A multi-agent heuristic task allocation method and system based on cooperative game

By employing a multi-agent heuristic task allocation method based on cooperative game theory, and utilizing real-time perception and dynamic updates from UAV sensors, the problem of insufficient consideration of resource and task coupling in UAV swarms is solved, achieving efficient task allocation and resource utilization.

CN117908577BActive Publication Date: 2024-11-19HARBIN ENG UNIV
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202311766597.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-21
Publication Date
2024-11-19
Estimated Expiration
2043-12-21

AI Technical Summary

Technical Problem

Existing drone swarm task allocation methods are inefficient when considering the coupling relationship between drone-borne resources and tasks, making it difficult to achieve efficient task allocation in specific scenarios.

Method used

A multi-agent heuristic task allocation method based on cooperative game theory is adopted. The set of neighboring UAVs is determined by the UAV's location information, external information is perceived in real time by sensors, the lead UAV is marked to broadcast tasks, and task allocation and dynamic updates are carried out through cooperative game theory algorithm to optimize resource utilization.

Benefits of technology

It improves the real-time performance and accuracy of drone swarm task allocation, enabling it to better adapt to dynamic environments and task changes, optimize resource utilization, and improve overall task execution efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117908577B_ABST
    Figure CN117908577B_ABST
Patent Text Reader

Abstract

A multi-agent heuristic task allocation method and system based on cooperative game, relating to the field of unmanned swarm technology. The method solves the problem that the coupling relationship between the resources carried by existing drones and tasks is less considered, which limits the efficiency of task allocation in specific scenarios. The method includes: establishing a set of neighboring drones based on drone location information; the drone relies on sensors to obtain external information and judge whether a new task has appeared; judging that a new task has been detected, the drone that finds the task is marked as the leader and broadcasts it as the initiator; the drone determines the final task allocation plan based on the task allocation algorithm; the drone performs related tasks in sequence according to the task sequence obtained by each, and continuously detects the environment and dynamically updates resources; if the task is completed, the alliance is disbanded, and the drone maintains the current flight state to perform tasks in the area; if the task is not completed, continue to execute the previous step. Applicable to the field of drone resource allocation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of unmanned cluster technology, and in particular to a task allocation method for a multi-agent cluster. Background Art

[0002] In recent years, with the rapid development of drone technology, drone swarms have become an important technical field. Especially under the conditions of information warfare, drone swarms play an increasingly important role in combat. Drone swarms can be used to perform various tasks, such as surveillance, search, search and rescue, etc. Most of these tasks have high requirements for real-time and efficiency, and also need to consider the balance between multiple technical indicators. The various intelligent agents included in the unmanned swarm share information, each intelligent agent can assign tasks and execute tasks, and there is no absolute hierarchy between intelligent agents, which means that there is no fixed division between leaders and executors. UAV swarm task allocation refers to determining a set of optimal action plans so that the drone swarm can complete the set task within a given time. However, the realization of this goal usually involves multiple conflicting goals, so it is necessary to weigh and choose between multiple goals.

[0003] For the above-mentioned task allocation problem, evolutionary optimization algorithms such as genetic algorithms or particle swarm algorithms are usually used to solve it. However, such algorithms require multiple rounds of iterations to find an approximate optimal solution, have poor real-time performance, and will experience dimensionality explosion as the number of drones increases. Therefore, such algorithms are more suitable for solving allocation problems with known global information. For dynamic task allocation problems, distributed algorithms based on market rules and auction theory are usually used to deal with sudden task situations that may occur at any time in the task environment. Such methods can reduce the solution time, and since the algorithm is distributed on drones, its robustness is relatively high. The above-mentioned task allocation method can obtain a reasonably allocated task sequence, but the coupling relationship between the resources carried by drones and tasks is less considered, which limits the efficiency of task allocation in specific scenarios. Therefore, how to design an allocation method that can both meet the task sequence relationship and fully consider the heterogeneity of tasks and resource heterogeneity is a technical problem that is urgently needed to be solved in this field.

[0004] Prior art CN113395676 discloses a method for UAV task collaboration based on overlapping alliance formation game, including constructing the UAV multi-task collaboration problem as a sequential overlapping alliance game model; using a bilateral mutual benefit criterion that is biased towards the utility of the entire alliance to evaluate the preference of the UAV for the alliance structure; using a taboo algorithm guided by preference gravity to optimize the allocation of UAV task resources under the overlapping alliance structure to form a stable alliance structure; optimizing the transmission strategy based on the update status of the current alliance structure and the UAV task resource allocation scheme and the current fading environment to maximize the utility of the UAV network in executing tasks. However, the above method uses a discrete resource heuristic matching method. Once the allocation is completed, the matching relationship between the target required resources and the UAV-mounted resources will not be released, which limits the possibility of matching better quality resources. Summary of the invention

[0005] Aiming at the problem that the coupling relationship between existing drone-carried resources and tasks is rarely considered, which limits the efficiency of task allocation in specific scenarios, the present invention proposes a multi-agent heuristic task allocation method based on cooperative game, and the specific scheme is as follows:

[0006] A multi-agent heuristic task allocation method based on cooperative game, the method comprising:

[0007] S1: Establish a set of neighboring drones based on drone location information;

[0008] S2: Each drone relies on its own sensors to obtain external information and determine whether a new task has occurred;

[0009] S3: If a new mission is detected, the UAV that discovered the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting;

[0010] S4: Each UAV determines the final task allocation plan according to the task allocation algorithm;

[0011] S5: Each UAV performs related tasks in sequence according to its own obtained task sequence, continuously detects the environment, and dynamically updates resources;

[0012] S6: If the mission is completed, the alliance is disbanded, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, continue to step S5.

[0013] Furthermore, a preferred method is proposed, wherein step S1 comprises:

[0014] The drone determines whether other drones are in its neighbor space based on the strength of the received signal. Once the drone is within the communication range of its neighbors, it establishes a two-way communication topology network with the neighbor’s drones and extracts the unique drone ID in the information packet.

