Multi-unmanned aerial vehicle multi-task allocation method based on improved contract net algorithm

Abstract

The invention provides a multi-unmanned aerial vehicle multi-task allocation method based on an improved contract net algorithm. The method comprises the steps of 1, designing evaluation indexes of task allocation; 2, constructing an exchange negotiation strategy based on a buying and selling negotiation strategy; 3, constructing a multi-unmanned aerial vehicle multi-task distribution auction strategy based on the evaluation indexes and the exchange negotiation strategy; according to the multi-unmanned aerial vehicle multi-task allocation method based on the improved contract network algorithm, the communication traffic of a traditional contract network algorithm is reduced, the problems that when multi-unmanned aerial vehicle multi-task allocation is carried out, the efficiency of the traditional contract network algorithm is reduced along with the increase of the number and scale of the unmanned aerial vehicles, and the allocation result quality is poor are solved, and the task allocation quality and efficiency are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of multi-unmanned aerial vehicle task scheduling and allocation, and particularly relates to a multi-unmanned aerial vehicle multi-task allocation method based on an improved contract net algorithm. BACKGROUND

[0002] With the continuous development of unmanned aerial vehicle related technologies, the types of unmanned aerial vehicles are also increasing, their functions are gradually improving and enriching, and the task fields of unmanned aerial vehicles are also expanding. Compared with a single unmanned aerial vehicle, a multi-unmanned aerial vehicle system has the characteristics of being mobile and flexible, and has strong adaptability, and can perform more complex, higher requirement and stronger confrontation tasks.

[0003] At present, the task allocation method takes the distributed solution algorithm as a research hotspot. Typical ones are the decentralized MDP (Dce-MDP) method, the market mechanism based method, the dynamic distributed constraint optimization problem method and the like. The market mechanism based method has good effects, such as the contract net protocol (CNP), the CBBA (Consensus-based Bundle Algorithm) algorithm, the auction algorithm and the like.

[0004] The contract net algorithm was proposed by Smith in the 1980s, which simulates the "bidding-tender-winning bid" mode of human beings in the market mechanism, and through communication and negotiation between multiple individuals meeting certain constraint conditions, the global optimal solution is pursued while the individual's own benefit is met, and it can be used for multi-unmanned aerial vehicle multi-task allocation. The unmanned aerial vehicle is used to represent the individual in the contract net model, and the unmanned aerial vehicle in the model can have three roles of "bidders", "tenders" and "winners". In the traditional contract net algorithm, a bidder is selected from all unmanned aerial vehicles according to certain rules in each round, and the remaining unmanned aerial vehicles meeting the tender conditions are all tenders. The bidding unmanned aerial vehicle is responsible for the auction in this round, and it takes out a task from its current task set to bid. All the tender unmanned aerial vehicles make bids according to the bidding task and the value of their own benefits. After the bidding, the bidder compares the bids of all the tenderers, and selects the unmanned aerial vehicle with the highest bid as the winner to complete the task ownership transfer.

[0005] The process of the traditional contract net algorithm is shown in Figure 1 The traditional contract net algorithm is widely used in the single task single winner task allocation model, but with the increase of the number of unmanned aerial vehicles, the efficiency of the traditional contract net algorithm for task allocation is reduced, and the allocation result is not ideal, and the following problems exist: 1) The bidding model is simple. In traditional contract network algorithms, the bidding mechanism is a "single task, single winner" model, meaning that in each round, the bidder only bids for a single task, selecting a winner from all bidders who meet the bidding criteria. Under this bidding model, it cannot be guaranteed that the task bid by the bidder is the optimal task that is most beneficial to the overall task allocation result.

[0006] 2) High communication volume during bidding and negotiation. As the overall number of drones used in the future continues to grow, the communication volume will increase rapidly due to the significant increase in the number of individual drones, leading to a decrease in the real-time performance and reliability of task allocation.

