multiple parameters, and multiple constraints. Its solution space grows exponentially with the number of assigned tasks.

[0004] In recent years, scholars have proposed a variety of research methods to solve the problem of real-time dynamic task flow allocation, mainly including centralized, distributed, and swarm intelligence-based task flow allocation methods. However, factors such as information coupling and limited communication resources in the task planning process will cause potential situational awareness errors in the UAV system, which will lead to inaccurate multi-UAV task assessments. The above real-time dynamic task flow allocation often faces the following problems: 1. The centralized task flow allocation method requires a central processor to collect information from all drone systems and calculate the best allocation strategy. Common algorithms include hybrid line planning, improved wolf pack search, dynamic task allocation, and cooperative task allocation based on dynamic programming. These algorithms have problems such as the calculation time increases with the increase in the number of targets and ignore the uncertainty brought by the complex dynamic environment in reality.

[0005] 2. Distributed task flow allocation methods can effectively reduce the computational burden of multi-UAV systems, and mainly focus on market-based programs, namely contract network algorithms and auction algorithms; but they lack consideration of uncertain factors such as constraints on UAVs and dynamic environments, such as external threats and external forces.

[0006] 3. Particle swarm optimization, genetic algorithm, artificial neural network and other swarm intelligence-based methods use a search-centric approach to obtain the optimal task flow allocation strategy. There are problems such as easy to ignore the real complex dynamic battlefield environment and external constraints of UAVs.

[0007] In summary, many current technical means have considered the task flow allocation problem of UAV systems, but their ability to handle the uncertainty and randomness faced in the real-time dynamic task allocation process is limited. Therefore, the present invention proposes a real-time dynamic distributed task flow allocation method based on an iterative strategy to solve the shortcomings of the above-mentioned prior art. Summary of the invention

[0008] The purpose of the present invention is to provide a real-time dynamic distributed task flow allocation method based on an iterative strategy, which adopts a dynamic allocation mechanism and reduces potential errors in the task flow allocation process while considering the dynamic iterative strategy, thereby meeting the real-time and dynamic requirements of task allocation in a distributed unmanned aerial vehicle system.

[0009] To achieve the above object, the present invention provides a real-time dynamic distributed task flow allocation method based on an iteration strategy, comprising the following steps: S1, the task allocation system publishes a subtask information set to the drone swarm; S2, the drone swarm bids for each subtask according to the subtask information set, and sends the bid price to the task allocation system; S3, the task allocation system selects the winning drone according to the bidding quotation, obtains the task allocation matrix, introduces a trial mechanism to check the task allocation results, adjusts the task allocation plan, and the subtasks that do not meet the allocation requirements enter the next stage; S4: After the initial task allocation, the drone sends task allocation requests at a fixed frequency. The subtasks of the new task and the subtasks that do not meet the allocation requirements in S3 form a new round of subtask sets. S5. Repeat S1-S4.

[0010] Preferably, in S1, the specific process of the task allocation system publishing the subtask information set includes: S11, initialize the drone attributes and drone group, the drone group is , n is the number of drones; S12, the task allocation system decomposes the combat task into several subtasks; S13. According to the constraints of a single UAV and subtasks, some infeasible subtasks are directly eliminated from the solution to obtain a subtask information set.

[0011] The subtask information set is ,in Indicates that the subtask set is not in Round auction; Indicated in r Round of auction j unassigned subtask set target locations; ， Indicated in r Round of auction j The target value of the unassigned subtask set; , Indicated in r Round of auction j unassigned subtask capability requirements; , Indicated in r Round of auction j Resource consumption of unassigned subtasks; , Indicates that the UAV system is executing r The request frequency of the subtask process of the round auction.

[0012] Preferably, in S2, the drone formation bidding for the subtask specifically includes the following steps: S21, the drone further removes unexecutable subtasks based on the declared information of the subtasks; S22. Match the UAV combat capability with the sub-task combat requirements to determine the bidding qualifications of the UAV; S23, the UAV performs performance evaluation on all matching subtasks and submits a request to the task allocation system based on the evaluation results; Preferably, in S21, the specific operation of eliminating the unexecutable subtasks is: establishing a proximity factor To describe the geographical dispersion of the drone swarm and subtasks, the proximity factor for, ; in, are the coordinates of the drone, are the coordinates of the subtask, h is the total number of subtasks, When the dispersion between drones and subtasks is greater than the threshold , then the subtask will be eliminated.

