Unmanned aerial vehicle cluster task allocation method and device based on ant lion optimization and medium

Abstract

The embodiment of the invention discloses an unmanned aerial vehicle cluster task allocation method and device based on ant lion optimization and a medium. The flexibility and comprehensiveness of unmanned aerial vehicle cluster task allocation can be improved. According to the method, an ant lion optimization algorithm is adopted to solve a task allocation result, so that model solving is realized through a multi-objective optimization strategy, and a random boundary strategy is applied to the ant lion optimization algorithm through a first formula, so that the diversity of the algorithm is increased, the global search capability of the algorithm is improved, and the diversity of solutions is ensured.

Description

Technical Field

[0001] The present application relates to the field of drone technology, and in particular to a method, device and medium for allocating drone cluster tasks based on antlion optimization. Background Art

[0002] As drones evolve from single platforms to cluster applications, the task allocation problem of drone clusters has become a major research focus. Current research on the task allocation of drone clusters mainly uses multiple indicators to evaluate the optimization results, that is, it faces a multi-objective optimization problem when solving the model.

[0003] However, current research on task allocation for drone swarms often transforms multi-objective optimization problems into single-objective optimization problems by weighting them in terms of model solving, resulting in an incomplete analysis of the problem. The task allocation results obtained by solving the problem will be affected by the subjective influence of the weights, which is not conducive to comprehensive decision-making, and thus leads to a lack of flexibility and comprehensiveness in the task allocation of drone swarms. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a method, device and medium for allocating tasks of drone clusters based on antlion optimization, which can improve the flexibility and comprehensiveness of drone cluster task allocation.

[0005] In a first aspect, an embodiment of the present application provides a method for allocating tasks of a drone cluster based on antlion optimization, the method comprising: Randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to represent the task allocation scheme of the drone cluster; According to the plurality of optimization objectives, an antlion optimization algorithm is executed to determine a Pareto solution set corresponding to the N ants, and in the process of executing the antlion optimization algorithm, a boundary around which the ants wander around the antlion in each iteration is determined according to the first formula; According to the Pareto solution set, determine the task allocation result of the drone cluster; The first formula is expressed as follows:

[0006] in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0007] A second aspect of an embodiment of the present application provides a drone cluster task allocation device based on antlion optimization, the device comprising: An initialization module is used to randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to characterize the task allocation scheme of the drone cluster; An algorithm execution module is used to execute the antlion optimization algorithm to determine the Pareto solution set corresponding to the N ants according to multiple optimization objectives, and in the process of executing the antlion optimization algorithm, determine the boundary of the ants walking around the antlion in each iteration according to the first formula; The result determination module is used to determine the task allocation result of the drone cluster according to the Pareto solution set; The first formula is expressed as follows:

[0008] in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0009] A third aspect of an embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the drone cluster task allocation method based on antlion optimization as described in the first aspect.

[0010] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of the drone cluster task allocation method based on antlion optimization as described in the first aspect are implemented.

[0011] In a fifth aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for allocating drone swarm tasks based on antlion optimization as described in the first aspect are implemented.

[0012] It can be seen from the above technical solution that compared with the single-objective optimization method based on weight strategy in traditional drone cluster task allocation, the present application adopts the ant lion optimization algorithm to solve the task allocation result, thereby realizing model solving with a multi-objective optimization strategy, so as to obtain a set of suboptimal solutions (i.e., Pareto solution set) that balance multiple optimization objectives. The distribution of the solution set can provide a reference for problem analysis and improve the comprehensiveness of problem analysis. The solution set can also provide a variety of solutions for flexible selection, which is conducive to comprehensive decision-making. In this way, the flexibility and comprehensiveness of drone cluster task allocation can be effectively improved; and the present application also designs a first formula to apply the random boundary strategy to the ant lion optimization algorithm, thereby increasing the diversity of the algorithm, improving the global search capability of the algorithm, and ensuring the diversity of solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 An implementation flow chart of a method for allocating tasks of drone swarms based on antlion optimization provided in an embodiment of the present application; Figure 2 A schematic diagram of a preference elite selection mechanism provided in an embodiment of the present application; Figure 3 A schematic diagram of mapping an algorithm code and a task allocation result provided in an embodiment of the present application; Figure 4 A schematic diagram of the structure of a drone cluster task allocation device based on antlion optimization provided in an embodiment of the present application; Figure 5 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0016] The rapid development of drone technology has attracted widespread attention around the world. As a technology with great potential, drones have shown great application prospects in the military, civilian and commercial fields.

[0017] With the advancement of technology and the reduction of costs, drones are gradually developing from single platforms to clustered applications. Among them, a clustered drone system refers to a system in which multiple drones work together to complete complex tasks by sharing information and assigning tasks.