[0015] Furthermore, a preferred method is proposed, wherein step S4 comprises:

[0016] S41: Initialize the alliance history structure taboo table, each drone history structure taboo table, allocation matrix, drone task sequence, drone alliance performing corresponding tasks and its corresponding task resources;

[0017] S42: Each UAV participates in the corresponding task alliance according to its own resources and calculates the benefits of the current task alliance structure;

[0018] S43: Determine whether the current task sequence allocation result of the UAV is valid and meets the task requirements;

[0019] S44: if the allocation result is valid, the current allocation result is added to the individual sequence history list and the alliance history structure list;

[0020] S45: Update the task sequence, available resource matrix, allocation matrix, task requirement matrix, the drone alliance that performs the corresponding task, its corresponding task resources and alliance structure of the local machine according to the importance ranking, and unify the above information through local broadcasting;

[0021] S46: Determine whether the current drone has a contribution in the mission alliance it participates in;

[0022] S47: Traverse each mission alliance in which the UAV participates and calculate the total mission resources of the current mission alliance.

[0023] Furthermore, a preferred method is proposed, wherein step S42 comprises:

[0024] Calculate the overall benefits of the task alliance structure after optimal allocation;

[0025] Calculate the relationship between currently allocated resources and resource requirements;

[0026] Calculate the benefits of drones performing mission sequences;

[0027] Calculate the cost of the drone to perform the mission;

[0028] Build an optimization model to optimize allocation with the goal of maximizing overall benefits;

[0029] Each UAV obtains the task resource gap status through inter-UAV communication;

[0030] Each drone calculates the degree of matching with the target based on the type and quantity of its remaining resources and the position relationship information with the target task;

[0031] According to the roulette rule, the task with the highest matching degree is selected from the UAV task sequence and the current type of resources is allocated to the task;

[0032] Update the local task sequence according to the task importance and obtain the task alliance structure after optimized allocation.

[0033] Furthermore, a preferred method is proposed, wherein the optimization model is constructed to optimize the allocation with the goal of maximizing the overall benefit, including:

[0034] Construct optimization model and constraints:

[0035]

[0036] By adjusting the alliance structure To maximize the overall benefit, the following constraints need to be met

[0037]

[0038]

[0039]

[0040] in, is the allocation result of the zth type of resources for UAV n to task m, is the available resource of the zth resource, Z is the type of resource required, N is the set of drones, is the number of resources required for each resource, m is the task index, M is the task set, Task n For task sequence.

[0041] Furthermore, a preferred method is proposed, wherein the calculation of the overall benefit under the current task alliance structure includes:

[0042]

[0043] Among them, U m (A m ) is the coalition benefit formed by the set of drones performing task m, C1 is a constant, β m To control the scope of benefit, C m (A m ) is the cost of executing task m, A m For task resources.

[0044] Furthermore, a preferred embodiment is proposed, wherein step S5 further comprises:

[0045] Each UAV extracts the location information and time window information of the target in its respective mission sequence, and uses the information as input information for path planning;

[0046] According to the current scenario requirements, select the appropriate local path planning method to meet the requirements of reaching the target while safely avoiding obstacles in the environment.

[0047] Based on the same inventive concept, the present invention also proposes a multi-agent heuristic task allocation system based on cooperative game, the system comprising:

[0048] A drone set determination unit, used to establish a neighboring drone set based on the drone position information;

[0049] The information acquisition unit is used by each drone to obtain external information through its own sensors to determine whether a new task has occurred;

[0050] The broadcast unit is used to determine if a new mission is detected. The UAV that discovers the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting.

[0051] A task allocation unit, used for each UAV to determine the final task allocation plan according to the task allocation algorithm;

[0052] The task execution unit is used for each UAV to perform related tasks in sequence according to the task sequence obtained by each UAV, and continuously detect the environment and dynamically update resources;

[0053] The judgment unit is used to disband the alliance if the mission is completed, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, the mission execution unit continues to execute.

[0054] Based on the same inventive concept, the present invention also proposes a computer-readable storage medium, characterized in that the computer-readable storage medium is used to store a computer program, and the computer program executes any of the above-mentioned multi-agent heuristic task allocation methods based on cooperative game.

[0055] Based on the same inventive concept, the present invention also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes a multi-agent heuristic task allocation method based on cooperative game according to any one of the above items.

[0056] The present invention is beneficial in that:

[0057] The present invention solves the problem that the coupling relationship between the resources carried by existing drones and tasks is rarely considered, thereby limiting the efficiency of task allocation in specific scenarios.

[0058] The present invention provides a multi-agent heuristic task allocation method based on cooperative game. By using the cooperative game method, the system can realize collaborative decision-making among multiple agents to allocate tasks more effectively. Cooperative game allows the formation of alliances between agents to jointly solve problems and improve overall efficiency. The method takes into account dynamic changes in the drone collection, such as the location of the drone and the emergence of tasks. Each drone perceives external information in real time through its own sensors so that it can quickly make adjustments when the task changes, ensuring the real-time and accuracy of task allocation. When a new task appears, by marking the drone that detects the task as a leader and setting it as a central node, timely discovery and organization of the task is achieved, which helps to start the task allocation process more efficiently. The method allows continuous detection of the environment and dynamic updating of resources. This real-time perception and update mechanism can make the system better adapt to changing tasks and environmental conditions. The condition for the dissolution of the alliance is the completion of the task, which helps to optimize resource utilization and improve overall efficiency.

[0059] The invention provides a multi-agent heuristic task allocation method based on cooperative game, which determines the set of neighboring drones according to the location information of drones to form a group of agents that can work together. The emergence of new tasks is detected by its own sensors, and the drone that detects the task is marked as the leader, which is set as the central node to initiate task broadcasting. The cooperative game algorithm is used to allocate tasks to ensure the reasonable allocation of tasks among agents and maximize the overall efficiency. The drone is dynamically updated according to environmental changes and task completion to ensure that the system has real-time response capabilities to task and environmental changes. By adopting the cooperative game algorithm, tasks can be allocated more flexibly, thereby improving the overall task execution efficiency. Considering the dynamics of drone sets and tasks, it has stronger adaptability and can better adapt to changing environmental and task conditions. By setting the alliance dissolution conditions, it is ensured that the alliance is dissolved in time after the task is completed, thereby optimizing the utilization of resources. Through the mechanism of real-time perception and task marking of drones, the system has higher real-time performance and accuracy, which helps to cope with complex and changing task scenarios.

