Distributed multi-unmanned aerial vehicle cooperative route planning method and system in complex environment

Abstract

The invention discloses a distributed multi-unmanned aerial vehicle cooperative flight path planning method and system in a complex environment. The method comprises the following steps: establishing an unmanned aerial vehicle path planning aggregation fitness function model based on a target constraint condition; a multi-population particle swarm algorithm is improved, multi-population dynamic division is carried out on the unmanned aerial vehicle path planning aggregation fitness function model, and multiple populations comprise a dominant population, a local particle swarm and an inferior population; optimizing the dominant population and the local particle swarm by adopting a preset Levy flight greedy optimizing strategy and a preset adaptive parameter updating strategy, and optimizing the disadvantage population by adopting a preset probabilistic hybrid variation strategy; and according to an optimization result, carrying out distributed multi-unmanned aerial vehicle cooperative flight path planning. The efficiency and safety of unmanned aerial vehicle cooperative flight path planning are improved, and it is ensured that multiple unmanned aerial vehicles can efficiently and safely complete tasks in a complex environment.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) trajectory planning, and in particular to a distributed multi-UAV collaborative trajectory planning method and system under complex environments. Background Art

[0002] Multi-rotor drones have broad application prospects, and are particularly suitable for high-risk application scenarios that lack effective prior information, such as power line inspections, accident site search and rescue, and environmental reconnaissance. In these environments, manual work processes are often cumbersome and time-consuming, and are often faced with sudden dangers, which not only creates obstacles for information collection, but also poses a potential threat to the lives of workers. Multi-rotor drones, with their excellent three-dimensional spatial motion capabilities, have broken through the limitations of complex environments on work and demonstrated great application potential.

[0003] With the complexity of flight missions and the diversification of battlefield environments, the efficiency of single UAV mission execution is constrained by its own rigid conditions. Multi-UAV collaborative technology is based on a single UAV, adding collaborative goals and time-space constraints to the UAV, so that multiple UAVs can work together to complete a complex task. The research on multi-UAV trajectory planning has far-reaching significance for guiding UAVs to perform reconnaissance, inspection and other tasks in the future, as well as for the development of UAV autonomous flight technology. With the increase in the number of drones, the difficulty of UAV path planning increases exponentially, and the quality of the path planning algorithm greatly affects the safety and efficiency of UAV work.

[0004] The particle swarm algorithm is a random optimization method based on swarm intelligence, which is inspired by the foraging behavior of bird flocks. Although the particle swarm algorithm has the advantages of easy implementation and strong parallelism, it also has potential problems, including slow convergence and easy to fall into local optimal solutions. In the research on these problems, some scholars proposed a rule of linear decreasing inertia weight at the level of parameter optimization, trying to achieve a balance between exploration and development. However, this method is not very effective in solving the problem of easy to fall into local optimal solutions. On the other hand, some research teams have tried to integrate genetic algorithms, wolf pack algorithms and particle swarm algorithms in order to improve the performance of the algorithms, but there are generally problems of cumbersome parameter adjustment and algorithm coupling. In addition, the decision-making process of most algorithms mainly adopts greedy decision-making, which ignores the dynamic characteristics of the aircraft, resulting in a decrease in global efficiency and a sharp increase in potential risks. Therefore, it is necessary to focus on proposing new optimization strategies and comprehensively consider the dynamic characteristics of drones to improve the global efficiency of the algorithm and reduce risks. Summary of the invention

[0005] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a distributed multi-UAV collaborative trajectory planning method and system in a complex environment, which can solve the problems mentioned in the background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a method for distributed multi-UAV collaborative trajectory planning in a complex environment, comprising:

[0010] Establishing an aggregated fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV spatiotemporal coordination constraints;

[0011] Improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, including dominant populations, local particle populations and disadvantaged populations;

[0012] The preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy are used to optimize the dominant population and the local particle swarm, and the preset probabilistic mixed mutation strategy is used to optimize the inferior population;

[0013] Based on the optimization results, distributed multi-UAV collaborative trajectory planning is carried out.

[0014] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under a complex environment described in the present invention, the UAV path length constraint includes:

[0015] The UAV path length constraint includes the sum of the length costs of different path segments. i The length cost function is as follows:

[0016]

[0017] Among them, P ij =(x ij ,y ij ,z ij ) are the path node coordinates, Represents the distance between two adjacent path nodes, i and j are path nodes, and n represents the total number of nodes.

[0018] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under complex environment described in the present invention, the UAV trajectory smoothness constraint includes:

[0019] The UAV trajectory smoothness constraint includes calculating the yaw angle and the climb angle, They are and Projection on the Oxy plane, The angle formed is the yaw angle ψ ij ,make is the unit vector of the z-axis, the projection vector is calculated as follows:

[0020]

[0021] The calculation formula of yaw angle is:

[0022]

[0023] Among them, the climb angle is the angle between the drone's ascending trajectory and the horizontal line. Its projection on the horizontal plane The angle formed is the climbing angle γ ij ;

[0024]

[0025] The UAV trajectory smoothness constraint is expressed as:

[0026]

[0027] Among them, ψ ij represents the yaw angle, γ ij Indicates the climb angle.

