An Unmanned Aerial Vehicle Cluster Task Allocation Method Based on Improved Contract Net Algorithm

Abstract

The present application discloses a method for UAV swarm task allocation based on an improved contract net algorithm. The method includes: establishing a target benefit model for UAV swarm task pre-allocation in combination with the UAV swarm operation environment information; establishing a loss cost model for UAV swarm task pre-allocation in combination with the UAV swarm operation environment information; establishing a UAV swarm collaborative task pre-allocation model based on the target benefit and loss cost models of task pre-allocation and in combination with the constraint conditions during the process of the UAV swarm executing tasks; establishing a loss cost model for UAV swarm task re-allocation in combination with the requirements of the task re-allocation problem; establishing a task allocation efficiency function model in combination with the UAV swarm task pre-allocation model and the task re-allocation loss cost model; optimizing the efficiency function model based on the improved contract net algorithm to obtain an optimal task allocation scheme. The present invention improves the effectiveness and real-time performance of UAV swarm task allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative control of UAV clusters, and particularly to a method for task allocation of UAV clusters based on an improved contract net algorithm. Background Art

[0002] Unmanned aerial vehicles (UAVs) originated in the military field and have the advantages of reusability, recyclability, no casualties, strong continuous working ability, and low full-life cycle cost. Due to the increasingly complex combat environment and the limited mission execution ability of a single UAV, the collaborative combat of multiple UAV clusters will become an important development trend in future UAV combat, and the collaborative combat of clusters can achieve complementary capabilities among UAVs, thereby improving the effectiveness of the entire system. An efficient and reasonable task allocation method is a prerequisite for giving full play to the advantages of the collaborative combat of multiple UAV clusters and realizing the effective utilization of resources.

[0003] At present, extensive exploration and research have been carried out on UAV cluster task allocation technology at home and abroad, and many research results can also achieve good goals in a static environment. However, with the increase in the number of UAVs, the existing solutions are inefficient, have a slow convergence speed, and currently, the methods basically do not consider the uncertainty of different mission objectives and unexpected situations that may occur at any time. Therefore, there is an urgent need for a method for task allocation of UAV clusters based on an improved contract net algorithm applicable to the dynamic transformation of the environment and tasks. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for task allocation of UAV clusters based on an improved contract net algorithm to solve the problem of resource waste or even mission failure caused by task conflicts in existing UAV clusters.

[0005] The present application provides a method for task allocation of UAV clusters based on an improved contract net algorithm, and the method includes the following steps:

[0006] Step 1: Combine the UAV cluster operation environment information to establish a target benefit model for pre-allocation of UAV cluster tasks;

[0007] Step 2: Combine the UAV cluster operation environment information to establish a loss cost model for pre-allocation of UAV cluster tasks;

[0008] Step 3: Based on the target benefit and loss cost models of task pre-allocation, add the constraint conditions in the process of UAV clusters executing tasks to establish a collaborative task pre-allocation model for UAV clusters;

[0009] Step 4: Combine the requirements of the task re-allocation problem to establish a loss cost model for task re-allocation of UAV clusters;

[0010] Step 5: Establish a task allocation efficiency function model based on the UAV cluster task pre-allocation model and the task re-allocation loss cost model;

[0011] Step 6: Optimize the efficiency function model based on the improved contract net algorithm to obtain the optimal task allocation plan.

[0012] Optionally, combine the UAV cluster operation environment information to establish a target revenue model for UAV cluster task pre-allocation, including:

[0013] Let P ij represent the probability of UAV U i successfully hitting target G i , and the corresponding value revenue is represented as I j . Then the expected value revenue when UAV U i hits target G i is expressed as: P ij ×I j . Then the target value function y1 when the UAV cluster conducts strikes is expressed as:

[0014]

[0015] where M is the number of UAVs, N is the number of targets to be struck, I j is the value of each sub-target, and the maximum target value is represented as W ij is the task allocation plan of the UAV;

[0016] The goal of task allocation is to obtain the maximum value of the value function which is expressed as: miny1 = 1 - y1.