[0060] The present invention is applicable to the field of unmanned aerial vehicle resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A flowchart of a multi-agent heuristic task allocation method based on cooperative game described in implementation mode 1;

[0062] Figure 2A schematic diagram of the corresponding relationship between the types and numbers of the mission required resources and the resources carried by the drone as described in Implementation Method 3;

[0063] Figure 3 This is a schematic diagram of the overall benefit comparison of the tasks described in Implementation Method 11;

[0064] Figure 4 A schematic diagram of the corresponding relationship between the tasks and the drones described in Implementation Example 11;

[0065] Figure 5 This is a schematic diagram of resource allocation utilization described in Implementation Method 11. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0067] Implementation method 1, see Figure 1 This embodiment describes a multi-agent heuristic task allocation method based on cooperative game, and the method includes:

[0068] S1: Establish a set of neighboring drones based on drone location information;

[0069] S2: Each drone relies on its own sensors to obtain external information and determine whether a new task has occurred;

[0070] S3: If a new mission is detected, the UAV that discovered the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting;

[0071] S4: Each UAV determines the final task allocation plan according to the task allocation algorithm;

[0072] S5: Each UAV performs related tasks in sequence according to its own obtained task sequence, continuously detects the environment, and dynamically updates resources;

[0073] S6: If the mission is completed, the alliance is disbanded, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, continue to step S5.

[0074] By using the cooperative game method, the system can achieve collaborative decision-making among multiple agents to allocate tasks more effectively. Cooperative games allow agents to form alliances to jointly solve problems and improve overall efficiency. The method takes into account dynamic changes in the drone collection, such as the location of drones and the emergence of tasks. Each drone perceives external information in real time through its own sensors so that it can quickly adjust when the task changes to ensure the real-time and accuracy of task allocation. When a new task appears, by marking the drone that detects the task as a leader and setting it as a central node, timely discovery and organization of the task is achieved, which helps to start the task allocation process more efficiently. The method allows continuous detection of the environment and dynamic updating of resources. This real-time perception and update mechanism can make the system better adapt to changing tasks and environmental conditions. The condition for the dissolution of the alliance is the completion of the task, which helps to optimize resource utilization and improve overall efficiency.

[0075] This implementation method forms a group of agents that can work together by determining the set of neighboring drones based on the drone location information. The emergence of new tasks is detected through its own sensors, and the drone that detects the task is marked as the leader, and it is set as the central node to initiate task broadcasting. The cooperative game algorithm is used for task allocation to ensure the reasonable allocation of tasks among agents and maximize the overall efficiency. The drone is dynamically updated according to environmental changes and task completion to ensure that the system has real-time response capabilities to task and environmental changes. By adopting a cooperative game algorithm, tasks can be allocated more flexibly, thereby improving the overall task execution efficiency. Considering the dynamic nature of drone sets and tasks, it has stronger adaptability and can better adapt to changing environmental and task conditions. By setting the alliance dissolution conditions, it is ensured that the alliance is disbanded in time after the task is completed, thereby optimizing the utilization of resources. Through the mechanism of real-time perception and task marking of drones, the system has higher real-time and accuracy, which helps to cope with complex and changing task scenarios.

[0076] Implementation method 2: This implementation method is a further limitation of the multi-agent heuristic task allocation method based on cooperative game described in implementation method 1. Step S1 includes:

[0077] The drone determines whether other drones are in its neighbor space based on the strength of the received signal. Once the drone is within the communication range of its neighbors, it establishes a two-way communication topology network with the neighbor’s drones and extracts the unique drone ID in the information packet.

[0078] This implementation method can automatically identify nearby drones by judging the strength of the received signal, without the need for static configuration in advance, and is more automated. By extracting the unique drone ID in the information packet, it can ensure that there will be no conflict or confusion when establishing the communication topology network, thereby enhancing the reliability of communication. Establishing a two-way communication topology network helps to achieve more flexible communication and information exchange, and provides a reliable communication foundation for subsequent collaborative work.

[0079] In this embodiment, the drone determines whether other drones are in its neighboring space by the strength of the received signal. Signal strength is usually related to distance, and a stronger signal indicates that the two drones are closer. A unique drone ID is extracted from the information packet to ensure that each drone has a unique identifier in the system to avoid information confusion or misjudgment. Once it is determined that two drones are within the communication range of their neighbors, they establish a two-way communication topology network. This can be achieved by exchanging communication information with each other and confirming the establishment of a communication connection. By automatically judging the strength of the signal, the automatic determination of neighboring drones is achieved, which reduces the work of manual configuration and improves the automation of the system. The reliability and stability of the system in terms of communication and information exchange are ensured by extracting the unique drone ID and establishing a two-way communication topology network. By establishing a communication topology network of neighboring drones, unnecessary communication interference can be reduced, making the system's communication more efficient and controllable.

[0080] Implementation method three, see Figure 2 This embodiment further defines the cooperative game-based multi-agent heuristic task allocation method described in Embodiment 1, and the step S4 includes:

[0081] S41: Initialize the alliance history structure taboo table, each drone history structure taboo table, allocation matrix, drone task sequence, drone alliance performing corresponding tasks and its corresponding task resources;

[0082] S42: Each UAV participates in the corresponding task alliance according to its own resources and calculates the benefits of the current task alliance structure;

[0083] S43: Determine whether the current task sequence allocation result of the UAV is valid and meets the task requirements;

[0084] S44: if the allocation result is valid, the current allocation result is added to the individual sequence history list and the alliance history structure list;

[0085] S45: Update the task sequence, available resource matrix, allocation matrix, task requirement matrix, the drone alliance that performs the corresponding task, its corresponding task resources and alliance structure of the local machine according to the importance ranking, and unify the above information through local broadcasting;

[0086] S46: Determine whether the current drone has a contribution in the mission alliance it participates in;

[0087] S47: Traverse each mission alliance in which the UAV participates and calculate the total mission resources of the current mission alliance.

