Unmanned aerial vehicle cluster cooperative combat method and system

Abstract

The invention discloses an unmanned aerial vehicle cluster cooperative combat method and system, and relates to the technical field of unmanned aerial vehicles. The method specifically comprises the steps that a plurality of unmanned aerial vehicles establish an unmanned aerial vehicle cluster, initialization is carried out through a preset communication protocol, and a communication network in the unmanned aerial vehicle cluster is established; allocating tasks to each unmanned aerial vehicle by using a task allocation algorithm based on artificial intelligence; a distributed path planning algorithm is adopted to plan an optimal flight path for each unmanned aerial vehicle, collision is avoided, and task execution efficiency is optimized; executing a task by applying a collaborative mechanism, and dynamically adjusting a self flight path and a task execution strategy to realize collaborative combat; and learning is carried out through real-time data and historical data, the cooperative combat strategy of the unmanned aerial vehicle cluster is optimized, and the capability of coping with a complex environment is improved. According to the method, intelligent distribution, path planning and communication coordination technologies are utilized, the combat effect of the unmanned aerial vehicle cluster is optimized, and the method has real-time performance, adaptability and high efficiency and can effectively deal with various challenges in a complex battlefield environment.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles (UAVs), and in particular to a UAV cluster collaborative combat method and system. Background Art

[0002] With the rapid development of drone technology, drone swarm collaborative operations have become an important research direction in the military and civilian fields, and have a wide range of applications, including reconnaissance, strike and electronic warfare; in the civilian field, drone swarms can be used for disaster response, environmental monitoring, logistics distribution, etc., greatly improving the efficiency and safety of mission execution. In the collaborative operations of existing drone swarms, due to the lack of an effective coordination mechanism, the task allocation and execution between drones are often not efficient enough, resulting in limited overall combat effectiveness; the resource allocation between different drones is often not fair, resulting in some drones being overloaded, while other drones have idle resources, resulting in uneven resource allocation; in addition, when drones perform tasks, conflicts such as path intersection and resource competition may occur, affecting the smooth progress of the mission.

[0003] In order to effectively improve the combat capability and flexibility of the cluster, it is necessary to develop a new method of UAV cluster collaborative combat. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for coordinated combat of drone swarms to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for coordinated combat of drone clusters, specifically comprising:

[0006] Step S1, multiple drones form a drone cluster and are initialized through a preset communication protocol to establish a communication network within the drone cluster;

[0007] Step S2, using an artificial intelligence-based task allocation algorithm to assign tasks to each drone;

[0008] Step S3, using a distributed path planning algorithm to plan the optimal flight path for each drone to avoid collisions and optimize mission execution efficiency;

[0009] Step S4, applying the coordination mechanism to perform the task, dynamically adjusting its own flight path and task execution strategy, achieving coordinated operations, and ensuring the coordination and consistency of the UAVs when performing the task;

[0010] Step S5, through learning real-time data and historical data, optimize the coordinated combat strategy of the drone cluster and improve the ability to cope with complex environments.

[0011] Furthermore, the artificial intelligence-based task allocation algorithm in step S2 adopts an ant colony algorithm, which is further as follows:

[0012] S21, clarify the mission and number of drones, and initialize pheromones;

[0013] S22, the drone selects the task to be performed based on the pheromone and heuristic information, and constructs the solution. The specific formula is:

[0014]

[0015] Among them, T ij (t) is the pheromone value between task i and drone j at time t, η ij is the heuristic information between task i and drone j, α and β are parameters that control the relative importance of pheromone and heuristic information;

[0016] S23. According to the set evaporation rate ρ, reduce the pheromone on the path. The calculation formula is:

[0017] T ij (t+1)=(1―ρ)·T ij (t);

[0018] The calculation formula for the de-enhancement pheromone constructed by the drone is:

[0019]

[0020] in, is the contribution of the kth UAV to the pheromone between task i and UAV j at time t, Defined as:

[0021]

[0022] Where Q is a constant, L k is the objective function value of the solution constructed by the kth UAV;

[0023] S24. The objective function value of the evaluation solution is selected to iterate the optimal solution. After the pheromone distribution becomes stable, the iteration is terminated and the final optimal solution is output to complete the task allocation of the drone cluster.

