Multi-unmanned aerial vehicle multi-target collaborative distribution method based on optimized ant colony algorithm

Abstract

The invention relates to a multi-unmanned aerial vehicle multi-target distribution method based on an optimized ant colony algorithm, and belongs to the field of aviation firepower control. The method comprises the steps of constructing a WTA model based on target value and overall damage effect, applying an optimized ant colony algorithm to solution of a multi-unmanned aerial vehicle multi-target distribution model, introducing a local search operator and a pheromone control mechanism, improving algorithm efficiency, reducing the risk of sinking into local optimum, efficiently solving a task distribution strategy, and improving unmanned aerial vehicle task execution efficiency. The method is suitable for unmanned aerial vehicle cluster cooperative task allocation.

Description

Technical Field

[0001] The present invention belongs to the field of aviation fire control, and in particular relates to a multi-UAV multi-target collaborative allocation method based on an optimized ant colony algorithm. Background Art

[0002] With the continuous advancement of UAV technology, multi-UAV systems have shown great potential in performing complex tasks. At the same time, efficient task allocation strategies are crucial to improving the performance of UAV systems. Multi-UAV multi-target collaborative allocation is to allocate multiple targets based on the overall mission requirements, comprehensively consider factors such as UAV performance, battlefield environment, and mission constraints, and determine the mission objectives of each aircraft to complete the mission efficiently. At present, the key to the study of multi-UAV multi-target collaborative allocation lies in the two main contents of problem modeling and algorithm solution. Summary of the invention

[0003] The technical problems to be solved by the present invention are:

[0004] In order to avoid the shortcomings of the prior art, the present invention provides a multi-UAV multi-target collaborative allocation method based on an optimized ant colony algorithm, which is used to achieve efficient allocation of UAV cluster collaborative tasks.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm, characterized by comprising:

[0007] Aiming at the problem of multi-UAV multi-target coordinated allocation, a multi-UAV multi-target allocation model is established, and a total value objective function of target value and overall damage effect is constructed;

[0008] Calculate the probability of damage based on the drone's strike capability, target value, target threat, strike distance cost, and cost of crossing the threat zone; calculate the expected degree of destruction based on the probability of damage and target value;

[0009] An optimized ant colony algorithm is used to solve the multi-UAV multi-objective allocation model. According to the local search and pheromone control mechanism, the optimal solution path is iteratively solved to obtain the optimal solution of the total value objective function and output the optimal allocation result.

[0010] A further technical solution of the present invention: The multi-UAV multi-target allocation model is as follows:

[0011]

[0012] In the above model, objective function (1) is the total value of the target to be destroyed, and constraint (2) indicates that the number of targets that each drone can actually track and attack cannot be greater than 1, and tracking and attacking can only be carried out under the condition that tracking and attacking is feasible;

[0013] Among them, x ij ∈{0,1},1≤i≤W,1≤j≤T,x ij =1 means that UAV i strikes target j, otherwise it does not strike; W is the number of UAVs, T is the number of targets, p ij is the probability of damage, v j represents the value of the jth target.

[0014] A further technical solution of the present invention: the calculation of the damage probability is as follows:

[0015] The value of each factor is limited to between 0 and 1, and the relative distance value is calculated as follows:

[0016]

[0017] in, is the distance cost value of drone i attacking target j, d ij represents the relative distance, D is the diameter of the combat area;

[0018] The calculation method of crossing threat cost is as follows:

[0019]

[0020] in, is the threat zone cost value of drone i attacking target j, k is the threat zone cost coefficient, Indicates the distance that the target-UAV line crosses the threat zone;

[0021] The damage probability is calculated as follows:

[0022]

[0023] Among them, k1, k2, and k3 represent the distance cost weight, threat area cost weight, target threat weight, and UAV strike capability weight, respectively.

[0024] A further technical solution of the present invention is as follows: the expected degree of destruction is calculated based on the damage probability and the target value, using the following formula:

[0025] d ij =v j *P ij (6).