[0018] Clustered UAV systems face a key problem, namely how to reasonably allocate tasks to achieve the best system performance. Current research on task allocation for UAV clusters mainly uses multiple indicators to evaluate the optimization results, that is, it faces a multi-objective optimization problem when solving the model.

[0019] However, in the current research work on the task allocation of drone swarms, in terms of model solving, multi-objective optimization problems are often converted into single-objective optimization problems by weighting based on weight strategies. It should be noted that the weight strategy refers to setting weights for multiple optimization objectives and converting them into single-objective optimization problems for solving. Although this method can simplify the solution and distinguish different optimization objectives by setting weights, unreasonable weight settings may cover up some optimization objectives and can only give a unique solution, which has poor flexibility. Therefore, the single-objective optimization method based on the weight strategy in the traditional drone swarm task allocation often does not analyze the problem comprehensively enough, and the task allocation results obtained by solving the problem will be subject to the subjective influence of the weight, which is not conducive to comprehensive decision-making, and thus leads to the lack of flexibility and comprehensiveness of the task allocation of drone swarms.

[0020] Based on the above analysis, in order to address the problem that the task allocation of drone clusters in related technologies lacks flexibility and comprehensiveness, the embodiments of the present application provide a method, device and medium for drone cluster task allocation based on antlion optimization, which can realize model solving with a multi-objective optimization strategy, and can effectively improve the flexibility and comprehensiveness of drone cluster task allocation. The present application also designs a random boundary strategy, an adaptive position update strategy, a preference elite selection mechanism and an individual encoding method for the antlion optimization algorithm, so as to increase the diversity of the algorithm, improve the algorithm's development capability and global search capability, and ensure the diversity of solutions, thereby making the antlion optimization algorithm more suitable for solving the task allocation problem of drone clusters.

[0021] First of all, in order to facilitate understanding of the technical solutions provided by the present application, the main technical concepts involved in the embodiments of the present application are briefly described below.

[0022] 1. Multi-objective optimization strategy Multi-objective optimization strategy refers to optimizing multiple conflicting optimization objectives at the same time. The output of the algorithm is not the only optimal solution, but a set of suboptimal solutions that balance multiple optimization objectives. The solution set distribution of multi-objective optimization strategy can provide a reference for problem analysis and can provide multiple solutions for flexible selection. It is a relatively excellent optimization strategy.

[0023] 2. Ant Lion Optimizer (ALO) ALO is a new nature-inspired algorithm proposed by Mirjalili in 2015. The algorithm simulates the foraging behavior of antlions trapping ants. The main steps include random walking of ants, antlions setting traps, ants sliding into traps, antlions preying on ants and antlions rebuilding traps. It is easy to understand, has good scalability and high flexibility, and is therefore widely favored by researchers and applied to computer science, engineering, energy, operations management and other fields.

[0024] 3. Pareto solution set The Pareto solution set is the core concept of multi-objective optimization. In multi-objective optimization, given two solutions x and y in the decision space, if x is better than y in at least one optimization objective and is not inferior to y in other optimization objectives, then x is said to dominate y; when there is no solution dominating x in the decision space, x is called a non-dominated solution, also known as the Pareto optimal solution. The Pareto solution set represents the set of all Pareto optimal solutions in the decision space.

[0025] See also Figure 1 As shown, it is an implementation flow chart of a method for allocating tasks of drone clusters based on antlion optimization provided in an embodiment of the present application. The method may include the following steps: Step S101: randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to characterize the task allocation scheme of the drone cluster.

[0026] In a specific implementation, N ants are randomly initialized as an initial population. For example, ant positions can be randomly generated within a given range of various parameters (such as drone serial number, task priority, etc.).

[0027] Step S102: According to multiple optimization objectives, the ant lion optimization algorithm is executed to determine the Pareto solution set corresponding to the N ants, and in the process of executing the ant lion optimization algorithm, the boundary of the ants wandering around the ant lion in each iteration is determined according to the first formula; wherein the first formula is expressed as follows:

[0028] in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0029] In the specific implementation, multiple optimization objectives are determined according to the actual scenario requirements, and the ant lion optimization algorithm is used to optimize the multiple optimization objectives. For example, the multiple optimization objectives are converted into multiple fitness values ​​through evaluation functions respectively, and then the ant lion optimization algorithm is executed with the lowest of the multiple fitness values ​​as the goal, and the task allocation scheme is repeatedly iterated to obtain a set of task allocation schemes with lower fitness values ​​(i.e., Pareto solution set).

[0030] In the process of executing the ant lion optimization algorithm, for the steps of simulating the ant lion to build a trap and the ants trapped in the ant lion trap and randomly walking, the present application introduces a random boundary strategy to achieve random walking. For example, the random walking position can be determined by the following formula:

[0031] in, t Indicates the current iteration number, T represents the maximum number of iterations, represents the random walk position, cumsum represents the cumulative sum of random walk steps, r represents the generating function of the random walk step length, and there is ,in rand Represents a random number between (0, 1).