[0017] Optionally, combine the UAV cluster operation environment information to establish a loss cost model for UAV cluster task pre-allocation, including:

[0018] Step 2-1: Determine the hit loss function y2:

[0019] Let Q ij represent the probability that UAV U i is hit by counterattack when attacking target G i . The corresponding value loss of the UAV is L i . Then the expected value loss caused by being hit by the enemy's counterattack when UAV U i attacks target G i is: Q ij ×L i . The hit loss function y2 during the strike of the UAV cluster is expressed as:

[0020]

[0021] Among them, M is the number of UAVs, N is the number of targets to be struck, and L i is the value of the UAV itself, represents the maximum value of the UAV itself, and W ij is the mission assignment plan of the UAV; the goal of mission assignment is to minimize the self-loss function, corresponding to obtaining miny2;

[0022] Step 2-2, determine the trajectory cost function y3:

[0023] Assume that the UAV U i starts from the starting point and executes n missions along the planned path. Then the total flight distance of the UAV U i is:

[0024]

[0025] Among them, D po is the flight distance of the UAV U i from the starting point to the first target, is the flight range of the UAV from the j-th target to the (j + 1)-th target on the route. Then the trajectory cost function f3 of the UAV formation is expressed as:

[0026]

[0027] Among them, M is the number of UAVs, represents the maximum flight range when the UAV executes the mission; the goal of mission assignment is to minimize the trajectory cost function miny3;

[0028] Step 2-3, determine the ammunition loss function y4:

[0029] When the UAV U i attacks the target G i the equipment model used is E i , and the corresponding construction cost is Then the ammunition loss function during the UAV swarm strike is expressed as:

[0030]

[0031] Among them, M is the number of UAVs, N is the number of targets to be struck, represents the maximum value of the ammunition construction cost, and W ij is the assignment plan. Then the goal of mission assignment is to minimize the ammunition loss function miny4.

[0032] Optionally, based on the target benefit and loss cost model of task pre - allocation, add the constraint conditions in the process of the UAV cluster executing tasks, and establish a collaborative task pre - allocation model for the UAV cluster, including:

[0033] Integrate the comprehensive objective value function, the hit loss function, the trajectory cost function, and the ammunition loss function, conduct multi - objective optimization, and obtain the overall evaluation function as:

[0034] minf(W)=(1 - y1(W),y2(W),y3(W),y4(W))

[0035] The overall evaluation function is the overall model for the collaborative task allocation of the UAV cluster;

[0036] The constraint conditions defined in the UAV cluster task allocation model are as follows:

[0037] (1) Tactical constraint c1:

[0038] To prevent situations that affect task effectiveness such as ineffective strikes and repeated strikes, it is stipulated that each target can only be executed once by one UAV:

[0039]

[0040] (2) Capacity constraint c2

[0041] To prevent the tasks assigned to UAVs from exceeding their own capabilities, it is stipulated that the number of tasks executed by a UAV is limited by the amount of ammunition it carries:

[0042]

[0043] Among them, A imax represents the maximum amount of resources that each UAV can carry;

[0044] (3) Task constraint c3

[0045] To prevent omissions in the task allocation process, it is stipulated that all targets need to be assigned tasks:

[0046]

[0047] Among them, N type represents the number of types of tasks executed;

[0048] Convert each objective function and constraint condition into a single - objective optimization problem through weighted summation, and give the comprehensive evaluation function Y(Task):

[0049] Y(Task)=σ1·y1 + σ2·y2 + σ3·y3 + σ4·y4

[0050] Among them, σ1, σ2, σ3, and σ4 are the weight coefficients of the four objective functions respectively, and σ1 + σ2 + σ3 + σ4 = 1.

[0051] Optionally, according to the requirements of the task reallocation problem, a loss cost model for the task reallocation of the UAV cluster is established, including:

[0052] Assume that the task sequence assigned to UAV Ui is S i ={G i1 , G i2 , …, G in}, and the communication cost function y5 is as follows:

[0053]

[0054] Among them, M is the number of UAVs, Com ij is the communication cost between UAV U i and UAV U j ; represents the maximum communication cost; Z ij is an M×N communication policy, and when the value is 1, it indicates that there is an auction behavior that requires communication.