[0026] A further technical solution of the present invention: the local search is specifically:

[0027] After one cycle is completed, the optimal solution of this cycle is obtained. For the current optimal solution, the 3-opt search strategy is adopted. UAV i randomly selects 3 unassigned targets as replacement objects. The probability of selecting target j is The calculation is as follows:

[0028]

[0029] If the expected destructive degree of the selected target calculated is greater than the minimum expected destructive degree of the original UAV i attacking the target, target j is used to replace the corresponding original target to form a new allocation strategy.

[0030] A further technical solution of the present invention: the pheromone control mechanism is specifically:

[0031] When pheromones are too concentrated in the local optimal path, the pheromone standard deviation becomes larger. When the pheromone standard deviation is greater than the tolerable degree, the pheromone is homogenized. At the same time, in order to take into account convergence, the tolerable degree gradually increases with the cycle iteration. After a cycle, for the i-th node, it is determined whether the pheromone standard deviation of the corresponding path exceeds the tolerable degree. If it exceeds, the path is punished, and additional pheromone compensation is given to the path with pheromone concentration lower than the average concentration. The formula for updating the tolerable degree after each cycle is as follows:

[0032] δ(t+1)=δ(t) / δ n (8)

[0033] Among them, δ n is the update coefficient.

[0034] A computer system, characterized in that it comprises: one or more processors, and a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned method.

[0035] A computer-readable storage medium, characterized in that it stores computer-executable instructions, which are used to implement the above method when executed.

[0036] A computer program product, characterized by comprising computer executable instructions, wherein the instructions are used to implement the above method when executed.

[0037] The beneficial effects of the present invention are:

[0038] The invention provides a multi-UAV multi-target collaborative allocation method based on an optimized ant colony algorithm, which comprehensively considers factors such as the UAV strike capability, target value, target threat, strike distance cost, and threat zone crossing cost, constructs a WTA model, applies the optimized ant colony algorithm to solve the multi-UAV multi-target allocation model, introduces a local search operator and a pheromone control mechanism, improves the algorithm efficiency, and reduces the risk of falling into a local optimal solution. It is suitable for the problem of collaborative task allocation of UAV clusters, can solve the problem of multi-UAV task allocation well, and has strong engineering realization value. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The drawings are only for the purpose of illustrating specific embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

[0040] Figure 1 Flow chart of the multi-UAV multi-target collaborative allocation method based on the optimized ant colony algorithm of the present invention.

[0041] Figure 2 Schematic diagram of the multi-UAV multi-target assignment problem.

[0042] Figure 3 Path representation graph.

[0043] Figure 4 Example 1 target allocation result diagram. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0045] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0046] The present invention provides a multi-UAV multi-target collaborative allocation method based on an optimized ant colony algorithm, which adopts the following steps:

[0047] (1) Aiming at the problem of coordinated allocation of multiple UAVs and multiple targets, the WTA mathematical model is established by comprehensively considering factors such as UAV strike capability, target value, target threat, strike distance cost, and cost of crossing the threat area, and constructing a total value objective function of target value and overall damage effect;

[0048] (2) Calculate the damage probability matrix based on the UAV's strike capability, target value, target threat, strike distance cost, and threat zone crossing cost; calculate the expected destructiveness matrix based on the damage probability matrix and target value.

[0049] (3) An optimized ant colony algorithm is used to solve the multi-UAV multi-objective allocation model. According to the local search operator and pheromone control mechanism, the optimal solution path is iteratively solved to obtain the optimal solution of the total value objective function and output the optimal allocation result.

[0050] The basic ant colony algorithm is used to solve the problem. After one cycle, a local search is performed in the neighborhood of the optimal solution obtained in this cycle to improve the current optimal solution. The pheromone is updated according to the improved current optimal solution to accelerate the convergence of the algorithm.