[0032] During the random walk, the algorithm simulates the process of the ant lion throwing sand into the trap, which gradually reduces the range of the ant's walk. This process can be expressed by the following formula:

[0033] in, c and d Indicates the upper and lower bounds of the individual's dimension values, c t and Respectively represent t The upper and lower bounds of the search range of each dimension value of the ant in the iteration, I Represents the boundary around which the ants wander.

[0034] In Classic ALO, I As the number of iterations increases, it becomes smaller and changes in a linear segmented trend. In the same round of iterations, the boundaries of all ants' random walks are exactly the same, which will reduce the diversity of the algorithm and is not conducive to the algorithm's search for the global optimal solution. To solve the above problems, the present invention introduces a random boundary strategy, that is, it is designed to use the first formula to determine I , thereby increasing the diversity of ants as they wander around the ant lion.

[0035] Optionally, the parameterw Satisfies the following relationship: when hour, w =0; when hour, w =2; when hour, w =3; when hour, w =4; when hour, w =5; when hour, w =6; in, t Indicates the current iteration number, T Indicates the maximum number of iterations.

[0036] Step S103: Determine the task allocation result of the drone cluster according to the Pareto solution set.

[0037] In specific implementation, a solution can be selected from the Pareto solution set according to the decision maker's preference and decoded to obtain the task allocation result of the drone cluster. Alternatively, a solution can be selected from the Pareto solution set according to its distribution and decoded to obtain the task allocation result of the drone cluster. This application does not impose any restrictions on this.

[0038] It can be seen from the above technical solution that compared with the single-objective optimization method based on weight strategy in traditional drone cluster task allocation, the present application adopts the ant lion optimization algorithm to solve the task allocation result, thereby realizing model solving with a multi-objective optimization strategy, so as to obtain a set of suboptimal solutions (i.e., Pareto solution set) that balance multiple optimization objectives. The distribution of the solution set can provide a reference for problem analysis and improve the comprehensiveness of problem analysis. The solution set can also provide a variety of solutions for flexible selection, which is conducive to comprehensive decision-making. In this way, the flexibility and comprehensiveness of drone cluster task allocation can be effectively improved; and the present application also designs a first formula to apply the random boundary strategy to the ant lion optimization algorithm, thereby increasing the diversity of the algorithm, improving the global search capability of the algorithm, and ensuring the diversity of solutions.

[0039] As a possible implementation, the method further includes: In the process of executing the ant lion optimization algorithm, the position of the ant is updated in each iteration according to the second formula, which is expressed as follows:

[0040] in, Indicates the updated ant position; Indicates t The vector obtained by the ants' random walk around the antlion selected based on roulette in the iteration; Indicates t The vector obtained by the ants' random walk around the elite antlion in the iteration; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0041] In this embodiment, the present application designs a second formula to apply the adaptive position strategy to the antlion optimization algorithm, thereby further increasing the diversity of the algorithm, improving the global search capability of the algorithm, and ensuring the diversity of solutions.

[0042] As a possible implementation, executing the antlion optimization algorithm to determine the Pareto solution set corresponding to the N ants according to multiple optimization objectives includes: If the maximum number of iterations is not reached, the iteration performs the following steps: Determining the fitness values ​​of the N ants according to the evaluation functions corresponding to the multiple optimization objectives; Dividing the N ants into sub-populations corresponding to the multiple optimization objectives, and taking the ants with the highest fitness value in each of the sub-populations as elite ant lions in each of the sub-populations; updating the Pareto solution set according to the positions of all individuals in the plurality of subpopulations, and ensuring that the solutions in the Pareto solution set are all non-dominated solutions during the updating; After updating the Pareto solution set, for each ant, a roulette strategy is used to select the antlion associated with the ant from the Pareto solution set according to sparsity; For each of the ants, determine the boundary of the ant's wandering around the ant lion according to the first formula, and based on the boundary of the ant's wandering around the ant lion, make the ant randomly wander around the ant lion associated with it and the elite ant lion of the sub-population to which it belongs; For each of the ants, the position of the ant is updated according to the second formula.

[0043] In this embodiment, reference Figure 2 The schematic diagram of the preferred elite selection mechanism shown in FIG. 1 shows that the present application divides the N ants into sub-populations corresponding to the multiple optimization targets, for example, according to the number of optimization targets. Obj Divide the N ants into Objsub-populations so that each sub-population only focuses on one optimization goal, and the ants with the highest fitness value in each sub-population are taken as the elite antlions in each sub-population, that is, the non-dominated solution with the best result on the corresponding optimization goal is selected as the elite antlion of the relevant sub-population, thereby applying the preferential elite selection mechanism to the antlion optimization algorithm, thereby further increasing the diversity of the algorithm, improving the global search ability of the algorithm, and ensuring the diversity of solutions.

