Unmanned aerial vehicle task allocation method based on Levy flight parrot optimization algorithm

Abstract

The invention discloses an unmanned aerial vehicle task allocation method based on a Levy flight parrot optimization algorithm, and the method comprises the following steps: S1, carrying out the modeling of an unmanned aerial vehicle task allocation scene based on a graph theory method, and building a task allocation mathematical model; s2, setting initialization parameters of a parrot optimization algorithm; s3, performing loop iteration on the optimization process: setting solving parameters and iteration ending conditions, recording a globally optimal solution of each iteration in the solution space, and performing parameter and position updating after one-time loop ending until the iteration ending conditions are met; and S4, outputting the distribution sequence, and calculating the minimum voyage cost of the whole process. On the premise of considering the voyage constraint of the unmanned aerial vehicle, the method is more suitable for the actual task condition, the chaotic mapping is used for optimizing the initialized population, the Levy flight strategy is adopted in the optimization process, the global search ability is higher, and the method is more suitable for solving the practical problem; the problem of unmanned aerial vehicle task allocation can be solved to a certain extent.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned aerial vehicle (UAV) mission planning, and mainly relates to a UAV mission allocation method based on the Levy Flying Parrot optimization algorithm. Background Art

[0002] In recent years, drone technology has achieved rapid development. Drones are small in size, low in cost, highly maneuverable, and highly safe and effective. They have been widely used in many fields, such as military, agriculture, logistics, disaster relief, environmental monitoring, and infrastructure inspection. The high flexibility, low cost, and versatility of drones make them an indispensable tool in modern society. However, with the increase in the number of drones and the complexity of missions, how to efficiently allocate and plan drone missions has become an important research topic. In order to improve the efficiency and accuracy of drone mission planning, optimization algorithms are widely used in task allocation and path planning. Optimization algorithms use mathematical models and calculation methods to find the best solution under given constraints, thereby maximizing mission efficiency, minimizing resource consumption, and improving the reliability of mission completion.

[0003] Task allocation based on optimization algorithms can be divided into two categories: one is based on traditional optimization techniques, and the other is a meta-inspired algorithm based on swarm intelligence. Traditional methods such as the Hungarian algorithm, artificial potential field method, and MILP algorithm can usually find the optimal task allocation results that meet the constraints, but the computational cost is very high, and the convergence speed is slow when the computational scale increases. Meta-inspired algorithms based on swarm intelligence are better at finding better solutions in large-scale problems. This type of algorithm guides the search by summarizing some rules or empirical knowledge, and can complete the optimization within a limited time. Task allocation algorithms inspired by swarm intelligence have the advantages of low computational complexity, fast convergence speed, and wide application range. Some typical swarm intelligence-inspired optimization algorithms currently under research include particle swarm optimization, wolf group algorithm, and ant colony algorithm. However, the relatively mature swarm intelligence algorithms currently proposed still have some problems to be solved.

[0004] General metaheuristic algorithms usually include two main phases: exploration and exploitation. Striking the right balance between these phases is a challenging but crucial task that directly affects the performance of the entire algorithm. In addition, some current algorithms may have limited optimization capabilities because they search according to certain rules and experience, and local optimal problems are prone to occur during the search and solution process. This problem is more likely to occur under complex constraints, when the solution space is very complex and there are multiple local optimal solutions. In addition, parameters have an important impact on the optimization performance of most algorithms. It is extremely challenging to determine the optimal parameters for a specific optimization problem. When the currently proposed algorithm lacks a comprehensive qualitative analysis and does not solve the problem of parameter sensitivity, it may be difficult to solve complex problems thoroughly and effectively. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for unmanned aerial vehicle task allocation based on the Levy Flying Parrot optimization algorithm. Under the premise of considering the range constraints of the unmanned aerial vehicle, the method is more in line with the actual task situation, and uses chaotic mapping optimization to initialize the population. In addition, the Levy flight strategy is adopted during the optimization process. It has stronger global search capabilities and is more suitable for solving practical problems. It can solve the problem of unmanned aerial vehicle task allocation to a certain extent.