[0007] 3) Limited negotiation strategies. In traditional contract network models, tenderers and bidders rely on simple buy-sell negotiation strategies, with both parties considering the inherent benefits and risks of the tendered task, resulting in poor quality of task allocation. In task allocation scenarios based on collaborative adversarial external tasks, the tenderers and bidders belong to the same multi-UAV system and are "comrades-in-arms," ​​so the benefits of the entire multi-UAV system should be considered, rather than just individual benefits.

[0008] In summary, when solving the task allocation problem involving multiple drones and multiple tasks, traditional contract network algorithms have shortcomings such as large communication volume, single and inefficient bidding mode, and poor optimization of allocation results. Summary of the Invention

[0009] The purpose of this invention is to provide a multi-task allocation method for multiple UAVs based on an improved contract network algorithm, which reduces the communication volume of the traditional contract network algorithm and overcomes the problems of reduced efficiency and poor allocation result quality of the traditional contract network algorithm when allocating multiple UAVs and tasks, thereby improving the quality and efficiency of task allocation. This application provides a method for multi-UAV multi-task allocation based on an improved contract network algorithm, the method comprising: Step 1: Design evaluation metrics for task allocation; Step 2: Construct an exchange negotiation strategy based on a buy-sell negotiation strategy; Step 3: Based on the evaluation indicators and the exchange negotiation strategy, construct a multi-UAV multi-task allocation bidding strategy.

[0010] Preferably, step 1 includes: Minimizing task execution time is used as the objective function for task allocation. When estimating tasks, the shortest time for a UAV to execute all tasks is calculated. When a UAV needs to bid for a task, the shortest path for each UAV to execute its task set is planned. At the same time, a task time cost table for each UAV to complete its own task set is provided. Preferably, each task in the task set has a corresponding time cost, and the value in the task time cost table represents the time cost that the UAV needs to pay to perform the corresponding task.

[0011] Preferably, the objective function for task allocation is to minimize task execution time; when estimating tasks, the shortest time for a UAV to execute all tasks is calculated; when a UAV needs to bid for a task, the shortest path for each UAV to execute its task set is planned; and a task time cost table for each UAV to complete its own task set is provided, including: drones Relative to its task sequence Tasks in The time cost is:

[0012] In the formula, Indicates drone Relative to the task The time cost, express Execute task sequence The time required Indicates from the task sequence Lieutenant General The new sequence after removal, express The time required to execute a new task sequence.

[0013] Preferably, step 2 includes: An exchange negotiation strategy was added to the buy-sell negotiation strategy of the Contract Network algorithm; Depending on the negotiation strategy, individuals in the contract network model submit two types of tenders: sales contracts and exchange contracts.

[0014] Preferably, the bidding document is defined as a seven-tuple. ; in, Indicates the bidding for drones, This indicates the number of remaining combat resources for the bidding drone. Indicates that a drone is being tendered. Indicates the task of bidding. This indicates the system's profit after the transaction is completed. This indicates the time required for the bid-tendered drone to complete its new mission sequence after the sale and purchase are completed. This indicates the time required for the tendered drone to complete its new mission sequence after the sale and purchase are completed.

[0015] Preferably, the exchange tender is defined as a seven-tuple. ; in, This indicates the first drone to be exchanged. This indicates the first drone used for the exchange mission. This indicates the second drone to be exchanged. This indicates that the second drone is used for the exchange mission. This indicates the system's benefit after the exchange is completed. This indicates the time required for the first drone to complete its new mission sequence after the swap. This indicates the time required for the second drone to complete its new mission sequence after the exchange.

[0016] Preferably, step 3 includes: 1) Bidding: With the objective function of minimizing the maximum task execution time of the UAV system, in each round, the bidder selects the UAV with the longest task execution time in the current multi-UAV system; the bidder adopts the auction strategy of "task group bidding" and publishes all tasks in its task set; 2) Bidding: Design a new contract format for "exchange of bids", including the generation of buy and sell bids and the generation of exchange bids; 3) Winning the bid: For valid bids generated during the bidding stage, the bidding drone will select the bid that maximizes the system benefit, i.e. optimizes the objective function the most, from all valid bids as the winning bid. The corresponding bidder becomes the winner, and the other bidders are considered unsuccessful bidders. For valid exchange bids generated during the bidding stage, the bid that maximizes the system benefit, i.e. optimizes the objective function the most, is selected as the winning bid, and the two corresponding drones will exchange tasks.