[0044] Exemplarily, the pseudo code of an antlion optimization algorithm provided in this application is as follows: Input: number of iterations T , Optimization target number Obj , Population P The number of individuals N ; Output: Population P The Pareto solution set composed of non-dominated solutions in PS ; 1. Random Initialization N ants as the initial population P; 2.WHILE (maximum number of iterations not reached) T ) 3. Calculate the population P The fitness values ​​of all individuals in ; 4. Based on the number of optimization targets Obj Will P Divide into Obj subpopulations, and the number of individuals in each subpopulation is N / Obj ; 5. The ants with the highest fitness value in each sub-population are saved as elite ant lions in that sub-population E ; 6. P All individuals in join PS ,renew PS ,ensure PS The solutions in are all non-dominated solutions; 7.FOR N Ants: 8. Use the roulette strategy to select PS Choose an antlion Antlion ; 9. Ants in each sub-population use random border strategies to surround Antlion and the E random walk of its subpopulation; 10. Use adaptive position update strategy to update the position of ants; 11.END FOR 12.END WHILE In step 6 of the pseudocode of the above antlion optimization algorithm, in each iteration, the algorithm will update all individuals in the population and compare the current population with PS Merge and select non-dominated solutions through Pareto dominant strategy and save them in PS middle.

[0045] For example, PS The pseudo code of the update algorithm is as follows: Input: original Pareto solution set PS , population P ; Output: New Pareto solution set PS_new ; 1. PS and P Merge into PS_tmp ; 2.for each x ∈ PS_tmp do 3.if x Not arbitrarily y ∈ PS_tmp ( y != x ) dominates then 4. x Marked as non-dominated solution; 5.end 6.end 7. Save all non-dominated solutions to PS_new middle In step 8 of the pseudocode of the Antlion optimization algorithm above, the concept of sparsity is used to determine the density of the location of a non-dominated solution. Simply put, the sparsity of a non-dominated solution is the number of other non-dominated solutions in the neighborhood of its location.

[0046] Exemplarily, the pseudo code of the sparsity algorithm is as follows: Input: Pareto solution set PS , optimization target number Obj ; Output: PS The sparsity of all solutions in S ; 1. Calculate the neighborhood radius of each target r ; 2. S Initialized to all 0s; 3.for each x ∈ PS do 4.for each y ∈ P.S.D. 5. If there is a distance in each optimization target ( x , y ) <rthen 6.S[ x ] = S[ x ] + 1; 7.end 8.end 9.end In getting PS After calculating the sparsity of all solutions in the sparsest region, the Ant Lion Optimization Algorithm uses a roulette wheel to select PS The solution in is used as the roulette antlion (i.e. the antlion to which the ant is associated).

[0047] Exemplarily, the pseudo code of the sparsity roulette algorithm is as follows: Input: Pareto solution set PS The sparsity of all solutions in S ; Output: The antlion number selected by the roulette wheel; 1. Calculate the inverse of sparsity and save it in X middle; 2.for each x ∈ X do 3. Calculate from X [0] to x The cumulative sum is stored in W middle; 4.end 5. Sum the reciprocals of all sparsity and multiply by random(0,1) as the selection probability p ; 6.for each w ∈ W do 7.if w > p then 8. Select w The index value is used as the antlion number to interrupt the loop; 9.end 10.end As a possible implementation method, in order to ensure the rationality of the task allocation of the drone cluster, the task benefit, drone threat cost and task time cost are respectively used as optimization targets (that is, the multiple optimization targets at least include: task benefit, drone threat cost and task time cost) to establish a multi-objective optimization model for the task allocation problem of the drone cluster. Among them, the task benefit refers to the value reward obtained by the drone when the task is successfully performed; the drone threat cost refers to the price paid by the drone to perform the task, the drone damage cost is related to the drone value and the probability of the drone being destroyed, and the probability of the drone being destroyed during different tasks is different; the task execution time refers to the total time taken by the drone cluster to complete all tasks, which can be described by the latest task end time in all drone task schedules.

[0048] In this embodiment, the present application adopts the ant lion optimization algorithm to optimize the task benefit, drone threat cost and task time cost, that is, taking these three as three optimization targets, converting the three into three fitness values ​​through evaluation functions respectively, and then executing the ant lion optimization algorithm with the lowest of these three fitness values ​​as the goal, repeatedly iterating to solve the task allocation plan in order to obtain a set of task allocation plans with lower three fitness values ​​(i.e., Pareto solution set).