[0055] Optionally, according to the UAV cluster task pre - allocation model and the task re - allocation loss cost model, a task allocation efficiency function model is established, including:

[0056] Combined with the objective value function y1, the hit loss function y2, the track cost function y3, and the ammunition loss function y4 in the task pre - allocation model, the overall efficiency function of task allocation is defined as:

[0057]

[0058] Among them, α and β are weight coefficients between 0 and 1, and α + β = 1;

[0059] Optionally, based on the improved contract net algorithm, the efficiency function model is optimized to obtain the optimal task allocation scheme, including:

[0060] Step 6 - 1: The tenderer transmits the task information to all members participating in the bidding;

[0061] Step 6 - 2: All members generate bids containing personal information on the premise that their personal capabilities meet the task requirements, and then submit them to the tenderer. If no one wins the bid, the algorithm ends;

[0062] Among them, the load - balancing buffer pool mechanism is introduced in the process of generating bids, allowing bidders to bid on multiple tasks and sign multiple contracts;

[0063] The UAV load rate is determined by the following method:

[0064] load i = nI(S i ), S i S i = {G i1 , G i2 , …, G in}

[0065]

[0066]

[0067] where load i is the payload of the UAV U i : I(S i ) is the corresponding effectiveness function; M is the number of UAVs, and n is the number of tasks executed; is the average load of all UAVs; LR i is the load rate of the UAV;

[0068] When the upper - layer control issues a new task as the bidder, the UAVs participating in the auction can participate in the bidding only when their load rates are lower than the average load rate;

[0069] Step 6 - 3: The bidder selects and authorizes the member with the highest evaluation level according to the bid. If the other party agrees, the contract is signed. The remaining bidders wait for the remaining work to be assigned. If they oppose, step 6 - 1 is executed;

[0070] Step 6 - 4: The winner successfully completes the task and feeds back the task information to the bidder. Otherwise, the bidder executes step 6 - 1.

[0071] This application also provides a system for the UAV swarm task allocation method based on the improved contract - net algorithm. The system includes:

[0072] A construction module, which is used to establish a target revenue model for the pre - allocation of UAV swarm tasks by combining the UAV swarm operation environment information; and to establish a loss cost model during the execution of tasks by the UAV swarm by combining the operation environment information; and to establish a collaborative task pre - allocation model for the UAV swarm by adding the constraint conditions during the execution of tasks based on the target revenue and loss cost models for task pre - allocation; and to establish a loss cost model for the task re - allocation of the UAV swarm by combining the requirements of the task re - allocation problem; and to establish a task allocation effectiveness function model according to the UAV swarm task pre - allocation model and the task re - allocation loss cost model;

[0073] A decision - making module is used to optimize the effectiveness function model based on the improved contract - net algorithm to obtain the optimal task allocation scheme.

[0074] Compared with the prior art, the remarkable advantages of the present invention are as follows: (1) In response to the requirements and constraints of the UAV mission pre-allocation problem in complex environments, the present invention constructs a corresponding mission pre-allocation model; in the case where the mission pre-allocation strategy cannot be executed, in view of the deficiencies of the traditional contract net algorithm, a buffer pool and load balancing strategy are introduced to ensure the load balancing of the allocation results; based on the improved contract net algorithm, a mission allocation model is constructed, making full use of the execution effect information and path information of UAVs for targets to achieve efficient allocation of collaborative missions of UAV swarms. (2) The UAV swarm mission allocation method based on the improved contract net algorithm provided by the present invention efficiently allocates various tasks to the UAV formation reasonably, making various performance indicators of the system optimal, giving play to the collaborative working efficiency of the UAV formation, and improving the effectiveness and real-time performance of UAV swarm mission allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 FIG. is a flowchart of the UAV swarm mission allocation method based on the improved contract net algorithm of the present invention.

[0076] Figure 2 FIG. is a flowchart of the improved contract net algorithm of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0077] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0078] Combined with Figure 1 , the present application provides a UAV swarm mission allocation method based on an improved contract net algorithm, including:

[0079] Step 1: Combine the UAV swarm operation environment information to establish an objective benefit model for UAV swarm mission pre-allocation;

[0080] Step 2: Combine the UAV swarm operation environment information to establish a loss cost model for UAV swarm mission pre-allocation;

[0081] Step 3: Based on the objective benefit and loss cost models of mission pre-allocation, add the constraint conditions in the process of UAV swarm executing missions to establish a collaborative mission pre-allocation model for UAV swarms;

[0082] Step 4: Combine the requirements of the mission re-allocation problem to establish a loss cost model for UAV swarm mission re-allocation;

[0083] Step 5: According to the UAV swarm mission pre-allocation model and the mission re-allocation loss cost model, establish a mission allocation efficiency function model;

[0084] Step 6: Optimize the efficiency function model based on the improved contract net algorithm to obtain the optimal mission allocation plan.