[0028] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under complex environment described in the present invention, the UAV safety constraints include:

[0029] The drone safety constraints include a given flight segment Considering the radius R of the drone and the danger distance S from the danger zone to the collision zone, the safety constraint of the drone in this flight segment is:

[0030]

[0031] In which, let K be the set of all obstacles, and let the center of the obstacle projection be C k , with a radius of R k, the radius of the drone is R, and the radius of the collision area is R c , the distance between the drone and the projection center is d k .

[0032] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under complex environment described in the present invention, the UAV spatiotemporal collaborative constraints include:

[0033] Assume that the flight speed range of the drone is The length of the mth track is L m , then the time interval for the drone to reach the target point is In order to ensure that multiple drones can reach the designated target point at the same time, it is necessary to ensure that the time interval T between any two drones reaching the target point m With T j If there is an intersection, then the calculation formula for the track m violating the time constraint is as follows:

[0034]

[0035] When the intersection of the flight time between any two drones is empty, it means that the two drones cannot meet the constraint of arriving at the same time, and the number of time constraint violations increases by 1. On the contrary, if the intersection of the flight time is a non-empty subset, it means that the two drones can achieve the condition of arriving at the same time, and the number of time violations increases by 0.

[0036] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under a complex environment described in the present invention, the UAV spatiotemporal collaborative constraints also include:

[0037] If the minimum distance between any two drones is d m,j , in order to ensure that the space constraints between UAVs are met, the calculation formula for the track m violating the space constraints is as follows:

[0038]

[0039] From the above formula, we can see that when the minimum distance between the tracks of any two drones is less than the allowed minimum safety distance d min , it means that there is no collision between the two tracks. At this time, the number of violations of the space constraint is increased by 0. Otherwise, the number of violations of the space constraint is increased by 1.

[0040] After calculating the number of times the track violates the time and space constraints, the total number of violations of the time and space constraints by track m is:

[0041] W m =WT m +WP m .

[0042] As a preferred solution of the distributed multi-UAV collaborative trajectory planning method under complex environment described in the present invention, wherein: the establishment of the UAV path planning aggregation fitness function model based on target constraints includes:

[0043] The UAV path planning aggregation fitness function model is expressed as:

[0044]

[0045] Among them, λ k is the weight coefficient and λ1+λ2+λ3=1, μ is the spatiotemporal coordination penalty coefficient.

[0046] In a second aspect, the present invention provides a distributed multi-UAV collaborative trajectory planning system in a complex environment, comprising:

[0047] A model building module is used to establish an aggregate fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV space-time coordination constraints;

[0048] An algorithm improvement module is used to improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, wherein the multiple populations include a dominant population, a local particle population, and a disadvantaged population;

[0049] The optimization module is used to optimize the dominant population and the local particle swarm by using the preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy, and to optimize the inferior population by using the preset probabilistic mixed mutation strategy;

[0050] The trajectory planning module is used to perform distributed multi-UAV collaborative trajectory planning based on the optimization results.

[0051] In a third aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned method when executing the computer program.

[0052] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the method described above when executed by a processor.

[0053] Compared with the prior art, the invention has the following beneficial effects: the invention proposes a method and system for distributed multi-UAV collaborative trajectory planning in a complex environment, establishes an aggregated fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV spatiotemporal collaborative constraints; improves the multi-population particle swarm algorithm, and dynamically divides the UAV path planning aggregated fitness function model into multiple populations, wherein the multiple populations include dominant populations, local particle populations, and inferior populations; adopts a preset Levy flight greedy optimization strategy and a preset adaptive parameter update strategy to optimize the dominant population and the local particle population, and adopts a preset probabilistic mixed mutation strategy to optimize the inferior population; and performs distributed multi-UAV collaborative trajectory planning based on the optimization results. The efficiency and safety of UAV collaborative trajectory planning are improved, and it is ensured that multiple UAVs can complete tasks efficiently and safely in a complex environment.

[0054] Specifically, by establishing an aggregated fitness function model for UAV path planning, multiple constraints can be comprehensively considered so that the planned trajectory not only meets the requirements of path length, trajectory smoothness and safety, but also meets the spatiotemporal coordination constraints. The improved multi-population particle swarm algorithm can effectively avoid falling into the local optimal solution and improve the global search capability, thereby obtaining a better trajectory planning result.

[0055] In addition, the preset Levy flight greedy optimization strategy and adaptive parameter updating strategy can further improve the optimization efficiency of the dominant population and the local particle swarm, while the probabilistic mixed mutation strategy helps the disadvantaged population to escape from the local optimum, increase population diversity and avoid premature convergence.

[0056] Ultimately, through distributed multi-UAV collaborative trajectory planning, it can ensure that multiple UAVs can effectively coordinate when performing tasks, avoid mutual interference and collision, and improve the success rate and efficiency of mission execution.

