Unmanned aerial vehicle cluster cooperation optimization method, device and equipment and readable storage medium

Abstract

The unmanned aerial vehicle cluster cooperation optimization method, device, equipment and readable storage medium take total social welfare maximization as the core target, dynamically adjust the alliance structure through the iteration optimization of alliance merger and split, and finally form a stable alliance with stable structure and balanced cost and benefit; relying on the task bidding submitted by each alliance, combining with the preset task constraint to complete the bidding alliance screening, ensuring that the alliance has task execution feasibility, while using the critical payment strategy to calculate the reward, balancing the interests of the task requester and the alliance, and protecting the individual rationality and the real price of the alliance; the bidding alliance executes the task according to the cooperation mechanism, collects and transmits the data, and the reward is distributed within the alliance according to the preset rule, forming a cooperation closed loop of "alliance optimization-task execution-reward incentive". The framework breaks through the inherent limitations of single unmanned aerial vehicle sensing range, energy reserve and equipment configuration, and effectively adapts to large-scale and differentiated sensing task demand.

Description

Technical Field

[0001] This application relates to the field of intelligent perception in unmanned aerial vehicle (UAV) swarms, specifically to a method, apparatus, device, and readable storage medium for optimizing UAV swarm collaboration. Background Technology

[0002] Currently, digital transformation in fields such as smart cities, emergency rescue, and environmental monitoring continues to deepen, and sensing tasks are rapidly developing towards large-scale deployment, high dynamic response, and high-precision data acquisition. For example, urban traffic situational awareness requires real-time data coverage of the entire road network, emergency rescue scenarios require rapid and comprehensive information collection of disaster areas, and environmental monitoring tasks need to meet the requirements for long-term, multi-dimensional, and accurate data acquisition. These application scenarios place stringent and diverse demands on the spatiotemporal coverage, data acquisition accuracy, and dynamic response speed of sensing tasks. Large-scale and differentiated sensing scenarios have become the core application direction of the industry. With its advantages of mobility, flexibility, and efficient deployment, drones have been widely used in various sensing tasks, becoming a key carrier to overcome the limitations of traditional sensing technologies.

[0003] In related technologies, drone-related solutions mostly focus on a simple matching mode of single drone and single task, adapting the performance parameters of a single drone to the task requirements, without involving the systematic design of multi-drone collaboration. Summary of the Invention

[0004] This application provides a method, apparatus, device, and readable storage medium for optimizing drone swarm collaboration, which can solve the technical problems in related technologies where single drone solutions are difficult to adapt to the needs of large-scale and differentiated sensing tasks, and lack a systematic design for multi-drone collaborative cooperation.

[0005] In a first aspect, embodiments of this application provide a method for optimizing drone swarm cooperation, the method comprising: Collect key parameters of the sensing mission and all UAV status information; Based on the collected key parameters of the sensing mission and all UAV status information, and with the goal of maximizing total social welfare, a stable alliance is formed through iterative optimization of alliance merging and splitting. Based on the task bidding of each alliance, the winning alliance is selected in combination with the task constraints, and the remuneration is calculated simultaneously using the critical payment strategy. The winning alliance collaborates to execute tasks and transmit collected data, with compensation distributed within the alliance according to rules, forming a closed-loop collaboration.

[0006] In conjunction with the first aspect, in one implementation, the collection of key parameters of the sensing task and all UAV status information includes: The sensing platform receives sensing tasks submitted by task requesters and collects information on the location, required sensor type, sensing quality requirements, duration, and lifespan of each task. Collect the location, remaining energy, sensor configuration, and flight speed of all drones.

[0007] In conjunction with the first aspect, in one implementation, the formation of a stable alliance based on collected key parameters of the sensing task and all UAV status information, with the goal of maximizing total social welfare, through iterative optimization of alliance merging and splitting, includes: The social welfare increment of each alliance performing corresponding tasks under the current division is calculated and summed to obtain the total social welfare of the current division; among which, the social welfare increment is comprehensively evaluated in combination with the perceived quality of alliances performing tasks, the rewards for alliances performing tasks, and the penalty factors of alliance size; For each alliance in the current division, try to merge with other alliances in turn to form a new division. Calculate the total social welfare of each alliance in the new division for performing the corresponding tasks. If the new total social welfare is higher than the current total social welfare, update the division and the total social welfare value. Repeat the operation until the total social welfare cannot be improved by merging. For each merged alliance, initialize the candidate split set and construct a new partition. Calculate the total social welfare of each alliance performing the corresponding task under the new partition. If the new total social welfare is higher than the current total social welfare, update the partition and the total social welfare value. Repeat the operation until the total social welfare can no longer be improved through splitting. Output a stable alliance that optimizes overall social welfare.

[0008] In conjunction with the first aspect, in one implementation, the step of selecting winning alliances based on task bidding from each alliance, combining task constraints and social welfare increments, and simultaneously calculating remuneration includes: Each alliance formulates and submits bidding information for each sensing task based on its actual cost of performing the task, and the bidding information is consistent with the actual cost. For each task, verify whether the participating bidding consortium meets the task constraints; If multiple alliances compete for the same task, the alliance whose bid meets the requirements and has the best social welfare increment will be selected as the winning bidder. If there is no competition, the task will be directly assigned to the suitable alliance. A critical payment strategy is used to calculate the remuneration of the winning consortium. After removing the winning consortium, the consortium formation and task allocation process is rerun to obtain alternative winning consortia. Based on the perceived quality, scale and bidding attributes of both consortia, a virtual bid value is calculated and the virtual bid value is determined as the remuneration amount.

[0009] In conjunction with the first aspect, in one implementation, the task constraints include that the remaining energy of each UAV in the alliance meets the total energy consumption of the task, the total time to complete the task does not exceed the task survival time, the perception quality of the task execution meets the task requirements, and at least one UAV in the alliance is equipped with the type of sensor required for the task.

