Heterogeneous UAV swarm pre-reconfiguration method and system based on hybrid genetic algorithm

Abstract

The present invention provides a heterogeneous UAV cluster pre-reconstruction method and system based on a hybrid genetic algorithm to support the need for dynamic fusion into a new formation when different business UAV formations are hard destroyed. It is applicable to when a UAV cluster formation is damaged, and the load carried by the remaining UAVs cannot support the cluster to complete its business, and the surviving UAVs can join the UAV resource pool. When the load of the nodes in the resource pool is sufficient, the UAV formation is rebuilt for the target task. When designing the UAV resource pool, the load parameters held by the UAV are considered and modeled. At the same time, a task-cluster double-layer clustering reconstruction scheme is proposed for large-scale mission scenarios, with the energy consumed by the UAV cluster to complete the mission target as the optimization target, and the UAV cluster carrying the load to meet the mission requirements as the constraint, so that the feasible task scheme assigned to the UAV formation obtained by the reconstruction under this scheme satisfies the minimum total energy consumption.

Description

Technical Field

[0001] The present invention belongs to the field of formation reconstruction of unmanned aerial vehicle systems, and in particular relates to a heterogeneous unmanned aerial vehicle cluster pre-reconstruction method and system based on a hybrid genetic algorithm. Background Art

[0002] With the continuous development of drone swarm-related fields, drone clusters have gradually demonstrated powerful mission capabilities on the battlefield. Drones carrying different payloads collaborate to form a multifunctional drone cluster that integrates reconnaissance, attack, and harassment, achieving system-level capability improvement.

[0003] Due to the high dynamics and uncertainty of the drone battlefield, when a drone cluster that has been assembled in advance is attacked, the drone formation designed for the target mission may encounter a series of situations such as damage to the drone payload, reduction in the drone formation, and insufficient battery capacity due to maneuvers to avoid attacks, resulting in the inability of the drone formation under static allocation to complete the assigned target mission. In a complex and urgent mission environment, ensuring the completion of the target mission is the primary decision-making factor in strategic decision-making. Cluster reconstruction is one of the important means to ensure that the damaged cluster completes the mission. The attacked cluster forms a new drone cluster with other clusters in the battlefield and re-allocates according to the mission payload. Therefore, it is of great significance to study a drone reconstruction scheme that can ensure the mission completion capability of the drone formation in a high-confrontation environment. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for pre-reconstructing a heterogeneous drone cluster based on a hybrid genetic algorithm to support the need for dynamic fusion into a new formation when drone formations of different services are hard destroyed. When a drone cluster is damaged by enemy hard destruction in a highly confrontational battlefield environment, the load carried by the remaining drones cannot support the cluster to complete its business, and the surviving drones can apply to join the drone resource pool. When the load of the nodes in the resource pool is sufficient, the drone formation can be rebuilt for the target task. When designing the drone cluster resource pool, the present invention considers the load parameters held by the drone and models them. On this basis, the present invention proposes a task-cluster double-layer clustering reconstruction scheme for large-scale mission scenarios, taking the energy consumed by the drone cluster to complete the mission target as the optimization target, and taking the drone cluster carrying the load to meet the mission requirements as the constraint, so that the feasible task scheme assigned to the drone formation obtained by the reconstruction under this scheme satisfies the minimum total energy consumption.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] A heterogeneous UAV cluster pre-reconfiguration method based on a hybrid genetic algorithm is characterized by comprising:

[0007] Cluster the tasks to be completed into multiple task clusters according to their geographical characteristics;

[0008] Add the drones that can participate in the reconstruction of the damaged heterogeneous drone cluster to the drone resource pool;

[0009] Through the hybrid genetic algorithm, the heterogeneous UAV cluster is reorganized according to the generated task clusters and UAV resource pool.