[0017] Preferably, the generation of the bidding document includes: Step 1: In addition to the tenderer, the system... After receiving the bidding task information, drones other than those listed above determine whether they meet the type constraints based on the task type, and then check whether their remaining combat resources meet the constraint on the number of tasks a single drone can perform. Drones that meet both constraints are considered bidders in this round of bidding. ; Step 2: Set the tenderer The values ​​in the task time cost table are the thresholds for generating bids and tenders for each bidding task. , for Relative to the tasks in its task set The time value of the task; each bidder submits a bid for the tendered task, which means "adding" the tendered task to their own task set and calculating the time required to complete the "new task sequence". Calculate the time required for the drone to complete its "new mission sequence" after it is "removed" from the bidding task. Finally, the calculations were performed to determine the relative performance of the "newly added" mission compared to the bidding drone. Task time value ;if This indicates that the system can complete the bidding task at a relatively low cost. At this point, the system's revenue is , The value represents the time saved by the system after a successful bidder generates a valid bid; if... but In this case, although the system failed to complete the task at a lower cost, it still optimized the objective function and generated a valid bid. The system's profit in this situation is... , The value represents the magnitude of the objective function optimization value after the bidder wins the bid; Step 3: Collect all valid bids and send them to the bidding drone. .

[0018] Preferably, the generation of the exchange tender document includes: Step 1: The "swap" strategy will not affect the number of tasks a single drone can perform. Therefore, all drones in the system that meet the type constraints will swap tasks with each other. Step 2: Identify the two drones to be swapped in this mission. and Then, in sequence Tasks in the task set The tasks in the task set are "swapped," and calculations are performed after each "swap." and Time to complete each "new task sequence" and Let the longer task execution time before the exchange be... The shorter task execution time is The execution time after the swap is and ,if and Both parties in the exchange can perform the same number of tasks and generate exchange tenders at a lower cost; in this case, the system's benefit is... , The value represents the time saved by the system after a successful swap. and The execution time of a previously shorter task has increased, but not more than the execution time of a previously longer task. In this case, the system's benefit is... , The value represents the extra time the system spends after a successful exchange, but overall the objective function is still optimized to generate the exchange tender. Step 3: Collect valid exchange tenders.

[0019] Beneficial technical effects of the present invention: This application aims to minimize the time required for a multi-UAV system to complete multiple tasks. It designs a multi-UAV multi-task allocation method based on an improved contract net algorithm, proposes an auction mode for task sets considering overall efficiency indicators, and introduces auction strategies that incorporate type constraints and set bidding thresholds. This effectively reduces the communication volume of UAV task allocation. Furthermore, it proposes a bidding method that adds an exchange strategy to the buying and selling strategy, further improving the overall optimization objective. The improved contract net algorithm proposed in this invention can obtain a better task allocation scheme faster than traditional contract net algorithms. Attached Figure Description

[0020] Figure 1 A flowchart of a traditional contract network algorithm provided for embodiments of this application; Figure 2 A schematic diagram of the exchange negotiation provided for an embodiment of this application; Figure 3 An initial allocation result diagram provided for an embodiment of this application; Figure 4 A task allocation result diagram of a traditional contract network provided for an embodiment of this application; Figure 5 A diagram showing the task allocation results of the improved contract network provided in this application embodiment. Detailed Implementation

[0021] Please see Figures 1-5 This invention provides a method for multi-UAV multi-task allocation based on an improved contract net algorithm, which can be used to make decisions on multi-UAV multi-task allocation for complex combat missions oriented towards penetration and strike under adversarial conditions, and realize the optimized allocation of multi-UAV multi-task execution.