[0010] In conjunction with the first aspect, in one implementation, the total energy consumption of the mission includes the flight energy consumption, hovering energy consumption, sensing energy consumption, communication energy consumption, and alliance collaboration energy consumption of the UAV. The total mission time includes flight time, sensing duration, and data transmission time.

[0011] In conjunction with the first aspect, in one implementation, the winning bid alliance collaboratively executes tasks and transmits collected data, with rewards distributed within the alliance according to rules, forming a collaborative closed loop, including: The winning consortium plans a coordinated flight path based on the mission location information to avoid flight conflicts and fly to the designated mission area; According to the required sensing range and accuracy, we will carry out data collection operations in a coordinated manner, verify the data quality in real time during the collection process, and re-collect any data that does not meet the requirements. The collected sensor data is transmitted to the sensing platform via encrypted transmission. The platform then processes and filters the data before submitting it to the task requester. The sensing platform deducts the relevant costs of the alliance's mission execution and formulates fair distribution rules based on the sensing contribution, energy consumption, and collaborative participation of each drone in the alliance. The rewards are distributed to each participating drone according to the distribution rules, which incentivizes them to continue participating in the collaboration and forms a closed loop.

[0012] Secondly, embodiments of this application provide a drone swarm cooperation optimization device, the drone swarm cooperation optimization device comprising: The information acquisition module is used to collect key parameters of the sensing mission and all UAV status information; The alliance formation optimization module is used to form stable alliances through iterative optimization of alliance merging and splitting, with the goal of maximizing total social welfare. The task matching and reward calculation module is used to select winning alliances based on task bidding from various alliances, combined with task constraints, and simultaneously calculate rewards using a critical payment strategy. The task execution and reward distribution module is used by the winning alliance to collaboratively execute tasks and transmit collected data. Rewards are distributed within the alliance according to rules, forming a collaborative closed loop.

[0013] Thirdly, embodiments of this application provide a drone swarm collaboration optimization device, which includes a processor, a memory, and a drone swarm collaboration optimization program stored in the memory and executable by the processor. When the drone swarm collaboration optimization program is executed by the processor, it implements the steps of the drone swarm collaboration optimization method as described in some of the above embodiments.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium storing a drone swarm cooperation optimization program, wherein when the drone swarm cooperation optimization program is executed by a processor, it implements the steps of the drone swarm cooperation optimization method as described in some of the above embodiments.

[0015] The beneficial effects of the technical solutions provided in this application include: With the maximization of total social welfare as its core objective, the framework dynamically adjusts the alliance structure through iterative optimization of alliance mergers and splits, ultimately forming a stable alliance with a balanced cost-benefit structure. Based on task bids submitted by each alliance and pre-defined task constraints, the winning alliance is selected to ensure the feasibility of task execution. A critical payment strategy is used to calculate compensation, balancing the interests of task requesters and alliances, and ensuring the rationality of individual alliances and the authenticity of bids. The winning alliance executes tasks, collects and transmits data according to a collaborative mechanism, and compensation is distributed within the alliance according to pre-defined rules, forming a collaborative closed loop of "alliance optimization - task execution - compensation incentives." This framework overcomes the inherent limitations of single UAV sensing range, energy reserves, and equipment configuration, effectively adapting to the needs of large-scale, differentiated sensing tasks. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the drone swarm collaboration optimization method in the embodiments of this application; Figure 2 This is a schematic diagram illustrating how the total social welfare changes with the maximum number of sensors per drone under different task types in the embodiments of this application. Figure 3 This is a schematic diagram illustrating the variation of total social welfare with average task quality requirements under different TTL conditions in the embodiments of this application; Figure 4 This is a schematic diagram showing how total social welfare changes with sensor quality weighting parameters under different payment adjustment coefficients in the embodiments of this application. Figure 5 This is a schematic diagram showing how the average sensing quality of a task changes with the number of drones under different task numbers in this application embodiment; Figure 6This is a schematic diagram showing the relationship between the average sensing quality of the task and the number of sensing tasks and the number of drones under different schemes in the embodiments of this application. Figure 7 This diagram illustrates the relationship between social welfare and the number of sensing tasks and drones under different schemes in the embodiments of this application. Figure 8 This is a schematic diagram of the hardware structure of the drone swarm collaboration optimization device involved in the embodiments of this application. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present application.

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0019] In a first aspect, embodiments of this application provide a method for optimizing drone swarm collaboration.

[0020] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the UAV swarm collaboration optimization method of this application. Figure 1 As shown, the optimization methods for drone swarm collaboration include: S100: Collects key parameters of the sensing mission and all UAV status information; S200: Based on the collected key parameters of the sensing mission and all UAV status information, with the goal of maximizing total social welfare, a stable alliance is formed through iterative optimization of alliance merging and splitting. S300: Based on the task bidding of each alliance, the winning alliance is selected by combining task constraints, and the critical payment strategy is used to calculate the remuneration simultaneously. S400: The winning alliance collaborates to execute tasks and transmit collected data, with compensation distributed within the alliance according to rules, forming a collaborative closed loop.