[0010] To optimize the above technical solutions, the specific measures taken also include:

[0011] Furthermore, the tasks to be completed are clustered into multiple task clusters according to the geographical characteristics of the tasks, specifically:

[0012] Tasks to be completed in the area represents the task set, N t Indicates the number of tasks;

[0013] According to the geographical characteristics of the tasks, a clustering algorithm based on the Gaussian mixture model is used to cluster the tasks to be completed into multiple task clusters. c k represents the kth task cluster, and the superscript T indicates transposition;

[0014] The formed task cluster contains the following information: the task sequence number c contained in the task cluster k =(c k1 ,c k2 ,...,c kL ) T And the number of loads required to complete the tasks in the task cluster c k1 ,c k2 ,...,c kL represents L task numbers, Ω 1 ,Ω 2 ,…,Ω M Indicates M type load.

[0015] Furthermore, the clustering algorithm based on the Gaussian mixture model is used to cluster the tasks to be completed into multiple task clusters, specifically:

[0016] The geographic location distribution of all tasks is represented as a Gaussian mixture model, and the probability density distribution function P(pos j,t |θ) is expressed as:

[0017]

[0018] In the formula, π k is the probability weight of the k-component, K is the number of task clusters, Ν(posj,t |μ k ,∑ k ) is based on ∑ k is the variance, μ k is a Gaussian distribution with a mean of POS j,t is the position coordinate of task j at time t, θ=(π 1 ,...,π K ,μ 1 ,…,μ K ,∑ 1 ,...,∑ K );

[0019] Introduce a latent variable z j , represents the task cluster number to which each task may be assigned in advance, and P(pos j,t |z j =k,θ)=Ν(pos j,t |μ k ,∑ k ), then the joint probability distribution P(pos,z|θ) of the position information of all tasks is expressed as:

[0020]

[0021] When zj=k, II(z j ,k)=1, otherwise 0;

[0022] Calculate the latent variable z j The expectation E(z) of the likelihood function under the current parameter θ j =k|pos j,t ,θ):

[0023]

[0024] Update the value of θ by maximizing the posterior distribution Q value:

[0025]

[0026] In the formula, θ n represents the value of the parameter θ in the nth round;

[0027] Update μ k and Σ k :

[0028]

[0029] In the formula, and represents the updated μ k and∑ k ;

[0030] Finally, according to Calculate the task cluster to which each task belongs.

[0031] Furthermore, the drones that can participate in the reconstruction of the damaged heterogeneous drone cluster are added to the drone resource pool, specifically:

[0032] The drone numbers that can participate in the reconstruction are Represents a set of numbers, N u is the total number of drones in the drone resource pool; drones that can participate in the reconstruction are added to the drone resource pool, and the load in the drone resource pool is represented as an M-dimensional vector Ω 1 ,Ω 2 ,...,Ω M represents the M type of load, and the load carried by UAV i is expressed as Ω i1 ,Ω i2 ,...,Ω iM Indicates the M-type payload carried by UAV i.

[0033] Furthermore, the hybrid genetic algorithm is used to reorganize the heterogeneous drone cluster for the generated task cluster and the drone resource pool, specifically:

[0034] The task sequence and drone combination in the task cluster are used as two chromosomes for real number encoding, and a mobile energy consumption model of the drone cluster based on the task sequence is constructed. The total energy consumption during the task process is taken as the algorithm optimization target, and drone clusters equal to the number of task clusters and the matching relationship between drone clusters and task clusters are obtained.

[0035] Furthermore, the algorithm optimization goal is:

[0036]

[0037] In the formula, ξ represents the task sequence, is a cluster of k drones, K is the number of mission clusters, Represents a single drone cluster s k energy consumption;

[0038] The following constraints exist during the allocation process:

[0039]

[0040] In the formula, C 1 With C 2 This indicates that the generated drone cluster can only be assigned to one task cluster, and a task cluster can only be completed by one drone cluster. 3 Indicates drone swarms kThe load carried needs to meet the task cluster c k Required load requirement, C 4 This shows that there is a two-dimensional distribution relationship between drone clusters and task clusters; x cs A two-dimensional variable representing the distribution relationship between drone clusters and task clusters, x cs =1 and x cs = 0 respectively indicates that the drone cluster s k Responsible and irresponsible task clusters c k A collection of tasks; Indicates drone swarms k The number of load resources available, Indicates that task cluster c is completed k The number of load resources required.

