Unmanned aerial vehicle task allocation method based on improved black wing plinary optimization algorithm

Through the improved Black-winged Kitchen optimization algorithm, combined with chaotic mapping, adaptive weights and Levi flight method, the problem that the UAV task allocation algorithm is prone to fall into local optimal solutions, and the global optimal solution capability of task allocation is improved and Adaptability enables efficient drone mission allocation.

CN118938948APending Publication Date: 2024-11-12UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF 0 Cites 6 Cited by

Patent Information

Application Number
CN202410980315.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing drone task allocation algorithm is prone to falling into local optimal solutions, difficult to deal with discrete task sequences, and poor adaptability to scenarios, making it impossible to effectively achieve efficient allocation of drone tasks.

Method used

The improved black-winged kite optimization algorithm is adopted to initialize the population by introducing a chaotic mapping strategy, combining the attack and migration behavior of the black-winged kite, using adaptive weights and Levi flight methods, and combining leader strategies, metaheuristic optimization algorithm is designed to Solve the problem of drone mission allocation.

Benefits of technology

The global optimal solution search capability of the drone task allocation problem is realized, and the local optimal solution can be jumped out of the algorithm, the adaptability and efficiency of the algorithm are improved, and a better task allocation sequence is obtained.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938948A_ABST
    Figure CN118938948A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle task allocation method based on an improved black-wing optimization algorithm, and is applied to the technical field of unmanned aerial vehicle task planning. The problems that an existing unmanned aerial vehicle task allocation method based on a swarm intelligent optimization algorithm is prone to falling into a local optimal solution, discrete task sequence solutions are difficult to process, and the adaptability of the algorithm to a scene is poor are solved. According to the method, firstly, an unmanned aerial vehicle task allocation model under voyage constraint is established based on task requirements of unmanned aerial vehicles; secondly, defining an initial parameter set of a black-wing supporting optimization algorithm, and adopting a chaotic mapping strategy in initialization of a black-wing supporting population; then, a model of a meta-heuristic algorithm is constructed according to the attack behavior and the migration behavior of the Heiwen. And finally, iterative solution is carried out through an optimization algorithm to obtain an optimal task allocation sequence, so that task allocation of the unmanned aerial vehicle is efficiently completed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) mission planning, and in particular relates to a UAV mission allocation technology. Background Art

[0002] Drones usually have high maneuverability and cost-effectiveness as well as good environmental adaptability and safety. Therefore, drones have a wide range of application scenarios in both military and civilian fields. For example, when performing reconnaissance missions, drones have the characteristics of high execution efficiency and fast response speed. However, the degree of autonomy of drones in performing tasks is limited by the algorithm of task allocation. In order to enable drones to efficiently complete task allocation and achieve better task results, it is necessary to design efficient and intelligent task planning methods based on specific mission scenarios and requirements. In recent years, for the problem of drone task allocation, using optimization algorithms to solve the optimal task allocation sequence of drones has become an important research direction.

[0003] Optimization algorithms can be divided into traditional optimization algorithms, random optimization algorithms, machine learning optimization algorithms, etc. Among them, traditional optimization algorithms such as exhaustive method, dynamic programming algorithm, greedy algorithm and other heuristic algorithms usually solve problems based on deterministic models and mathematical rules, so the effect of the algorithm is greatly limited by specific scenarios. Machine learning optimization algorithms such as gradient descent algorithm and neural network algorithm have large computational complexity, so the operation time is long. Random optimization algorithms such as simulated annealing algorithm and swarm intelligence algorithm are algorithms based on probability and randomness. By setting multiple candidate solutions and introducing random perturbations to search for the optimal solution in the solution space. Among them, the swarm intelligence algorithm in the random optimization algorithm is an optimization algorithm based on swarm intelligence. Through multiple search agents in the group, the optimal solution is searched according to a specific random position update strategy during the iteration process, so that the swarm intelligence algorithm can realize the optimization process on large-scale problems. Swarm intelligence algorithm is a meta-heuristic algorithm, including ant colony algorithm, particle swarm algorithm, etc. However, the existing UAV task allocation method based on swarm intelligence optimization algorithm is still: 1) It is easy to fall into the local optimal solution. 2) It is difficult to handle discrete task sequence solutions. 3) The algorithm has poor adaptability to the scene. Therefore, it is necessary to comprehensively consider the constraints of the scene and add constraints when designing the adaptive function to complete the task allocation. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method for assigning unmanned aerial vehicle tasks based on an improved black kite optimization algorithm. According to the task requirements, a model for assigning unmanned aerial vehicle tasks under range constraints is established, a chaotic mapping strategy is introduced in the process of population initialization, a model of a meta-heuristic optimization algorithm is established through the attack and predation behaviors of black kites, an adaptive weight method is introduced in the local search stage, i.e., the attack behavior stage of black kites, and a Levy flight method is introduced in the global search stage, i.e., the migration behavior stage of black kites. Finally, the optimization solution of the unmanned aerial vehicle task assignment problem is realized.