[0022] This invention provides a multi-UAV multi-task allocation method based on an improved contract network algorithm, which reduces negotiation communication and improves task allocation efficiency. The invention is summarized as follows: 1) Construct a task set bidding strategy to promote global optimization of task allocation. This invention employs a "task set bidding" approach to address the shortcomings of "single task bidding." In this bidding model, the bidder submits bids for all tasks within its task set, and each bidder can submit bids for all tasks within that set. In each round of bidding, the bidder can successfully sell at most one task from its task set, completing the auction for a single task while considering overall interests.

[0023] 2) Increase bidding constraints to reduce the number of bidders and decrease communication volume. Given the continued growth in the number of drones used in the future, this invention introduces two strategies—"type constraints" and "setting a bidding threshold"—to address the issue of high communication volume. Drones are selected for bidding based on the type of task. This means that drones of appropriate types are chosen to participate in the bidding process. Task allocation based on type constraints can significantly reduce the number of drones participating in the bidding, thereby reducing the communication volume during the bidding process. Furthermore, this invention sets a bidding threshold for the bidder, using the bidder's benefit from performing the bidding task as the bidding threshold. When the benefit of all bidders for the bidding task is lower than the bidder's own benefit from performing the bidding task, the bidder will not respond to the bidders.

[0024] 3) Improve negotiation strategies and enhance the quality of distribution plans. This invention adds an "exchange negotiation" strategy to the traditional "buyer-seller negotiation" strategy. "Buyer-seller negotiation" occurs between the tenderer and the bidder, requiring the bidder to have surplus combat capabilities to execute new missions that might lead to a successful bid. Other UAVs in the contract network model without surplus combat capabilities are disqualified from bidding, thus losing the possibility of improving the mission execution efficiency of these UAVs. This invention designs an "exchange negotiation" strategy. "Exchange" occurs between any two UAVs of the same type in the system (denoted as UAV1 and UAV2), meaning UAV1 and UAV2 swap missions. "Exchange" does not increase the number of additional UAV combat resources, but it may improve the overall mission allocation efficiency of the system after the exchange.

[0025] A diagram of the exchange negotiation is attached. Figure 2 As shown in the figure. The original UAV1 mission set is... and The original UAV2 mission set was and Both sides' tasks are concentrated and After the exchange, the UAV1 mission set is and The UAV2 mission set is and After the exchange, UAV1 and UAV2 did not consume any additional combat resources, but their mission execution efficiency was improved.

[0026] In this embodiment of the application, a method for multi-UAV multi-task allocation based on an improved contract network algorithm is provided, including the following steps: Step 1: Design evaluation metrics for task allocation This invention uses minimizing task execution time as the objective function for task allocation. When estimating tasks, it calculates the shortest time for a UAV to execute all tasks. When UAVs are required to bid on tasks, it plans the shortest path for each UAV to execute its task set, and simultaneously provides a task time cost table for each UAV to complete its task set. Each task in the task set has a corresponding time cost, and the values ​​in the time cost table represent the time cost that UAV needs to incur to execute the corresponding task.

[0027] The specific description is as follows: Drone Relative to its task sequence Tasks in The time cost is

[0028] In the formula, Indicates drone Relative to the task The time cost, express Execute task sequence The time required Indicates from the task sequence Lieutenant General The new sequence after removal, express The time required to execute a new task sequence.

[0029] Example: Let the task set of UAV1 be... The shortest time task sequence for UAV1 is calculated as follows: ,calculate The time cost compared to the UAV1 drone is... By removing the original task sequence and calculating the time difference between the original and new task sequences executed by UAV1, we can obtain... The value of .

[0030] Step 2: Construct an exchange negotiation strategy based on a buy-sell negotiation strategy This invention adds an exchange negotiation strategy to the traditional contract network algorithm's buyer-seller negotiation strategy. Depending on the negotiation strategy, individuals in the contract network model can propose two types of bids: buy-sell contracts and exchange contracts.

[0031] The bidding document is defined as a seven-tuple. .