[0021] In this embodiment, key parameters of the sensing task and all UAV status information are first systematically collected to provide comprehensive data support for subsequent collaborative decision-making. Based on the aforementioned collected information, with the maximization of total social welfare as the core objective, the alliance structure is dynamically adjusted through iterative optimization of alliance merging and splitting. This integrates the collaborative capabilities of heterogeneous UAVs and avoids efficiency losses caused by imbalances in alliance size, ultimately forming a stable alliance with a stable structure and balanced cost and benefit. Based on the task bids submitted by each alliance and combined with preset task constraints, the winning alliance is selected to ensure that the alliance has the feasibility of task execution. At the same time, a critical payment strategy is used to calculate remuneration, balancing the interests of task requesters and alliances, and ensuring the rationality of individual alliances and the authenticity of bids. The winning alliance executes the task, collects and transmits data according to the collaborative mechanism, and the remuneration is distributed within the alliance according to preset rules, forming a collaborative closed loop of "alliance optimization - task execution - remuneration incentive". This framework breaks through the inherent limitations of a single UAV's sensing range, energy reserves, and equipment configuration, effectively adapting to the needs of large-scale and differentiated sensing tasks. It ensures optimal overall social welfare through alliance iterative optimization, improves execution reliability through task constraint screening, achieves interest balance through critical payment strategies, and ensures data acquisition quality and transmission security through a closed-loop collaboration mechanism. At the same time, it incentivizes UAVs to continuously participate in collaboration through fair reward distribution, significantly improving the efficiency of sensing task execution and the stability of system collaboration.

[0022] Furthermore, in one embodiment, S100 includes the following steps: S101: The sensing platform receives sensing tasks submitted by the task requester and collects the location, required sensor type, sensing quality requirements, duration and lifespan of each task. S102: Collects the location, remaining energy, sensor configuration, and flight speed of all drones.

[0023] In this embodiment, the sensing platform acts as the collaboration hub, receiving sensing tasks submitted by task requesters and collecting core requirement parameters for each task, such as location, required sensor type, sensing quality requirements, duration, and survival time. Simultaneously, it collects key status information from all participating drones, including location, remaining energy, sensor configuration, and flight speed. By comprehensively and accurately collecting core information related to task requirements and drone capabilities, it provides standardized and highly reliable data support for the entire collaboration process, including alliance formation optimization, task matching and selection, constraint verification, and reward calculation. This ensures that subsequent decisions are made accurately based on the actual situation of the task and the drones, avoiding collaboration imbalances caused by missing or biased information. It effectively guarantees the scientific nature and execution effectiveness of the entire drone swarm collaboration optimization process, laying a solid data foundation for overcoming the limitations of single drone adaptation and meeting the needs of large-scale, differentiated sensing tasks.

[0024] Furthermore, in one embodiment, step S200 includes the following steps: S201: Calculate the social welfare increment of each alliance performing the corresponding task under the current division, and sum them up to obtain the total social welfare of the current division; among which, the social welfare increment is comprehensively evaluated in combination with the perceived quality of the alliance's task performance, the reward for the alliance's task performance, and the alliance size penalty factor; S202: For each alliance in the current division, try to merge with other alliances in turn to form a new division. Calculate the total social welfare of each alliance performing the corresponding tasks under the new division. If the new total social welfare is higher than the current total social welfare, update the division and the total social welfare value. Repeat the operation until the total social welfare cannot be improved by merging. S203: For each merged alliance, initialize the candidate split set and construct a new partition. Calculate the total social welfare of each alliance performing the corresponding task under the new partition after the split. If the new total social welfare is higher than the current total social welfare, update the partition and the total social welfare value. Repeat the operation until the total social welfare cannot be improved by splitting. S204: Output the stable alliance with the optimal total social welfare.

[0025] In this embodiment, the social welfare increment of each alliance performing corresponding tasks under the current alliance division is calculated, and the social welfare increments of all alliances are summed to obtain the total social welfare of the current division. The social welfare increment is comprehensively evaluated by integrating the perceived quality of alliance task performance, the reward for alliance task performance, and the alliance size penalty factor. For each alliance in the current division, it is sequentially attempted to merge with other alliances to construct a new alliance division. The total social welfare of each alliance performing corresponding tasks under the new division is calculated. If the new total social welfare is higher than the current total social welfare, the alliance division and the total social welfare value are updated. The above merging and calculation operations are repeated until the total social welfare cannot be improved through alliance merging. For each merged alliance, a candidate split set is initialized and a new alliance is constructed accordingly. The process involves dividing the alliances and calculating the total social welfare of each alliance in the newly divided alliances for performing corresponding tasks. If the new total social welfare is higher than the current total social welfare, the alliance division and total social welfare value are updated. This process of dividing and calculating alliances is repeated until the total social welfare can no longer be improved through alliance division. Finally, the stable alliance with the optimal total social welfare is output. Through the progressive process of "incremental calculation - merging optimization - splitting optimization - optimal output", the collaborative capabilities of heterogeneous drones are integrated, while avoiding the problems of soaring coordination costs due to excessively large alliance size or insufficient scale to meet task requirements. This ensures that the formed alliance structure is stable and has the optimal total social welfare, providing a reliable alliance foundation for efficient matching and execution of subsequent tasks, and significantly improving the adaptability of drone swarms to large-scale and differentiated sensing tasks.

[0026] Furthermore, in one embodiment, step S300 includes the following steps: S301: Each alliance shall formulate and submit bidding information for each sensing task based on its actual cost of performing the task, and the bidding information shall be consistent with the actual cost. S302: For each task, verify whether the participating bidding alliance meets the task constraints; S303: If multiple alliances compete for the same task, select the alliance whose bid meets the requirements and has the best social welfare increment as the winning bidder. If there is no competition, the task will be directly assigned to the suitable alliance. S304: Use a critical payment strategy to calculate the remuneration of the winning consortium. After removing the winning consortium, rerun the consortium formation and task allocation process to obtain alternative winning consortia. Calculate the virtual bid value based on the perceived quality, scale, and bidding attributes of both consortia, and determine the virtual bid value as the remuneration amount.