[0051] In the process of cyclic iteration and local search, in order to reduce the situation of falling into local optimality, a pheromone control mechanism is introduced. When pheromones are too concentrated in the local optimal path, the pheromone standard deviation becomes larger. When the pheromone standard deviation is greater than the tolerable level, the pheromone is homogenized. The pheromone is used to adjust the dispersion of the solution in the solution space, so as to control the convergence speed of the solution, effectively jump out of the local optimal point, and obtain the optimal allocation result.

[0052] The above steps are as follows:

[0053] 1. Mathematical model of multi-UAV multi-target collaborative allocation

[0054] Consider the problem of coordinated allocation of multiple targets by multiple drones with different initial states, such as Figure 2 As shown in the figure, the initial state of the drone is different, and the target state is also different. The battlefield threat mainly considers the no-fly zone. Considering the drone's strike capability, target value, target threat, strike distance cost, and cost of crossing the threat zone, target allocation is performed to determine which drones attack which targets, so as to achieve the purpose of improving the overall attack effectiveness and minimizing the attack cost.

[0055] The mathematical model of this problem is as follows:

[0056]

[0057] Among them, x ij∈{0,1},1≤i≤W,1≤j≤T. The description of the model parameters is shown in Table 1. In this model, the objective function (1) is the total value of the target to be destroyed, and the constraint (2) indicates that the number of targets that each drone can actually track and attack cannot be greater than 1, and tracking and attacking can only be carried out under the condition that tracking and attacking is feasible.

[0058] Table 1 Description of WTA mathematical model parameters

[0059]

[0060] (II) Calculation of damage probability matrix and expected damage matrix

[0061] The damage probability in the objective function of the multi-UAV multi-target collaborative allocation mathematical model is calculated as follows, taking into account factors such as the UAV strike capability, the length of the threat zone path that the UAV-target line passes through, the relative distance between the UAV and the target, and the target threat level to calculate the damage probability. The value of each factor is limited to between 0 and 1, and the relative distance value is calculated as follows:

[0062]

[0063] in, is the distance cost value of drone i attacking target j, d ij represents the relative distance, and D is the diameter of the combat area.

[0064] The calculation method of crossing threat cost is as follows:

[0065]

[0066] in is the threat zone cost value of drone i attacking target j, k is the threat zone cost coefficient, which is between 0.1 and 0.25. Indicates the distance that the target-drone line crosses the threat zone.

[0067] The damage probability is calculated as follows:

[0068]

[0069] Among them, k1, k2, and k3 represent the distance cost weight, threat area cost weight, target threat weight, and drone strike capability weight, respectively, and their values ​​are between 0 and 1.

[0070] Expected destruction degree of the target d ij The calculation is as follows:

[0071] d ij =v j *P ij (14)

[0072] (III) Optimizing the ant colony algorithm solution

[0073] 1. Representation of the solution and path of the ant colony algorithm for the task allocation model

[0074] According to the mathematical model of the above multi-UAV multi-target collaborative allocation problem, the optimized ant colony algorithm is applied to solve the model.

[0075] In order to optimize the ant colony algorithm solution, the solution of the multi-UAV multi-objective collaborative allocation problem model is expressed in list form, as shown in Table 2.

[0076] Table 2 Solution representation

[0077]

[0078] Note: In the table Indicates the number of drones that hit W The target number is j, T W Indicates the actual number of targets hit by UAV k.

[0079] The path is represented as Figure 3 As shown. Figure 3 It can be seen that a path contains W+1 nodes in total, and there are multiple path selection methods between every two nodes. The ant starts from a node, selects different attack targets according to certain rules, and then reaches the next node. Define the subpath from node i to a target j of drone i as j i .

[0080] 2. Path selection rules

[0081] In order to reduce the computational complexity of the algorithm in dealing with infeasible solutions, constraints are used in the solution generation process to eliminate the interference of infeasible solutions, ensure that the generated solutions are always within the feasible domain, avoid operations on infeasible solutions, and thus improve the efficiency of the algorithm. In the iterative search process, constraints are used to ensure the feasibility of the solution. At the same time, in order to reduce the blindness of the algorithm, the heuristic construction method is used instead of the random construction method to improve the reliability of the initial solution and accelerate the convergence speed. For the WTA problem of W drones and T targets, when the i-th drone selects a specific attack target, it randomly selects the target as the attack object based on the target's expected destructiveness and the amount of path pheromone. The expected destructiveness of drone i to target j is shown in formula (6).