[0032] in, Indicates the bidding for drones, This indicates the number of remaining combat resources for the bidding drone. Indicates that a drone is being tendered. Indicates the task of bidding. This indicates the system's profit after the transaction is completed. This indicates the time required for the bid-tendered drone to complete its new mission sequence after the sale and purchase are completed. This indicates the time required for the tendered drone to complete its new mission sequence after the sale and purchase are completed.

[0033] The exchange tender document is defined as a seven-tuple. .

[0034] in, This indicates the first drone to be exchanged. This indicates the first drone used for the exchange mission. This indicates the second drone to be exchanged. This indicates that the second drone is used for the exchange mission. This indicates the system's benefit after the exchange is completed. This indicates the time required for the first drone to complete its new mission sequence after the swap. This indicates the time required for the second drone to complete its new mission sequence after the exchange.

[0035] Step 3: Construct a multi-drone, multi-task allocation bidding strategy The bidding strategy is the core of the Contract Network algorithm, which consists of three steps: bidding, tendering, and awarding. This invention improves upon the three steps of bidding, tendering, and awarding in the traditional Contract Network algorithm, thereby increasing resource utilization in multi-UAV systems and obtaining optimal task allocation results.

[0036] 1) Bidding: With the objective function of minimizing the maximum task execution time of the unmanned aerial vehicle (UAV) system, in each round, the bidder selects the UAV with the longest task execution time among the current multi-UAV systems. The bidder adopts an auction strategy of "task group bidding," issuing all tasks in its task set.

[0037] 2) Bidding: Design a new contract format for "bid exchange". This includes generating buy and sell bids and generating exchange bids. Generating buy and sell bids involves the following three steps: Step 1: In addition to the tenderer, the system... After receiving the bidding task information, drones other than those listed above determine whether they meet the type constraints based on the task type, and then check whether their remaining combat resources meet the constraint on the number of tasks a single drone can perform. Drones that meet both constraints are considered bidders in this round of bidding. .

[0038] Step 2: Set the tenderer The values ​​in the task time cost table are the thresholds for generating bids and tenders for each bidding task. , for Relative to the tasks in its task set The time value of the task. Each bidder submits a bid for the tendered task, which means "adding" the tendered task to their own task set and calculating the time required to complete the "new task sequence". Calculate the time required for the drone to complete its "new mission sequence" after it is "removed" from the bidding task. Finally, the calculations were performed to determine the relative performance of the "newly added" mission compared to the bidding drone. Task time value .if This indicates that the system can complete the bidding task at a relatively low cost. At this point, the system's revenue is , The value represents the time saved by the system after a bidder wins the bid, which can then generate a valid bid.

[0039] if but In this case, although the system failed to complete the task at a lower cost, it still optimized the objective function and generated a valid bid. The system's profit in this situation is... , The value represents the magnitude of the objective function optimization value after the bidder wins the bid.

[0040] Step 3: Collect all valid bids and send them to the bidding drone. .

[0041] The process of generating exchange tender documents involves the following three steps: Step 1: The "swap" strategy will not affect the number of tasks a single drone can perform. Therefore, all drones in the system that meet the type constraints can swap tasks with each other.

[0042] Step 2: Identify the two drones to be swapped in this mission. and Then, in sequence Tasks in the task set The tasks in the task set are "swapped," and calculations are performed after each "swap." and Time to complete each "new task sequence" and Let the longer task execution time before the exchange be... The shorter task execution time is The execution time after the swap is and ,if and Both parties can perform the same number of tasks at a lower cost, and can generate an exchange tender.

[0043] The system's profit at this time is , The value represents the time saved by the system after a successful swap. If and The execution time of a previously shorter task has increased, but not more than the execution time of a previously longer task. In this case, the system's benefit is... , The value represents the extra time the system spends after a successful exchange, but overall the objective function is still optimized, and an exchange tender can be generated.

[0044] Step 3: Collect valid exchange tenders.

[0045] 3) Winning the bid.