[0027] In this embodiment, each alliance formulates and submits bidding information consistent with the actual cost of its own task execution for each sensing task. For each sensing task, it verifies whether the participating alliances meet the preset task constraints. If multiple alliances compete for the same sensing task, the alliance with the best social welfare increment that meets the requirements is selected as the winning bidder. If there is no competition, the sensing task is directly assigned to the suitable alliance. The remuneration of the winning alliance is calculated using a critical payment strategy. The winning alliance is removed first, and then the alliance formation and task allocation process is rerun to obtain the candidate winning alliances. Then, a virtual bid value is calculated based on the perception quality, alliance size, and bidding attributes of the winning alliance and the candidate winning alliances. This virtual bid value is determined as the final remuneration amount of the winning alliance. This process ensures the authenticity of bids and cost transparency through the submission of genuine bidding information. Combined with task constraint verification, it eliminates unsuitable alliances from the source, ensuring the feasibility of task execution. Through differentiated competitive screening and direct allocation without competition, it achieves efficient and accurate matching between alliances and sensing tasks. Relying on the critical payment strategy and comparing and calculating remuneration with alternative alliances, it not only satisfies the rational demands of individual alliance members but also balances the interests of task requesters and alliances, effectively incentivizing alliances to continuously participate in collaboration and significantly improving the fairness, economy, and execution efficiency of task allocation.

[0028] Furthermore, in one embodiment, in S300, the task constraints include the remaining energy of each UAV in the alliance meeting the total energy consumption of the task, the total time to complete the task not exceeding the task survival time, the perception quality of the task execution meeting the task requirements, and at least one UAV in the alliance being equipped with the sensor type required for the task.

[0029] In this embodiment, the task constraints explicitly require the alliance to simultaneously meet four types of adaptation requirements: First, the remaining energy of each drone in the alliance must be sufficient to support the total energy consumption required to execute the task, ensuring no interruption due to energy depletion during task execution. Second, the total time to complete the task must not exceed the task's preset lifespan, adapting to the task's dynamic response timeliness requirements. Third, the perception quality of the alliance's task execution must meet the task's clearly defined quality requirements, ensuring the effectiveness and accuracy of data collection. Fourth, at least one drone in the alliance must be equipped with the sensor type required for the task, ensuring it possesses the core capabilities to perform the sensing task. Through the collaborative verification of these multi-dimensional constraints, alliances meeting the task execution conditions are selected from four core dimensions: resource guarantee, timeliness adaptation, quality compliance, and capability matching. This avoids collaboration failures caused by unsuitable alliances participating in the bidding process, ensuring that the winning alliance can stably and reliably complete the sensing task, providing crucial guarantees for the accuracy of task allocation and the efficiency of the collaboration process.

[0030] Furthermore, in one embodiment, the total energy consumption of the mission includes the flight energy consumption, hovering energy consumption, sensing energy consumption, communication energy consumption, and alliance collaboration energy consumption of the UAV; the total mission time includes flight time, sensing duration, and data transmission time.

[0031] In this embodiment, the total task energy consumption explicitly covers the core energy consumption types throughout the entire process of UAVs performing sensing tasks. This includes flight energy consumption to reach the task area, hovering energy consumption while operating within the task area, sensing energy consumption for data acquisition, communication energy consumption for transmitting acquired data, and alliance collaboration energy consumption during multi-UAV collaborative operations. This achieves comprehensive coverage of task execution energy consumption. The total task time includes the flight time of the UAV from its current location to the task area, the sensing duration for completing data acquisition as required by the task, and the data transmission time for uploading the acquired data to the sensing platform, fully encompassing the key time stages of task execution. By accurately decomposing and clearly defining the total task energy consumption and total time, precise data is provided for cost accounting, bidding price formulation, and task constraint verification for alliance task execution. This avoids problems such as alliance profit imbalance and task overtime caused by incomplete energy consumption and time statistics, ensuring the scientific nature of task allocation and reward calculation, and further improving the reliability and economy of UAV swarm collaborative sensing task execution.

[0032] Furthermore, in one embodiment, step S400 includes the following steps: S401: The consortium plans a collaborative flight path based on the mission location information to avoid flight conflicts and fly to the designated mission area; S402: According to the required sensing range and accuracy of the task, coordinate to carry out data acquisition operations, verify data quality in real time during the acquisition process, and re-acquire data that does not meet the requirements; S403: The collected sensor data is transmitted to the sensing platform via encrypted transmission. The platform processes and filters the data before submitting it to the task requester. S404: The sensing platform deducts the relevant costs of mission execution by the alliance and formulates fair distribution rules based on the sensing contribution, energy consumption and collaborative participation of each drone in the alliance. S405: Distribute rewards to each participating drone according to the distribution rules to incentivize their continued participation in collaboration and form a closed loop.

[0033] In this embodiment, the winning consortium plans a collaborative flight path based on the previously acquired mission location information to avoid flight conflicts, and then flies to the designated mission area. According to the mission's preset sensing range and accuracy requirements, they collaboratively conduct data collection operations, verifying data quality in real time during the collection process and re-collecting any data that does not meet the standards. The collected sensor data is uploaded to the sensing platform via encrypted transmission. The sensing platform then organizes and filters the received data before submitting it to the mission requester. The sensing platform deducts the relevant costs incurred by the consortium in executing the mission and formulates a fair and reasonable reward distribution rule based on the sensing contribution, energy consumption, and collaborative participation of each drone within the consortium. The calculated reward is distributed to each participating drone according to this rule, effectively incentivizing drones to continue participating in subsequent collaborations, forming a complete collaborative closed loop, ensuring the continuity and efficiency of the sensing mission execution, while also ensuring that data collection quality meets standards, transmission is secure and reliable, and reward distribution is fair, further consolidating the stability and sustainability of the consortium's collaboration.