[0005] The technical solution adopted by the present invention is: a method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm, comprising:

[0006] S1. According to the UAV task allocation scenario, a UAV task allocation model under range constraints is established;

[0007] S2. Use the Black Kite optimization algorithm to solve the UAV task allocation model established in step S1 to obtain the final UAV task allocation plan; the specific implementation process includes the following sub-steps:

[0008] S21, defining the initial parameter set of the black kite optimization algorithm, and initializing the black kite population based on chaotic mapping;

[0009] S22. A metaheuristic algorithm model was constructed based on the attack and migration behaviors of black kites. An adaptive weight method was introduced in the algorithm's depth search phase. A Levy flight method was introduced in the algorithm's breadth search phase and the Levy flight method was combined with a leader strategy.

[0010] S23. Iterate and solve through optimization algorithm to obtain the optimal task allocation sequence.

[0011] Beneficial effects of the present invention: The present invention takes drones as the objects of task planning, and models the scenarios of drones such as material transportation or regional reconnaissance in reality as task allocation problems. First, a task allocation model is established according to the task scenario; secondly, the initial population position is initialized using chaotic mapping, and then, according to the black kite optimization algorithm, a metaheuristic algorithm model is established based on the black kite behavior, and an adaptive weight method is introduced in the deep search stage of the algorithm. In the breadth search stage of the algorithm, the Levy flight method is introduced and the Levy flight method is combined with the leader strategy; finally, the optimization algorithm is iteratively solved to obtain the optimal task allocation sequence. The drone task allocation is completed. Therefore, the proposed drone task allocation method based on the improved black kite optimization algorithm has a strong global optimal solution search capability and the ability to jump out of the local optimal solution for complex task planning problems, and can accurately search for the global optimal solution. Compared with the existing technology, the method of the present invention has the following advantages:

[0012] 1) This paper proposes a UAV task allocation method based on an improved Black Kite optimization algorithm. Compared with traditional swarm intelligence algorithms such as ant colony algorithm and particle swarm algorithm, the novel Black Kite algorithm is introduced to solve the UAV task allocation problem, and the feasibility and effectiveness of the Black Kite optimization algorithm in solving the UAV task allocation problem are proved, which broadens the application scenarios of the Black Kite optimization algorithm.

[0013] 2) The present invention introduces a chaotic mapping strategy in the initialization stage. Compared with some swarm intelligence algorithms that use uniform distribution to randomly generate the initial position of the population, the present invention performs chaotic mapping on the generated initial population position. The chaotic mapping guides the algorithm to optimize the initial search path in the solution space in a nonlinear manner, and the unpredictability of the chaotic mapping can increase the diversity of individuals in the population and avoid premature convergence of the convergence curve.

[0014] 3) The present invention introduces an adaptive weight method in the local search stage of the algorithm. Through the characteristic that the adaptive weight factor decreases as the number of iterations increases, the search speed and search accuracy of the optimization algorithm are dynamically adjusted. In the early stage of the algorithm search, the quality of the solution is poor, and a larger adaptive weight factor can increase the search speed of the algorithm; in the later stage of the algorithm search, the quality of the solution is good, and a smaller adaptive weight factor can increase the search accuracy of the algorithm and enhance the local search ability of the algorithm.