[0024] Further, step S3 adopts A * The algorithm performs path planning, further as follows:

[0025] S31, initialize the node and determine the starting point and end point;

[0026] S32. Use the Euclidean distance as the heuristic function to calculate the straight-line distance between the target node and the current node. The calculation formula is:

[0027]

[0028] Among them, (x g ,y g ,z g ) is the coordinate of the target node, (x n ,y n ,z n ) is the coordinate of the current node;

[0029] S33, calculate the cost and obtain the path cost, the calculation formula is:

[0030] f(n)=g(n)+h(n);

[0031] Among them, g(n) is the actual cost from the starting point to the current node;

[0032] S34, if n is the target node, the path planning is completed and the path is constructed; if n is not the target node, the neighbor nodes of n are expanded, and the cost is calculated for each neighbor node m, and the formula is:

[0033] g(m)=g(n)+d(n,m);

[0034] f(m)=g(m)+h(m);

[0035] Among them, d(n,m) is the edge weight; if neighbor m has not been explored or a better path is found, its cost is updated and put into the priority queue to realize the path planning of the drone cluster.

[0036] Furthermore, step S4 uses game theory to conduct coordinated combat of drone clusters, specifically:

[0037] S41, select Shapley value as the cooperative game model and conduct modeling;

[0038] S42. Calculate the Shapley value φ of each UAV i , get the contribution of the drone, the formula is:

[0039]

[0040] Among them, v is the characteristic function, N is the set of all drones, and n is the number of drones;

[0041] S43. Based on the real-time information and the results of Shapley value, the unmanned swarm makes real-time adjustments to the path, flight speed and mission execution strategy for coordinated operations.

[0042] Furthermore, step S5 is further as follows: analyzing the Shapley value results to determine the contribution and benefit distribution of each drone; and adopting the Q-learning learning method and the coordinated combat strategy of the drone cluster.

[0043] The present invention also provides a UAV cluster cooperative combat system, comprising:

[0044] Drone swarm module, including multiple drones;

[0045] Communication network module, used to realize real-time information exchange and sharing between drones;

[0046] The cooperative game module uses the Shapley value cooperative game model to calculate the contribution and benefit distribution of each drone;

[0047] The path planning module dynamically adjusts the path of the drone based on real-time information and cooperative game results;

[0048] The mission execution module dynamically adjusts the UAV's mission execution strategy based on the results of the cooperative game;

[0049] Learning mechanism module, through machine learning methods, the drone learns from historical data and experience to improve its decision-making model and strategy selection.

[0050] Beneficial effects of the present invention: This scheme is initialized through a preset communication protocol to ensure the stability and efficiency of the internal communication network of the drone cluster, providing a solid foundation for subsequent task allocation and collaborative operations, and utilizing an artificial intelligence-based task allocation algorithm to intelligently allocate tasks to each drone according to the drone's performance, task requirements and real-time status, thereby improving the efficiency and flexibility of task execution; applying a collaborative mechanism to execute tasks, dynamically adjusting its own flight path and task execution strategy, achieving collaborative operations between drones, ensuring coordination and consistency when executing tasks, and improving overall combat effectiveness; through strategy optimization, identifying conflicts and resolving them in a timely manner, ensuring that the strategy of each drone is fairly executed; learning through real-time data and historical data, continuously optimizing the collaborative combat strategy of the drone cluster, improving the ability to cope with complex environments, and enabling the drone cluster to better adapt to battlefield changes and mission requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is the overall flow chart of the present invention.

[0052] Figure 2 It is a flow chart of step S2 of the present invention.