[0034] In summary, the technical solution for optimizing drone swarm collaboration provided in the embodiments of this application is described in its entirety as follows: I. Core Definitions and Problem Objectives This technical solution addresses the core technical challenges of single drones being limited by their sensing range, energy reserves, and equipment configuration, making them unsuitable for large-scale, differentiated sensing tasks. It also addresses the lack of systematic design for multi-drone collaborative systems in existing technologies. The core objective is to integrate alliance-forming game theory with preference-based centralized auction mechanisms to construct a comprehensive, systematic collaborative framework encompassing "alliance formation - task allocation - reward calculation - collaborative closed loop." This framework aims to stabilize the alliance structure, improve task allocation efficiency, ensure fair reward calculation, and achieve a closed-loop collaborative process. Ultimately, it overcomes the limitations of single drones, meets the needs of large-scale, differentiated sensing tasks, balances the interests of task requesters and drone alliances, and improves sensing task execution efficiency, data quality, and system sustainability.

[0035] The core setup is as follows: the sensing platform acts as the collaboration hub, responsible for task distribution, alliance selection, and reward calculation; there are I heterogeneous drones and M sensing tasks. Each task can be executed collaboratively by a drone alliance. Each drone joins only one alliance, and the alliances are mutually exclusive and cover all participating drones. The alliance structure is defined as follows: The task set is defined as ={1,…,m,…,M} where As the nth alliance, it overcomes the limitations of single drone capabilities through collaborative operations, adapting to large-scale, differentiated sensing needs.

[0036] The complete model of task m is ,in The coordinates of the mission location. For duration, For the required sensor type, To perceive quality requirements, For other auxiliary parameters, The task time to live (TTL).

[0037] The key quantification is defined as follows: alliance Perceived quality of task m , Let be a binary variable (1 represents the sensor k required for mission m when UAV i is equipped). m (0 indicates not equipped). Sensor k equipped for drone i m Perceived quality; alliance Total cost of executing task m , These are the flight energy consumption, hovering energy consumption, sensing energy consumption, and communication energy consumption of drone i. Cost of alliance collaboration (non-zero when alliance size ≥ 2). , , These represent the unit energy cost, unit sensor cost, and unit transmission cost of drone i, respectively. alliance The social welfare increment of task m λ is the perceived quality weighting coefficient, and ω is the cost adjustment coefficient. The reward for performing missions for the alliance, where δ is the size penalty coefficient. For the alliance The number of drones, For the penalty intensity parameter; Total social welfare =∑( ∈ )∑(m∈ ) ,in For alliance structure (including all participating alliances) ), The task set (containing all sensor tasks m to be executed) is superimposed. All alliances in China The social welfare increment corresponding to the task is used to quantify the overall benefits of the entire collaborative system.

[0038] II. Step-by-step breakdown of technical solutions Step 1: System Initialization and Information Collection 1. The sensing platform receives M sensing tasks submitted by the task requester and collects the location of each task. Required sensor type Perceived quality requirements Duration Survival time Core parameters are used to form a task set. ={1,…,m,…,M}; 2. Simultaneously collect the location and remaining energy of all drones. i (t), Sensor configuration (including) Parameters, flight speed, and other status information are used to form a drone ensemble. ={1,…,i,…,I}; 3. Standardize the collected task parameters and UAV status information to provide accurate data support for subsequent alliance formation, task allocation, and constraint verification.

[0039] Step Two: Alliance Formation Optimization 1. Initialize the drone alliance partition set Z={ (T is the total number of partitions), each partition satisfies Based on the structural requirements, calculate the current partition. Total social welfare =∑( ∈ )∑(m∈ ) (in For the current alliance structure, (for task sets) 2. Execute the merge mechanism: Traverse the current partition. Each alliance S in l other alliances (l≠l') Merging to form a new division (Maintaining the definition of the alliance structure), calculate the total social welfare Ω of the newly partitioned area. )=∑( ∈ )∑(m∈ ) If Ω( )>Ω( (If total social welfare achieves an incremental increase), then update = Ω ( )=Ω( Repeat this process until it becomes impossible to improve overall social welfare through merging; 3. Let the result of the merge be... merge ={S1,…,S g ,…,S G It still conforms to the definition of an alliance structure; 4. Execute the splitting mechanism: traversal merg For each alliance Sg, initialize the candidate split set and construct a new partition. (Maintaining the characteristics of the alliance structure), calculate the total social welfare Ω after the split ( )=∑( ∈ )∑(m∈ ) If Ω( )>Ω( (If total social welfare achieves an incremental increase), then update = Ω ( )=Ω( Repeat this process until it becomes impossible to improve overall social welfare through further splitting. 5. Output the stable alliance partition that optimizes total social welfare Π*= This division is the optimal alliance structure to adapt to subsequent task allocation, solving the problems of alliance size adaptation and stability.

[0040] Step 3: Task Assignment and Bidding Processing 1. Sensor platform broadcast task set Detailed information about each alliance in the alliance structure Π* For each task m, submit a bid value based on the actual cost of executing the task. = (The bid value is consistent with the actual cost); 2. Iterate through each alliance With task m, preset Calculate the corresponding payment price for the successful bidder. and social welfare gains = (can be determined) ; 3. Initialize the number of tasks to be assigned, x = | | Total social welfare W=0, Distribution scheme X= ( =1 indicates alliance S l Assign to task m, otherwise assign 0); 4. When Executes in a loop: ① Each unassigned alliance selects the highest social welfare gain for all its corresponding unassigned tasks. and the corresponding task m; ② If multiple alliances compete for the same task m, the task will be assigned to the alliance with the lowest bid value and the best social welfare increment; if there is no competition, the task will be directly assigned to the suitable alliance. ③ Update allocation scheme X, mark =1, total social welfare summed with the currently distributed welfare gain. ; ④ Remove the assigned alliance from the set to be assigned. And task m, update ; 5. Output the maximum total social welfare And the allocation scheme X, and at the same time convert the allocation scheme X into the index matrix Z, =1 and =1 is equivalent to both representing an alliance. The winning alliance and corresponding task are clearly defined for the winning bidder of task m, so as to achieve precise matching between tasks and alliances.