[0041] Furthermore, the drone cluster s k The energy consumption is:

[0042]

[0043] In the formula, is the task cluster c assigned to the cluster k The total number of tasks, For clusters in task sequences The energy consumption generated by moving from the lth task to the l+1th task in is the energy consumption of the cluster in task j, It is the energy consumption of the drone due to waiting.

[0044] Furthermore, the drone cluster s k The energy consumption is calculated by the following formula:

[0045]

[0046] Where N is the UAV cluster s k The number of drones in V wait Indicates the speed at which the drone hovers at the mission location. It is a drone swarm k The time it takes to complete task j, Represents the drone cluster s k The time spent hovering at task j; Represents the drone cluster s k The energy consumption from flying from task j to task j', For drone swarms k The speed of flight, Represents the drone cluster s k The time it takes to fly from task j to task j', according to Calculated; P(V wait )and The energy consumption of the drone cluster is calculated according to the following formula:

[0047]

[0048] Where P 0 With P i They represent the blade profile power and induced power in the hovering state, U tip and U 0 They represent the rotor tip speed and the rotor average speed in the hovering state, respectively, and d 0 and s represent the fuselage drag ratio and rotor solidity respectively, ρ and A represent the air density and rotor disc area respectively, and v is the speed of the UAV cluster.

[0049] Further, and T skj Calculated according to the following formula:

[0050]

[0051] Where, d j,j′ is the distance from task j to task j'; η j is the amount of data information that needs to be transmitted for task j, W is the spectrum bandwidth resources available between UAVs, It is a drone swarm k The number of drones involved in the communication, It is a drone swarm k The transmission power, is the channel fading parameter, β 0 is the channel gain at a distance of 1m, d j,r is the communication distance between the drone cluster and the relay communication drone, σ 2 Represents the ambient power spectral density.

[0052] In addition, the present invention also provides a heterogeneous UAV cluster pre-reconfiguration system based on a hybrid genetic algorithm, which is characterized by comprising:

[0053] A target task clusterer is used to cluster the tasks to be completed into multiple task clusters according to the geographical characteristics of the tasks;

[0054] The drone resource pool is used to accommodate drones that can participate in the reconstruction of a heterogeneous drone cluster after it is damaged;

[0055] The task allocator is used to reorganize the heterogeneous UAV cluster based on the generated task clusters and the UAV resource pool through a hybrid genetic algorithm.

[0056] The beneficial effects of the present invention are as follows: the present invention discloses a tactical scenario in which a heterogeneous drone cluster is reorganized into multiple new drone cluster formations by reorganizing with other damaged clusters after being damaged by an attack, so that a new drone cluster can be formed by reorganizing with the remaining drone formations. In the reconstruction process, the payload resources carried by the drones are modeled into a drone resource pool, taking into account the heterogeneity of different drones, and the different payloads and battery capacities carried. The present invention reorganizes the damaged cluster into a new drone formation to complete the task by rescheduling the payload resources in the drone resource pool, which is more practical for the tactical scenario, and also improves the sustainable combat capability of the entire drone cluster in the face of a highly confrontational battlefield environment, and enhances the combat dynamic adjustment capability and radiation range of the drone cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a pre-reconstructed scene diagram of a heterogeneous UAV swarm formation according to an embodiment of the present invention.

[0058] Figure 2 It is a pre-reconstruction strategy diagram of a heterogeneous UAV cluster formation based on a hybrid genetic algorithm according to an embodiment of the present invention.

[0059] Figure 3 It is a flowchart of a hybrid genetic algorithm according to an embodiment of the present invention.

