Cluster distributed cooperative task allocation method capable of being used for denial environment

Abstract

The invention provides a cluster distributed cooperative task allocation method capable of being used for a denial environment, and relates to the technical field of task planning of an unmanned aerial vehicle cluster. According to the task allocation method provided by the invention, the function complexity is reduced by delimiting the task allocation range, so that the task allocation method can be applied to a large-scale unmanned aerial vehicle cluster; the cooperative task is decomposed through the dividing and conquering thought and then is allocated to the method for internal role division, so that the method can be used for task allocation of the cooperative task; through the architecture of distributed calculation and centralized global decision, the calculation pressure and time consumption of task allocation are reduced while the global task allocation effect is ensured; through a perfect denial environment processing strategy, the method can be used for task allocation in a denial environment.

Description

Technical Field

[0001] The present invention relates to the technical field of task planning for unmanned aerial vehicle clusters, and in particular provides a cluster distributed collaborative task allocation method that can be used in a denial environment. Background Art

[0002] After discovering a target, a drone swarm generates several combat missions targeting the target, which require specific drones to execute in order to build a complete kill chain. Currently, this is mostly done manually, but as the size of the swarm continues to expand, manual work is becoming increasingly difficult. There are also optimization methods such as auction algorithms that automatically assign tasks, but the process often requires multiple information interactions within the cluster, which makes it difficult to use in a denial environment. As the size of the drone increases, the task assignment algorithm becomes more and more time-consuming, making it difficult to use for large-scale cluster task assignment. In addition, the collaborative relationship between tasks is not considered during assignment, making it difficult to use for collaborative task assignment.

[0003] Therefore, there is an urgent need for a fast collaborative task allocation method that can be used in denied environments. Summary of the invention

[0004] The purpose of this application is to provide a fast collaborative task allocation method that can be used in a denied environment. By defining the allocation scope, distributed cost calculation, collaborative task decomposition and decision-making division, and denied environment adaptation, it ensures the global effect of task allocation while reducing the time consumption of task allocation, and solves the problem of collaborative task allocation for large-scale drone clusters in a denied environment.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a cluster distributed collaborative task allocation method that can be used in a denied environment, the method comprising:

[0006] S1: Calculate the scope of task allocation;

[0007] S2: Send a quotation request to the drones participating in the task allocation. After receiving the quotation request, the drone calculates the task execution cost and returns the cost information to the relay;

[0008] S3: After receiving the cost information, the relay stores it to form a task cost matrix;

[0009] S4: Calculate the task value according to the target type and combine it with the task cost to obtain the task benefit matrix;

[0010] S5: Initialize the profit matrix, where the profit is the difference between the task revenue and the quotation;

[0011] S6: The drones in the drone list start bidding in order. Based on the new bidding information, the profit matrix of all drones for the task is updated. The task is successfully bid by the drone with the highest bid. Until all drones have completed their bidding, the drone with the successful bid is paired with the task.

[0012] S7: Traverse all tasks to determine whether they are collaborative tasks. If they are collaborative tasks, redivide the work according to the collaborative task requirements. The collaborative task time is calculated based on the estimated completion time of the task, and task instructions are issued based on the task allocation results.

[0013] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has the following technical features: S1 includes calculating the task allocation range by screening the drones.

[0014] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has the following technical features: screening the drones includes first screening the drone types and then screening based on the distance between the drones and the targets.

[0015] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has the following technical feature: the task execution cost C calculated in S2 is:

[0016] C=p 1 d+p 2 t+p 3 s

[0017] Among them, p is the weight coefficient, d is the distance cost, t is the time cost, and s is the threat cost.

[0018] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has such a technical feature that S2 also includes that if a certain UAV fails to return the task execution cost within 100 cycles, the task execution cost is calculated by the relay machine instead, and the relay machine uses the straight-line distance between the lost UAV and the target as the task execution cost of the UAV.

[0019] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has the following technical features: the task is successfully bid by a drone with a high bid, including:

[0020] The drone that starts bidding will bid for the task with the highest profit. If the drone's profit for all tasks is not greater than 0, it will quit bidding. Otherwise, it will increase its bid. The amount of the price increase is calculated based on the profit of the task and the profit of the second-best task.

[0021] If another drone has bid for the task being bid, and the current bid is higher, the previous bidder will fail, and the drone that failed will start bidding again.

[0022] If no other drone has bid for the current bid task, the next drone in the drone list will start bidding.