[0041] Step 4: Compensation Calculation and Constraint Verification 1. Initialize the optimal payoff vector For each winning alliance ( =1 means For the successful bidder (the party awarded task m), the execution fee calculation is as follows: ①Remove the winning bid alliance Build a new drone ensemble ; ② with The algorithm for forming the alliance in step two is used as input to obtain the new alliance structure and the index matrix. Find the candidate winning alliance corresponding to task m. ; ③ Calculate the virtual bid value Determine the optimal payment for the winning consortium. ; 2. Perform multi-dimensional task constraint verification to ensure that the winning alliance meets all adaptation requirements: ① Energy Constraint: The energy constraint is that the remaining energy ei(t) of each UAV in the winning alliance is greater than or equal to the total energy consumption (including collaboration cost) for performing the task. ②Time Constraint: The time constraint is that the total time for the UAV to complete the mission (flight time + perception duration + data transmission time) ≤ mission survival time. ; ③ Quality Constraints: Quality constraints refer to the actual perceived quality of the alliance's mission execution. Task Requirements ; ④ Sensor constraints: The sensor constraint is that at least one drone in the winning consortium must be equipped with the type of sensor required for the mission. ; ⑤ Allocation constraint: Each task can be assigned to at most one federation, and each federation can execute at most one task at a time; ⑥ Economic attribute constraints: Economic attribute constraints are to meet the requirement of a true price quote ( ) and individual rationality (alliance utility) ).

[0042] Step 5: Task Execution and Closed-Loop Collaboration 1. The winning alliance is based on task location information. Plan coordinated flight paths, avoid flight conflicts, and fly to the designated mission area; 2. According to the required sensing range and accuracy of the task. Collaborate to carry out data collection operations, verify data quality in real time during the collection process, and re-collect data that does not meet the standards; 3. The collected sensor data is uploaded to the sensor platform via encrypted transmission. The platform then processes and filters the data before submitting it to the task requester. 4. The sensing platform, after deducting the relevant costs of the alliance's mission execution, will base its calculations on the sensing contributions of each drone within the alliance (and...). Positive correlation), energy consumption input (with) Develop fair distribution rules based on positive correlation and collaborative participation. 5. Distribute rewards to participating drones according to the allocation rules to incentivize drones to continue participating in the collaboration, forming a closed-loop collaboration mechanism of "alliance optimization - task matching - reward incentive" to ensure the sustainability of the system.

[0043] III. Combination Figures 2-7 Explanation Figure 2 : A diagram illustrating how total social welfare changes with the maximum number of sensors per drone under different task types; This graph corresponds to the alliance formation optimization in step two and the task allocation in step three, verifying the impact of the number of sensors on the alliance's adaptability and total social welfare. The horizontal axis represents the maximum number of sensors per drone (3-10), and the vertical axis represents the total social welfare. The variable condition is that the number of task types = 10, 15, and 20 scenarios (corresponding to the task sets). (Differences in scale). As the number of sensors increases, the alliance... The coverage capability of mid-mission sensor types is enhanced, and the redundancy cost is balanced through the scale penalty mechanism in step two, so the total social welfare increases steadily and then levels off; when the number of sensors reaches 8-10, the alliance's sensing capability is sufficient. As task requirements increase, marginal benefits diminish. The greater the number of task types, the higher the overall social welfare, confirming the effectiveness of the collaborative advantages of heterogeneous sensors in the task allocation process of step three, and demonstrating the adaptability of the "alliance formation-task allocation" collaborative design to large-scale differentiated tasks.

[0044] Figure 3 : A schematic diagram illustrating how total social welfare changes with average task quality requirements under different time-to-live (TTL) conditions; This diagram corresponds to the task parameter acquisition in step one (TTL is...). The core parameters of the task (and the alliance's perceived quality optimization in step two) are used to verify the moderating effect of task quality requirements and TTL parameters on total social welfare. The horizontal axis represents the average task quality requirement (0.5-0.68), and the vertical axis represents total social welfare, with variables including TTL=30s, 40s, and 50s scenarios. Increased task quality requirements drive the alliance to improve perceived quality through optimization in step two, while a higher TTL enhances the quality gain effect. Furthermore, the reward calculation mechanism in step four ensures cost stability. Therefore, total social welfare steadily increases, and the higher the TTL, the higher the curve. This result verifies the supporting role of the completeness of task parameter collection in step one for subsequent optimization, and the effectiveness of this scheme in... The ability to adapt to the quality requirements of differentiated tasks.

[0045] Figure 4 : A schematic diagram illustrating the change in total social welfare with sensor quality weighting parameters under different payment adjustment coefficients; This graph corresponds to the compensation calculation mechanism in step four, verifying the impact of the balance between the core parameter λ (sensor quality weighting parameter) and ω (payment adjustment coefficient) on total social welfare. The horizontal axis represents the sensor quality weighting parameter (0.65-0.8), and the vertical axis represents total social welfare. The variable conditions are three scenarios with payment adjustment coefficients of 0.3, 0.5, and 0.7. The larger λ is, the higher the weight of perceived quality in welfare calculation, and the positive contribution of the alliance's perceived quality is amplified, thus welfare continues to rise. The larger ω is, the stronger the negative impact of cost expenditure, thus the smaller the payment adjustment coefficient, the higher the welfare level. This graph confirms the design logic of "balance between quality weighting and payment adjustment" in the compensation calculation formula in step four, providing experimental basis for parameter configuration and ensuring the balance of interests between task requesters and the alliance.