[0006] The objective of the present invention is achieved through the following technical solution: a method for allocating unmanned aerial vehicle tasks based on the Levy Flying Parrot optimization algorithm, comprising the following steps:

[0007] S1. Model the UAV task allocation scenario based on graph theory, determine the fitness function and its constraints, and establish a mathematical model for task allocation. The sub-steps are as follows:

[0008] S11. Assume that the mission scenario graph G is a square area of ​​m×m, there are n tasks to be performed in G, the union of the n tasks is denoted as N, and the drone must perform all the tasks after taking off;

[0009] S12, use x(i,j) to indicate that the order of task execution is from task i to task j; i,j∈V,i≠j; where x(i,j)=1 indicates that the execution order is i→j, otherwise it is 0;

[0010] S13, define the edge set E = {e ij}, where e ij represents the line segment with task point i and task point j as endpoints;

[0011] S14. Use d(i,j) to represent the distance from task i to task j, d(i,j)=d(j,i);

[0012] S15. Define that there are x dangerous areas in the task scenario G, and the union of the x dangerous areas is denoted as D; D = D i , 1≤i≤k, the drone cannot fly into the dangerous area when performing the mission;

[0013] S16. Define the UAV's cruising range as W;

[0014] S17. According to the principle of minimizing the distance, the objective function and its constraints are established as follows:

[0015]

[0016] S2. Set the initialization parameters of the Parrot optimization algorithm;

[0017] S3, iterate the optimization process in a loop: set solution parameters and iteration end conditions, record the global optimal solution of each iteration in the solution space, that is, the optimal position; update the parameters and position after one cycle; until the iteration end conditions are met, execute S4;

[0018] S4. Iterate and solve to obtain the optimal allocation of UAVs to multiple tasks, output the allocation sequence, and calculate the minimum flight cost of the entire process.

[0019] The present invention is aimed at the specific object of drones, and abstracts the scenarios where drones need to perform multiple tasks such as urban logistics distribution or battlefield search as task allocation problems. First, a mathematical model of the task allocation problem is established, and based on the shortest flight constraint, a solution is designed based on the Levy Flying Parrot optimization algorithm. This algorithm is not only based on the Levy flight strategy, but also can make the optimization problem better jump out of the local optimum and has extremely high optimization performance. At the same time, in the process of initializing the population, the Logistic chaotic mapping is also combined to optimize the originally randomly generated population, increase the randomness, better simulate the distribution of the population in the entire solution space, improve the global search ability, and also make the algorithm converge faster, reducing the calculation time and calculation cost on large-scale problems. Therefore, the drone task allocation method based on the Levy Flying Parrot optimization algorithm proposed in the present invention has great advantages over traditional algorithms when facing realistic complex problems, which are mainly reflected in the following aspects:

[0020] 1. The present invention proposes a novel UAV task allocation method based on the Levy flying parrot optimization algorithm. Compared with the traditional ant colony algorithm, wolf group algorithm, particle swarm algorithm, etc., the present invention introduces a new optimization algorithm based on swarm intelligence: the parrot algorithm.

[0021] 2. In the current related research, most of the swarm intelligence optimization algorithms use uniform distribution to initialize the population. This method cannot simulate the complex distribution of the individual natural environment, and will also reduce the randomness of the algorithm's global search and slow the convergence speed. To address this problem, the present invention uses chaotic mapping for initialization, which increases the randomness and unpredictability of the population and enables the algorithm to converge faster.

[0022] 3. In the optimization mechanism of the Parrot optimization algorithm, unlike the traditional optimization algorithm which finds it difficult to achieve the right balance between the exploration and development stages, the Parrot algorithm of the present invention has no clear distinction between the task exploration and development (E&P) stages, but enhances the optimization capability.