[0088] This implementation provides the necessary foundation for the subsequent task allocation process by initializing various data structures and tables. By calculating the benefits of each agent when participating in the task alliance, information is provided for the agent to select an appropriate task alliance. Through the benefit calculation in game theory, the contribution of each agent in different alliances is evaluated. Each agent can selectively join the task alliance to maximize the overall benefit. By judging and updating the task allocation results, ensure that the generated task allocation plan is effective and meets the task requirements. The method of effectiveness judgment and importance sorting is adopted so that the task allocation results of each agent are screened and updated. Generate an optimized and validity-verified task allocation plan to improve overall efficiency. By calculating the contribution of the agent in its task alliance and the total task resources of the task alliance, the actual effect of task execution is evaluated. Through the contribution calculation in game theory, the contribution of each agent to the task alliance is understood; by calculating the total task resources of the task alliance, the overall resource situation of the task alliance is understood. Optimize the task allocation plan to ensure that the task can be effectively executed as required.

[0089] Implementation mode 4: This implementation mode is a further limitation of the multi-agent heuristic task allocation method based on cooperative game described in implementation mode 3, and the step S42 includes:

[0090] Calculate the overall benefits of the task alliance structure after optimal allocation;

[0091] Calculate the relationship between currently allocated resources and resource requirements;

[0092] Calculate the benefits of drones performing mission sequences;

[0093] Calculate the cost of the drone to perform the mission;

[0094] Build an optimization model to optimize allocation with the goal of maximizing overall benefits;

[0095] Each UAV obtains the task resource gap status through inter-UAV communication;

[0096] Each drone calculates the degree of matching with the target based on the type and quantity of its remaining resources and the position relationship information with the target task;

[0097] According to the roulette rule, the task with the highest matching degree is selected from the UAV task sequence and the current type of resources is allocated to the task;

[0098] Update the local task sequence according to the task importance and obtain the task alliance structure after optimized allocation.

[0099] This implementation method ensures that the task allocation scheme can achieve the maximum benefit as a whole by calculating the overall benefit under the task alliance structure after the optimized allocation. The cooperative game concept of game theory is used to consider the overall benefit of multi-agent collaborative work, so as to better evaluate the quality of task allocation. By analyzing the relationship between the current allocated resources and resource requirements, it helps to evaluate whether the drone meets the basic requirements of the task. Ensure that the task is fully met and avoid waste or shortage of resources. By calculating the benefits brought by the drone's execution of the task sequence and the cost required to execute the task, the benefits and costs of task execution are fully considered. By comprehensively considering the actual value and execution cost of the task, each task is evaluated for its contribution to the overall benefit. An optimization model is constructed with the goal of maximizing the overall benefit to ensure that the task allocation scheme is carried out under the global optimal situation. Mathematical modeling and optimization algorithms are used to find the optimal solution for task allocation. Improve the intelligence and efficiency of task allocation. Through inter-machine communication between drones, the task resource gap status is obtained in real time to improve the real-time and accuracy of task allocation. Ensure that the task allocation scheme is adaptable and flexible in a dynamic environment. Update the task sequence of the local machine according to the order of task importance to ensure that key tasks are given priority. With limited resources, ensure that key tasks are met with priority and improve overall robustness and execution efficiency.

[0100] Implementation mode 5: This implementation mode is a further limitation of the multi-agent heuristic task allocation method based on cooperative game described in Implementation mode 4. The local sequence rearrangement method based on task importance is used to optimize the allocation and obtain the task alliance structure after the optimized allocation, including:

[0101] Construct optimization model and constraints:

[0102]

[0103] By adjusting the alliance structure In order to maximize the overall benefit, the following constraints need to be met:

[0104]

[0105]

[0106]

[0107] in, is the allocation result of the zth type of resources for UAV n to task m, is the available resource of the zth resource, Z is the type of resource required, N is the set of drones, is the number of resources required for each resource, m is the task index, M is the task set, Task n For task sequence.

[0108] The local sequence rearrangement method based on task importance in this embodiment can respond to task priorities more flexibly, ensure that critical tasks are given higher priority, and improve the robustness and execution efficiency of the overall system. By rearranging the local sequence, the task alliance structure is optimized, the complexity of the global search space is avoided, and the practicality and execution efficiency of the algorithm are improved. By considering the importance of tasks in the task sequence, the system is more adaptable and can flexibly adjust task allocation in a dynamic environment to cope with changes in tasks and resources. An optimization model is constructed, which is committed to maximizing the overall benefit by adjusting the task alliance structure to ensure that the task allocation plan is optimal on a global scale.

[0109] The local sequence rearrangement method based on task importance in this embodiment integrates this information into the task allocation process by sorting and considering the importance of tasks. This helps to improve the flexibility of the system for tasks and ensure that the system pays more attention to the execution of key tasks. Compared with global search, the local sequence rearrangement method pays more attention to the local adjustment of the current task sequence. Such a strategy is more feasible in practice, especially in large-scale systems, reducing the computational complexity. A series of constraints are introduced, including the relationship between resource availability, resource demand and task sequence. These constraints are the basis for ensuring that the task allocation scheme is reasonable and feasible. By adjusting the task alliance structure, it aims to maximize the overall benefit. This ensures the efficient execution of the multi-agent system under limited resources and task constraints. Considering the importance of tasks and local adjustments, the system can adapt to the dynamic requirements of tasks more flexibly, improving robustness and adaptability. Compared with global search, the local sequence rearrangement method reduces the computational complexity to a certain extent, improves the practicality of the algorithm, and is suitable for large-scale systems. The introduction of constraints ensures that the task allocation scheme is carried out under the condition of sufficient resources and reasonable demand, avoiding unrealistic or infeasible task allocation.