[0015] 4) The present invention introduces the Lévy flight method in the global search phase of the algorithm. The Lévy flight is a random movement method that combines a short step with a high probability and a long step with a low probability. The improved algorithm of the present invention combines the Lévy flight with the leader strategy, so that individuals in the population can better jump out of the local optimal solution while being guided by the leader, thereby enhancing the algorithm's ability to search for the global optimal solution.

[0016] 5) The present invention improves the way the black kite optimization algorithm updates the global optimal solution, and designs a leader competition mechanism based on the leader strategy. The leader competition mechanism compares the best individual in the population obtained after the current iteration with the leader of this iteration, and selects the individual with a better fitness value as the population leader of the next iteration. Through the leader competitor mechanism, it is ensured that the leader of the population position update strategy and the optimal solution output at the end of the algorithm iteration are always the best individuals in the population up to the current iteration. Thereby improving the ability of the algorithm to search for the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The flowchart of the method for allocating unmanned aerial vehicle tasks based on adaptive weight and Levy flying black-winged kite optimization algorithm of the present invention is as follows;

[0018] Figure 2It is a flow chart of the iterative solution part of the optimization algorithm based on adaptive weight and Levy flying black-winged kite of the present invention;

[0019] Figure 3 Assigning a scene graph to a task of an embodiment;

[0020] Figure 4 The optimal task execution sequence scheme of the embodiment;

[0021] Figure 5 It is a graph of algorithm convergence of the embodiment. DETAILED DESCRIPTION

[0022] To facilitate those skilled in the art to understand the technical content of the present invention, the present invention is further explained below with reference to the accompanying drawings.

[0023] like Figure 1 As shown, the present invention is a method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm, comprising the following steps:

[0024] S1. By establishing the objective function and constraints, the UAV task allocation scenario is modeled and the mathematical model of unmanned task allocation is constructed. The implementation process of this step specifically includes the following sub-steps:

[0025] S11. Regarding the scenario of drone task allocation, assume that the task scenario graph G is an area of ​​size m×n, and there are k tasks waiting to be executed in the task scenario graph G. Let the set of tasks be recorded as S={s i}, i = 1, 2...k, where S includes all pending tasks and the pending tasks are not repeated, i.e., s i ≠s j , i≠j, i, j=1, 2...k.

[0026] S12, set obstacles, that is, no-fly zones R, and drones cannot pass through the no-fly zones. The obstacle setting includes the center position of the obstacle (x R ,y R ) and the size of the obstacle.

[0027] S13. Define a task execution matrix X, where X is a k×k matrix. X(i, j) indicates that the task execution order is from task i to task j, where i, j=1, 2...k, i≠j. When X(i, j)=1, it indicates that the task execution order is task i→task j. When the task execution order does not satisfy task i→task j, X(i, j)=0.

[0028] S14, define the location set A of the task points = {a i}, a i Represents the position of task point i in the task allocation scenario.

[0029] S15. Define the lines between the positions of the task points as edges, and the edge set E = {e ij}, i, j = 1, 2...k, i ≠ j, where e ij Represents the line between the position of task point i and the position of task point j.

[0030] S16. Let d(i, j) be the distance between the position of task point i and the position of task point j, d(i, j) = d(i, j).

[0031] S17. According to the established mission scenario, define the fitness value F of the objective function as the UAV’s mileage W, that is, F = W, and the UAV’s maximum mileage is W max .

[0032] S18. According to the scene constraints, task requirements and the principle of minimizing the fitness function, the objective function and its constraints are established as follows:

[0033] minΣd(i, j)×X(i, j), i, j=1, 2...k, i≠j,

[0034]

[0035]

[0036]

[0037] Σd(i, j)×X(i, j)≤W max , i, j = 1, 2... k, i ≠ j

[0038] Where E = {e ij} is an edge set, R represents a no-fly zone, It means that the generated path does not pass through the no-fly zone, that is, there is no intersection, which is a constraint condition; Is an empty set.

[0039] S2. Set the initialization parameters of the adaptive weight and Levy flying black kite optimization algorithm, including population size, number of iterations, problem dimension and other parameters:

[0040] S21. Define the initial population size as pop, the maximum number of iterations of the population as T, and the dimension of the problem to be solved as dim.