[0023] 4. In the behavioral part of the algorithm, random exploration is performed through the Levy flight strategy, and the random step size that obeys the Levy distribution is introduced to adjust the position of the search point so as to better explore the potential solution space during the search process. This randomness and long-distance movement help the algorithm jump out of the local minimum point, while increasing the diversity of the search space, so as to better explore the area of ​​the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The flowchart of the method for allocating unmanned aerial vehicle tasks based on the Levy Flying Parrot optimization algorithm of the present invention;

[0025] Figure 2 is a flow chart of the iterative portion of the present invention;

[0026] Figure 3 It is a task scenario diagram of an embodiment;

[0027] Figure 4 The optimal task execution sequence scheme of the embodiment;

[0028] Figure 5 It is a graph of algorithm convergence of the embodiment. DETAILED DESCRIPTION

[0029] In order to make the use method, solution and advantages of the present invention more clearly understood, the specific implementation method and operation steps of the present invention are further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] like Figure 1 As shown, a method for allocating unmanned aerial vehicle tasks based on the Levy Flying Parrot optimization algorithm of the present invention comprises the following steps:

[0031] S1. Model the UAV task allocation scenario based on graph theory, determine the fitness function and its constraints, and establish a mathematical model for task allocation. The sub-steps are as follows:

[0032] S11. Assume that the mission scenario graph G is a square area of ​​m×m, there are n tasks to be performed in G, the union of the n tasks is denoted as N, and the drone must perform all the tasks after taking off;

[0033] S12, use x(i,j) to indicate that the order of task execution is from task i to task j; i,j∈V,i≠j; where x(i,j)=1 indicates that the execution order is i→j, otherwise it is 0;

[0034] S13, define the edge set E = {e ij}, where e ij represents the line segment with task point i and task point j as endpoints;

[0035] S14. Use d(i,j) to represent the distance from task i to task j, d(i,j)=d(j,i);

[0036] S15. Define that there are x dangerous areas in the task scenario G, and the union of the x dangerous areas is denoted as D; D = D i , 1≤i≤k, the drone cannot fly into the dangerous area when performing the mission;

[0037] S16. Define the UAV's cruising range as W;

[0038] S17. According to the principle of minimizing the distance, the objective function and its constraints are established as follows:

[0039]

[0040] S2, set the initialization parameters of the parrot optimization algorithm, including population size, number of iterations, type of chaotic mapping, etc.; the specific iteration process is as follows Figure 2 As shown, there are the following sub-steps:

[0041] S21, define the initial population size G, the number of population iterations is T;

[0042] S22, Initialize population position: randomly generate G parrot individuals, all parrot individuals form an initialized population X:

[0043]

[0044] Where lb and ub are the lower and upper bounds of the solution space; the solution space refers to the interval composed of all feasible solutions, which refers to the order of tasks to be traversed in the present invention. rand(0,1) is a random number in the interval [0,1], represents the initial position of the ith parrot; in the present invention, a task execution sequence is regarded as a parrot individual;

[0045] S23, chaotic mapping to reinitialize the population position: For the population X initialized in S22, use the Logistic chaotic mapping strategy to reinitialize:

[0046]

[0047] μ is an adjustable parameter. When μ∈[3.57,4], the mapping will exhibit chaotic behavior, that is, the final state of X will not be repeated, but will be taken over the entire interval with equal probability.

[0048] S3, iterate the optimization process in a loop: set solution parameters and iteration end conditions, record the global optimal solution of each iteration in the solution space, that is, the optimal position; update the parameters and position after one cycle ends; execute S4 until the iteration end conditions are met.

[0049] The above step S3 has the following sub-steps:

[0050] S31. Each individual in the parrot population randomly performs one of the four behaviors: foraging, staying, communicating, and fear of strangers.

[0051] Before performing these four actions, the following parameters need to be updated: represents the average position within the population at the tth iteration, expressed by the following formula:

[0052]

[0053] At the same time, set a variable St:

[0054] St = randi([1,4])

[0055] For any individual in an iteration, different St values ​​will be randomly generated. When St=1, the foraging behavior is performed and the process goes to S32; when St=2, the staying behavior is performed and the process goes to S33; when St=3, the communication behavior is performed and the process goes to S34; when St=4, the fear of strangers behavior is performed and the process goes to S35;

[0056] S32. Foraging behavior: In the foraging behavior of parrots, they mainly estimate the approximate location of the food by observing the location of the food or considering the location of the owner, and then fly to their respective locations. Therefore, the position movement follows the following equation:

[0057]

[0058] Indicates the current location. Indicates the location of subsequent updates, Represents the average position within the current population;

[0059] Levy(dim) represents Levy distribution, which is used to describe the flight of parrots; X best It indicates the position of the optimal solution found from initialization to the current number of iterations, and also indicates the current position of the master; t indicates the current number of iterations, and T indicates the maximum number of iterations. When t<T, iteration will be performed. Represents movement based on its position relative to its owner. It means that a single parrot individual updates its own position by observing the position of the entire population and further determining the direction of the food.