[0049] Optionally, the task benefit includes: a reconnaissance reward corresponding to the reconnaissance task, an attack reward corresponding to the attack task, and an evaluation reward corresponding to the evaluation task; the evaluation function corresponding to the task benefit is expressed as follows:

[0050] in, Represents the value of the evaluation function corresponding to the task benefit; M Indicates the target maximum value corresponding to the task benefit; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; P ik Indicates drone i Execute the task k Capacity parameters; S ik Indicates drone i Execute the task k The probability of being destroyed when Vk M Representation Task k value.

[0051] In the specific implementation, for application scenarios where the mission types include reconnaissance, attack and evaluation, the mission benefits are divided into three types according to the mission type: reconnaissance reward (i.e., the amount of information about the object obtained by the drone), attack reward, and evaluation reward; among them, the purpose of the reconnaissance mission is to confirm the type and identity information of the object to reduce the uncertainty of the object; the purpose of the attack mission is to destroy the enemy object; the purpose of the evaluation mission is to obtain the damage effect information after attacking the object. The three types of rewards are related to the mission value, the mission capability parameters of the drone, and the probability of the drone being destroyed. In order to minimize the corresponding optimization target, this application designs the reward optimization function (i.e., the evaluation function corresponding to the mission benefit) as the loss value of the corresponding optimization target. The lower the value of the reward optimization function, the better the task completion.

[0052] Optionally, the evaluation function corresponding to the drone threat cost is expressed as follows:

[0053] in, Represents the value of the evaluation function corresponding to the drone threat cost; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; S ik Indicates drone i Execute the task k The probability of being destroyed when V i U Indicates drone i value.

[0054] Optionally, the evaluation function corresponding to the task time cost is expressed as follows:

[0055] in, Indicates the value of the evaluation function corresponding to the task time cost; x ik represents the decision variable, when the task k Assigned to drone ihour, x ik =1, otherwise x ik =0; Indicates the task k End time; N V represents the total number of drones; N M Indicates the total number of tasks.

[0056] As a possible implementation method, in order to improve the rigor of UAV cluster task allocation, task timing constraints are incorporated into the model.

[0057] Taking the application scenario where the task types include reconnaissance, attack, and assessment as an example, the task timing constraints are defined as follows:

[0058] in, t s Indicates the task start time. t e Indicates the end time of the task; the above formula means that on the same object, the start time of the strike task needs to be later than the end time of the reconnaissance task, and the start time of the evaluation task needs to be later than the end time of the strike task.

[0059] It is understandable that in the process of iterative optimization using the Antlion Optimization Algorithm, the task timing constraints will limit the feasible domain of the solution, ensuring that the solution obtained by the algorithm meets the task timing constraints. Finally, after the algorithm iteration is completed, by decoding the solutions in the Pareto solution set, a task allocation plan containing different task types can be obtained.

[0060] Optionally, mission time window constraints, mission integrity constraints, UAV collaboration constraints, UAV range constraints, and UAV attack capability constraints can also be incorporated into the model.

[0061] As a possible implementation, the position of the ant is determined according to the sum of the drone sequence number and the task priority; and the task allocation result of the drone cluster is determined according to the Pareto solution set, including: Selecting a Pareto solution from the Pareto solution set as a task allocation vector, and sending it to each drone in the drone cluster, so that each drone determines the task allocation result according to the task allocation vector and the association between the task priority and the task sequence, wherein the task allocation result includes: the type of each task and the drone that performs each task; Each element in the task allocation vector is represented by a real number, the integer part of the real number is used to indicate the UAV that performs the task corresponding to the relevant element, and the decimal part is used to indicate the priority of the task corresponding to the relevant element.

[0062] In this embodiment, the drone cluster task allocation method provided in the present application can be executed by a pre-set central node (set according to the actual scenario). After obtaining the task allocation vector, the central node will send the task allocation vector to all drones in the cluster so that the drone cluster can work together.

[0063] For example, taking an application scenario including a drone cluster and multiple objects, each object corresponds to three tasks: reconnaissance, strike and evaluation. The drone cluster needs to start from the initial position and perform reconnaissance, strike and evaluation tasks on all objects in turn according to the task allocation results.

[0064] When the drone swarm performs a task, only one drone can perform a task on the same object at the same time, and the three tasks on the object must be performed in strict accordance with the order of "reconnaissance-strike-assessment". In addition, each drone must ensure that the flight distance does not exceed its maximum range and the number of strikes does not exceed its own ammunition load. It is understandable that this application scenario is only an example, and the method proposed in this application is not limited to this application scenario.

[0065] In the above application scenario, the task allocation result is represented by a task allocation vector. The encoding length of the vector is the number of objects multiplied by the number of task types. Each element in the vector corresponds to a task on each object. Each element is represented by a real number, where the integer part of the real number represents the drone that performs the task, and the decimal part represents the priority of the task. The smaller the real number, the higher the priority. Since the three tasks on the object must be performed strictly in the order of "reconnaissance-strike-evaluation", the tasks on the same object are determined to be "reconnaissance-strike-evaluation" from high to low priority, thereby determining the correlation between task priority and task timing.