[0060] Figure 4 Schematic diagram of chromosome encoding of the hybrid genetic algorithm according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0062] In one embodiment, the present invention provides a heterogeneous UAV cluster pre-reconfiguration method based on a hybrid genetic algorithm, such as Figure 1 As shown, the method is applied to multiple damaged drone clusters. The system enters the available payload of the drone cluster into the drone resource pool and reorganizes the drones in the resource pool into multiple new task formations according to the target tasks.

[0063] like Figure 2 As shown, the heterogeneous UAV cluster pre-reconstruction method includes the following steps:

[0064] (1) Target task clustering: The total number of tasks to be completed in the region is The task clusterer clusters the target tasks to be completed into multiple task clusters according to the geographical characteristics of the target tasks. The formed task cluster contains the following information: the task sequence number c contained in the task cluster k =(ck1 .c k2 ,...,c kL ) T And the number of loads required to complete the tasks within the cluster Where L is the number of tasks in the corresponding task cluster.

[0065] The task clusterer uses a clustering algorithm based on a Gaussian mixture model, which can fully consider the overlapping fuzzy areas between different task clusters. In the clustering algorithm process, the preset target task location is Gaussian distributed in the area, then the geographical location distribution of all tasks can be expressed as a mixed Gaussian model, and the probability density distribution function of the jth task can be expressed as:

[0066]

[0067] In the formula, π k is the probability weight of the k-component, K is the number of task clusters, Ν(pos j,t |μ k ,∑ k ) is based on ∑ k is the variance, μ k is a Gaussian distribution with a mean of POS j,t is the position coordinate of task j at time t, θ=(π 1 ,...,π K ,μ 1 ,...,μ K ,Σ 1 ,...,Σ K ).

[0068] Introduce a latent variable z j , which is represented by the pre-estimated task cluster number to which each task node may be assigned, and P(pos j,t |z j =k,θ)=Ν(pos j,t |μ k ,Σ k ). The joint probability distribution of the location information of all target tasks can be expressed as Specifically, it can be further refined into When z j = k, II(z j ,k)=1, otherwise it is 0.

[0069] The clustering algorithm is divided into two steps: estimation and likelihood function maximization. In the estimation step, the latent variable z is calculated. j The expectation E of the likelihood function under the current parameter θ:

[0070]

[0071] In the likelihood function maximization step, the θ value is updated by maximizing the posterior distribution Q value:

[0072]

[0073] In the formula, θ n represents the value of the parameter θ in the nth round;

[0074] renew

[0075]

[0076] In the formula, and represents the updated μ k and∑ k .

[0077] Finally, the task cluster to which each target task belongs is

[0078] (2) Establishment of drone resource pool: After a heterogeneous drone cluster is damaged, the drones that can participate in the reconstruction apply to join the drone resource pool to participate in resource integration. After integration, the drones that can be used for reconstruction are numbered Where N u is the total number of drones in the resource pool. A drone resource pool is established based on the types of tasks undertaken by heterogeneous drones, such as attack payload resources, communication payload resources, and reconnaissance payload resources. The payload in the drone resource pool can be represented as an M-dimensional vector Where M is the total number of load types. Specifically, the load carried by UAV i can be expressed as

[0079] (3) Reconstruction of heterogeneous drone clusters: A hybrid genetic algorithm is used to reconstruct heterogeneous drone clusters based on the constructed task clusters and drone resource pools. During the reconstruction process, the task sequence and the combination of drones in the task cluster are used as two chromosomes for real number encoding. In combination with the needs of drone formations, a mobile energy consumption model of the drone cluster based on the task sequence is constructed, and the total energy consumption during the task process is used as the algorithm optimization target. When the algorithm ends, a drone cluster equal to the number of task clusters will be obtained, as well as a matching relationship between the drone cluster and the task cluster.