[0057] In summary, the present invention is not only technologically innovative, but also has significant practical value in practical applications, and provides an effective solution for the collaborative operation of multiple UAVs in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0059] Figure 1A method flow chart of a method and system for distributed multi-UAV collaborative trajectory planning in a complex environment provided by an embodiment of the present invention;

[0060] Figure 2 A flowchart of an improved multi-objective particle swarm algorithm for a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of updating reference particle information of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by one embodiment of the present invention;

[0062] Figure 4 A schematic diagram of the yaw and climb angles of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0063] Figure 5 A schematic diagram of the collision threat range of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0064] Figure 6 A Levy flight trajectory of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0065] Figure 7 A schematic diagram of the hybrid particle generation principle of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0066] Figure 8 A multi-UAV path map planned by a traditional particle swarm algorithm in a simulation experiment of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0067] Fig. 9 A multi-UAV path map planned by an optimized particle swarm algorithm with adaptive factors in a simulation experiment of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0068] Fig.10 A multi-UAV path map planned by an improved multi-objective particle swarm algorithm in a simulation experiment of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by an embodiment of the present invention;

[0069] Fig.11 A comparison diagram of fitness iteration curves of three algorithms in the same environment in a simulation experiment of a distributed multi-UAV cooperative trajectory planning method and system under a complex environment provided by one embodiment of the present invention;

[0070] Fig.12 An internal structure diagram of a computer device of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0072] Example 1

[0073] Reference Figure 1-Figure 12 , which is the first embodiment of the present invention, and provides a method and system for distributed multi-UAV collaborative trajectory planning in a complex environment, including:

[0074] Before describing the embodiments of the present application in detail, some related concepts are first explained for the sake of clarity.

[0075] Levy flight greedy optimization strategy: Levy flight is a random search mode whose step size follows the Levy distribution (a heavy-tailed distribution). This property enables the search process to effectively explore the global scope and conduct detailed development in the local area. In optimization problems, Levy flight can help the algorithm jump out of the local optimal solution and improve the global search capability. The greedy optimization strategy usually refers to selecting the current best solution or the direction closest to the best solution to move in each iterative step. Combined with Levy flight, this strategy means that when searching using Levy flight, it will tend to move in the direction of the currently known best solution to accelerate the convergence process.

[0076] Fitness function: A fitness function is a function used to evaluate the quality of a solution in an optimization problem. In this case, the fitness function combines multiple constraints (such as path length, trajectory smoothness, safety, spatiotemporal coordination, etc.) to quantify the quality of the UAV trajectory planning solution. The goal of optimization is usually to find one or more solutions that maximize or minimize the value of the fitness function. The design of the fitness function is critical to the performance of the algorithm, as it determines which solutions will be considered more "fit" or better.

[0077] Multi-swarm PSO algorithm: Particle swarm optimization (PSO) is a heuristic search algorithm that imitates the behavior of bird flocks or fish schools to find the optimal solution. Multi-swarm PSO algorithm refers to dividing the particle swarm into multiple different groups, each of which may have different initialization positions, velocity update rules or other parameter settings. This method can increase the diversity and global search capabilities of the algorithm. In your description, you mentioned that the particle swarm is divided into dominant populations, local particle swarms and inferior populations, and each group adopts a different optimization strategy to better balance the global search and local refinement process.

[0078] In related technologies, there are some challenges, such as communication delays, dynamic obstacles, environmental changes, and other issues that drones may encounter when performing missions.

[0079] This application provides a method that can effectively solve the above-mentioned problems. Next, we will combine multiple embodiments to elaborate on how to implement the distributed multi-UAV collaborative trajectory planning method in this complex environment;

[0080] Figure 1 A method flow chart of a distributed multi-UAV collaborative trajectory planning method and system in a complex environment is shown, including:

[0081] S101, establishing a UAV path planning aggregation fitness function model based on target constraints, where the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV space-time coordination constraints;

[0082] In the embodiments of the present application, in order to ensure the best planning effect, it is necessary to select appropriate evaluation criteria according to different application scenarios. Multi-UAV collaborative trajectory planning is an NP-hard problem. The distributed method divides the multi-machine problem into two parts: single machine and collaborative planning. It is necessary to comprehensively consider various constraints to establish an aggregate fitness function as a path evaluation model to evaluate the effect of the path planned by the algorithm. The length of the UAV flight path is the primary consideration in solving the path planning problem, so a UAV path length constraint is established.

[0083] The UAV path length constraint includes the sum of the length costs of different path segments, and calculates a certain path X i The length cost function is as follows:

[0084]

[0085] Among them, P ij =(x ij ,y ij ,z ij ) are the path node coordinates, Represents the distance between two adjacent path nodes, i and j are path nodes, and n represents the total number of nodes.