[0013] Preferably, in S22, the specific method for determining the drone bidding qualification is: At least one of the drone's electronic reconnaissance capability, flight capability, electronic jamming capability, firepower strike capability and other capabilities meets the requirements of the sub-task.

[0014] Preferably, in S23, the specific operation of the drone's performance evaluation of the subtask is: The drone calculates the efficiency values ​​of all matching subtasks and sorts the efficiency values ​​to obtain the total efficiency set , select the subtask with the highest efficiency value b Submit bids to the task allocation system ,in, It is the first drone to complete the subtask b The total efficiency value.

[0015] Preferably, S3 specifically includes: S31, calculate the j The scenario gain and loss set of subtasks and drone performance set ,in, B The bidding quotation set for the subtask by the drone bidding set, and the task allocation system and The weight of j The basis for winning the bid for each subtask and obtaining the task allocation matrix ; S32. Introduce a trial mechanism into the mission planning framework. The trial agency determines the deviation between the arrival time TOA and the estimated arrival time ETA of the UAV mission. If the deviation If the threshold is exceeded, the drone repeats the S2 process and bids again; the trial mechanism is: ; S33. Calculate the overall efficiency of the drone swarm and determine whether the overall efficiency meets expectations. If not, continue to adjust the allocation plan of the drone swarm and the subtask set until the expectations are met.

[0016] Preferably, in S4, idle drones that are not performing tasks bid for a new round of subtask sets.

[0017] The advantages and positive effects of the real-time dynamic distributed task flow allocation method based on iterative strategy described in the present invention are: 1. The task allocation method described in the present invention has a small amount of calculation, which can save trouble and seize the initiative in time-sensitive air combat, and realize the demand for real-time collaborative task allocation. It overcomes the disadvantage that the traditional method of solving multi-parameter and multi-constraint polynomials requires huge calculation costs; it can find a satisfactory solution for task flow allocation in a short time.

[0018] 2. The addition of the trial mechanism ensures that the auction algorithm will not have large errors, thereby maintaining the feasibility of the allocation results; it overcomes the shortcomings of traditional auction algorithms such as potential errors and inaccurate task evaluation; in task flow allocation, the auction algorithm based on iterative strategy has better algorithm stability.

[0019] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 The following is a flowchart of task allocation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In the description of the present invention, it should be noted that the terms "upper", "lower", "inside", "outside", etc. indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, or the positions or positional relationships in which the invented product is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In the description of the present invention, it should also be noted that, unless otherwise clearly specified and limited, the terms "setting", "installation", and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0022] In this application, unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of this application. In the event of any inconsistency, the meaning described in this specification or the meaning derived from the contents recorded in this specification shall prevail. In addition, the terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application. In order to accurately describe the technical content in this application and to accurately understand the present invention, the following explanations or definitions are given to the terms used in this specification before describing the specific embodiments: The embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0023] like Figure 1 A real-time dynamic distributed task flow allocation method based on an iteration strategy includes the following steps: S1. The task allocation system publishes a subtask information set to the drone swarm.

[0024] The specific process of the task allocation system publishing subtask information sets includes: S11, initialize the drone attributes and drone group, the drone group is , n The number of drones.

[0025] S12. Based on the mission environment and mission objectives, the mission allocation system decomposes the combat mission and divides the combat mission into several subtasks.

[0026] S13. According to the constraints of a single UAV and subtasks, some infeasible subtasks are directly eliminated from the solution to obtain a subtask information set.

[0027] The subtask information set is ,in Indicates that the subtask set is not in Round auction; Indicated in r Round of auction j unassigned subtask set target locations; ， Indicated in r Round of auction j The target value of the unassigned subtask set; , Indicated in r Round of auction j unassigned subtask capability requirements; , Indicated in r Round of auction j Resource consumption of unassigned subtasks; , Indicates that the UAV system is executing r The request frequency of the subtask process of the round auction.

[0028] S2. The drone swarm bids for each subtask based on the subtask information set and sends the bid quotation to the task allocation system.