[0082] The probability q that drone i chooses target j as the target to be attacked ij The expression is as follows:

[0083]

[0084] in, is subpath j i , α is the information heuristic factor, β is the expected heuristic factor, usually α=1, β=0.5. From formula (7), we can see that when the attack feasibility f ij =0, the probability of hitting the target is also 0, ensuring that the unhittable target will not be hit, thus ensuring the feasibility of the solution.

[0085] 3. Pheromone computing

[0086] The information index update rule adopts the optimal ant update strategy commonly used in the ant colony algorithm: after one cycle is completed, the pheromone is updated according to the current optimal solution and optimal path, as shown in formula (8) and formula (9):

[0087]

[0088] in, is subpath j i The concentration of pheromone, ρ is the volatility coefficient of pheromone, and the value is ρ = 0.05. best is the objective function value of the optimal ant, and ∧ is the inspiration factor.

[0089] 4. Local search accelerates convergence

[0090] After one cycle is completed, the optimal solution of this cycle is obtained. For the current optimal solution, the 3-opt search strategy is adopted. UAV i randomly selects 3 unassigned targets as replacement objects. The probability of selecting target j is The calculation is as follows:

[0091]

[0092] If the expected destructive degree of the selected target calculated is greater than the minimum expected destructive degree of the original UAV i attacking the target, target j is used to replace the corresponding original target to form a new allocation strategy.

[0093] 5. Pheromone control mechanism

[0094] In the process of loop iteration and local search, in order to reduce the situation of falling into the local optimum, the pheromone control mechanism is introduced. When the pheromone is too concentrated in the local optimal path, the pheromone standard deviation becomes larger. When the pheromone standard deviation is greater than the tolerable degree, the pheromone is homogenized. When a cycle is completed, for the i-th node, it is determined whether the pheromone standard deviation of the corresponding path exceeds the tolerable degree. If it exceeds, the path is punished, and the path with pheromone concentration lower than the average concentration is compensated with additional pheromone. The formula for updating the tolerable degree after each cycle is as follows:

[0095] δ(t+1)=δ(t) / δn (19)

[0096] Among them, δ n is the update coefficient.

[0097] The effects of this application are described in detail below in conjunction with simulation.

[0098] Example 1: Target allocation problem for 16 drones with 4 targets. Target allocation is performed. The initial position, speed, initial position of the target and the position of the threat zone of the drone are shown in Table 3, Table 4 and Table 5 respectively. The target allocation results are shown in Figure 3 .

[0099] Table 3 UAV initial position and speed information

[0100] Drone Number Position coordinates / m Speed ​​m / s Drone Number Position coordinates / m Speed ​​m / s 1 (1000,2000) 30 9 (5300,650) 30 2 (1300,8000) 30 10 (750,3700) 30 3 (4000,1000) 30 11 (1600,850) 30 4 (8400,1500) 30 12 (12500,1250) 30 5 (2000,1000) 30 13 (1500,1900) 30 6 (400,6500) 30 14 (1450,700) 30 7 (7800,900) 30 15 (2100,1300) 30 8 (1600,2100) 30 16 (12000,1200) 30

[0101] Table 4 Target initial position and velocity information

[0102] Target Number Position coordinates / m Speed ​​m / s Target Number Position coordinates / m Speed ​​m / s 1 (15000,17500) 0 3 (7000,13000) 0 2 (13500,6000) 0 4 (8540,15000) 0

[0103] Table 5 Threat zone information

[0104] Threat Zone Number Center coordinates / m Radius / m Threat Zone Number Center coordinates / m Radius / m 1 (4000,10000) 1500 7 (15000,10000) 1000 2 (12000,14000) 1500 8 (5000,16000) 1500 3 (8000,4000) 1500 9 (16000,4000) 1500 4 (4000,4000) 500 10 (11000,3500) 1000 5 (7000,11000) 1000 11 (16000,13500) 1200 6 (10000,8000) 1800 12 (10000,17500) 1000

[0105] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should be included in the protection scope of the present invention.