[0060] The Levy distribution is obtained based on the following rule, where γ is assigned a value of 1.5:

[0061]

[0062] The parameters μ and v follow a normal distribution with a mean of 0 and a variance of dim, and the parameter σ can be calculated using the gamma function.

[0063] S33. Staying behavior. The Piroa Morena parrot is a highly social creature. Its staying behavior mainly consists of suddenly flying to any part of the owner's body, where it remains stationary for a period of time. This process is represented by:

[0064]

[0065] ones(1,dim) represents a vector of all 1s of dimension dim, X best Levy(dim) indicates the process of flying to the owner, rand(0,1) and ones(1,dim) indicates randomly stopping at a certain part of the owner's body;

[0066] S34. Communication behavior. As a natural social animal, it is characterized by close communication within the group. This communication behavior includes flying to the group and communicating without flying to the group. In the Pilora Morina parrot, it is assumed that the probability of these two behaviors is equal, and the average position of the current group is used to symbolize the center of the group. This behavior is represented by:

[0067]

[0068] in It means that individuals join the parrot group to communicate;

[0069] represents the process in which the individual flies away immediately after communication. Both behaviors are feasible, so P is randomly generated in the range of [0,1] to achieve this;

[0070] S35. Fearful behavior towards strangers. Generally speaking, birds show an innate fear of strangers, and the Pilora Morena parrot is no exception. In order to find a safe environment, they keep a distance from unfamiliar people and seek safety from their owners, which is expressed as:

[0071]

[0072] It shows the process of the Piroa Morena parrot flying back to its owner.

[0073] It indicates the process of moving away from strangers.

[0074] In each iteration process, the optimal solution in the current iteration number will be recorded. After the iteration is completed, the optimal solution of this iteration process will be compared with the current global optimal solution. If the fitness function value of the current optimal solution is less than the fitness function value of the global optimal solution, the global optimal solution will be updated. Otherwise, the global optimal solution will not be updated, and the next iteration will continue until the maximum number of iterations is reached, and the global optimal solution obtained will be used as the optimal allocation plan.

[0075] S4. Iterate and solve to obtain the optimal allocation of UAVs to multiple tasks, output the allocation sequence, and calculate the minimum flight cost of the entire process.

[0076] In this embodiment, the task scene is a 100×100 square area G, and there are 7 unreachable areas D={D i}, i=1,2,3,4,5,6,7, D1 to D4 are four rectangular danger areas, the four vertex coordinates of D1 are {(5,15),(15,15),(5,25),(15,25)}, which is a square area with a side length of 10, the four vertex coordinates of D2 are {(10,70),(35,70),(35,90),(10,90)}, which is a rectangular area with a length of 25 and a width of 20. The four vertex coordinates of D3 are {(60,20),(75,20),(75,45),(60,45)}, which is a rectangular area with a length of 15 and a width of 25. The four vertex coordinates of D4 are {(80,70),(90,70),(90,90),(80,90)}, which is a rectangular area with a length of 10 and a width of 20. D5 is an irregular danger zone, and its four vertices have coordinates {(30,30), (42,52), (38,63), (15,40)}. D6 is a circular danger zone, and its center has coordinates (60,70) and radius 15. D7 is a triangular zone, and its three points have coordinates {(80,10), (95,20), (80,30),}. There are 20 tasks to be performed in G, and each task location corresponds to a two-dimensional coordinate. The task scene is as follows: Figure 3 As shown, the UAV range W is 1000.

[0077] In this example, the total number of iterations of the optimization model is T = 2000, the initial population size is N = 300, and the chaotic mapping control parameter is selected as μ = 4. According to the above task scenarios and parameter settings, the optimal task allocation sequence is calculated as follows: Figure 4 As shown, the algorithm convergence curve is as follows Figure 5 As shown in the figure, it can be seen that the allocation sequence converges to the global optimum. The simulation results confirm that the UAV task allocation method based on the Levy Flying Parrot optimization algorithm can achieve the optimal allocation of UAV tasks.