[0080] In the process of heterogeneous UAV cluster reconstruction, the UAV cluster calculates the energy consumption value corresponding to the task allocation of the chromosome and the UAV cluster scheduling relationship in this iteration according to the task sequence of each round in the algorithm. Among them, the UAV cluster s is calculated according to the following formula k Energy consumption value:

[0081]

[0082] Where, L ck is the task cluster c assigned to the cluster k The total number of tasks, is the task consumption generated by the cluster in task j, is the energy consumption of the drone due to waiting, For clusters in task sequences The energy consumption generated by moving from the lth task to the l+1th task.

[0083] The energy consumption formula of drone cluster is:

[0084]

[0085] Where P 0 With P i They represent the blade profile power and induced power in the hovering state, U tip and U 0 They represent the rotor tip speed and the rotor average speed in the hovering state, respectively, and d 0 and s represent the fuselage drag ratio and rotor solidity respectively, ρ and A represent the air density and rotor disc area respectively, and v is the speed of the UAV cluster.

[0086] According to the energy consumption formula of the drone cluster, the drone cluster s can be calculated. k Energy consumption from task j to task j' in, Represents the drone cluster s k The time it takes to fly from task j to task j', d j,j′ is the distance from task j to task j', For drone swarms k The speed of flight. It is a drone swarm k The time it takes to complete task j, η j is the amount of data information that needs to be transmitted for task j, W is the spectrum bandwidth resources available between UAVs, It is a drone swarm k The number of drones involved in the communication, It is a drone swarm k The transmission power, is the channel fading parameter, β 0 is the channel gain at a distance of 1m, d j,r is the communication distance between the drone cluster and the relay communication drone, σ 2 Represents the ambient power spectral density.

[0087] In the optimization stage based on the hybrid genetic algorithm, the total energy consumption value of all drone clusters is taken as the optimization target. It can be specifically expressed as: The following constraints exist during the allocation process:

[0088]

[0089] In the formula, C 1 With C 2 This indicates that the generated drone cluster can only be assigned to one task cluster, and a task cluster can only be completed by one drone cluster. 3 Indicates drone swarms k The load carried needs to meet the task cluster c k Required load requirement, C 4 This shows that there is a two-dimensional distribution relationship between drone clusters and task clusters; x cs A two-dimensional variable representing the distribution relationship between drone clusters and task clusters, x cs =1 and x cs = 0 respectively indicates that the drone cluster s k Responsible and irresponsible task clusters c k A collection of tasks; Indicates drone swarms k The number of load resources available, Indicates that task cluster c is completed k The number of load resources required.

[0090] like Figure 3 As shown in the figure, the hybrid genetic algorithm proposed in the present invention is composed of an annealing mechanism, a genetic algorithm and a taboo search algorithm. In view of the early solution space search process of the genetic algorithm, the annealing algorithm mechanism is introduced to enhance the algorithm's ability to jump out of the wide-area search for the optimal solution. In view of the weak local search in the late stage of the genetic algorithm, the taboo search mechanism is introduced to enhance the optimal value positioning ability of the neighborhood search.

[0091] like Figure 4 As shown, in the hybrid genetic algorithm proposed by the present invention, the chromosome adopts real number coding, and the chromosome is divided into two chromosomes in total. The chromosome segment is the chromosome with the drone number, and the chromosome length is the number of drones, and the segment is the chromosome with the task number, and the chromosome length is the total number of tasks. In each round, each individual in the population corresponds to a chromosome containing three segments. The system obtains the matching relationship between the corresponding task sequence and the drone cluster based on the two segment chromosomes and calculates the total energy consumption value.

[0092] In another embodiment, the present invention proposes a heterogeneous UAV cluster pre-reconfiguration system based on a hybrid genetic algorithm, characterized by comprising:

[0093] A target task clusterer is used to cluster the tasks to be completed into multiple task clusters according to the geographical characteristics of the tasks;

[0094] The drone resource pool is used to accommodate drones that can participate in the reconstruction of a heterogeneous drone cluster after it is damaged;

[0095] The task allocator is used to reorganize the heterogeneous UAV cluster based on the generated task clusters and the UAV resource pool through a hybrid genetic algorithm.