[0053] Figure 3 This is a flow chart of step S3 of the present invention. DETAILED DESCRIPTION

[0054] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0055] Reference Figure 1 , the present invention provides a UAV cooperative combat method, the specific steps are as follows: step S1, multiple UAVs form a UAV cluster, and initialize through a preset communication protocol to establish a communication network within the UAV cluster;

[0056] Step S2, using an artificial intelligence-based task allocation algorithm to assign tasks to each drone;

[0057] Step S3, using a distributed path planning algorithm to plan the optimal flight path for each drone to avoid collisions and optimize mission execution efficiency;

[0058] Step S4, applying the coordination mechanism to perform the task, dynamically adjusting its own flight path and task execution strategy, achieving coordinated operations, and ensuring the coordination and consistency of the UAVs when performing the task;

[0059] Step S5, through learning real-time data and historical data, optimize the coordinated combat strategy of the drone cluster and improve the ability to cope with complex environments.

[0060] The UAV swarm collaborative combat method of the present invention realizes efficient collaborative combat of UAV swarms through intelligent task allocation, dynamic path planning and real-time communication collaboration. Compared with the traditional single-machine combat mode, this method optimizes the combat effect of UAV swarms by using advanced intelligent allocation, path planning and communication coordination technologies, which can significantly improve the task execution efficiency, flexibility and combat success rate of the swarm, while reducing the risk of mission failure.

[0061] In a further embodiment, the artificial intelligence-based task allocation algorithm adopts an ant colony algorithm, which is further:

[0062] S21, clarify the mission and number of drones, and initialize pheromones;

[0063] S22, the drone selects the task to be performed based on the pheromone and heuristic information, and constructs the solution. The specific formula is:

[0064]

[0065] Among them, T ij (t) is the pheromone value between task i and drone j at time t, η ijis the heuristic information between task i and drone j, α and β are parameters that control the relative importance of pheromone and heuristic information;

[0066] S23. According to the set evaporation rate ρ, reduce the pheromone on the path. The calculation formula is:

[0067] T ij (t+1)=(1―ρ)·T ij (t);

[0068] The calculation formula for the de-enhancement pheromone constructed by the drone is:

[0069]

[0070] in, is the contribution of the kth UAV to the pheromone between task i and UAV j at time t, Defined as:

[0071]

[0072] Where Q is a constant, L k is the objective function value of the solution constructed by the kth UAV;

[0073] S24. The objective function value of the evaluation solution is selected to iterate the optimal solution. After the pheromone distribution becomes stable, the iteration is terminated and the final optimal solution is output to complete the task allocation of the drone cluster.

[0074] In a further embodiment, using A * The algorithm performs path planning, further as follows:

[0075] S31, initialize the node and determine the starting point and end point;

[0076] S32. Use the Euclidean distance as the heuristic function to calculate the straight-line distance between the target node and the current node. The calculation formula is:

[0077]

[0078] Among them, (x g ,y g ,z g ) is the coordinate of the target node, (x n ,y n ,z n ) is the coordinate of the current node;

[0079] S33, calculate the cost and obtain the path cost, the calculation formula is:

[0080] f(n)=g(n)+h(n);

[0081] Among them, g(n) is the actual cost from the starting point to the current node;

[0082] S34, if n is the target node, the path planning is completed and the path is constructed; if n is not the target node, the neighbor nodes of n are expanded, and the cost is calculated for each neighbor node m, and the formula is:

[0083] g(m)=g(n)+d(n,m);

[0084] f(m)=g(m)+h(m);

[0085] Among them, d(n,m) is the edge weight; if neighbor m has not been explored or a better path is found, its cost is updated and put into the priority queue to realize the path planning of the drone cluster.

[0086] In this embodiment, A * The algorithm performs path planning and obtains the straight-line distance between the coordinates of the target node and the coordinates of the current node according to the Euclidean distance, minimizing the error to the greatest extent, so as to realize the distributed path planning of the UAV in three-dimensional space, avoid collisions and optimize the execution efficiency of the task.