[0066] Reference Figure 3 The diagram shows the mapping between the algorithm coding and the task allocation results. Taking the application scenario where two drones go to two objects to complete tasks as an example, the task allocation result is [1.2837, 2.8449, 2.5364, 1.0482, 2.4619, 1.2984], which means that task 1 on object 1 and tasks 4 and 6 on object 2 are assigned to drone 1. The priority of task 1 on object 1 is 0.2837, the priority of task 4 on object 2 is 0.0482, and the priority of task 6 on object 2 is 0.2984. The execution order is task 4 first, then task 1, and finally task 6. The same is true for drone 2.

[0067] For object 1, since the priorities of its corresponding tasks 1, 2, and 3 are 0.2837, 0.8449, and 0.5364, respectively, the two drones can determine that task 1 is a reconnaissance task, task 2 is an assessment task, and task 3 is a strike task based on the correlation between task priority and task sequence. The same is true for object 2.

[0068] Based on the above embodiments, the present application takes into account that the ant lion optimization algorithm still has problems such as low optimization accuracy, easy to fall into local optimality, and unbalanced exploration and development capabilities. On the basis of the traditional multi-objective ant lion optimization algorithm (Multi-Objective AntLion Optimizer, MOALO), a random boundary strategy, an adaptive position update strategy and a preference elite selection mechanism are adopted to obtain a multi-objective multi-population self-adaptive ant lion optimization algorithm (Multi-objective Multi-population Self-adaptive Ant Lion Optimizer, MMSALO) to increase the diversity of the algorithm, improve the development capability and global search capability of the algorithm, and ensure the diversity of understanding.

[0069] Therefore, this application proposes a multi-task and multi-constraint task allocation method for drone clusters based on antlion optimization, considering multiple task types and multiple constraints, encoding the antlion position as the sum of the drone serial number and the task priority, and solving the task allocation result with MMSALO. Compared with traditional multi-objective antlion optimization algorithm, multi-objective particle swarm optimization algorithm, multi-objective genetic algorithm (NSGA-II), multi-objective evolutionary algorithm based on decomposition (Multi-Objective Evolutionary Algorithm based on Decomposition, MOEA / D) and other classic multi-objective optimization algorithms, the method provided in this application has improved the hyper-volume (HV) of the Pareto solution set, that is, a better set of task allocation solutions can be obtained.

[0070] It should be noted that, for the method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited by the described order of actions, because according to the embodiments of the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present application.

[0071] The present application also provides a drone cluster task allocation device based on antlion optimization, such as Figure 4As shown, the device comprises: An initialization module is used to randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to characterize the task allocation scheme of the drone cluster; An algorithm execution module is used to execute the antlion optimization algorithm to determine the Pareto solution set corresponding to the N ants according to multiple optimization objectives, and in the process of executing the antlion optimization algorithm, determine the boundary of the ants walking around the antlion in each iteration according to the first formula; The result determination module is used to determine the task allocation result of the drone cluster according to the Pareto solution set; The first formula is expressed as follows:

[0072] in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0073] Optionally, the device further comprises: The position updating module is used to update the position of the ant in each iteration according to the second formula during the execution of the ant lion optimization algorithm. The second formula is expressed as follows:

[0074] in, Indicates the updated ant position; Indicates t The vector obtained by the ants' random walk around the antlion selected based on roulette in the iteration; Indicates t The vector obtained by the ants' random walk around the elite antlion in the iteration; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

[0075] Optionally, the algorithm execution module includes: The execution submodule is used to iteratively execute the following steps if the maximum number of iterations has not been reached: Determining the fitness values ​​of the N ants according to the evaluation functions corresponding to the multiple optimization objectives; Dividing the N ants into sub-populations corresponding to the multiple optimization objectives, and taking the ants with the highest fitness value in each of the sub-populations as elite ant lions in each of the sub-populations; updating the Pareto solution set according to the positions of all individuals in the plurality of subpopulations, and ensuring that the solutions in the Pareto solution set are all non-dominated solutions during the updating; After updating the Pareto solution set, for each ant, a roulette strategy is used to select the antlion associated with the ant from the Pareto solution set according to sparsity; For each of the ants, determine the boundary of the ant's wandering around the ant lion according to the first formula, and based on the boundary of the ant's wandering around the ant lion, make the ant randomly wander around the ant lion associated with it and the elite ant lion of the sub-population to which it belongs; For each of the ants, the position of the ant is updated according to the second formula.