[0029] The specific steps for the drone formation to bid for the subtask include: S21. The drone further eliminates the subtasks that cannot be executed based on the declared information of the subtasks. When two subtasks are far apart, a drone cannot successively execute the subtasks of two targets.

[0030] The specific operation of eliminating unexecutable subtasks is to establish a proximity factor To describe the geographical dispersion of the drone swarm and subtasks, the proximity factor for, ; in, are the coordinates of the drone, are the coordinates of the subtask, h is the total number of subtasks. When the dispersion between drones and subtasks is greater than the threshold , then the subtask will be eliminated.

[0031] S22. Match the UAV’s combat capability with the sub-task combat requirements to determine the UAV’s bidding eligibility.

[0032] For different combat subtasks, the subtask information set is Sub-mission capabilities for drones With different requirements, the drone can find the corresponding task by matching the drone's reconnaissance capability, flight capability, electronic jamming capability, firepower strike capability and other capabilities with the sub-task requirements.

[0033] In the process of subtask allocation, the UAV’s electronic reconnaissance capability, flight capability, and other capabilities must at least meet the requirements of the first i Subtask requirements, namely , otherwise the drone is not eligible to bid.

[0034] S23. The drone performs performance evaluation on all matching subtasks and calculates the scenario benefit loss set and the drone effectiveness set.

[0035] Drone Computing j The scenario benefit loss set and drone effectiveness set of each subtask. j The scenario benefit loss set of subtasks , Indicates The scenario benefit loss value of a drone for the jth subtask is calculated as follows: ;

[0036] in, Indicates A drone against j The situation assessment index of the first target is j Target The ratio of the situational assessment index of the drones, Indicates the set constant value.

[0037] No. j The drone performance set of subtasks , is the task allocation matrix The function is calculated as follows: ;

[0038] in, Indicates r Task Assignment Matrix The multi-task reward function under Indicates r Task Assignment Matrix The multi-task cost function under .

[0039] Furthermore, we can obtain the total efficiency set of this round of auction: ,in, B The bid price set by the drone bidding set for this subtask, and Represents the weight coefficient.

[0040] S3. The task allocation system selects the winning drone based on the bidding price, obtains the task allocation matrix, introduces a trial mechanism to check the task allocation results, adjusts the task allocation plan, and the subtasks that do not meet the allocation requirements enter the next stage.

[0041] Specifically include: S31. If all unassigned tasks in this round meet the assignment constraints, calculate the first j The scenario benefit loss set of subtasks and drone performance set , and the total efficiency set . Select the subtask with the highest efficiency value b Submit bids to the task allocation system ,in, It is Complete the sub-task with a drone b The task allocation system concentrates drones according to the total efficiency. and The weight of j The basis for winning the bid for each subtask and obtaining the task allocation matrix .

[0042] S32. In order to eliminate the potential errors of traditional auctions, a trial mechanism is introduced into the task planning framework to achieve task allocation based on iterative strategies. The trial agency determines the deviation between the arrival time TOA and the estimated arrival time ETA of the drone to perform the task. If the deviation If the threshold is exceeded, the drone repeats the S2 process and bids again.

[0043] The trial mechanism is as follows: ; This trial mechanism is used to check whether the task allocation plan is correctly valued. If the task allocation plan is overestimated, the drone needs to be re-evaluated and the auction process repeated. from The mission was won, but the trial mechanism found that the mission was overestimated, and the drone The drone will be notified Repeat the bidding process.

[0044] S33, calculate the overall efficiency of the drone group, and determine whether the overall efficiency meets expectations. If not, continue to adjust the allocation plan of the drone group and the subtask set (i.e. repeat S2 and S3 processes) until it reaches expectations. If there are subtasks that do not meet the allocation requirements, they will be marked as unallocated tasks and enter the next stage.

[0045] S4. According to the trigger mechanism of task flow allocation, select the time-based trigger mode. After the initial task allocation, the drone sends a task allocation request at a fixed frequency. If the current subtask has not been completed, the drone continues to perform the task; if the current subtask has been completed, a new task is assigned to the drone. When a new task appears, nearby drones and idle drones provide bidding quotations for the new task, and decide which drone to perform the task based on the quotations, so as to realize dynamic adjustment of task allocation. In the new task environment, the emerging combat subtasks and the unassigned tasks in the third phase form a new round of subtask sets, and repeat S1-S4 for task allocation.