[0087] In a further embodiment, game theory is used to conduct coordinated operations of drone clusters, specifically:

[0088] S41, select Shapley value as the cooperative game model and conduct modeling;

[0089] S42. Calculate the Shapley value φ of each UAV i , get the contribution of the drone, the formula is:

[0090]

[0091] Among them, v is the characteristic function, N is the set of all drones, and n is the number of drones;

[0092] S43. Based on the real-time information and the results of the Shapley value, the unmanned swarm makes real-time adjustments to the path, flight speed and mission execution strategy to coordinate operations, including:

[0093] Collision detection, the formula is:

[0094]

[0095] Among them, COllision(p i ,p j ) is collision detection;

[0096] Path planning, the formula is:

[0097] Replan(p i )=A * (s i ,g i ,O);

[0098] Task redistribution, the formula is:

[0099] Reassign(u i ,t j )=Assign(u i ,t j )>CurrentPriority(u i );

[0100] Task priority adjustment, the formula is:

[0101] AdjustPriority(t i )=Priority(t i )×Factor(t i );

[0102] For each drone u i , receive information from other drones, check whether the path or mission needs to be adjusted based on the received information, and if the path needs to be adjusted, perform path replanning Replan (p i ), if you need to adjust the task, execute task redistribution Reassign(u i ,y j ) is used to realize the dynamic path adjustment and mission execution strategy of UAVs in collaborative operations, ensuring the coordination and consistency of UAVs when performing tasks.

[0103] In a further embodiment, the contribution and benefit distribution of each drone are determined by analyzing the Shapley value results; the cooperative combat strategy of the drone cluster is specifically as follows using the Q-learning method:

[0104] S51, analyze the Shapley value result, and select action a according to the current state s and Q table Q(s,a);

[0105] Execute the selected action s and observe the next state s′ and reward r;

[0106] Update the Q table Q(s,a), the formula is:

[0107] Q(a,a)←Q(s,a)+α[r+γmax a′ Q(a′,s′) ― Q(s,a)];

[0108] Among them, α is the learning rate and γ is the discount factor;

[0109] S52. According to the Q table Q(s,a), select the optimal action a * , the formula is:

[0110] a * = argmax a Q(s,a);

[0111] According to the optimal action a of all drones * , formulate collaborative combat strategies, and the collaborative optimization formula is:

[0112]

[0113] S53, real-time detection and identification of potential path intersections and resource contention conflicts C ij , through collaborative optimization strategy to resolve conflicts and ensure that the strategy of each drone is implemented fairly. The specific formula is:

[0114] ConflictResolution=ResolveConflict(C ij ,CooperativeStrategy);

[0115] S54, evaluate the performance of the UAV cluster in terms of task completion, coordination efficiency, and safety, etc. The formula is:

[0116] P=EvaluatePerformance(CooperativeStrategy);

[0117] Adjust the Q table parameters according to the performance evaluation results to improve the model performance. The feedback learning formula is:

[0118] Q(s,a)=FeedbackLearning(P,Q(s,a));

[0119] The above process is iterated cyclically, and the drones continuously adjust their coordination strategies based on real-time information and mission requirements, and improve the coordination capabilities of the drone cluster through continuous feedback and learning.

[0120] The present invention also provides a UAV cluster cooperative combat system, comprising:

[0121] The drone cluster module includes multiple drones; the communication network module is used to realize real-time information exchange and sharing between drones; the cooperative game module adopts the Shapley value cooperative game model to calculate the contribution and benefit distribution of each drone; the path planning module dynamically adjusts the path of the drone according to real-time information and cooperative game results; the task execution module dynamically adjusts the task execution strategy of the drone according to the cooperative game results; the learning mechanism module uses machine learning methods to enable the drone to learn from historical data and experience and improve its decision-making model and strategy selection.

[0122] The cooperative game-based UAV swarm collaborative combat system provided by the solution of the present invention can realize efficient collaboration, intelligent decision-making and dynamic adjustment of UAV swarms. Through the cooperative game model, it ensures that the contribution of each UAV is fairly rewarded, and improves the efficiency and safety of task execution. At the same time, through machine learning methods, the performance and adaptability of the system are continuously improved.