[0096] In this embodiment, the working principles and workflows of the target task clusterer, the UAV resource pool and the task allocator are the same as those of the heterogeneous UAV cluster pre-reconstruction method based on a hybrid genetic algorithm in Embodiment 1, and therefore will not be described in detail here.

[0097] The above are only preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A heterogeneous UAV cluster pre-reconfiguration method based on a hybrid genetic algorithm, characterized in that: include: Cluster the tasks to be completed into multiple task clusters according to their geographical characteristics; Add the drones that can participate in the reconstruction of the damaged heterogeneous drone cluster to the drone resource pool; Through the hybrid genetic algorithm, the heterogeneous UAV cluster is reorganized based on the generated task clusters and UAV resource pool; The tasks to be completed are clustered into multiple task clusters according to the geographical characteristics of the tasks, specifically: Tasks to be completed in the area represents the task set, N t Indicates the number of tasks; According to the geographical characteristics of the tasks, a clustering algorithm based on the Gaussian mixture model is used to cluster the tasks to be completed into multiple task clusters. The cluster set c = (c1, c2, ..., c k ) T , c k represents the kth task cluster, and the superscript T indicates transposition; The formed task cluster contains the following information: the task sequence number c contained in the task cluster k =(c k1 ,c k2 ,…,c kL ) T And the number of loads required to complete the tasks in the task cluster c k1 ,c k2 ,…,c kL represents L task numbers, Ω1,Ω2,…,Ω M Indicates M type load; The clustering algorithm based on Gaussian mixture model is used to cluster the tasks to be completed into multiple task clusters, specifically: The geographic location distribution of all tasks is represented as a Gaussian mixture model, and the probability density distribution function P(pos j,t |θ) is expressed as: In the formula, π k is the probability weight of the k-component, K is the number of task clusters, Ν(pos j,t |μ k ,… k ) is based on ∑ k is the variance, μ k is a Gaussian distribution with a mean of POS j,t is the position coordinate of task j at time t, θ=(π1,...,π K ,μ1,...,μ K ,∑1,...,∑ K ); Introduce a latent variable z j , represents the task cluster number to which each task may be assigned in advance, and P(pos j,t |z j =k,θ)=Ν(pos j,t |μ k ,∑ k ), then the joint probability distribution P(pos,z|θ) of the position information of all tasks is expressed as: When z j = k, II(z j ,k)=1, otherwise 0; Calculate the latent variable z j The expectation E(z) of the likelihood function under the current parameter θ j =k|pos j,t ,θ): Update the value of θ by maximizing the posterior distribution Q value: In the formula, θ n represents the value of the parameter θ in the nth round; Update μ k and∑ k : In the formula, and represents the updated μ k and∑ k ; Finally, according to Calculate the task cluster to which each task belongs.

2. The method for pre-reconfiguration of heterogeneous UAV clusters based on a hybrid genetic algorithm as claimed in claim 1, characterized in that: The adding of the drones that can participate in the reconstruction after the heterogeneous drone cluster is damaged into the drone resource pool is specifically as follows: The drone numbers that can participate in the reconstruction are Represents a set of numbers, N u is the total number of drones in the drone resource pool; drones that can participate in the reconstruction are added to the drone resource pool, and the load in the drone resource pool is represented as an M-dimensional vector Ω1,Ω2,...,Ω M represents the M type of load, and the load carried by UAV i is expressed as Ω i1 ,Ω i2 ,...,Ω iM Indicates the M-type payload carried by UAV i.

3. The heterogeneous UAV cluster pre-reconfiguration method based on hybrid genetic algorithm according to claim 1, characterized in that: The hybrid genetic algorithm is used to reorganize the heterogeneous UAV cluster for the generated task cluster and the UAV resource pool, specifically: The task sequence and drone combination in the task cluster are used as two chromosomes for real number encoding, and a mobile energy consumption model of the drone cluster based on the task sequence is constructed. The total energy consumption during the task process is taken as the algorithm optimization target, and drone clusters equal to the number of task clusters and the matching relationship between drone clusters and task clusters are obtained.