[0085] Specifically, in step 1, for the UAV swarm mission allocation method based on the improved contract net algorithm according to claim 1, it is characterized in that, by combining the UAV swarm operation environment information, a target benefit model for UAV swarm mission pre-allocation is established, including:

[0086] Let P ij represent the probability of UAV U i successfully hitting target G i , and the corresponding value benefit is expressed as I j . Then, the expected value benefit when UAV U i hits target G i is expressed as: P ij ×I j . Then, the target value function y1 when the UAV swarm conducts the strike is expressed as:

[0087]

[0088] where M is the number of UAVs, N is the number of targets to be struck, I j is the value of each sub-target, and the maximum target value is expressed as W ij is the UAV mission allocation plan;

[0089] The goal of mission allocation is to obtain the maximum value of the value function expressed as: miny1 = 1 - y1.

[0090] In step 2, by combining the UAV swarm operation environment information, a loss cost model for UAV swarm mission pre-allocation is established, including:

[0091] Step 2-1, determine the hit loss function y2:

[0092] Let Q ij represent the probability that UAV U i is hit by counterattack when attacking target G i . The corresponding value loss of the UAV is L i . Then, the expected value loss caused by being hit by the enemy's counterattack when UAV U i attacks target G i is: Q ij ×L i . The hit loss function y2 during the strike of the UAV swarm is expressed as:

[0093]

[0094] where M is the number of UAVs, N is the number of targets to be struck, and L i is the value of the UAV itself, Represents the maximum self - value of the UAV, W ij is the task allocation scheme for the UAV; the goal of task allocation is to minimize the self - loss function, corresponding to obtaining miny2;

[0095] Step 2 - 2, determine the trajectory cost function y3:

[0096] Assume the UAV U i starts from the starting point and executes n tasks along the planned path, then the UAV U i 's total flight distance is:

[0097]

[0098] where D po is the flight distance of the UAV U i from the starting point to the first target, is the flight range of the UAV from the j - th target to the (j + 1) - th target on the flight path, then the trajectory cost function f3 of the UAV formation is expressed as:

[0099]

[0100] where M is the number of UAVs, represents the maximum flight range when the UAV executes tasks; the goal of task allocation is to minimize the trajectory cost function miny3;

[0101] Step 2 - 3, determine the ammunition loss function y4:

[0102] The UAV U i uses the equipment model E i when attacking the target G i , and the corresponding construction cost is Then the ammunition loss function during the UAV swarm strike is expressed as:

[0103]

[0104] where M is the number of UAVs, N is the number of targets to be attacked, represents the maximum value of the ammunition construction cost, W ij allocation scheme, then the goal of task allocation is to minimize the ammunition loss function miny4.

[0105] In step 3, based on the target benefit and loss cost model of task pre - allocation, add the constraint conditions during the process of the UAV swarm executing tasks, and establish a UAV swarm collaborative task pre - allocation model, including:

[0106] Based on the comprehensive objective value function, the hit loss function, the track cost function, and the ammunition loss function, multi-objective optimization is carried out to obtain the overall evaluation function as follows:

[0107] minf(E)=(1 - y1(W), y2(W), y3(W), y4(W))

[0108] The overall evaluation function is the overall model for the collaborative task allocation of the UAV swarm;

[0109] The constraint conditions defined in the UAV swarm task allocation model are as follows:

[0110] (1) Tactical constraint c1:

[0111] To prevent situations that affect mission effectiveness such as ineffective strikes and repeated strikes, it is stipulated that each target can only be tasked by one UAV once:

[0112]

[0113] (2) Capacity constraint c2

[0114] To prevent the tasks assigned to a UAV from exceeding its own capabilities, it is stipulated that the number of tasks performed by a UAV is limited by the amount of ammunition it carries:

[0115]

[0116] Among them, A imax represents the maximum amount of resources that each UAV can carry;

[0117] (3) Task constraint c3

[0118] To prevent omissions during the task allocation process, it is stipulated that all targets need to be assigned tasks:

[0119]

[0120] Among them, N type represents the number of types of tasks performed;

[0121] By weighted summation, each objective function and constraint condition are converted into a single-objective optimization problem, and the comprehensive evaluation function Y(Task) is given:

[0122] Y(Task)=σ1·y1 + σ2·y2 + σ3·y3 + σ4·y4

[0123] Among them, σ1, σ2, σ3, σ4 are the weight coefficients of the four objective functions, and σ1 + σ2 + σ3 + σ4=1.