[0086] In the embodiment of the present application, limited internal energy and short flight time are common problems encountered by drones when performing tasks. If the planned path has too many twists and turns, the drone will frequently accelerate and decelerate during flight, and the accelerated energy consumption will compress the effective operation time. Therefore, it is hoped that the path is as smooth as possible. Therefore, the drone trajectory smoothness constraint is established, and the drone trajectory smoothness constraint includes:

[0087] The smoothness constraints of the drone trajectory include calculating the yaw angle and the climb angle. They are and Projection on the Oxy plane, The angle formed is the yaw angle ψ ij ,make is the unit vector of the z-axis, the projection vector is calculated as follows:

[0088]

[0089] The calculation formula of yaw angle is:

[0090]

[0091] Among them, the climb angle is the angle between the drone's ascending trajectory and the horizontal line. Its projection on the horizontal plane The angle formed is the climbing angle γ ij ;

[0092]

[0093] The UAV trajectory smoothness constraint is expressed as:

[0094]

[0095] Among them, ψ ij represents the yaw angle, γ ij Indicates the climb angle.

[0096] In the embodiment of the present application, the planned path of the drone will inevitably have errors due to different reasons such as the working environment and positioning accuracy. The closer the path is to obstacles, the greater the potential risk, so the danger around the path is also considered as an evaluation factor. Therefore, drone safety constraints are established, and drone safety constraints include:

[0097] Drone safety constraints include a given flight segment Considering the radius R of the drone and the danger distance S from the danger zone to the collision zone, the safety constraint of the drone in this flight segment is:

[0098]

[0099] In which, let K be the set of all obstacles, and let the center of the obstacle projection be C k , with a radius of R k , the radius of the drone is R, and the radius of the collision area is R c , the distance between the drone and the projection center is d k .

[0100] Assume that the number of drones performing the target strike mission is N, and the alternative track of the mth drone is denoted as ph m =(x m,0 ,x m,1 ,,x m,n-1 ,x m,n ), where x m,k (k=1,2,,n) represents the kth track point in the mth alternative track, and the flight time of the UAV is recorded as T m , the flight path of each drone is as follows:

[0101]

[0102] The multi-UAV collaborative trajectory planning model is as follows:

[0103]

[0104] Among them, Workforce represents the spatial scope of planning, ph m ∩ph j =Null indicates the spatial constraints between UAVs. If there is no intersection between the tracks of any two UAVs, there is no spatial collision. m ∩T j ∩,,T N =Non-Null indicates the UAV time coordination constraint. If the flight time intervals of all UAVs have an intersection, they can reach the designated target point at the same time.

[0105] In the embodiment of the present application, during the design of the information interaction layer, the number of times the trajectory violates the time constraint and the space constraint is selected as the criterion for selecting the optimal collaborative trajectory. Therefore, it is necessary to process the time constraint and the space constraint in the designed information interaction layer. Therefore, the UAV spatiotemporal collaborative constraint is established, and the UAV spatiotemporal collaborative constraint includes:

[0106] Assume that the flight speed range of the drone is The length of the mth track is L m , then the time interval for the drone to reach the target point is In order to ensure that multiple drones can reach the designated target point at the same time, it is necessary to ensure that the time interval T between any two drones reaching the target pointm With T j If there is an intersection, then the calculation formula for the track m violating the time constraint is as follows:

[0107]

[0108] When the intersection of the flight time between any two drones is empty, it means that the two drones cannot meet the constraint of arriving at the same time, and the number of time constraint violations increases by 1. On the contrary, if the intersection of the flight time is a non-empty subset, it means that the two drones can achieve the condition of arriving at the same time, and the number of time violations increases by 0.

[0109] UAV spatiotemporal coordination constraints also include:

[0110] If the minimum distance between any two drones is d m,j , in order to ensure that the space constraints between UAVs are met, the calculation formula for the track m violating the space constraints is as follows:

[0111]

[0112] From the above formula, we can see that when the minimum distance between the tracks of any two drones is less than the allowed minimum safety distance d min , it means that there is no collision between the two tracks. At this time, the number of violations of the space constraint is increased by 0. Otherwise, the number of violations of the space constraint is increased by 1.

[0113] After calculating the number of times the track violates the time and space constraints, the total number of violations of the time and space constraints by track m is:

[0114] W m =WT m +WP m .

[0115] The establishment of the UAV path planning aggregation fitness function model based on target constraints includes:

[0116] The UAV path planning aggregation fitness function model is expressed as:

[0117]

[0118] Among them, λ k is the weight coefficient and λ1+λ2+λ3=1, μ is the time-space coordination penalty coefficient. The present application has found through multiple experiments that setting the weight parameters λ1=0.7, λ2=0.15, λ3=0.15 can make the planned path meet the needs of the research object and application environment of the present application for the navigation route.

[0119] It should be noted that the benefit of S101 is that it can ensure that drones have high coordination and safety when performing tasks. By optimizing the number of times the trajectory violates the time and space constraints, the risk of collision between drones can be effectively reduced, and the efficiency and success rate of mission execution can be improved.