[0046] Therefore, the real-time dynamic distributed task flow allocation method based on iterative strategy described in the present invention adopts a dynamic allocation mechanism to reduce potential errors in the task flow allocation process while considering the dynamic iterative strategy, thereby meeting the real-time and dynamic requirements of task allocation of distributed unmanned aerial vehicle systems.

[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A real-time dynamic distributed task flow allocation method based on iterative strategy, characterized in that: The following steps are involved: S1, the task allocation system publishes a subtask information set to the drone swarm; S2, the drone swarm bids for each subtask according to the subtask information set, and sends the bid price to the task allocation system; S3, the task allocation system selects the winning drone according to the bidding quotation, obtains the task allocation matrix, introduces a trial mechanism to check the task allocation results, adjusts the task allocation plan, and the subtasks that do not meet the allocation requirements enter the next stage; S4: After the initial task allocation, the drone sends task allocation requests at a fixed frequency. The subtasks of the new task and the subtasks that do not meet the allocation requirements in S3 form a new round of subtask sets. S5. Repeat S1-S4.

2. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 1, characterized in that: In S1, the specific process of the task allocation system publishing the subtask information set includes: S11, initialize the drone attributes and drone group, the drone group is , n is the number of drones; S12, the task allocation system decomposes the combat task into several subtasks; S13. According to the constraints of a single UAV and subtasks, some infeasible subtasks are directly eliminated from the solution to obtain a subtask information set; The subtask information set is ,in Indicates that the subtask set is not in Round auction; Indicated in r Round of auction j unassigned subtask set target locations; ， Indicated in r Round of auction j The target value of the unassigned subtask set; , Indicated in r Round of auction j unassigned subtask capability requirements; , Indicated in r Round of auction j Resource consumption of unassigned subtasks; , Indicates that the UAV system is executing r The request frequency of the subtask process of the round auction.

3. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 2, characterized in that: In S2, the drone formation bidding for the subtask specifically includes the following steps: S21, the drone further removes unexecutable subtasks based on the declared information of the subtasks; S22. Match the UAV combat capability with the sub-task combat requirements to determine the bidding qualifications of the UAV; S23. The UAV performs performance evaluation on all matching subtasks and submits a request to the task allocation system based on the evaluation results.

4. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 3, characterized in that: In S21, the specific operation of eliminating the unexecutable subtasks is: establishing a proximity factor To describe the geographical dispersion of the drone swarm and subtasks, the proximity factor for, ； in, are the coordinates of the drone, are the coordinates of the subtask, h is the total number of subtasks, When the dispersion between drones and subtasks is greater than the threshold , then the subtask will be eliminated.

5. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 4, characterized in that: In S22, the specific method for determining the bidding qualification of the drone is: At least one of the drone's electronic reconnaissance capability, flight capability, electronic jamming capability and firepower strike capability meets the requirements of the sub-task.

6. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 5, characterized in that: In S23, the specific operation of the drone's performance evaluation of the subtask is: The drone calculates the efficiency values ​​of all matching subtasks and sorts the efficiency values ​​to obtain the total efficiency set , select the subtask with the highest efficiency value b Submit bids to the task allocation system ,in, It is Complete the sub-task with a drone b The total efficiency value.

7. A real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 6, characterized in that: The S3 specifically includes: S31, calculate the j The scenario benefit loss set of subtasks and drone performance set ,in, B The bidding set for the subtask is set by the drone bidding set. The task allocation system collects the drones according to the scenario benefit loss set and the drone effectiveness set. and The weight of j The basis for winning the bid for each subtask and obtaining the task allocation matrix ; S32. Introduce a trial mechanism into the mission planning framework. The trial agency determines the deviation between the arrival time TOA and the estimated arrival time ETA of the UAV mission. If the deviation If the threshold is exceeded, the drone repeats the S2 process and bids again; the trial mechanism is: ； S33. Calculate the overall efficiency of the drone swarm and determine whether the overall efficiency meets expectations. If not, continue to adjust the allocation plan of the drone swarm and the subtask set until the expectations are met.

8. The real-time dynamic distributed task flow allocation method based on iterative strategy according to claim 7 is characterized by: In S4, idle drones that are not performing tasks bid for a new round of subtask sets.

Images