[0046] For valid bids generated during the bidding phase, the bidding drone will select the bid that maximizes the system benefit (i.e., optimizes the objective function) from all valid bids as the winning bid, and the corresponding bidder will become the successful bidder. Other bidders will be rejected. For valid exchange bids generated during the bidding phase, the bid that maximizes the system benefit (i.e., optimizes the objective function) will be selected as the winning bid, and the corresponding two drones will then exchange tasks.

[0047] In other embodiments of this application: it is assumed that there are 20 drones on the battlefield and 30 missions. All drones fly at a speed of 50 m / s, and the information of each drone and mission is shown in the table below.

[0048] Table 1. Unmanned Aerial Vehicle Information Table

[0049] Table 2 Task Information Table

[0050] The initial allocation follows the proximity principle, directly assigning tasks to the nearest UAV that meets the task payload requirements. The initial allocation results are shown in the attached figure. Figure 3 As shown in the attached figure. Based on the initial task allocation results, the traditional contract net algorithm and the improved contract net algorithm are used to optimize the initial task allocation results, respectively. Figure 4 and attached Figure 5 As shown.

[0051] The comparison between the traditional contract net algorithm and the improved contract net algorithm is shown in Table 3 below.

[0052] Table 3 Comparison between Traditional Contract Network and Improved Contract Network

[0053] The optimized mission sequences and execution times for each UAV are shown in Tables 4 and 5, respectively.

[0054] Table 4. Drone mission sequence and execution schedule for traditional contract networks

[0055] Table 5. Improved Contract Network: Unmanned Aerial Vehicle (UAV) Mission Sequence and Execution Schedule

[0056] As shown in the table and figures above, the improved contract net algorithm can effectively improve the objective function. Under the conditions of satisfying the constraints of the number of tasks, type constraints, and cooperation constraints of a single UAV, it can obtain a better task allocation scheme than the traditional contract net algorithm. After optimization, the maximum task execution time of the multi-UAV system using the traditional contract net algorithm is 371.5s, which is the time required for UAV ID 2 to execute its task sequence. After optimization, the maximum task execution time of the multi-UAV system using the improved contract net algorithm is 349.4s, which is the time required for UAV ID 12 to execute its task sequence. This shows that the improved contract net algorithm has a stronger optimization capability than the traditional contract net algorithm when optimizing the objective function.

Claims

1. A method for multi-UAV multi-task allocation based on an improved contract network algorithm, characterized in that, The method includes: Step 1: Design evaluation metrics for task allocation; Step 2: Construct an exchange negotiation strategy based on a buy-sell negotiation strategy; Step 3: Based on the evaluation indicators and the exchange negotiation strategy, construct a multi-UAV multi-task allocation bidding strategy.

2. The method according to claim 1, characterized in that, Step 1 includes: Minimizing task execution time is used as the objective function for task allocation. When estimating tasks, the shortest time for a UAV to execute all tasks is calculated. When a UAV needs to bid for a task, the shortest path for each UAV to execute its task set is planned. At the same time, a task time cost table for each UAV to complete its own task set is provided.

3. The method according to claim 2, characterized in that, Each task in the task set has a corresponding time cost, and the value in the task time cost table represents the time cost that the UAV needs to pay to perform the corresponding task.

4. The method according to claim 3, characterized in that, The objective function for task allocation is to minimize task execution time. When estimating tasks, the shortest time for a UAV to execute all tasks is calculated. When UAVs need to bid on tasks, the shortest path for each UAV to execute its task set is planned. A task time cost table for each UAV to complete its own task set is also provided, including: drones Relative to its task sequence Tasks in The time cost is: In the formula, Indicates drone Relative to the task The time cost, express Execute task sequence The time required Indicates from the task sequence Lieutenant General The new sequence after removal, express The time required to execute a new task sequence.

5. The method according to claim 1, characterized in that, Step 2 includes: An exchange negotiation strategy was added to the buy-sell negotiation strategy of the Contract Network algorithm; Depending on the negotiation strategy, individuals in the contract network model submit two types of tenders: sales contracts and exchange contracts.