[0078] 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 protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. The UAV task allocation method based on Levy Flying Parrot optimization algorithm is characterized by: The following steps are involved: S1. Model the UAV task allocation scenario based on graph theory, determine the fitness function and its constraints, and establish a mathematical model for task allocation. The sub-steps are as follows: S11. Assume that the mission scenario graph G is a square area of ​​m×m, there are n tasks to be performed in G, the union of the n tasks is denoted as N, and the drone must perform all the tasks after taking off; S12, use x(i,j) to indicate that the execution order of tasks is from task i to task j; i,j∈V,i≠j; where x(i,j)=1 indicates that the execution order is i→j, otherwise it is 0; S13. Define edge set E = {e ij }, where e ij represents the line segment with task point i and task point j as endpoints; S14. Use d(i,j) to represent the distance from task i to task j, d(i,j)=d(j,i); S15. Define that there are x dangerous areas in the task scenario G, and the union of the x dangerous areas is denoted as D; D = D i , 1≤i≤k, the drone cannot fly into the dangerous area when performing the mission; S16. Define the UAV's cruising range as W; S17. According to the principle of minimizing the distance, the objective function and its constraints are established as follows: S2. Set the initialization parameters of the Parrot optimization algorithm; S3, iterate the optimization process in a loop: set solution parameters and iteration end conditions, record the global optimal solution of each iteration in the solution space, that is, the optimal position; update the parameters and position after one cycle; until the iteration end conditions are met, execute S4; S4. Iterate and solve to obtain the optimal allocation of UAVs to multiple tasks, output the allocation sequence, and calculate the minimum flight cost of the entire process.

2. The method for allocating unmanned aerial vehicle tasks based on the Levy Flying Parrot optimization algorithm according to claim 1 is characterized in that: In S2, there are the following sub-steps: S21, define the initial population size G, the number of population iterations is T; S22, Initialize population position: Randomly generate G parrot individuals, all parrot individuals form an initialized population X: Where lb and ub are the lower and upper bounds of the solution space; rand(0,1) is a random number in the interval [0,1]. represents the initial position of the ith parrot; a task execution order as a parrot individual; S23, chaotic mapping to reinitialize the population position: For the population X initialized in S22, use the Logistic chaotic mapping strategy to reinitialize:

3. The method for allocating unmanned aerial vehicle tasks based on the Levy Flying Parrot optimization algorithm according to claim 1 is characterized in that: In the above-mentioned S3, there are the following sub-steps: S31. Each individual in the parrot population randomly performs one of the four behaviors: foraging, staying, communicating, and fear of strangers. Before performing these four actions, the following parameters need to be updated: represents the average position within the population at the tth iteration, expressed by the following formula: At the same time, set a variable St: St = randi([1,4]) For any individual in an iteration, different St values ​​will be randomly generated. When St=1, the foraging behavior is performed and the process goes to S32; when St=2, the staying behavior is performed and the process goes to S33; when St=3, the communication behavior is performed and the process goes to S34; when St=4, the fear of strangers behavior is performed and the process goes to S35; S32, foraging behavior, position movement follows the following equation: Indicates the current location. Indicates the location of subsequent updates, Represents the average position within the current population; Levy(dim) represents Levy distribution, which is used to describe the flight of parrots; X best It represents the position of the optimal solution found from initialization to the current number of iterations; t represents the current number of iterations; S33, stay behavior, expressed as: ones(1,dim) represents a vector of all 1s of dimension dim, X best Levy (dim) indicates the process of flying to the host, rand(0,1)·ones(1,dim) means randomly stopping at a certain part of the owner's body; S34, communication behavior, expressed as: S35. Fear of strangers, expressed as: In each iteration process, the optimal solution in the current iteration number will be recorded. After the iteration is completed, the optimal solution of this iteration process will be compared with the current global optimal solution. If the fitness function value of the current optimal solution is less than the fitness function value of the global optimal solution, the global optimal solution will be updated. Otherwise, the global optimal solution will not be updated, and the next iteration will continue until the maximum number of iterations is reached, and the global optimal solution obtained will be used as the optimal allocation plan.

Images