[0124] In step 4, according to the requirements of the task reallocation problem, a loss cost model for the task reallocation of the UAV cluster is established, including:

[0125] Assume that the task sequence assigned to UAV Ui is S i ={G i1 ,G i2 ,…,G in}}, and the communication cost function y5 is as follows:

[0126]

[0127] where M is the number of UAVs, Com ij is the communication cost between UAV U i and UAV U j , represents the maximum cost of communication; Z ij is an M×M communication policy, and a value of 1 indicates that there is an auction behavior that requires communication.

[0128] In step 5, according to the UAV cluster task pre-allocation model and the task reallocation loss cost model, a task allocation efficiency function model is established, including:

[0129] Combined with the target value function y1, the hit loss function y2, the track cost function y3, and the ammunition loss function y4 in the task pre-allocation model, the overall efficiency function of the task allocation is defined as:

[0130]

[0131] where α and β are weight coefficients between 0 and 1, and α + β = 1.

[0132] In step 6, based on the improved contract net algorithm, the efficiency function model is optimized to obtain the optimal task allocation scheme, including:

[0133] Step 6-1: The tenderer transmits the task information to all participating bidders

[0134] Step 6-2: All members generate bids containing personal information on the premise that their personal capabilities meet the task requirements, and then submit them to the tenderer. If no one wins the bid, the algorithm ends;

[0135] Among them, the load balancing buffer pool mechanism is introduced in the process of generating bids, allowing bidders to bid on multiple tasks and sign multiple contracts;

[0136] The UAV load rate is determined by the following method:

[0137] load i = nI(S i ),S i Si = {G i1 , G i2 , …, G in}

[0138]

[0139]

[0140] Among them, load i is the payload of the UAV U i : I(S i ) is the corresponding effectiveness function; M is the number of UAVs, and n is the number of tasks executed; is the average load of all UAVs; LR i is the load rate of the UAV;

[0141] When the upper - layer control, as the tenderer, issues a new task, the UAVs participating in the auction can participate in the bidding only when their load rates are lower than the average load rate;

[0142] Step 6 - 3: The tenderer selects and authorizes the member with the highest evaluation level according to the tender document. If the other party agrees, the contract is signed. The remaining tenderers wait for the remaining work to be assigned. If they object, step 6 - 1 is executed;

[0143] Step 6 - 4: The winning bidder successfully completes the task and feeds back the task information to the tenderer. Otherwise, the tenderer executes step 6 - 1.

[0144] The flow chart of the improved contract - net algorithm is as Figure 2 shown.

[0145] This application also provides a system for the UAV swarm task allocation method based on the improved contract - net algorithm, characterized in that the system includes:

[0146] A construction module, used to establish a target revenue model for the pre - allocation of UAV swarm tasks by combining the UAV swarm operation environment information; and used to establish a loss cost model during the execution of tasks by the UAV swarm by combining the operation environment information; and based on the target revenue and loss cost models of the pre - allocation of tasks, adding the constraint conditions during the execution of tasks by the UAV swarm to establish a collaborative task pre - allocation model for the UAV swarm; and combining the requirements of the task re - allocation problem to establish a loss cost model for the task re - allocation of the UAV swarm; and establishing a task allocation effectiveness function model according to the UAV swarm task pre - allocation model and the task re - allocation loss cost model;

[0147] A decision - making module is used to optimize the effectiveness function model based on the improved contract - net algorithm to obtain the optimal task allocation plan.

[0148] Correspondingly, the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method provided by the present application are performed.

[0149] The present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method provided by the present application are performed.

[0150] In view of the requirements and constraints of the UAV mission pre-allocation problem in complex environments, the present invention constructs a corresponding mission pre-allocation model; in the case where the mission pre-allocation strategy cannot be executed, in view of the deficiencies of the traditional contract net algorithm, a buffer pool and a load balancing strategy are introduced to ensure the load balancing of the allocation results; a mission allocation model is constructed based on the improved contract net algorithm, making full use of the execution effect information and path information of the UAVs for the targets, and realizing the efficient allocation of the collaborative tasks of the UAV cluster. The UAV cluster mission allocation method based on the improved contract net algorithm provided by the present invention efficiently allocates various tasks to the UAV formation reasonably, making the various performance indicators of the system optimal, giving play to the collaborative working efficiency of the UAV formation, and improving the effectiveness and real-time performance of the UAV cluster mission allocation.