[0110] Implementation 6: This implementation is a further limitation of the multi-agent heuristic task allocation method based on cooperative game described in Implementation 4. The calculation of the overall benefit under the current task alliance structure includes:

[0111]

[0112] Among them, U m (A m ) is the coalition benefit formed by the set of drones performing task m, C1 is a constant, βm To control the scope of benefit, C m (A m ) is the cost of executing task m, A m For task resources.

[0113] This implementation method can globally evaluate the pros and cons of the current task allocation scheme by calculating the overall benefit under the current task alliance structure, which helps to find the global optimal solution. It provides a basis for decision-making, so that trade-offs and adjustments can be made according to the overall benefit during the task allocation process, so as to better meet the overall target requirements. Through quantitative overall benefit evaluation, unnecessary calculations and searches can be avoided, thereby improving calculation efficiency.

[0114] The multi-agent task allocation method based on cooperative game in this embodiment generally involves cooperation and competition among participants. The calculation of the overall benefit under the current task alliance structure is based on game theory, considering the contribution of cooperation between various agents to the overall benefit. It involves comparing and weighing the benefits of various task allocation schemes. In the cooperative game of agents, by adjusting the allocation of tasks, maximizing the overall benefit becomes the optimization goal. By calculating the overall benefit under the current task alliance structure, it aims to find the optimal task allocation plan so that the entire system can complete the task with the highest efficiency. Through the evaluation of global benefits, the purpose is to improve the overall performance of the entire multi-agent system and ensure the best task execution effect under limited resources. Through the calculation of overall benefits, the utilization of resources can be better optimized, resource waste and task delays can be avoided, thereby improving the efficiency and performance of the drone.

[0115] Implementation method 7: This implementation method is a further limitation of the multi-agent heuristic task allocation method based on cooperative game described in implementation method 1, and step S5 further includes:

[0116] Each UAV extracts the location information and time window information of the target in its respective mission sequence, and uses the information as input information for path planning;

[0117] According to the current scenario requirements, select the appropriate local path planning method to meet the requirements of reaching the target while safely avoiding obstacles in the environment.

[0118] In this implementation, each drone extracts the target location and time window information according to its own task sequence, making the path planning more personalized and helping to meet the time and space requirements of each task to the greatest extent. Continuous detection of the environment and dynamic update of resources enable the system to adapt to changes in the environment and resource status in real time, improving the flexibility and adaptability of the system. Selecting the appropriate local path planning method according to the current scene requirements makes the path planning more intelligent, and can ensure safe avoidance of obstacles in the environment while meeting the target.

[0119] In this embodiment, each drone extracts the location information and time window information of the target from its own task sequence, and uses this information as the input of path planning. This ensures that the specific requirements of each task are taken into account in the path planning process. Each drone continuously detects the environment and updates resource information in real time, including the location of obstacles, environmental changes, etc. This is based on the principle of real-time information update, which ensures the real-time perception of changes by the drone. According to the current scene requirements, select appropriate local path planning methods such as A* algorithm, D algorithm, etc., and make intelligent choices according to the specific environment and task requirements in actual situations. Through personalized path planning, each drone can perform tasks more efficiently, meet the time window requirements of the target on time, and improve the overall task execution efficiency. The purpose of dynamic resource update and environmental detection is to enable the system to adapt to changes in the environment in real time, ensure that the latest environmental information is taken into account in the path planning process, and reduce the occurrence of unexpected events. Select an appropriate local path planning method to ensure that the drone can safely avoid obstacles in the environment when performing tasks, and improve the safety of task execution.

[0120] Embodiment 8: A multi-agent heuristic task allocation system based on cooperative game described in this embodiment, the system includes:

[0121] A drone set determination unit, used to establish a neighboring drone set based on the drone position information;

[0122] The information acquisition unit is used by each drone to obtain external information through its own sensors to determine whether a new task has occurred;

[0123] The broadcast unit is used to determine if a new mission is detected. The UAV that discovers the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting.

[0124] A task allocation unit, used for each UAV to determine the final task allocation plan according to the task allocation algorithm;

[0125] The task execution unit is used for each UAV to perform related tasks in sequence according to the task sequence obtained by each UAV, and continuously detect the environment and dynamically update resources;

[0126] The judgment unit is used to disband the alliance if the mission is completed, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, the mission execution unit continues to execute.

[0127] Implementation method nine, a computer-readable storage medium described in this implementation method, wherein the computer-readable storage medium is used to store a computer program, and the computer program executes a multi-agent heuristic task allocation method based on cooperative game as described in any one of implementation methods one to seven.

[0128] Embodiment 10. A computer device described in this embodiment includes a memory and a processor, wherein a computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes a multi-agent heuristic task allocation method based on cooperative game as described in any one of embodiments 1 to 7.

[0129] Implementation 11: This implementation provides a specific example of a multi-agent heuristic task allocation method based on cooperative game described in Implementation 1, and is also used to explain Implementation 2 to Implementation 7. Specifically:

[0130] Step 1: Each drone establishes a set of neighboring drones based on the estimated information of the positions of surrounding drones and exchanges various status information including its own position, speed, and available mission resources with them through local broadcast;

[0131] Step 1 specifically includes the following steps:

[0132] Step 1.1: Each drone n∈{1,...,N} determines whether the other drones are within its communication range based on the strength of the received signal and extracts the drone ID in the information packet;

[0133] Step 1.2: Once the drone is within the communication range of its neighbor, a two-way communication topology network is established with it;

[0134] Step 2: Each drone relies on its own sensors to collect external information and detect changes in the external environment to determine whether a new task has occurred;

[0135] Step 2 specifically includes the following steps:

[0136] When the onboard infrared sensor or other equipment detects that the target has entered the executable mission range, it will trigger the task information collection state and jump to step 3, otherwise jump to step 7;