[0041] S22, initialize the position of the population. Set the upper and lower limits of the population initial position, and randomly generate black-winged kite individuals P uniformly distributed in the population initial position generation area. i :

[0042] P i=L+r×(UL), 1≤i≤pop

[0043] Where L and U are the upper and lower bounds of the solution space, r is a random vector uniformly distributed in the interval [0, 1], and pop is the population size. The initial population position can be expressed in the following matrix form:

[0044]

[0045] Where dim is the dimension size of the problem assigned to a given task.

[0046] S23, realize the chaotic mapping of the initial population position. Perform chaotic mapping on the initial population position obtained after initialization in sub-step S22, and use the a-order Chebyshev chaotic mapping strategy to initialize the population position again. The formula of the a-order Chebyshev chaotic mapping strategy is as follows:

[0047] P i '

[0048] a4

[0049] The value of a is usually 4.

[0050] S3, the optimization algorithm is used for iterative solution, such as Figure 2 As shown, the specific process includes the following:

[0051] Set the parameters and inputs required by the algorithm; initialize the population and calculate the fitness; set the start and end conditions of the algorithm's cyclic iterative solution; simulate the attack and migration behaviors of the black kite population in nature to construct a position update strategy, so that in each iteration, the candidate solutions in the population update their positions in the solution space after the attack and migration behaviors, and calculate and record the optimal fitness value of the population in this iteration; compare the optimal fitness value of the population in this iteration with the optimal fitness value of the population recorded in the previous iteration, and select the better fitness value and the corresponding solution; finally, output the global optimal solution at the end of the iteration.

[0052] S31. According to the leader strategy, select the best black-winged kite individual in the previous iteration as the leader, that is, The black kite optimization algorithm introduces a leader strategy in the migration phase of sub-step S33 by simulating the characteristic of the black kite population following the guidance of the leader, ensuring that the algorithm can effectively utilize the current best solution and guide the search direction.

[0053] S32, deep search stage. As a predator of small mammals and insects in the grassland, the black kite adjusts the angle of its wings and tail according to the wind speed during flight, hovers quietly to observe its prey, and then quickly dives to attack. The black kite optimization algorithm uses this attack behavior as a deep search method. The attack strategies of the black kite can be divided into two types: strategy one: hovering in the air and waiting for attack; strategy two: hovering in the air and looking for prey. Set p = 0.9 as a fixed constant value, generate a random number r, r∈[0, 1], when p < r, execute attack strategy one, and simulate the position update method of this attack strategy as follows:

[0054]

[0055] in In order to enhance the local optimization ability of black kite individuals, the algorithm of the present invention introduces an adaptive weight method in the deep search stage of the black kite optimization algorithm, integrating the adaptive weight factor with the attack strategy. The formula of the adaptive weight w is Where t is the current iteration number and T is the maximum iteration number. The position update method of attack strategy 1 with adaptive weight factor is:

[0056]

[0057] When p ≥ r, execute attack strategy 2, and simulate the position update method of this attack strategy as follows:

[0058]

[0059] Where r is a uniformly distributed random number and r∈[0,1]. The position update method of attack strategy 2 with adaptive weight factor is:

[0060]

[0061] After the position update in the deep exploration phase is completed, each black-winged kite individual P i Compare the fitness values ​​of the previous iteration t-1 with the current iteration t, and select the solution with a better fitness value as the black kite individual P in this iteration t. i The solution.

[0062] S33, breadth search stage. Bird migration behavior is a complex behavior affected by environmental factors such as food supply and climate. In order to adapt to seasonal changes, many birds migrate from cold areas to warm areas in winter to obtain better resources and living environment. Bird migration is usually led by a leader, and the leader's guiding ability has an important impact on the migration of the population. The black kite optimization algorithm proposes a hypothesis based on the migration behavior of birds: if the fitness value of the current population individual Less than the fitness value of a random population individual It means that the leader is no longer suitable to lead the migration direction of the population, so the leader will give up the leadership of the population and join the migrating population; on the contrary, if the fitness value of the current population individual is Greater than the fitness value of a random population individual The leader will continue to guide the population to the destination. Through the leader strategy, the population can dynamically select a better leader to ensure the success rate of migration. The black kite optimization algorithm improves the efficiency of the algorithm's search for the current area by introducing a leader strategy in the breadth search phase and effectively balances the dynamic relationship between breadth search and depth search. The migration behavior of the black kite population is simulated and the mathematical model is established as follows:

[0063] when When , the leader gives up the leadership of the population and joins the migrating population. For this migration strategy, the Black Kite Optimization Algorithm introduces the Cauchy mutation strategy. The Cauchy mutation strategy is a probability distribution strategy that can help the algorithm jump out of the local optimal solution and improve the algorithm's ability to find the global optimal solution. The specific position update method is as follows:

[0064]

[0065] in, represents the leader of the black-winged kite population in the jth dimension in the tth iteration; and denote the position of the i-th black-winged kite in the j-th dimension in the t-th and t+1-th iterations respectively; represents the fitness value of the current position of the i-th black-winged kite in the j-th dimension in the t-th iteration; represents the fitness value of a random black kite in the population in the jth dimension in the tth iteration. Cauthy(0,1) represents the Cauchy mutation, and the probability density function of the one-dimensional Cauchy distribution is:

[0066]

[0067] Where: χ represents the sample value in the probability density function of the one-dimensional Cauchy distribution, δ represents the scale parameter of the half width at half the peak of the probability density function, and μ represents the position parameter that defines the peak position of the distribution.

[0068] When δ=1,μ=0, its probability density function is in standard form:

[0069]

[0070] Cauthy(0,1) in the Black Kite optimization algorithm uses the inverse cumulative distribution function of the Cauchy distribution, and the formula is:

[0071]

[0072] Where r is a uniformly distributed random number and r∈[0,1].

[0073] when When the leader continues to guide the population to the destination, for this migration strategy, the Levy flight method is introduced based on the adaptive weight and Levy flight black kite optimization algorithm and combined with the leader strategy. The Levy flight method can describe the foraging behavior of birds. When food is scarce in a certain area, the flock of birds will migrate to a new area and forage in this area. Therefore, the Levy flight method is a random movement method composed of a short step with a high probability and a long step with a low probability. The present invention combines the Levy flight method with the leader strategy, so that individuals in the population can better jump out of the local optimal solution while being guided by the leader, thereby enhancing the ability of the population to search for the global optimal solution. The method of this position update strategy is:

[0074]

[0075] Where α is the step size control factor, and Levy(β) is the random search path of Levy flight. The combination of Levy flight mode and leader strategy is as follows:

[0076]

[0077] in Γ(x) is the gamma function, also known as Euler's second integral. The gamma function is defined in the real number domain as: The mean is 0 and the variance is Normal distribution of represents the position of the i-th black-winged kite individual on the j-th dimension in the t-th iteration; Represents the position of the leader in the jth dimension in the tth iteration; β∈[0,2] is a random number; usually the value of the step size control factor α is 0.01. In the present invention, the algorithm search adopts an adaptive step size adjustment method. In the early stage of the algorithm search, a larger step size control factor can improve the search speed of the population; in the later stage of the algorithm search, a smaller step size control factor can improve the search accuracy of the population. Adaptive step size control factor Where t is the current number of iterations and T is the maximum number of iterations.

[0078] After the position update in the breadth exploration phase is completed, each black-winged kite individual P i Compare the fitness values ​​of the previous iteration t-1 with the current iteration t, and select the solution with a better fitness value as the black kite individual P in this iteration t. i The solution.

[0079] S34, update the global optimal solution. To solve the problem of updating the position of the global optimal solution, a leader competition mechanism is designed based on adaptive weight and Levy flying black kite optimization algorithm. The fitness value of With the leaders in this iteration The fitness value of In comparison, a better solution is selected as the population leader for the next iteration The expression is:

[0080]

[0081] S4. Solve to obtain the optimal UAV task allocation sequence and calculate the corresponding adaptability, and output the final UAV task allocation plan.