Claims

1. A method for UAV swarm mission assignment based on an improved contract net algorithm, characterized in that It includes the following steps: Step 1: Combine the information of the UAV cluster operation environment to establish an objective benefit model for the pre-allocation of UAV cluster tasks; Step 2: Combine the information of the UAV cluster operation environment to establish a loss cost model for the pre-allocation of UAV cluster tasks; Step 3: Based on the objective benefit and loss cost models of task pre-allocation, add the constraint conditions during the execution of tasks by the UAV cluster to establish a collaborative task pre-allocation model for the UAV cluster; Step 4: Combine the requirements of the task re-allocation problem to establish a loss cost model for the task re-allocation of the UAV cluster; Step 5: According to the UAV cluster task pre-allocation model and the task re-allocation loss cost model, establish a task allocation efficiency function model; Step 6: Optimize the efficiency function model based on the improved contract net algorithm to obtain the optimal task allocation plan; Optimizing the efficiency function model based on the improved contract net algorithm to obtain the optimal task allocation plan, including: Step 6-1: The tenderer transmits the task information to all members participating in the bidding; Step 6-2: All members generate bids containing personal information on the premise that their personal capabilities meet the task requirements, and then submit them to the tenderer. If no one wins the bid, the algorithm ends; Among them, the load balancing buffer pool mechanism is introduced in the process of generating bids, allowing bidders to bid for multiple tasks and sign multiple contracts; The UAV load rate is determined by the following method: load i = nI(S i ), S i S i = {G i1 , G i2 , …, G in} Among them, load i is the payload of the drone U i : I(S i ) is the corresponding effectiveness function; M is the number of drones, and n is the number of tasks executed; is the average load of all drones; LR i is the load rate of the drone; When the upper-level control issues a new task as the tenderer, the UAVs participating in the auction can participate in the bidding only when their load rates are lower than the average load rate; Step 6-3: The bidder selects and authorizes the member with the highest evaluation level according to the bid. If the other party agrees, the contract is signed. The remaining bidders wait for the remaining work to be assigned. If they object, execute Step 6-1; Step 6-4: The winner successfully completes the task and feeds back the task information to the tenderer. Otherwise, the tenderer executes Step 6-1.

2. The method for task allocation of an unmanned aerial vehicle cluster based on an improved contract net algorithm according to claim 1, wherein Combining the information of the UAV cluster operation environment to establish an objective benefit model for the pre-allocation of UAV cluster tasks, including: Denote P ij represent the drone U i strike target G i the probability of success, and the corresponding value gain is denoted as I j , then the drone U i against target G i The expected value gain during the strike is denoted as: P ij ×I j , then the target value function y1 of the drone swarm during the strike is denoted as: Among them, M is the number of UAVs, N is the number of targets to be struck, and I j is the value of each sub-target, and the maximum target value is expressed as W ij is the mission assignment plan of the UAVs; The goal of task allocation corresponds to obtaining the maximum value of the value function It is expressed as: miny1 = 1 - y1.

3. The method for task allocation of an unmanned aerial vehicle cluster based on an improved contract net algorithm according to claim 1, characterized in that, Combining the information of the UAV cluster operation environment to establish a loss cost model for the pre-allocation of UAV cluster tasks, including: Step 2-1, determine the hit loss function y2: Denote Q ij represent the probability that the drone U i is counterattacked and hit when attacking the target G i and the corresponding value loss of the drone is L i , then for the drone U i when attacking the target G i the expected value loss caused by being counterattacked and hit by the enemy is: Q ij ×L i , the hit loss function y2 of the drone swarm during the strike process is expressed as: Among them, M is the number of UAVs, N is the number of targets to be attacked, and L i is the value of the UAV itself, represents the maximum value of the UAV itself, W ij is the mission assignment plan of the UAV; the goal of mission assignment is to minimize the self-loss function, corresponding to obtaining miny2; Step 2-2, determine the track cost function y3: Suppose the drone U i starts from the starting point and executes n tasks along the planned path, then the drone U i 's total flight distance is: Among them, D po is the flight distance of the UAV U i from the starting point to the first target, is the flight range of the UAV from the j-th target to the (j + 1)-th target on the route. Then, the trajectory cost function f3 of the UAV formation is expressed as: where M is the number of UAVs, represents the maximum flight range of the UAVs when performing tasks; the goal of task allocation is to minimize the trajectory cost function miny3; Step 2-3, determine the ammunition loss function y4: Drone U i Attack target G i The equipment model used when is E i , and the corresponding construction cost is Then the ammunition loss function during the drone swarm strike is expressed as: Where M is the number of UAVs and N is the number of targets to be struck. represents the maximum value of the ammunition cost, W ij For the allocation plan, the goal of task allocation is to minimize the ammunition loss function miny4.