[0137] Step 3: If a new task is detected, the UAV that discovers the task is marked as the leader and set as a temporary central node to play the role of the initiator and broadcast the status information of the new task m∈{1,...,M} such as location, time window information, priority, etc., and execute step 4;

[0138] Step 4: Each UAV executes the task allocation algorithm to obtain the final task allocation plan until the overall task benefit does not change;

[0139] Step 4 specifically includes the following steps:

[0140] Step 4.1: Assume that the set of drones that perform tasks is N = {1,...,n,...,N}, and the set of tasks to be performed is M = {1,...,m,...,M}. Except for the task requirements, the other attributes of each task are represented by an array. <V m ,D m ,β m ,x m >, respectively representing the value of task m, the threat level to the drone, the importance level, and the time decay factor. Each task requirement matrix Z represents the type of resource required. represents the number of resources required for each resource, and the importance of the task is β = {β1,...,β m ,...,β M}, the available resource matrix for each drone Indicates the available resources for the zth resource. Initialize the taboo table of the alliance history structure Tabu S Empty, each drone history structure taboo table Tabu n Empty, initial allocation matrix is empty, Represents the result vector of the drone's assignment to task m, and the drone's task sequence Task n Is empty, the drone alliance Mem(A m ) and its corresponding task resources is empty, It represents the number of resources allocated by the drone to the zth resource in the mth task, where the drone member performing task m is defined by the following formula:

[0141]

[0142] Step 4.2: Each drone selects part of its own resources to participate in the corresponding task alliance and calculates the benefits;

[0143] Step 4.2 specifically includes the following steps:

[0144] Step 4.2.1: Calculate the overall benefit under the current task alliance structure:

[0145]

[0146] Among them, C m(A m ) represents the cost associated with executing the task and is calculated by the following formula:

[0147] C m (A m )=ω1r1(A m )+ω2r2(A m )-ω3r3(A m ),ω1,ω2,ω3>0

[0148] Step 4.2.2: Calculate the relationship between the current allocated resources and the resource demand r1(A m ):

[0149]

[0150]

[0151]

[0152] r1(A m )=exp(k1E1)-exp(k2E2),k1,k2>0

[0153] Step 4.2.3: Calculate the benefit r2(A m ):

[0154]

[0155]

[0156] Step 4.2.4: Calculate the cost r3(A m ):

[0157]

[0158] Step 4.2.5: Define the optimization model and constraints for this problem as follows:

[0159]

[0160] By adjusting the alliance structure In order to maximize the overall benefit, the following constraints need to be met:

[0161] Constraint 1:

[0162] Constraint 2:

[0163] Constraint 3:

[0164] The above inequalities represent the following constraints:

[0165] Constraint 1: The maximum allocatable resource of each drone cannot exceed the total resources it carries;

[0166] Constraint 2: The resource allocation of each drone for the mission cannot be empty;

[0167] Constraint 3: The resource requirements of each task need to be met;

[0168] It should be noted that constraint 3 cannot always be met, so the concept of penalty function is introduced here to guide the alliance building process.

[0169] In order to solve the above optimization problem, the method described in this embodiment also models it as a cooperative game, and its mathematical model can be expressed as G=<N,S,U,Θ,P> , where N is the set of drones participating in the game, S is the formed alliance structure, U is the alliance utility, Θ is the switching criterion when the alliance utility changes, and P is the resource allocation matrix of the drones. For fairness, the individual utility of the drone is determined according to the proportion of its resource contribution to the task. The operation of switching from an existing resource structure to another resource structure is defined as and The former indicates the withdrawal of resources, while the latter indicates the entry of resources. Furthermore, the changed alliance structure can be defined as The alliance structure before the change was

[0170] In order to maximize the resource matching degree, the method described in this embodiment uses a switching method based on collective interests, which can be expressed as That is, the alliance structure can be updated as long as the overall benefit after the switch is greater than the overall benefit before the switch.

[0171] In order to improve the allocation efficiency, this implementation adopts a heuristic allocation method based on the resource matching degree. Without loss of generality, assume that the current iteration number is k, and drone n allocates resources to the kth resource of task m.

[0172] Step 4.2.6: The UAVs obtain the task resource gap status through inter-machine communication;

[0173]

[0174] Step 4.2.7: Each drone calculates its matching degree with the target based on its own remaining resource types and quantities, its positional relationship with the target task, and other information. in:

[0175]

[0176] It represents the distance calculation method between two adjacent task points, which can be calculated in a variety of ways. In this implementation, the Euclidean distance is used for calculation;

[0177] Step 4.2.8: Select the task with the highest matching degree in its task sequence according to the roulette rule and assign the current type of resources to the task:

[0178]

[0179] Step 4.2.9: In order to prevent conflicts between UAV tasks, a local sequence re-arrangement method based on task importance is designed, that is, urgent tasks are executed first, and the task sequence of the local machine is updated according to the importance order. n ;

[0180] Step 4.3: Determine whether the current task sequence is valid and meets the task requirements;

[0181] Step 4.3 specifically includes the following steps:

[0182] Step 4.3.1: Check whether the existing allocation results meet the constraints in step 4.2.5. If they meet the constraints, execute steps 4.3.2 to 4.3.3. Otherwise, execute step 4.3.4.