[0120] S102, improve the multi-swarm particle swarm algorithm, and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, including dominant populations, local particle populations, and inferior populations;

[0121] It should be noted that the multi-population PSO algorithm is actually a local PSO algorithm based on a special neighborhood topology. Compared with fixed division of multi-populations, dynamic random recombination can avoid excessive restriction of particle freedom and improve the efficiency of individual information exchange. Fixed division of populations is based on the idea of ​​classification, using the mean and standard deviation of fitness values ​​as measurements, and dividing populations into multiple levels by evaluating the positional relationship of particles. Its defects are manifested in the later stages of iteration. Excellent populations will attract more individuals to join, which will lead to excessive aggregation of particles and difficulty in information exchange between populations.

[0122] Therefore, on this basis, this application uses the median value of the ascending fitness value as the boundary to divide the sub-population, and implements dynamic adjustments in each generation. The smaller the fitness value, the "dominant particle", and the larger the "inferior particle"; then an equal number of particles are extracted from the two to form a local PSO model, and the three population sizes are balanced, namely the "close neighborhood" (dominant population) closer to the optimal solution, the "reorganization domain" (local particle population) of the local PSO model, and the "alienation domain" (inferior population). The particles in the "close neighborhood" have a smaller fitness value, and need to inherit the location information of the optimal solution in the population, reduce the step size for a more detailed search; while the particles in the "alienation domain" should expand the "pace", while approaching the global extreme value, explore new optimal solutions around to improve mining capacity; as a local model, the "reorganization domain" will dynamically adjust the diversity of the population, drawing on individual experience and sharing global information. Close neighborhood: These particles should take a smaller step size to explore the space near the local optimal solution more finely. Recombination domain: These particles should use the information of the global best solution while maintaining a certain degree of randomness to promote knowledge exchange among the population. Alienation domain: These particles should take larger steps to explore new areas far away from the current best solution.

[0123] It should be noted that the benefit of S102 is that it can significantly improve the efficiency and quality of drone path planning. By dynamically dividing multiple populations, it is possible to effectively avoid excessive aggregation of particles, thereby improving the diversity among populations. This diversity helps avoid falling into local optimal solutions and can better explore global optimal solutions. In addition, by dynamically adjusting the population size, it is possible to ensure that each subpopulation maintains an appropriate size during the search process, thereby improving the convergence speed of the algorithm and the quality of the solution.

[0124] S103, using a preset Levy flight greedy optimization strategy and a preset adaptive parameter update strategy to optimize the dominant population and the local particle swarm, and using a preset probabilistic mixed mutation strategy to optimize the inferior population;

[0125] In an embodiment of the present application, after dynamically dividing multiple populations according to the fitness of each generation of particles, the improved algorithm introduces different parallel update strategies based on the characteristics of the dominant group and the mixed group: the introduction of Levy flight and greedy strategy is conducive to a more detailed and effective search for the dominant group; the fusion of cosine function and adaptive strategy is used to balance the local mining and global exploration capabilities of the mixed group.

[0126] In the embodiment of the present application, in order to solve the problem that the dominant group has no effective update, the improved algorithm first introduces Levy flight. Figure 6 As shown in the figure, Lévy flight is a random walk with a heavy-tailed probability distribution of step length, which means that there is a relatively high probability of large steps in the process of random walking, which helps to enhance the activity and jumping ability of particles, expand the search range of particles while conducting a more detailed search, and effectively avoid being attracted by local optimal values.

[0127] Levy flight is a non-Gaussian random process whose step size follows the Levy distribution, which is a simple power function:

[0128] Levy(λ)~μ=k -λ ,1<λ≤3

[0129] The Levy distribution is very complex and difficult to implement, so the random search path of the Levy flight can be simulated using the Mantegna algorithm as:

[0130]

[0131] Among them, the value of β is generally 1<β<3, μ and υ are two normally distributed random numbers in the interval [0,1], and the variance σ μ and σ v Determined by the following formula:

[0132]

[0133]

[0134] Therefore, the position update formula based on Levy flight is:

[0135]

[0136] in, and x g(k) are the global optimal particle position and the global optimal particle position of the population after Levy flight update at the kth iteration, α is the step size control factor, and Γ is the gamma function.

[0137] Although Levy flight can make particles escape from the local optimum, it cannot guarantee that the updated particle position is better than the original position. Based on this, this application introduces the evaluation strategy of the greedy algorithm to determine whether to update the optimal particle position. The implementation process is shown in the following formula:

[0138]

[0139] When the updated position is better than the original position, the position is updated, otherwise the original position is retained. The greedy algorithm meets the classification purpose of multiple populations, ensuring that the particles of the dominant group can inherit the position information of the optimal solution in the population, avoiding meaningless position updates, and realizing the positive evolution of the dominant group.

[0140] As the "reorganization domain" of the local PSO model, the hybrid group should absorb individual experience and share global information. Therefore, the present application introduces the cosine function to improve the conventional adaptive function.

[0141] Ideally, a larger ω in the early stage of the algorithm enables particles to quickly search the entire solution space at a higher speed and find the global optimal solution in a short time; as the number of iterations increases, ω begins to gradually decrease, and the search step size of each particle also decreases accordingly, thereby enhancing the algorithm's local optimization ability and accelerating convergence.