4. The method for allocating tasks of an unmanned aerial vehicle cluster based on an improved contract net algorithm according to claim 1, wherein Based on the objective benefit and loss cost models of task pre-allocation, add the constraint conditions during the execution of tasks by the UAV cluster to establish a collaborative task pre-allocation model for the UAV cluster, including: Integrate the objective value function, hit loss function, track cost function, and ammunition loss function to conduct multi-objective optimization to obtain the overall evaluation function as: minf(W)=(1-y1(W),y2(W),y3(W),y4(W)) The overall evaluation function is the overall model for the collaborative task allocation of the UAV cluster; The constraint conditions defined in the UAV cluster task allocation model are as follows: (1) Tactical constraint c1: To prevent situations that affect task effectiveness such as ineffective strikes and repeated strikes, it is stipulated that each target can only be executed by one UAV once for a task: (2) Capacity constraint c2 To prevent the tasks assigned to the UAVs from exceeding their capabilities, it is stipulated that the number of tasks executed by the UAVs is limited by the amount of ammunition they carry: Among them, A imax represents the maximum amount of resources that each drone can carry; (3) Task constraint c3 To prevent omissions in the task assignment process, it is stipulated that all targets need to be assigned tasks: Among them, N type represents the number of types of tasks to be executed; By weighted summation, each objective function and constraint condition are converted into a single-objective optimization problem, and a comprehensive evaluation function Y(Task) is given: Y(Task) = σ1·y1 + σ2·y2 + σ3·y3 + σ4·y4 where σ1, σ2, σ3, and σ4 are the weight coefficients of the four objective functions, and σ1 + σ2 + σ3 + σ4 = 1.

5. The method for unmanned aerial vehicle cluster task allocation based on the improved contract net algorithm according to claim 1, characterized in that, According to the requirements of the task reallocation problem, a loss cost model for the UAV cluster task reallocation is established, including: Suppose the task sequence assigned to the UAV Ui is S i ={G i1 ,G i2 ,…,G in}, and the communication cost function y5 is as follows: Among them, M is the number of drones, Com ij is the communication cost between the drone U i and the drone U j ; represents the maximum cost of communication; Z ij is an M×M communication policy, and a value of 1 indicates that there is an auction behavior that requires communication.​ 6. The method for allocating tasks of an unmanned aerial vehicle cluster based on an improved contract net algorithm according to claim 1, characterized in that According to the UAV cluster task pre-allocation model and the task reallocation loss cost model, a task allocation efficiency function model is established, including: Combined with the objective value function y1, the hit loss function y2, the track cost function y3, and the ammunition loss function y4 in the task pre-allocation model, the overall efficiency function of task allocation is defined as: where α and β are weight coefficients between 0 and 1, and α + β = 1.

7. A system for the method of UAV swarm task allocation based on the improved contract net algorithm according to any one of claims 1 to 6, characterized in that, The system includes: A construction module for establishing a target benefit model for the UAV cluster task pre-allocation by combining the UAV cluster operation environment information; and a construction module for establishing a loss cost model during the UAV cluster task execution by combining the operation environment information; and a construction module for adding the constraint conditions during the UAV cluster task execution based on the target benefit and loss cost models of the task pre-allocation to establish a UAV cluster collaborative task pre-allocation model; and establishing a loss cost model for the UAV cluster task reallocation according to the requirements of the task reallocation problem; and establishing a task allocation efficiency function model according to the UAV cluster task pre-allocation model and the task reallocation loss cost model; A decision module for optimizing the efficiency function model based on the improved contract net algorithm to obtain the optimal task allocation plan.

8. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 6.

Images