[0123] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes in form and details may be made without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A drone swarm collaborative combat method, characterized in that: Specifically include: Step S1, multiple drones form a drone cluster and are initialized through a preset communication protocol to establish a communication network within the drone cluster; Step S2, using an artificial intelligence-based task allocation algorithm to assign tasks to each drone; Step S3, using a distributed path planning algorithm to plan the optimal flight path for each drone to avoid collisions and optimize mission execution efficiency; Step S4, applying the coordination mechanism to perform the task, dynamically adjusting its own flight path and task execution strategy, achieving coordinated operations, and ensuring the coordination and consistency of the UAVs when performing the task; Step S5, through learning real-time data and historical data, optimize the coordinated combat strategy of the drone cluster and improve the ability to cope with complex environments.

2. The UAV swarm collaborative combat method according to claim 1, characterized in that: The artificial intelligence-based task allocation algorithm in step S2 adopts an ant colony algorithm, which is further as follows: S21, clarify the mission and number of drones, and initialize pheromones; S22, the drone selects the task to be performed based on the pheromone and heuristic information, and constructs the solution. The specific formula is: Among them, T ij (t) is the pheromone value between task i and drone j at time t, η ij is the heuristic information between task i and drone j, α and β are parameters that control the relative importance of pheromone and heuristic information; S23. According to the set evaporation rate ρ, reduce the pheromone on the path. The calculation formula is: T ij (t+1)=(1―ρ)·T ij (t); The calculation formula for the de-enhancement pheromone constructed by the drone is: in, is the contribution of the kth UAV to the pheromone between task i and UAV j at time t, Defined as: Where Q is a constant, L k is the objective function value of the solution constructed by the kth UAV; S24. The objective function value of the evaluation solution is selected to iterate the optimal solution. After the pheromone distribution becomes stable, the iteration is terminated and the final optimal solution is output to complete the task allocation of the drone cluster.

3. The UAV swarm collaborative combat method according to claim 1, characterized in that: Step S3 uses A * The algorithm performs path planning, further as follows: S31, initialize the node and determine the starting point and end point; S32. Use the Euclidean distance as the heuristic function to calculate the straight-line distance between the target node and the current node. The calculation formula is: Among them, (x g ,y g ,z g ) is the coordinate of the target node, (x n ,y n ,z n ) are the coordinates of the current node; S33, calculate the cost and obtain the path cost, the calculation formula is: f(n)=g(n)+h(n); Among them, g(n) is the actual cost from the starting point to the current node; S34, if n is the target node, the path planning is completed and the path is constructed; if n is not the target node, the neighbor nodes of n are expanded, and the cost is calculated for each neighbor node m, and the formula is: g(m)=g(n)+d(n,m); f(m)=g(m)+h(m); Among them, d(n,m) is the edge weight; if neighbor m has not been explored or a better path is found, its cost is updated and put into the priority queue to realize the path planning of the drone cluster.

4. The UAV swarm collaborative combat method according to claim 1, characterized in that: Step S4 uses game theory to conduct coordinated combat of drone clusters, specifically: S41, select Shapley value as the cooperative game model and conduct modeling; S42. Calculate the Shapley value φ of each drone i , get the contribution of the drone, the formula is: Among them, v is the characteristic function, N is the set of all drones, and n is the number of drones; S43. Based on the real-time information and the results of Shapley value, the unmanned swarm makes real-time adjustments to the path, flight speed and mission execution strategy for coordinated operations.

5. The UAV swarm collaborative combat method according to claim 1, characterized in that: Step S5 is further as follows: analyzing the Shapley value results to determine the contribution and benefit distribution of each drone; and adopting the Q-learning method to develop a collaborative combat strategy for the drone cluster.

6. A drone swarm collaborative combat system, characterized in that: include: Drone swarm module, including multiple drones; Communication network module, used to realize real-time information exchange and sharing between drones; The cooperative game module uses the Shapley value cooperative game model to calculate the contribution and benefit distribution of each drone; The path planning module dynamically adjusts the path of the drone based on real-time information and cooperative game results; The mission execution module dynamically adjusts the UAV's mission execution strategy based on the results of the cooperative game; Learning mechanism module, through machine learning methods, the drone learns from historical data and experience to improve its decision-making model and strategy selection.

Images