[0142]

[0143] Among them, w min 、w max is the minimum and maximum value of the inertia weight, k max is the total number of iterations.

[0144] When c1 is large, a large number of particles will gather in the local dominant area; when c2 is small, the particles will converge to the local minimum too early and fall into the local optimum. The adaptive asynchronous learning factor is introduced to balance the global search and local search capabilities of the algorithm to ensure the convergence of the algorithm itself.

[0145]

[0146]

[0147]

[0148] Among them, c min 、c maxare fixed parameters, which are the minimum and maximum values ​​of the learning factor. c1 and c2 represent the learning factors, which are used to control the particle in the current optimal position (personal best position). ) and the global optimal position (global optimal position G gd ) These two values ​​will change as the number of iterations k increases. Represents the updated value of the velocity of the i-th particle in the d-th dimension, which determines the moving direction and distance of the particle in the next iteration. Indicates the inertia weight, which indicates the degree of memory of historical information of the particle. This value will also change with the number of iterations. r1 and r2 represent random numbers, which are used to introduce randomness so that the particle will not move completely along a deterministic trajectory. id Indicates the position of the i-th particle in the d-th dimension at the current moment. and G gd Denote the personal best position and the global best position found by the i-th particle at the k-th iteration, respectively.

[0149] The introduction of the adaptive parameters of the cosine function can effectively address the imbalance between the global development ability and the local search ability of the particle swarm algorithm itself, achieve local meticulous development, enhance the ability of the algorithm to jump out of the local minimum, and improve the convergence speed.

[0150] In step S4, since there is less information about particles with retention value in the disadvantaged group and the population is far away from the optimal solution of the problem, the improved algorithm adopts a strategy of probabilistic mixed mutation to improve it.

[0151] First, the mixed particles are introduced to share the advantage information to assist the mutation of the disadvantaged group particles to quickly find the optimal solution. Figure 6 ,remember For the k-th iteration of the mixed particle, each dimension is randomly selected from a dimension corresponding to the current historical optimal position (Pbest) of each particle, and multiple dimensions of the same particle are not repeatedly selected.

[0152] Each particle is subjected to probabilistic mutation. When the individual mutation judgment is established, the particle mutation operation can be executed, otherwise it is updated in the form of adaptive parameters. The significance of probabilistic operation is to make the particles mutate with a greater probability in the initial stage. After multiple iterations, the probability of the judgment formula being established gradually decreases. At this time, the particles are also closer to the optimal position, and the probability of particle mutation is gradually reduced.

[0153]

[0154] GV(x)=x×(1+N(0,1))

[0155] Among them, r and N(0,1) are random numbers between [0,1], x is the original parameter value, and GV(x) is the mutated value.

[0156] Based on this, an improved speed update formula can be proposed as:

[0157]

[0158] Among them, c3 is the hybrid learning factor, and r3 is a random number between (0,1).

[0159] It should be noted that hybrid particles, as traction factors, guide the particle speed update, which can effectively solve the problem of falling into the local optimum. At the same time, their own excellence also makes the particles evolve in a better direction. Adding probabilistic selection can not only give particles more search opportunities, but also ensure that they do not escape from the suboptimal solution area. This optimization method is very suitable for particles in the disadvantaged group to perform volatility search in the solution space.

[0160] S104: Based on the optimization results, distributed multi-UAV collaborative trajectory planning is performed.

[0161] In the embodiment of the present application, the purpose of the distributed multi-UAV collaborative trajectory planning is to ensure that the UAV group can efficiently and safely complete the trajectory planning when performing tasks. First, according to the optimization results, the UAV group is divided into multiple subgroups, each of which is responsible for the trajectory planning of a specific area. A collaborative mechanism based on the particle swarm optimization algorithm is adopted within each subgroup to ensure that the UAVs can coordinate with each other and avoid conflicts when performing tasks.

[0162] In the collaborative trajectory planning process, each UAV updates its trajectory information in real time according to its own position, speed and mission requirements. Through the particle swarm optimization algorithm, the UAV can dynamically adjust the trajectory to adapt to the changing environment and mission requirements. At the same time, the UAVs exchange information in real time through the wireless communication network, share each other's trajectory information and environmental perception data, and thus achieve the global optimal trajectory planning.

[0163] In addition, in order to ensure the safety of drones when performing tasks, a trajectory planning mechanism based on safety constraints is introduced. During the trajectory planning process, drones need to consider safety constraints such as flight altitude, flight speed, and obstacle avoidance distance to ensure that there will be no collisions or dangerous situations when performing tasks. By introducing safety constraints, drones can achieve efficient and flexible trajectory planning while ensuring safety.

[0164] Ultimately, through distributed multi-UAV collaborative trajectory planning, the UAV swarm can complete the task efficiently and safely, improving the efficiency and reliability of task execution. At the same time, the collaborative mechanism based on the particle swarm optimization algorithm can ensure that the UAVs can coordinate with each other when performing tasks, avoid conflicts, and achieve global optimal trajectory planning.