[0046] Figure 5 : Schematic diagram showing how the average sensing quality of a task changes with the number of drones under different task numbers; This graph corresponds to the optimization of the alliance formation in step two and the task execution in step five, verifying the effect of the number of drones on improving the alliance's sensing capabilities. The horizontal axis represents the number of drones (30-120), and the vertical axis represents the average sensing quality of the task. The variable conditions are three scenarios with 10, 15, and 20 tasks (corresponding to...). (Differences in scale). Step two's alliance resource adaptation scheme optimizes the drone and... The matching relationship between tasks avoids resource redundancy or insufficiency, so as the number of drones increases, the sensing quality continues to improve; the fewer the number of tasks, the more fully the sensing capabilities of each drone can be utilized for the task, and the higher the growth rate. When the number of drones reaches 120, the sensing quality of the M=10 scenario is close to 2.0, which reflects the effectiveness of the alliance formation in step two and the collaborative improvement of sensing quality in task execution in step five, breaking through the perception limitations of a single drone.

[0047] Figure 6 : A schematic diagram showing the relationship between the average sensing quality of the mission and the number of sensing missions and drones under different schemes; This figure compares our proposed solution with three traditional solutions: GATA, SPSA, and FPSA, comprehensively verifying the synergistic optimization effect of steps two through five. Figure 6 Subgraph (a) (with a fixed number of 20 drones) corresponds to the task allocation in step three. This solution dynamically adapts to the alliance structure. With sensor resources, avoid The increased number of tasks dilutes the sensing capabilities, resulting in a steady increase in sensing quality. In contrast, traditional solutions, lacking a dynamic optimization mechanism, experience stagnant quality growth. Figure 6 Subgraph (b) in (the number of tasks is fixed at 20, corresponding to...) (With a fixed scale) Corresponding to the formation of the alliance in step two, this solution efficiently utilizes the newly added drone resources and reduces scale penalty losses, thus achieving a much higher increase in sensor quality than traditional solutions. This figure demonstrates the optimization advantages of this solution throughout the entire process of "alliance formation - task allocation - task execution," ensuring that the average sensor quality of each task remains at a leading level and adapting to the needs of large-scale tasks.

[0048] Figure 7 A diagram illustrating the relationship between social welfare and the number of sensing tasks and drones under different schemes; This diagram compares this scheme with the traditional scheme to verify the role of the core mechanisms in steps two through four in safeguarding social welfare. Figure 7 Subgraph (a) (with a fixed number of drones of 20) corresponds to the task allocation in step three and the reward calculation in step four. This scheme controls the growth of costs and penalties by optimizing the balance between quality weighting, payment adjustment and alliance size penalty. Therefore, social welfare is maintained in a high positive range, while the traditional scheme suffers from a decline in welfare due to the lack of a coordination mechanism. Figure 7 Subgraph (b) in (the number of tasks is fixed at 20, corresponding to...) (Fixed size) Corresponding to the formation of the alliance in step two, this scheme dynamically adjusts the efficiency of resource input and avoids the uncontrolled expansion of the alliance structure Π through gradient size penalties. Therefore, social welfare remains at a stable positive level, while the traditional scheme suffers from a continuous decline in welfare due to uncontrolled alliance expansion. This diagram fully demonstrates the effectiveness of the "alliance formation-task allocation-reward calculation" collaborative mechanism of this scheme in welfare optimization, regardless of... With drones Regardless of changes in scale, the overall benefits of the system can be guaranteed.

[0049] Secondly, this application also provides a drone swarm collaboration optimization device, which includes: an information acquisition module for collecting key parameters of sensing tasks and all drone status information; an alliance formation optimization module for forming stable alliances through iterative optimization of alliance merging and splitting with the goal of maximizing total social welfare; a task matching and reward calculation module for selecting winning alliances based on task bidding of each alliance and combining task constraints, and simultaneously calculating rewards using a critical payment strategy; and a task execution and reward distribution module for winning alliances to collaboratively execute tasks, transmit collected data, and distribute rewards within the alliance according to rules, forming a collaborative closed loop.

[0050] The functions of each module in the above-mentioned UAV swarm collaboration optimization device correspond to the steps in the above-mentioned UAV swarm collaboration optimization method embodiment, and their functions and implementation processes will not be described in detail here.

[0051] Thirdly, embodiments of this application provide a drone swarm collaboration optimization device, which can be a device with data processing capabilities such as a personal computer (PC), a laptop, or a server.

[0052] Reference Figure 8 , Figure 8 This is a schematic diagram of the hardware structure of the UAV swarm collaboration optimization device involved in the embodiments of this application. In the embodiments of this application, the UAV swarm collaboration optimization device may include a processor, a memory, a communication interface, and a communication bus.

[0053] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0054] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting components within the UAV swarm collaboration optimization device, as well as interfaces used for interconnecting the UAV swarm collaboration optimization device with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0055] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0056] The processor can be a general-purpose processor, which can call the UAV swarm cooperation optimization program stored in memory and execute the UAV swarm cooperation optimization method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the UAV swarm cooperation optimization program is called can be referred to in the various embodiments of the UAV swarm cooperation optimization method of this application, and will not be repeated here.

[0057] Those skilled in the art will understand that Figure 8The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0058] Fourthly, embodiments of this application also provide a readable storage medium.

[0059] The present application stores a drone swarm cooperation optimization program on a readable storage medium, wherein when the drone swarm cooperation optimization program is executed by a processor, it implements the steps of the drone swarm cooperation optimization method as described above.

[0060] The method implemented when the UAV swarm cooperation optimization program is executed can be referred to in various embodiments of the UAV swarm cooperation optimization method of this application, and will not be repeated here.