[0082] like Figure 3 As shown, the task scene in this embodiment is a 50×50 square area G, and there are 5 obstacle areas R={R i}, i = 1, 2, 3, 4, 5, the navigation path of the drone cannot pass through the obstacle area. Among them, R1, R2, and R3 are three rectangular obstacle areas. R1 is a rectangular area with a center of (11.5, 17.5), a width of 3, and a height of 5. R2 is a rectangular area with a center of (32, 38), a width of 4, and a height of 6. R3 is a square area with a center of (43, 15), a width of 6, and a height of 6. R4 is a circular area with a center of (30, 20) and a radius of 4. R5 is a right semicircular area with a center of (10, 40) and a radius of 5. There are 15 tasks to be executed in the mission scene G. Each task point corresponds to a coordinate position in the two-dimensional mission scene G. The coordinate positions of the 15 tasks to be executed are randomly generated according to a uniform distribution. The maximum cruising range of the drone is W max is 300.

[0083] In this embodiment, the number of iterations T of the adaptive weight and Levy flying black kite optimization algorithm is 5000, the number of individuals in the population pop is 500, and the dimension dim of the solution to the problem to be solved corresponds to the number of tasks to be performed is 15. The chaotic mapping method used in initialization is the a-order Chebyshev chaotic mapping.

[0084] According to the above mission scenarios and parameter settings, the optimal solution of the UAV task allocation scheme calculated based on the adaptive weight and Levy flying black kite optimization algorithm is as follows: Figure 4 As shown, the convergence result of the algorithm convergence curve is as follows Figure 5The feasibility and effectiveness of the UAV task allocation method based on adaptive weight and Levy flying black kite optimization algorithm are verified through the simulation results of this embodiment.

[0085] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm, characterized in that: Including: S1. Establish a UAV mission allocation model under range constraints according to the UAV mission allocation scenario. S2. Use the black-winged kite optimization algorithm to solve the UAV mission allocation model established in step S1 to obtain the final UAV mission allocation plan. The specific implementation process includes the following sub-steps: S21. Define the initial parameter set of the black-winged kite optimization algorithm and initialize the black-winged kite population based on chaotic mapping. S22. Construct a meta-heuristic algorithm model according to the attack behavior and migration behavior of the black-winged kite. An adaptive weight method is introduced in the depth search stage of the meta-heuristic algorithm, and the Lévy flight method is introduced in the breadth search stage of the meta-heuristic algorithm and the Lévy flight method is combined with the leader strategy. S23. Iteratively solve by optimizing the meta-heuristic algorithm to obtain the optimal task allocation sequence.

2. A method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm according to claim 1, characterized in that: The implementation process of step S1 includes the following sub-steps: S11. Assume that the task allocation scene graph G is an area of ​​size m×n. There are k tasks waiting to be executed in the task scene graph G. The set of tasks is denoted as S={s i }, i = 1, 2...k, where S includes all pending tasks and the pending tasks are not repeated, that is, s i ≠s j , i≠j, i, j=1, 2...k; S12, set up obstacles, that is, no-fly zones R. The setting of obstacles includes the center position of the obstacles (x R ,y R ) and the size of obstacles; S13. Define the task execution matrix X, where X is a k×k matrix, and X(i, j) represents the task execution order from task i to task j, where i, j = 1, 2...k, i≠j. When X(i, j) = 1, it means the task execution order is task i→task j. When the task execution order does not satisfy task i→task j, then X(i, j) = 0. S14, define the location set A of the task points = {a i }, a i Represents the position of task point i in the task allocation scenario; S15. Define the lines between the positions of the task points as edges, and the edge set E = {e ij }, i, j = 1, 2...k, i ≠ j, where e ij represents the line between the position of task point i and the position of task point j; S16. Let d(i, j) be the distance between the position of task point i and the position of task point j, and d(i, j) = d(j, i). S17. According to the task allocation scenario, define the fitness value of the objective function as the UAV’s mileage W, that is, F = W, and the UAV’s maximum mileage is W max ; S18. According to the scenario constraints, task requirements and the principle of minimizing the fitness function, establish the objective function and its constraints as:

3. A method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm according to claim 2, characterized in that: The implementation process of step S21 is as follows: S211. Define the initial population size as pop, the maximum number of iterations of the population as T, and the dimension of the problem to be solved as dim. S212, initialize the position of the population; set the upper and lower limits of the population initial position, and randomly generate black-winged kite individuals P uniformly distributed in the population initial position generation area i : P i =L+r×(UL), 1≤i≤pop Where L and U are the upper and lower bounds of the solution space respectively, r is a random vector uniformly distributed in the interval [0, 1], and pop is the population size. The initial population position is represented in the following matrix form: Where dim is the dimension size of the given task allocation problem. S213. Implement the chaotic mapping of the initial population position; perform chaotic mapping on the initial population position obtained after initialization in sub-step S212, and re-initialize the population position using the a-order Chebyshev chaotic mapping strategy.