Claims

1. A multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm, characterized in that: include: Aiming at the problem of multi-UAV multi-target coordinated allocation, a multi-UAV multi-target allocation model is established, and a total value objective function of target value and overall damage effect is constructed; Calculate the probability of damage based on the drone's strike capability, target value, target threat, strike distance cost, and cost of crossing the threat zone; calculate the expected degree of destruction based on the probability of damage and target value; An optimized ant colony algorithm is used to solve the multi-UAV multi-objective allocation model. According to the local search and pheromone control mechanism, the optimal solution path is iteratively solved to obtain the optimal solution of the total value objective function and output the optimal allocation result.

2. According to claim 1, a multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm is characterized in that: The multi-UAV multi-target allocation model is as follows: In the above model, objective function (1) is the total value of the target to be destroyed, and constraint (2) indicates that the number of targets that each drone can actually track and attack cannot be greater than 1, and tracking and attacking can only be carried out under the condition that tracking and attacking is feasible; Among them, x ij ∈{0,1},1≤i≤W,1≤j≤T,x ij =1 means that UAV i strikes target j, otherwise it does not strike; W is the number of UAVs, T is the number of targets, p ij is the probability of damage, v j represents the value of the jth target.

3. According to claim 2, a multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm is characterized in that: The calculation of the damage probability is as follows: The value of each factor is limited to between 0 and 1, and the relative distance value is calculated as follows: in, is the distance cost value of drone i attacking target j, d ij represents the relative distance, D is the diameter of the combat area; The calculation method of crossing threat cost is as follows: in, is the threat zone cost value of drone i attacking target j, k is the threat zone cost coefficient, Indicates the distance that the target-UAV line crosses the threat zone; The damage probability is calculated as follows: Among them, k1, k2, and k3 represent the distance cost weight, threat area cost weight, target threat weight, and UAV strike capability weight, respectively.

4. According to claim 3, a multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm is characterized in that: The expected degree of destruction is calculated based on the damage probability and the target value using the following formula: d ij =v j *P ij (6)。 5. According to claim 1, a multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm is characterized in that: The local search is specifically: After one cycle is completed, the optimal solution of this cycle is obtained. For the current optimal solution, the 3-opt search strategy is adopted. UAV i randomly selects 3 unassigned targets as replacement objects. The probability of selecting target j is The calculation is as follows: If the expected destructive degree of the selected target calculated is greater than the minimum expected destructive degree of the original UAV i attacking the target, target j is used to replace the corresponding original target to form a new allocation strategy.

6. According to claim 1, a multi-UAV multi-target collaborative allocation method based on optimized ant colony algorithm is characterized in that: The pheromone control mechanism is specifically: When pheromones are too concentrated in the local optimal path, the pheromone standard deviation becomes larger. When the pheromone standard deviation is greater than the tolerable value, the pheromone is homogenized. At the same time, in order to take into account convergence, the tolerable value gradually increases with the loop iteration. When a loop ends, for the i-th node, it is determined whether the pheromone standard deviation of the corresponding path exceeds the tolerable value. If it exceeds, the path will be penalized, and additional pheromone compensation will be given to the path where the pheromone concentration is lower than the average concentration; the formula for updating the tolerance after each cycle is as follows: δ(t+1)=δ(t) / δ n (8) Among them, δ n is the update coefficient.

7. A computer system, characterized in that include: One or more processors, and a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method of claim 1.

8. A computer-readable storage medium, characterized in that Computer executable instructions are stored, and when the instructions are executed, they are used to implement the method of claim 1.

9. A computer program product, characterized in that The invention comprises computer executable instructions, which are used to implement the method of claim 1 when being executed.

Images