[0061] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0062] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0063] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0064] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0065] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0066] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0067] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for optimizing cooperation of a UAV cluster, characterized in that, The drone swarm collaboration optimization method includes: Collect key parameters of the sensing mission and all UAV status information; Based on the collected key parameters of the sensing mission and all UAV status information, and with the goal of maximizing total social welfare, a stable alliance is formed through iterative optimization of alliance merging and splitting. Based on the task bidding of each alliance, the winning alliance is selected in combination with the task constraints, and the remuneration is calculated simultaneously using the critical payment strategy. The winning alliance collaborates to execute tasks and transmit collected data, with compensation distributed within the alliance according to rules, forming a closed-loop collaboration. Based on the collected key parameters of the sensing mission and the status information of all UAVs, and with the goal of maximizing total social welfare, a stable alliance is formed through iterative optimization of alliance merging and splitting, including: The social welfare increment of each alliance performing corresponding tasks under the current division is calculated and summed to obtain the total social welfare of the current division; among which, the social welfare increment is comprehensively evaluated in combination with the perceived quality of alliances performing tasks, the rewards for alliances performing tasks, and the penalty factors of alliance size; For each alliance in the current division, try to merge with other alliances in turn to form a new division. Calculate the total social welfare of each alliance performing the corresponding tasks under the new division. If the new total social welfare is higher than the current total social welfare, update the division and the total social welfare value. Repeat the operation until the total social welfare cannot be improved by merging. For each merged alliance, initialize the candidate split set and construct a new partition. Calculate the total social welfare of each alliance performing the corresponding task under the new partition. If the new total social welfare is higher than the current total social welfare, update the partition and the total social welfare value. Repeat the operation until the total social welfare can no longer be improved through splitting. Output a stable alliance that maximizes overall social welfare; The process involves bidding for tasks from various alliances, selecting winning alliances based on task constraints and social welfare increments, and simultaneously calculating remuneration, including: Each alliance formulates and submits bidding information for each sensing task based on its actual cost of performing the task, and the bidding information is consistent with the actual cost. For each task, verify whether the participating bidding consortium meets the task constraints; If multiple alliances compete for the same task, the alliance whose bid meets the requirements and has the best social welfare increment will be selected as the winning bidder. If there is no competition, the task will be directly assigned to the suitable alliance. A critical payment strategy is used to calculate the remuneration of the winning consortium. After removing the winning consortium, the consortium formation and task allocation process is rerun to obtain alternative winning consortia. Based on the perceived quality, scale and bidding attributes of both consortia, a virtual bid value is calculated and the virtual bid value is determined as the remuneration amount.

2. The UAV swarm collaboration optimization method as described in claim 1, characterized in that, The collection of key parameters for the sensing task and all UAV status information includes: The sensing platform receives sensing tasks submitted by task requesters and collects information on the location, required sensor type, sensing quality requirements, duration, and lifespan of each task. Collect the location, remaining energy, sensor configuration, and flight speed of all drones.

3. The UAV swarm cooperation optimization method as described in claim 1, characterized in that, The task constraints include that the remaining energy of each UAV in the alliance meets the total energy consumption of the task, the total time to complete the task does not exceed the task survival time, the perception quality of the task execution meets the task requirements, and at least one UAV in the alliance is equipped with the type of sensor required for the task.

4. The UAV swarm collaboration optimization method as described in claim 3, characterized in that, The total energy consumption of the mission includes the UAV's flight energy consumption, hovering energy consumption, sensing energy consumption, communication energy consumption, and alliance collaboration energy consumption. The total mission time includes flight time, sensing duration, and data transmission time.

5. The UAV swarm collaboration optimization method as described in claim 1, characterized in that, The winning consortium collaborates to execute tasks and transmit collected data, with compensation distributed within the consortium according to rules, forming a closed-loop collaboration mechanism, including: The winning consortium plans a coordinated flight path based on the mission location information to avoid flight conflicts and fly to the designated mission area; According to the required sensing range and accuracy, we will carry out data collection operations in a coordinated manner, verify the data quality in real time during the collection process, and re-collect any data that does not meet the requirements. The collected sensor data is transmitted to the sensing platform via encrypted transmission. The platform then processes and filters the data before submitting it to the task requester. The sensing platform deducts the relevant costs of the alliance's mission execution and formulates fair distribution rules based on the sensing contribution, energy consumption, and collaborative participation of each drone in the alliance. The rewards are distributed to each participating drone according to the distribution rules, which incentivizes them to continue participating in the collaboration and forms a closed loop.

6. An unmanned aerial vehicle cluster cooperation optimization apparatus, characterized in that, It includes the UAV swarm cooperation optimization method as described in any one of claims 1-5, and the UAV swarm cooperation optimization device further includes: The information acquisition module is used to collect key parameters of the sensing mission and all UAV status information; The alliance formation optimization module is used to form stable alliances through iterative optimization of alliance merging and splitting, with the goal of maximizing total social welfare. The task matching and reward calculation module is used to select winning alliances based on task bidding from various alliances, combined with task constraints, and simultaneously calculate rewards using a critical payment strategy. The task execution and reward distribution module is used by the winning alliance to collaboratively execute tasks and transmit collected data. Rewards are distributed within the alliance according to rules, forming a collaborative closed loop.

7. A drone swarm collaboration optimization device, characterized in that, The UAV swarm collaboration optimization device includes a processor, a memory, and a UAV swarm collaboration optimization program stored in the memory and executable by the processor, wherein when the UAV swarm collaboration optimization program is executed by the processor, it implements the steps of the UAV swarm collaboration optimization method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a drone swarm cooperation optimization program, wherein when the drone swarm cooperation optimization program is executed by a processor, it implements the steps of the drone swarm cooperation optimization method as described in any one of claims 1 to 5.

Images