[0165] In summary, the present invention proposes a method for distributed multi-UAV collaborative trajectory planning in a complex environment, establishes an aggregated fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV spatiotemporal collaborative constraints; improves the multi-population particle swarm algorithm, and dynamically divides the UAV path planning aggregated fitness function model into multiple populations, wherein the multiple populations include dominant populations, local particle populations, and inferior populations; adopts a preset Levy flight greedy optimization strategy and a preset adaptive parameter update strategy to optimize the dominant population and the local particle population, and adopts a preset probabilistic mixed mutation strategy to optimize the inferior population; and performs distributed multi-UAV collaborative trajectory planning based on the optimization results. The efficiency and safety of UAV collaborative trajectory planning are improved, and it is ensured that multiple UAVs can complete tasks efficiently and safely in a complex environment.

[0166] Specifically, by establishing an aggregated fitness function model for UAV path planning, multiple constraints can be comprehensively considered so that the planned trajectory not only meets the requirements of path length, trajectory smoothness and safety, but also meets the spatiotemporal coordination constraints. The improved multi-population particle swarm algorithm can effectively avoid falling into the local optimal solution and improve the global search capability, thereby obtaining a better trajectory planning result.

[0167] In addition, the preset Levy flight greedy optimization strategy and adaptive parameter updating strategy can further improve the optimization efficiency of the dominant population and the local particle swarm, while the probabilistic mixed mutation strategy helps the disadvantaged population to escape from the local optimum, increase population diversity and avoid premature convergence.

[0168] Ultimately, through distributed multi-UAV collaborative trajectory planning, it can ensure that multiple UAVs can effectively coordinate when performing tasks, avoid mutual interference and collision, and improve the success rate and efficiency of mission execution.

[0169] In summary, the present invention is not only technologically innovative, but also has significant practical value in practical applications, and provides an effective solution for the collaborative operation of multiple UAVs in complex environments.

[0170] This embodiment also provides a distributed multi-UAV collaborative trajectory planning system in a complex environment, including:

[0171] The model building module is used to establish the UAV path planning aggregation fitness function model based on the target constraint conditions. The target constraint conditions include the UAV path length constraint, the UAV trajectory smoothness constraint, the UAV safety constraint and the UAV space-time coordination constraint;

[0172] The algorithm improvement module is used to improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, including dominant populations, local particle populations and disadvantaged populations;

[0173] The optimization module is used to optimize the dominant population and the local particle swarm by using the preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy, and to optimize the inferior population by using the preset probabilistic mixed mutation strategy;

[0174] The trajectory planning module is used to perform distributed multi-UAV collaborative trajectory planning based on the optimization results.

[0175] The above-mentioned unit modules may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to the above-mentioned modules.

[0176] This embodiment also provides a computer device, which may be a terminal, and its internal structure diagram may be as shown in FIG. Fig.12 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a distributed multi-UAV collaborative trajectory planning method in a complex environment is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse.

[0177] This embodiment further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0178] Establish an aggregate fitness function model for UAV path planning based on target constraints. The target constraints include UAV path length constraint, UAV trajectory smoothness constraint, UAV safety constraint, and UAV space-time coordination constraint.

[0179] Improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, including dominant populations, local particle populations and disadvantaged populations;

[0180] The preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy are used to optimize the dominant population and the local particle swarm, and the preset probabilistic mixed mutation strategy is used to optimize the inferior population;

[0181] Based on the optimization results, distributed multi-UAV collaborative trajectory planning is carried out.

[0182] Example 2

[0183] Reference Figure 2-Figure 11 , which is an embodiment of the present invention, provides a distributed multi-UAV collaborative trajectory planning method and system in a complex environment. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0184] Taking a quadrotor drone as the object, the simulation verification was carried out in the MATLAB / Simulink environment. In order to verify the feasibility of the improved algorithm, a relevant three-dimensional simulation environment was established and a fitness function was designed. The improved multi-objective particle swarm optimization (IDMPSO) was compared with the traditional PSO and the improved particle swarm optimization with adaptive factors (IPSOAF). The specific parameters of the simulation are shown in Table 1.

[0185] Table 1: Simulation parameters.

[0186]

[0187] The simulation environment was built in an area of ​​1000*1000*20, with multiple column-shaped obstacles randomly generated and four drones set up. The fitness function uses the flight distance as the main evaluation basis, and adds a danger penalty mechanism based on the distance to the obstacle. The experiment tested three algorithms in the same scenario, generating the flight path of each drone and the fitness value convergence curve. The effects of each algorithm are shown in the figure below. Figure 8 , 9 , 10, and 11.

[0188] Traditional PSO is prone to fall into local optimal solutions during the search process. The improved algorithm, by introducing more flexible exploration strategies, achieves more precise path adjustments with stronger local mining capabilities and more efficient information utilization capabilities, thereby effectively capturing the global characteristics of the problem, accelerating convergence, and jumping out of the local optimum.

[0189] The experimental results show that, under different conditions, the improved algorithm can effectively increase the convergence speed, escape from the local optimum, and improve the efficiency of path planning, so that the UAV can reach the target point safely, stably and quickly.