[0183] Step 4.3.2: Check whether the current task sequence exists in the individual history task list Tabu n If it exists, it is considered invalid and step 4.3.4 is executed;

[0184] Step 4.3.3: Check whether the current overall task resource allocation structure exists in the alliance structure history list Tabu S If it exists, it is considered invalid and step 4.3.4 is executed;

[0185] Step 4.3.4: Abandon the current allocation result and jump to step 4.2.6;

[0186] Step 4.4: If the allocation result is valid, add the current allocation result to the individual sequence history list And a list of alliance structures and update the alliance structure;

[0187] Step 4.5: Update the local task sequence Task according to the importance order n , Available Resources Matrix Allocation Matrix Mission Requirements Matrix The UAV alliance Mem(Am ), its corresponding task resources;

[0188] Step 4.6: Determine whether the current drone has a low contribution in the mission alliance it participates in;

[0189] Step 4.6.1: Traverse the UAV's task list and calculate whether the existing resources to perform the task meet the task resource requirements;

[0190] Step 4.6.2: If there is a type of resource that has met the mission resource requirements, then traverse all drones participating in the mission, calculate the resource redundancy of the current drone in this type and compare it with the set exit threshold;

[0191] Step 4.6.3: If the redundancy exceeds the threshold, the contribution of such task resources among all drones executing the current task is calculated. If the resource contribution of the current drone is the smallest, the exit task list is updated;

[0192] Step 4.7: Traverse each mission alliance in which the UAV participates and calculate the total mission resources of the current mission alliance;

[0193] Step 4.7.1: Initialize the resource change matrix to an all-0 matrix;

[0194] Step 4.7.2: Check whether each type of task resource meets the task requirements. If so, randomly withdraw certain resources from the excess resources.

[0195] Step 4.7.3: Check whether the current drone's resources meet the exit resource number. If so, update the resource change.

[0196] Step 4.7.4: Update its own resource allocation matrix;

[0197] Step 4.8: Repeat steps 4.2 to 4.7 until the overall mission benefit does not change, where ε is an arbitrarily small positive number;

[0198] |S k -S k-1 |≤ε

[0199] Step 5: Each UAV performs related tasks in turn according to its respective task sequence and maintains continuous detection of the environment and dynamic update of resources;

[0200] Step 5 specifically includes the following steps:

[0201] Step 5.1: Each UAV extracts the location information and time window information of the target in its respective mission sequence and uses it as the input information of the path planning module;

[0202] Step 5.2: According to the current scenario requirements, select an appropriate real-time local path planning method to meet the requirements of reaching the target while safely avoiding obstacles in the environment;

[0203] Step 6: If the task is completed, the alliance will be disbanded. If not, proceed to step 5.

[0204] Step 7: The drone maintains the current flight status and performs the mission in the area;

[0205] The beneficial effects of the present invention are further illustrated below in conjunction with specific embodiments.

[0206] Assume that there are multiple randomly distributed drones and mission target points in an area of ​​400m×400m. Each drone can carry four consumable resources and two non-consumable resources. The type of resource required for each mission target is equal to the type carried by the drone, and the mission target has a series of attributes. The goal of this embodiment is to minimize resource redundancy while maximizing mission utility. Without loss of generality, it is assumed that the locations of all drones and missions are randomly generated within the mission area, and the number of resources owned and required are randomly generated within a specified range. Table 1 gives the parameter range and displays the statistical data of the experimental results.

[0207] Table 1 Key simulation parameters

[0208] <![CDATA[Maximum number of iterations Iter max > 500 <![CDATA[Task value V m > 30-50 <![CDATA[Task threat level D m > 0.1-0.3 <![CDATA[Time decay factor χ m > 0.1-0.2 <![CDATA[Task priority β m > 0.1-0.5 <![CDATA[Threshold T for exiting the alliance r > 0.2 <![CDATA[Carry resource range C n > {[4,6],[3,6]} <![CDATA[Required resource range L m > {[8,10],[6,8]} <![CDATA[Speed range v n > [10,15]m / s <![CDATA[Unit distance consumption δ0]]> 0.2

[0209] from Figure 3 It can be seen that the statistical results of the total utility of the alliance under multiple experiments show that this method is more effective, with a median value of about 240, which is about 10% higher than that of other methods. In addition, the distribution of the solution of this method does not have any outliers and is relatively stable and concentrated.

[0210] Figure 4 The “UAV” marked in the figure represents the task sequence, and the “S” marked represents the set of drones that perform the task. It can be seen that the drones are assigned to various task points and can meet the requirements of performing the task.

[0211] Figure 5 In the middle column chart, the part above 0% reflects the matching degree between the onboard resources and the task resources, and the part below 0% reflects the resource redundancy rate. It should be noted that in this example, the number of drones is 8 and the number of tasks is 6, which means that the total onboard resources of the drones are less than the required resources for the tasks, so it is difficult to achieve a 100% matching rate. It is worth noting that our method reduces the redundancy rate under the condition of insufficient resources, better allocates them to the corresponding tasks, and improves the overall resource matching degree.

[0212] In summary, the cooperative game-based heuristic task allocation method for UAV systems proposed in the present invention can obtain a task sequence that meets the task requirements. Moreover, compared with the prior art, the method described in this embodiment adopts a resource contribution threshold and a method of random mobilization of remaining resources. Partial reconstruction of the established resource-task matching relationship can reduce resource redundancy, improve resource utilization, and obtain higher overall alliance benefits.

[0213] Although the preferred embodiments of the present disclosure have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present disclosure.