[0023] The cluster distributed collaborative task allocation method for a denied environment provided by the present invention also has the following technical feature: the redivision of labor in S7 includes: decomposing the collaborative task into a corresponding number of stand-alone tasks according to the required number of drones.

[0024] The cluster distributed collaborative task allocation method for a denial environment provided by the present invention also has such a technical feature that the method further includes: after the relay machine loses connection, the drone closest to the center position of all drones is used as a temporary relay node to coordinate the task allocation process instead of the relay machine.

[0025] Beneficial Effects

[0026] The task allocation method provided by the present invention reduces functional complexity by defining the task allocation scope, so that it can be used for large-scale drone clusters; the collaborative tasks are decomposed and allocated after internal role division through the divide-and-conquer idea, so that it can be used for task allocation of collaborative tasks; through distributed computing and a centralized global decision-making architecture, the global task allocation effect is ensured while reducing the computational pressure and time consumption of task allocation; through a complete denied environment processing strategy, it can be used for task allocation in a denied environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 A flowchart of a method provided by an embodiment of the present invention;

[0029] Figure 2 A functional architecture diagram of the method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The present application is further described in detail below in conjunction with the accompanying drawings and examples, but it should be noted that these implementation methods are not limitations of the present application, and equivalent changes or substitutions in functions, methods, or structures made by ordinary technicians in the field based on these implementation methods are all within the scope of protection of the present application.

[0031] In the description of the embodiments of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the creation of the present application 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 creation of the present application.

[0032] In addition, the terms "first", "second", "third", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the invention of this application, unless otherwise specified, the meaning of "plurality" is two or more.

[0033] The terms "installed", "connected" and "connected" 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 an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the invention of this application can be understood by specific circumstances.

[0034] like Figure 1-2 As shown, an embodiment of the present application provides a cluster distributed collaborative task allocation method that can be used in a denied environment, the method comprising:

[0035] S1: Calculate the scope of task allocation;

[0036] S2: Send a quotation request to the drones participating in the task allocation. After receiving the quotation request, the drone calculates the task execution cost and returns the cost information to the relay;

[0037] S3: After receiving the cost information, the relay stores it to form a task cost matrix;

[0038] S4: Calculate the task value according to the target type and combine it with the task cost to obtain the task benefit matrix;

[0039] S5: Initialize the profit matrix, where the profit is the difference between the task revenue and the quotation;

[0040] S6: The drones in the drone list start bidding in order. Based on the new bidding information, the profit matrix of all drones for the task is updated. The task is successfully bid by the drone with the highest bid. Until all drones have completed their bidding, the drone with the successful bid is paired with the task.

[0041] S7: Traverse all tasks to determine whether they are collaborative tasks. If they are collaborative tasks, redivide the work according to the collaborative task requirements. The collaborative task time is calculated based on the estimated completion time of the task, and task instructions are issued based on the task allocation results.

[0042] The method provided in the above embodiment performs collaborative task allocation of large-scale clusters in a denied environment through task allocation scope definition, distributed cost calculation, auction-style task allocation, collaborative task decomposition and role division, and denied environment adaptation process.

[0043] In some embodiments, S1 includes calculating a task allocation range by screening the drones.

[0044] In some embodiments, screening the drones includes first screening the drone types and then screening by the distance between the drones and the target.

[0045] In some embodiments, task machines participating in task allocation are screened through task constraints, the impact range of the task allocation function is limited, and the number of interactions between the relay machine and the task machine is reduced, thereby reducing bandwidth occupancy and effectively reducing the complexity of the task allocation process, thereby ensuring the real-time performance of the task allocation process.

[0046] Mission constraints include drone type constraints and distance constraints. Drone type constraints are generated by the mission plan generation module. Drones that do not meet drone type constraints will not participate in this mission allocation. Distance constraints are configuration parameters, and the default is 20km. Only mission aircraft with a distance less than 20km from the target will participate in this mission allocation.

[0047] In some embodiments, the task execution cost C calculated in S2 is:

[0048] C=p 1 d+p 2 t+p 3 s

[0049] Among them, p is the weight coefficient, d is the distance cost, t is the time cost, and s is the threat cost.