[0190] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0191] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments 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.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0192] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer 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 device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing 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.

[0193] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising 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.

[0194] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing 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.

[0195] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0196] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A distributed multi-UAV collaborative trajectory planning method in a complex environment, characterized in that: include: Establishing an aggregated fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV spatiotemporal coordination constraints; Improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, including dominant populations, local particle populations and disadvantaged populations; The preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy are used to optimize the dominant population and the local particle swarm, and the preset probabilistic mixed mutation strategy is used to optimize the inferior population; Based on the optimization results, distributed multi-UAV collaborative trajectory planning is carried out.

2. The method for coordinated trajectory planning of distributed multi-UAVs in a complex environment as claimed in claim 1, characterized in that: The UAV path length constraints include: The UAV path length constraint includes the sum of the length costs of different path segments. i The length cost function is as follows: Among them, P ij =(x ij ,y ij ,z ij ) are the path node coordinates, Represents the distance between two adjacent path nodes, i and j are path nodes, and n represents the total number of nodes.

3. The method for distributed multi-UAV collaborative trajectory planning in a complex environment as claimed in claim 2, characterized in that: The UAV trajectory smoothness constraints include: The UAV trajectory smoothness constraint includes calculating the yaw angle and the climb angle, They are and Projection on the Oxy plane, The angle formed is the yaw angle ψ ij ,make is the unit vector of the z-axis, the projection vector is calculated as follows: The calculation formula of yaw angle is: Among them, the climb angle is the angle between the drone's ascending trajectory and the horizontal line. Its projection on the horizontal plane The angle formed is the climbing angle γ ij ; The UAV trajectory smoothness constraint is expressed as: Among them, ψ ij represents the yaw angle, γ ij Indicates the climb angle.

4. The method for coordinated trajectory planning of distributed multi-UAVs in a complex environment as claimed in claim 3 is characterized in that: The drone safety constraints include: The drone safety constraints include a given flight segment Considering the radius R of the drone and the danger distance S from the danger zone to the collision zone, the safety constraint of the drone in this segment is: In which, let K be the set of all obstacles, and let the center of the obstacle projection be C k , with a radius of R k , the radius of the drone is R, and the radius of the collision area is R c , the distance between the drone and the projection center is d k .

5. The method for coordinated trajectory planning of distributed multi-UAVs in a complex environment as claimed in claim 4, characterized in that: The UAV space-time coordination constraints include: Assume that the flight speed range of the drone is The length of the mth track is L m , then the time interval for the drone to reach the target point is In order to ensure that multiple drones can reach the designated target point at the same time, it is necessary to ensure that the time interval T between any two drones reaching the target point m With T j If there is an intersection, then the calculation formula for the track m violating the time constraint is as follows: When the intersection of the flight time between any two drones is empty, it means that the two drones cannot meet the constraint of arriving at the same time, and the number of time constraint violations increases by 1. On the contrary, if the intersection of the flight time is a non-empty subset, it means that the two drones can achieve the condition of arriving at the same time, and the number of time violations increases by 0.

6. The method for coordinated trajectory planning of distributed multi-UAVs in a complex environment as claimed in claim 5, characterized in that: The UAV space-time coordination constraints also include: If the minimum distance between any two drones is d m,j , in order to ensure that the space constraints between UAVs are met, the calculation formula for the track m violating the space constraints is as follows: From the above formula, we can see that when the minimum distance between the tracks of any two drones is less than the allowed minimum safety distance d min , it means that there is no collision between the two tracks. At this time, the number of violations of the space constraint is increased by 0. Otherwise, the number of violations of the space constraint is increased by 1. After calculating the number of times the track violates the time and space constraints, the total number of violations of the time and space constraints by track m is: W m =WT m +WP m 。 7. The method for coordinated trajectory planning of distributed multi-UAVs in a complex environment as claimed in claim 6, characterized in that: The establishment of the UAV path planning aggregation fitness function model based on the target constraint condition includes: The UAV path planning aggregation fitness function model is expressed as: Among them, λ k is the weight coefficient and λ1+λ2+λ3=1, μ is the spatiotemporal coordination penalty coefficient.

8. A distributed multi-UAV collaborative trajectory planning system in a complex environment, characterized by: include: A model building module is used to establish an aggregate fitness function model for UAV path planning based on target constraints, wherein the target constraints include UAV path length constraints, UAV trajectory smoothness constraints, UAV safety constraints, and UAV space-time coordination constraints; An algorithm improvement module is used to improve the multi-swarm particle swarm algorithm and dynamically divide the UAV path planning aggregation fitness function model into multiple populations, wherein the multiple populations include a dominant population, a local particle population, and a disadvantaged population; The optimization module is used to optimize the dominant population and the local particle swarm by using the preset Levy flight greedy optimization strategy and the preset adaptive parameter update strategy, and to optimize the inferior population by using the preset probabilistic mixed mutation strategy; The trajectory planning module is used to perform distributed multi-UAV collaborative trajectory planning based on the optimization results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images