4. The method for pre-reconfiguration of heterogeneous UAV clusters based on a hybrid genetic algorithm as claimed in claim 3, characterized in that: The optimization goal of the algorithm is: In the formula, ξ represents the task sequence, is a cluster of k drones, K is the number of mission clusters, Represents a single drone cluster s k energy consumption; The following constraints exist during the allocation process: In the formula, C1 and C2 indicate that the generated drone cluster can only be assigned to one task cluster, and a task cluster can only be completed by one drone cluster. C3 indicates that the drone cluster s k The load carried needs to meet the task cluster c k The required load requirement,C4 indicates that there is a two-dimensional distribution relationship between the UAV cluster and the task cluster;x cs A two-dimensional variable representing the distribution relationship between drone clusters and task clusters, x cs =1 and x cs = 0 respectively indicates that the drone cluster s k Responsible and irresponsible task clusters c k A collection of tasks; Indicates drone swarms k The number of load resources available, Indicates that task cluster c is completed k The number of load resources required.

5. The method for pre-reconfiguration of heterogeneous UAV clusters based on a hybrid genetic algorithm as claimed in claim 4, characterized in that: The drone swarms k The energy consumption is: In the formula, is the task cluster c assigned to the cluster k The total number of tasks, For clusters in task sequences The energy consumption generated by moving from the lth task to the l+1th task in is the energy consumption of the cluster in task j, It is the energy consumption of the drone due to waiting.

6. The method for pre-reconfiguration of heterogeneous UAV clusters based on a hybrid genetic algorithm as claimed in claim 5, characterized in that: The drone swarms k The energy consumption is calculated by the following formula: Where N is the UAV cluster s k The number of drones in V wait Indicates the speed at which the drone hovers at the mission location. It is a drone swarm k The time it takes to complete task j, Represents the drone cluster s k The time spent hovering at task j; Represents the drone cluster s k The energy consumption from flying from task j to task j', For drone swarms k The speed of flight, Represents the drone cluster s k The time it takes to fly from task j to task j', according to Calculated; P(V wait )and The energy consumption of the drone cluster is calculated according to the following formula: In the formula, P0 and P i They represent the blade profile power and induced power in the hovering state, U tip and U0 represent the rotor tip speed and the average rotor speed in the hovering state, respectively, d0 and s represent the fuselage drag ratio and the rotor solidity, respectively, ρ and A represent the air density and the rotor disc area, respectively, and v is the speed of the UAV cluster.

7. The method for pre-reconfiguration of heterogeneous UAV clusters based on a hybrid genetic algorithm as claimed in claim 6, characterized in that: and T skj Calculated according to the following formula: Where, d j,j′ is the distance from task j to task j'; η j is the amount of data information that needs to be transmitted for task j, W is the spectrum bandwidth resources available between UAVs, It is a drone swarm k The number of drones involved in the communication, It is a drone swarm k The transmission power, is the channel fading parameter, β0 is the channel gain when the distance is 1m, and d j,r is the communication distance between the drone cluster and the relay communication drone, σ 2 Represents the ambient power spectral density.

8. A heterogeneous UAV swarm pre-reconfiguration system based on a hybrid genetic algorithm, characterized in that: The method for pre-reconfiguration of a heterogeneous UAV cluster as claimed in any one of claims 1 to 7 comprises: A target task clusterer is used to cluster the tasks to be completed into multiple task clusters according to the geographical characteristics of the tasks; The drone resource pool is used to accommodate drones that can participate in the reconstruction of a heterogeneous drone cluster after it is damaged; The task allocator is used to reorganize the heterogeneous UAV cluster based on the generated task clusters and the UAV resource pool through a hybrid genetic algorithm.

Images