[0076] Optionally, the multiple optimization objectives include at least: mission benefit, drone threat cost and mission time cost; the mission benefit includes: reconnaissance reward corresponding to the reconnaissance mission, attack reward corresponding to the attack mission, and evaluation reward corresponding to the evaluation mission; the evaluation function corresponding to the mission benefit is expressed as follows:

[0077] in, Represents the value of the evaluation function corresponding to the task benefit; M Indicates the target maximum value corresponding to the task benefit; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; P ik Indicates drone i Execute the task k Capacity parameters; S ik Indicates drone i Execute the task k The probability of being destroyed when V k M Indicates the task k value.

[0078] Optionally, the multiple optimization objectives include at least: mission benefit, drone threat cost and mission time cost; the evaluation function corresponding to the drone threat cost is expressed as follows:

[0079] in, Represents the value of the evaluation function corresponding to the drone threat cost; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; S ik Indicates drone i Execute the task k The probability of being destroyed when V i U Indicates drone i value.

[0080] Optionally, the multiple optimization objectives include at least: mission benefit, drone threat cost and mission time cost; the evaluation function corresponding to the mission time cost is expressed as follows:

[0081] in, Indicates the value of the evaluation function corresponding to the task time cost; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; Representation Task k End time; N V represents the total number of drones; N M Indicates the total number of tasks.

[0082] Optionally, the result determination module includes: A determination submodule is used to select a Pareto solution from the Pareto solution set as a task allocation vector, and send it to each drone in the drone cluster, so that each drone determines the task allocation result according to the task allocation vector and the association between the task priority and the task timing, and the task allocation result includes: the type of each task and the drone that performs each task; Each element in the task allocation vector is represented by a real number, the integer part of the real number is used to indicate the UAV that performs the task corresponding to the relevant element, and the decimal part is used to indicate the priority of the task corresponding to the relevant element.

[0083] Optionally, the parameter w Satisfies the following relationship: when hour, w =0; when hour, w =2; when hour, w =3; when hour, w =4; when hour, w =5; when hour, w =6; in, t Indicates the current iteration number, T Indicates the maximum number of iterations.

[0084] It can be seen from the above technical solution that compared with the single-objective optimization method based on weight strategy in traditional drone cluster task allocation, the present application adopts the ant lion optimization algorithm to solve the task allocation result, thereby realizing model solving with a multi-objective optimization strategy, so as to obtain a set of suboptimal solutions (i.e., Pareto solution set) that balance multiple optimization objectives. The distribution of the solution set can provide a reference for problem analysis and improve the comprehensiveness of problem analysis. The solution set can also provide a variety of solutions for flexible selection, which is conducive to comprehensive decision-making. In this way, the flexibility and comprehensiveness of drone cluster task allocation can be effectively improved; and the present application also designs a first formula to apply the random boundary strategy to the ant lion optimization algorithm, thereby increasing the diversity of the algorithm, improving the global search capability of the algorithm, and ensuring the diversity of solutions.

[0085] The present application also provides an electronic device, referring to Figure 5 , Figure 5 Schematic diagram of an electronic device proposed in an embodiment of the present application. Figure 5 As shown, the electronic device 100 includes: a memory 110 and a processor 120. The memory 110 and the processor 120 are connected via a bus communication. A computer program is stored in the memory 110. The computer program can be run on the processor 120 to implement the steps in the drone cluster task allocation method based on ant lion optimization disclosed in the embodiment of the present application.

[0086] The embodiment of the present application also provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the drone cluster task allocation method based on antlion optimization as disclosed in the embodiment of the present application is implemented.

[0087] The embodiments of the present application also provide a computer program product, including a computer program / instruction, which, when executed by a processor, implements the drone cluster task allocation method based on antlion optimization as disclosed in the embodiments of the present application.

[0088] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0089] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, devices or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt 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.

[0090] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, systems, devices, storage media, and program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal 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 A device that provides the functions specified in a block or multiple blocks.

[0091] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0092] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement 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.

[0093] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0094] The above is a detailed introduction to the method, device and medium for allocating tasks of drone clusters based on antlion optimization provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technicians in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for allocating tasks of drone swarms based on antlion optimization, characterized in that: The method comprises: Randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to represent the task allocation scheme of the drone cluster; According to the plurality of optimization objectives, an antlion optimization algorithm is executed to determine a Pareto solution set corresponding to the N ants, and in the process of executing the antlion optimization algorithm, a boundary around which the ants wander around the antlion in each iteration is determined according to the first formula; According to the Pareto solution set, determine the task allocation result of the drone cluster; The first formula is expressed as follows: in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

2. The method according to claim 1, characterized in that The method further comprises: In the process of executing the ant lion optimization algorithm, the position of the ant is updated in each iteration according to the second formula, which is expressed as follows: in, Indicates the updated ant position; Indicates t The vector obtained by the ants' random walk around the antlion selected based on roulette in the iteration; Indicates t The vector obtained by the ants' random walk around the elite antlion in the iteration; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