[0050] In the above embodiment, the task execution cost is calculated using distributed computing, and each drone calculates the cost of executing its own task separately, which can effectively reduce the computing pressure of the relay machine, thereby supporting real-time task allocation within a large-scale drone cluster. At the same time, due to the distributed computing of the task cost, the task cost calculation process allows the use of a more complex and accurate method. Therefore, we use the real-time route planning method to deduce the task execution process to calculate the task execution cost. The calculation result is closer to the actual situation and more accurate. Real-time route planning can also output the estimated time consumed to execute the task, which is convenient for the subsequent calculation of the collaborative task time for the collaborative task and ensures the time consistency of the collaborative task.

[0051] Supports user-defined task cost indicators, including range, time consumption, number of threats along the way, etc., and supports subsequent expansion according to user needs, so that the task allocation results can better meet user needs. By default, the range is used as the task cost indicator. If the mission aircraft's endurance is insufficient, the task cost is set to 200, giving the maximum task cost, and the UAV will exit the task allocation. The task cost of non-photoelectric detection and attack aircraft is increased by 100, and photoelectric detection and attack aircraft are used first.

[0052] In some embodiments, it is assumed that there are N different drones {U 1 ,U 2 ,......U N}, M tasks {T 1 ,T 2 ,......T M}, define the task allocation matrix X N×M :

[0053]

[0054] Each drone is only allowed to perform one mission at a time, and each mission is only allowed to be performed by one drone, that is, there are the following constraints:

[0055]

[0056] The objective function of the multi-UAV collaborative task allocation problem is defined as:

[0057] J(X)=X.*C

[0058] Among them, C is an N×M allocation basis matrix, c i j means U i Execute Task T j The cost.

[0059] C=p 1 d+p 2 t+p 3 s+…

[0060] Among them, p is the weight coefficient, d is the distance cost, t is the time cost, and s is the threat cost, and it can be continuously expanded according to user needs.

[0061] In some embodiments, S2 also includes: if a certain drone does not return the task execution cost within 100 cycles, the task execution cost is calculated by the relay machine instead, and the relay machine uses the straight-line distance between the lost drone and the target as the task execution cost of the drone.

[0062] In some embodiments, if the task machine loses connection when issuing instructions to the task machine according to the task assignment result after the task assignment is completed, the idle task machine closest to the target is dispatched to perform the task. If the task machine also loses connection, the above process is repeated until the task instruction is successfully issued.

[0063] In some embodiments, the task being successfully bid by a drone with a high bid includes:

[0064] The drone that starts bidding will bid for the task with the highest profit. If the drone's profit for all tasks is not greater than 0, it will quit bidding. Otherwise, it will increase its bid. The amount of the price increase is calculated based on the profit of the task and the profit of the second-best task.

[0065] If another drone has bid for the task being bid, and the current bid is higher, the previous bidder will fail, and the drone that failed will start bidding again.

[0066] If no other drone has bid for the current bid task, the next drone in the drone list will start bidding.

[0067] In the above embodiment, the task allocation adopts an auction algorithm. The auction algorithm is to bid for the task that the drone is interested in, and the task is allocated to the drone with the highest bid. For drones that did not win the bid (other drones bid higher), you can make a higher bid or change the task you want to perform and bid again. When the drone finds that the revenue of all tasks is less than 0, it stops bidding until all drones bid for the task or stop bidding, and the task allocation process ends. For drones that have bid for the task, the combination of drones and tasks forms the final task allocation plan. The iterative process of the auction algorithm can make the indicators better at each iteration. Compared with the traditional combinatorial optimization algorithm, it converges faster, takes less time to allocate, is not easily limited to the local optimum, and has a better allocation effect.

[0068] The value of a task is determined by the target type. Time-sensitive targets have a higher value, which is 100, while non-time-sensitive targets have a lower value, which is 50. Time-sensitive targets are attacked first. The cost of a task represents the cost of executing the task, which is calculated by the distributed task cost. The income of a task is the difference between the value of the task and the cost of the task.

[0069] income=value-cost

[0070] The profit of a task is the difference between the task revenue and the current quotation of the task.

[0071] profit = income-offer

[0072] The drone will quote the task with the highest profit, and the markup value π is calculated by the following formula

[0073] π=profit-(income s -offer s )+ε

[0074] Among them, profit is the profit of this task, income s -offer s is the profit of the suboptimal task, ε is a constant parameter to avoid too small price increase each time, generally taken as 1 / n, and n is the number of drones.