[0214] Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

[0215] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0216] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0217] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0218] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A multi-agent heuristic task allocation method based on cooperative game, characterized in that: The method comprises: S1: Establish a set of neighboring drones based on drone location information; S2: Each drone relies on its own sensors to obtain external information and determine whether a new task has occurred; S3: If a new mission is detected, the UAV that discovered the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting; S4: Each UAV determines the final task allocation plan according to the task allocation algorithm; S5: Each UAV performs related tasks in sequence according to its own obtained task sequence, continuously detects the environment, and dynamically updates resources; S6: If the mission is completed, the alliance is disbanded, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, continue to step S5; The step S4 comprises: S41: Initialize the alliance history structure taboo table, each drone history structure taboo table, allocation matrix, drone task sequence, drone alliance performing corresponding tasks and its corresponding task resources; S42: Each UAV participates in the corresponding task alliance according to its own resources and calculates the benefits of the current task alliance structure; S43: Determine whether the current task sequence allocation result of the UAV is valid and meets the task requirements; S44: if the allocation result is valid, the current allocation result is added to the individual sequence history list and the alliance history structure list; S45: Update the task sequence, available resource matrix, allocation matrix, task requirement matrix, the drone alliance that performs the corresponding task, its corresponding task resources and alliance structure of the local machine according to the importance ranking, and unify the above information through local broadcasting; S46: Determine whether the current drone has a contribution in the mission alliance it participates in; S47: traverse each mission alliance in which the UAV participates, and calculate the total mission resources of the current mission alliance; The step S42 comprises: Calculate the overall benefits of the task alliance structure after optimal allocation; Calculate the relationship between currently allocated resources and resource requirements; Calculate the benefits of drones performing mission sequences; Calculate the cost of the drone to perform the mission; Build an optimization model to optimize allocation with the goal of maximizing overall benefits; Each UAV obtains the task resource gap status through inter-UAV communication; Each drone calculates the degree of matching with the target based on the type and quantity of its remaining resources and the position relationship information with the target task; According to the roulette rule, the task with the highest matching degree is selected from the UAV task sequence and the current type of resources is allocated to the task; Update the local task sequence according to the task importance and obtain the task alliance structure after optimal allocation; The optimization model is constructed to optimize the allocation with the goal of maximizing the overall benefit, including: Build the optimization model: in, The benefits of a coalition formed for a collection of drones performing mission m; By adjusting the alliance structure To maximize the overall benefit, the following constraints need to be met in, For drones To the task The allocation result of the zth type of resources, For the z The available resources of the resource, For the type of resources needed, For drone collection, is the demand quantity for each resource, For the task index, For the task set, For task sequence.

2. According to the cooperative game-based multi-agent heuristic task allocation method of claim 1, it is characterized in that: The step S1 comprises: The drone determines whether other drones are in its neighbor space based on the strength of the received signal. Once the drone is within the communication range of its neighbors, it establishes a two-way communication topology network with the neighbor’s drones and extracts the unique drone ID in the information packet.

3. The multi-agent heuristic task allocation method based on cooperative game according to claim 1, characterized in that: The overall benefit of the task alliance structure after the calculation of the optimal allocation includes: in, The benefit of the alliance formed by the collection of drones performing mission m, is a constant, To control the scope of benefits, is the cost of executing task m, For task resources.

4. The multi-agent heuristic task allocation method based on cooperative game according to claim 1 is characterized in that: The step S5 further comprises: Each UAV extracts the location information and time window information of the target in its respective mission sequence and uses the information as input information for path planning; According to the current scenario requirements, select the appropriate local path planning method to meet the requirements of reaching the target while safely avoiding obstacles in the environment.

5. A multi-agent heuristic task allocation system based on cooperative game, characterized in that: The system comprises: A drone set determination unit, used to establish a neighboring drone set based on the drone position information; The information acquisition unit is used by each drone to obtain external information through its own sensors to determine whether a new task has occurred; The broadcast unit is used to determine if a new mission is detected. The UAV that discovers the mission is marked as the leader and is set as a temporary central node to act as the initiator for broadcasting. A task allocation unit, used for each UAV to determine the final task allocation plan according to the task allocation algorithm; The task execution unit is used for each UAV to perform related tasks in sequence according to the task sequence obtained by each UAV, and continuously detect the environment and dynamically update resources; The judgment unit is used to disband the alliance if the mission is completed, and the drone maintains the current flight status to perform the mission in the area; if the mission is not completed, the mission execution unit continues to execute; The task allocation unit comprises: S41: Initialize the alliance history structure taboo table, each drone history structure taboo table, allocation matrix, drone task sequence, drone alliance performing corresponding tasks and its corresponding task resources; S42: Each UAV participates in the corresponding task alliance according to its own resources and calculates the benefits of the current task alliance structure; S43: Determine whether the current task sequence allocation result of the UAV is valid and meets the task requirements; S44: if the allocation result is valid, the current allocation result is added to the individual sequence history list and the alliance history structure list; S45: Update the task sequence, available resource matrix, allocation matrix, task requirement matrix, the drone alliance that performs the corresponding task, its corresponding task resources and alliance structure of the local machine according to the importance ranking, and unify the above information through local broadcasting; S46: Determine whether the current drone has a contribution in the mission alliance it participates in; S47: traverse each mission alliance in which the UAV participates, and calculate the total mission resources of the current mission alliance; The step S42 comprises: Calculate the overall benefits of the task alliance structure after optimal allocation; Calculate the relationship between currently allocated resources and resource requirements; Calculate the benefits of drones performing mission sequences; Calculate the cost of the drone to perform the mission; Build an optimization model to optimize allocation with the goal of maximizing overall benefits; Each UAV obtains the task resource gap status through inter-UAV communication; Each drone calculates the degree of matching with the target based on the type and quantity of its remaining resources and the position relationship information with the target task; According to the roulette rule, the task with the highest matching degree is selected from the UAV task sequence and the current type of resources is allocated to the task; Update the local task sequence according to the task importance and obtain the task alliance structure after optimal allocation; The optimization model is constructed to optimize the allocation with the goal of maximizing the overall benefit, including: Build the optimization model: in, The benefits of a coalition formed for a collection of drones performing mission m; By adjusting the alliance structure To maximize the overall benefit, the following constraints need to be met in, For drones To the task The allocation result of the zth type of resources, For the z The available resources of the resource, For the type of resources needed, For drone collection, is the demand quantity for each resource, For the task index, For the task set, For task sequence.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program executes the multi-agent heuristic task allocation method based on cooperative game as described in any one of claims 1-4.

7. A computer device, characterized in that: It includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor runs the computer program stored in the memory, the processor executes a multi-agent heuristic task allocation method based on cooperative game according to any one of claims 1-4.

Citation Information

Patent Citations

  • Unmanned aerial vehicle task cooperation method based on overlapping alliance formation game

    CN113395676A

  • Intelligent unmanned cluster layered and distributed task planning decision-making method based on mixed game

    CN116542470A

  • Heterogeneous unmanned aerial vehicle group task allocation method based on alliance formation game

    CN116880540A

  • Dynamic task allocation in an autonomous multi-UAV mission

    US20170131727A1