4. A method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm according to claim 3, characterized in that: The formula of the a-order Chebyshev chaotic mapping strategy is as follows:

5. A method for allocating unmanned aerial vehicle tasks based on an improved black kite optimization algorithm according to claim 4, characterized in that: The implementation process of the depth search stage in step S22 is as follows: A1. There are two attack strategies for the black-winged kite. Strategy 1: Hover in the air and wait for an attack. Strategy 2: Hover in the air and search for prey. Set p as a fixed constant value, generate a random number r, r∈[0, 1]. When p < r, execute attack strategy 1, and the position update method simulating this attack strategy is as follows: in A2. The formula for the adaptive weight w is: Where t is the current iteration number and T is the maximum iteration number. A3. The position update method of attack strategy 1 incorporating the adaptive weight is: A4. When p ≥ r, execute attack strategy 2, and the position update method simulating this attack strategy is as follows: Where r is a uniformly distributed random number and r∈[0, 1]. A5. The position update method of attack strategy 2 incorporating the adaptive weight factor is: A6. After the position update in the deep exploration phase is completed, each black-winged kite individual P i Compare the fitness values ​​of the previous iteration t-1 with the current iteration t, and select the solution with a better fitness value as the black kite individual P in this iteration t. i The solution.

6. The method for allocating unmanned aerial vehicle tasks based on the improved black kite optimization algorithm according to claim 5 is characterized in that: The implementation process of the breadth search phase in step S22 is: B1. If the fitness value of the current population individual Less than the fitness value of a random population individual The leader gives up the leadership of the population, joins the migrating population, introduces the Cauchy mutation strategy, and updates the position as follows: in, represents the leader of the black-winged kite population in the jth dimension in the tth iteration; and denote the position of the i-th black-winged kite in the j-th dimension in the t-th and t+1-th iterations respectively; represents the fitness value of the current position of the i-th black-winged kite in the j-th dimension in the t-th iteration; represents the fitness value of a random black kite in the population in the jth dimension in the tth iteration; Cauthy (0, 1) represents Cauchy mutation; B2. If the fitness value of the current population individual Greater than or equal to the fitness value of a random population individual The leader continues to guide the population to the destination. For this migration strategy, the Levy flight method is introduced and combined with the leader strategy; the position update strategy is as follows: Where α is the step size control factor, Levy(β) is the random search path of Levy flight; Here’s how the Levi flying method is combined with the leader’s strategy: in Γ(x) is the gamma function, also known as Euler's second integral. The gamma function is defined in the real number domain as: represents the position of the i-th black-winged kite individual on the j-th dimension in the t-th iteration; represents the position of the leader in the jth dimension in the tth iteration; β∈[0,2] is a random number; α is the step size control factor; B3. After the position update in the breadth exploration phase is completed, each black-winged kite individual P i Compare the fitness values ​​of the previous iteration t-1 with the current iteration t, and select the solution with a better fitness value as the black kite individual P in this iteration t. i The solution.

7. The method for allocating unmanned aerial vehicle tasks based on the improved black kite optimization algorithm according to claim 6 is characterized in that: Where t is the current number of iterations and T is the maximum number of iterations.

Citation Information

Cited By

  • Image encryption and decryption method and system based on waterfowl-plant hyperchaotic system

    CN119484722A

  • Internet-of-vehicles computing unloading and resource allocation method based on improved black-wing plinary optimization algorithm

    CN119815417A

  • Unmanned aerial vehicle three-dimensional path planning method based on multi-strategy improved black-wing optimization algorithm

    CN120628130A

  • Unloading method in ultra-dense millimeter wave MEC network

    CN121037916A

  • Vehicle intelligent navigation method and system

    CN121117659A