[0075] If other drones have already bid for the task, and a drone has made a higher bid this time, the original drone will lose the bid and need to make a higher bid or change the task to bid. When the profit of the drone for all targets is less than 0, the drone will withdraw from the auction. When all drones have obtained the task or withdraw from the auction, the task allocation ends. For all drones that have won the task, such as U i The task to be performed is T i , by all U i Tasks won in the auction i The task allocation matrix X of the task allocation problem can be combined N×M , which is the optimal task allocation solution.

[0076] In some embodiments, the redivision of labor in S7 includes: decomposing the collaborative task into a corresponding number of single-machine tasks according to the number of drones required, for example, a collaborative task that requires two drones to attack and one drone to evaluate is decomposed into two attack tasks and one evaluation task. After the single-machine task allocation, the tasks are redistributed among the drones that undertake the same collaborative task, and are allocated to the task machines from near to far from the target in sequence according to the execution sequence of the tasks to meet the timing constraints of the collaborative task. For example, if the collaborative evaluation task is an accompanying evaluation, the drone that performs the evaluation task needs to arrive first and shoot the entire attack process. The drone closest to the target among the three drones performs the evaluation task, and the other two drones perform the attack task.

[0077] The time consistency of collaborative tasks means that the drones performing collaborative tasks must agree on a time to start performing the collaborative tasks at the same time to ensure the execution effect of the collaborative tasks. The task cost has calculated the time required for the drone to perform the task, which includes the time to reach the target point, and the execution time of the collaborative task can be determined based on this. The estimated time for all drones of the same collaborative task to reach the target point is calculated, and the latest arrival time is added with a 5-second time buffer as the execution time of the collaborative task to meet the time consistency requirements of the collaborative task.

[0078] In some embodiments, the method further includes: after the relay loses connection, the drone closest to the center of all drones is used as a temporary relay node to coordinate the task allocation process instead of the relay. The stability and robustness of the task allocation process are ensured through a perfect exception handling mechanism.

[0079] The above is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application. The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and variations can be made without departing from the technical principles of the present application, and these improvements and variations should also be regarded as the protection scope of the present application.

Claims

1. A cluster distributed collaborative task allocation method that can be used in a denial environment, characterized in that: The method comprises: S1: Calculate the scope of task allocation; S2: Send a quotation request to the drones participating in the task allocation. After receiving the quotation request, the drone calculates the task execution cost and returns the cost information to the relay; S3: After receiving the cost information, the relay stores it to form a task cost matrix; S4: Calculate the task value according to the target type and combine it with the task cost to obtain the task benefit matrix; S5: Initialize the profit matrix, where the profit is the difference between the task revenue and the quotation; S6: The drones in the drone list start bidding in order. Based on the new bidding information, the profit matrix of all drones for the task is updated. The task is successfully bid by the drone with the highest bid. Until all drones have completed their bidding, the drone with the successful bid is paired with the task. S7: Traverse all tasks to determine whether they are collaborative tasks. If they are collaborative tasks, redivide the work according to the collaborative task requirements. The collaborative task time is calculated based on the estimated completion time of the task, and task instructions are issued based on the task allocation results.

2. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The S1 includes calculating the task allocation range by screening the drones.

3. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 2, characterized in that: Screening of drones includes first screening by drone type and then screening by the distance between the drone and the target.

4. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The calculation task execution cost C in S2 is: C=p1d+p2t+p3s Among them, p is the weight coefficient, d is the distance cost, t is the time cost, and s is the threat cost.

5. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The S2 also includes that if a certain UAV does not return the task execution cost within 100 cycles, the task execution cost is calculated by the relay machine instead, and the relay machine uses the straight-line distance between the lost UAV and the target as the task execution cost of the UAV.

6. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The task successfully bid by the drone with the highest bid includes: The drone that starts bidding will bid for the task with the highest profit. If the drone's profit for all tasks is not greater than 0, it will quit bidding. Otherwise, it will increase its bid. The amount of the price increase is calculated based on the profit of the task and the profit of the second-best task. If another drone has bid for the task being bid, and the current bid is higher, the previous bidder will fail, and the drone that failed will start bidding again. If no other drone has bid for the current bid task, the next drone in the drone list will start bidding.

7. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The re-division of work in S7 includes: decomposing the collaborative task into a corresponding number of single-machine tasks according to the required number of drones.

8. The cluster distributed collaborative task allocation method applicable to a denied environment according to claim 1, characterized in that: The method further includes: after the relay loses connection, using the drone closest to the center of all drones as a temporary relay node to coordinate the task allocation process instead of the relay.

Images