6. The method according to claim 5, characterized in that, The bidding document is defined as a seven-tuple. ; in, Indicates the bidding for drones, This indicates the number of remaining combat resources for the bidding drone. Indicates that a drone is being tendered. Indicates the task of bidding. This indicates the system's profit after the transaction is completed. This indicates the time required for the bid-tendered drone to complete its new mission sequence after the sale and purchase are completed. This indicates the time required for the tendered drone to complete its new mission sequence after the sale and purchase are completed.

7. The method according to claim 5, characterized in that, The exchange tender document is defined as a seven-tuple. ; in, This indicates the first drone to be exchanged. This indicates the first drone used for the exchange mission. This indicates the second drone to be exchanged. This indicates that the second drone is used for the exchange mission. This indicates the system's benefit after the exchange is completed. This indicates the time required for the first drone to complete its new mission sequence after the swap. This indicates the time required for the second drone to complete its new mission sequence after the exchange.

8. The method according to claim 5, characterized in that, Step 3 includes: 1) Bidding: With the objective function of minimizing the maximum task execution time of the UAV system, in each round, the bidder selects the UAV with the longest task execution time in the current multi-UAV system; the bidder adopts the auction strategy of "task group bidding" and publishes all tasks in its task set; 2) Bidding: Design a new contract format for "exchange of bids", including the generation of buy and sell bids and the generation of exchange bids; 3) Winning the bid: For valid bids generated during the bidding stage, the bidding drone will select the bid that maximizes the system benefit, i.e. optimizes the objective function the most, from all valid bids as the winning bid. The corresponding bidder becomes the winner, and the other bidders are considered unsuccessful bidders. For valid exchange bids generated during the bidding stage, the bid that maximizes the system benefit, i.e. optimizes the objective function the most, is selected as the winning bid, and the two corresponding drones will exchange tasks.

9. The method according to claim 8, characterized in that, The generation of the bidding document includes: Step 1: In addition to the tenderer, the system... After receiving the bidding task information, drones other than those listed above determine whether they meet the type constraints based on the task type, and then check whether their remaining combat resources meet the constraint on the number of tasks a single drone can perform. Drones that meet both constraints are considered bidders in this round of bidding. ; Step 2: Set the tenderer The values ​​in the task time cost table are the thresholds for generating bids and tenders for each bidding task. , for Relative to the tasks in its task set The time value of the task; each bidder submits a bid for the tendered task, which means "adding" the tendered task to their own task set and calculating the time required to complete the "new task sequence". Calculate the time required for the drone to complete its "new mission sequence" after it is "removed" from the bidding task. Finally, the calculations were performed to determine the relative performance of the "newly added" mission compared to the bidding drone. Task time value ;if This indicates that the system can complete the bidding task at a relatively low cost. At this point, the system's revenue is , The value represents the time saved by the system after a successful bidder generates a valid bid; if... but In this case, although the system failed to complete the task at a lower cost, it still optimized the objective function and generated a valid bid. The system's profit in this situation is... , The value represents the magnitude of the objective function optimization value after the bidder wins the bid; Step 3: Collect all valid bids and send them to the bidding drone. .

10. The method according to claim 8, characterized in that, The generation of the exchange tender document includes: Step 1: The "swap" strategy will not affect the number of tasks a single drone can perform. Therefore, all drones in the system that meet the type constraints will swap tasks with each other. Step 2: Identify the two drones to be swapped in this mission. and Then, in sequence Tasks in the task set The tasks in the task set are "swapped," and calculations are performed after each "swap." and Time to complete each "new task sequence" and Let the longer task execution time before the exchange be... The shorter task execution time is The execution time after the swap is and ,if and Both parties in the exchange can perform the same number of tasks and generate exchange tenders at a lower cost; in this case, the system's benefit is... , The value represents the time saved by the system after a successful swap. and The execution time of a previously shorter task has increased, but not more than the execution time of a previously longer task. In this case, the system's benefit is... , The value represents the extra time the system spends after a successful exchange, but overall the objective function is still optimized to generate the exchange tender. Step 3: Collect valid exchange tenders.

Images