3. The method according to claim 2, characterized in that The step of executing the antlion optimization algorithm according to the multiple optimization objectives to determine the Pareto solution set corresponding to the N ants includes: If the maximum number of iterations is not reached, the iteration performs the following steps: Determining the fitness values ​​of the N ants according to the evaluation functions corresponding to the multiple optimization objectives; Dividing the N ants into sub-populations corresponding to the multiple optimization objectives, and taking the ants with the highest fitness value in each of the sub-populations as elite ant lions in each of the sub-populations; updating the Pareto solution set according to the positions of all individuals in the plurality of subpopulations, and ensuring that the solutions in the Pareto solution set are all non-dominated solutions during the updating; After updating the Pareto solution set, for each ant, a roulette strategy is used to select the antlion associated with the ant from the Pareto solution set according to sparsity; For each of the ants, determine the boundary of the ant's wandering around the ant lion according to the first formula, and based on the boundary of the ant's wandering around the ant lion, make the ant randomly wander around the ant lion associated with it and the elite ant lion of the sub-population to which it belongs; For each of the ants, the position of the ant is updated according to the second formula.

4. The method according to claim 3, characterized in that: The multiple optimization objectives include at least: mission benefit, UAV threat cost and mission time cost; the mission benefit includes: reconnaissance reward corresponding to the reconnaissance mission, attack reward corresponding to the attack mission, and evaluation reward corresponding to the evaluation mission; the evaluation function corresponding to the mission benefit is expressed as follows: in, Represents the value of the evaluation function corresponding to the task benefit; M Indicates the target maximum value corresponding to the task benefit; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; P ik Indicates drone i Execute the task k Capacity parameters; S ik Indicates drone i Execute the task k The probability of being destroyed when V k M Indicates the task k value.

5. The method according to claim 3, characterized in that: The multiple optimization objectives include at least: mission benefit, drone threat cost and mission time cost; the evaluation function corresponding to the drone threat cost is expressed as follows: in, Represents the value of the evaluation function corresponding to the drone threat cost; N V represents the total number of drones; N M Indicates the total number of tasks; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; S ik Indicates drone i Execute the task k The probability of being destroyed when V i U Indicates drone i value.

6. The method according to claim 3, characterized in that The multiple optimization objectives include at least: mission benefit, drone threat cost and mission time cost; the evaluation function corresponding to the mission time cost is expressed as follows: in, Indicates the value of the evaluation function corresponding to the task time cost; x ik represents the decision variable, when the task k Assigned to drone i hour, x ik =1, otherwise x ik =0; Indicates the task k End time; N V represents the total number of drones; N M Indicates the total number of tasks.

7. The method according to any one of claims 1 to 6, characterized in that: The position of the ant is determined according to the sum of the drone sequence number and the task priority; and the task allocation result of the drone cluster is determined according to the Pareto solution set, including: Selecting a Pareto solution from the Pareto solution set as a task allocation vector, and sending it to each drone in the drone cluster, so that each drone determines the task allocation result according to the task allocation vector and the association between the task priority and the task timing, and the task allocation result includes: the type of each task and the drone that performs each task; Each element in the task allocation vector is represented by a real number, the integer part of the real number is used to indicate the UAV that performs the task corresponding to the relevant element, and the decimal part is used to indicate the priority of the task corresponding to the relevant element.

8. The method according to any one of claims 1 to 6, characterized in that: The parameters w Satisfies the following relationship: when hour, w =0; when hour, w =2; when hour, w =3; when hour, w =4; when hour, w =5; when hour, w =6; in, t Indicates the current iteration number, T Indicates the maximum number of iterations.

9. A drone swarm task allocation device based on antlion optimization, characterized in that: The device comprises: An initialization module is used to randomly initialize N ants, where N is a positive integer, and the positions of the ants are used to characterize the task allocation scheme of the drone cluster; An algorithm execution module is used to execute the antlion optimization algorithm to determine the Pareto solution set corresponding to the N ants according to multiple optimization objectives, and in the process of executing the antlion optimization algorithm, determine the boundary of the ants walking around the antlion in each iteration according to the first formula; The result determination module is used to determine the task allocation result of the drone cluster according to the Pareto solution set; The first formula is expressed as follows: in, I Represents the boundary of the ants around the ant lion; parameter w The value of is determined according to the current number of iterations t, and 10 w As the number of iterations increases, it shows a segmented exponential increasing trend; T Indicates the maximum number of iterations; rand Represents a random number between (0,1).

10. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the method for allocating drone swarm tasks based on antlion optimization as described in any one of claims 1 to 8 is implemented.

Images