CN116009569A - Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene - Google Patents

Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene Download PDF

Info

Publication number
CN116009569A
CN116009569A CN202111218159.7A CN202111218159A CN116009569A CN 116009569 A CN116009569 A CN 116009569A CN 202111218159 A CN202111218159 A CN 202111218159A CN 116009569 A CN116009569 A CN 116009569A
Authority
CN
China
Prior art keywords
unmanned aerial
aerial vehicle
task
chromosome
target
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111218159.7A
Other languages
Chinese (zh)
Inventor
郑洪源
张康良
陈珍
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202111218159.7A priority Critical patent/CN116009569A/en
Publication of CN116009569A publication Critical patent/CN116009569A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

Aiming at the problem of cooperative task planning of heterogeneous multi-unmanned aerial vehicle under SEAD task scene, the invention provides a novel multi-type gene chromosome genetic algorithm. The algorithm not only considers the isomerism, kinematics and resource constraint of the unmanned aerial vehicle, but also considers the time sequence priority constraint among tasks. Considering the situation that the unmanned aerial vehicle is sunk into infinite mutual waiting deadlock caused by time priority constraint and the sequence of the unmanned aerial vehicle executing tasks, the execution process of the tasks is abstracted into a directed graph, and a deadlock detection and release algorithm is designed based on the directed graph. In addition, specific genetic operators, such as initialization, fitness, crossover and mutation, are designed to ensure the feasibility of chromosomes during evolution. The algorithm also introduces a Dubins curve and path extension strategy to generate the real flight path of the unmanned aerial vehicle. The method provided by the invention can distribute reasonable task sequences and generate the real flight path of the unmanned aerial vehicle cluster on the premise of lower task completion time for the unmanned aerial vehicle cluster in a limited time.

Description

Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene
Technical Field
The invention relates to the field of heterogeneous multi-unmanned aerial vehicle task planning in SEAD character scenes, which is mainly used for distributing reasonable task sequences and generating real flight paths of unmanned aerial vehicle clusters on the premise of low task completion time for the unmanned aerial vehicle clusters in a limited time.
Background
In recent years, unmanned aerial vehicle technology has been widely developed. However, due to the low viability and limited payload of single-frame drones, it is difficult to accomplish the compacting enemy air defense (sea) task. Therefore, the cooperative task execution of multiple unmanned aerial vehicles is an unavoidable trend. Task allocation has become one of the most challenging problems in multi-unmanned systems due to the complexity of the task environment, the differences between the members of the multi-unmanned system, the complexity of the task demands, and the computational complexity.
Disclosure of Invention
The invention aims to: some previous studies of mission allocation for drones have considered only a single type of mission or a homogeneous drone. Previous methods have not been directly applicable to the sea task scenario due to lack of consideration for multiple types of tasks and heterogeneous drones. Therefore, how to effectively complete the data acquisition task in a limited communication range through unmanned aerial vehicle track optimization becomes a main technical problem.
In order to solve the technical problem, the invention provides a heterogeneous multi-unmanned aerial vehicle task planning method based on a multi-type gene chromosome genetic algorithm in an SEAD task scene, which can distribute reasonable task sequences and generate a real flight path of an unmanned aerial vehicle cluster on the premise of lower task completion time for the unmanned aerial vehicle cluster in a limited time.
The technical scheme is as follows: in order to achieve the technical effects, the technical scheme provided by the invention is as follows:
the heterogeneous multi-unmanned aerial vehicle task planning method based on the multi-type gene chromosome genetic algorithm in the SEAD task scene is characterized by comprising the following steps:
(1) And constructing a system combination optimization model, wherein the model mainly aims at the collaborative combat problem of the heterogeneous fixed wing unmanned aerial vehicle cluster, and takes the limitation of limited resources (such as ammunition resources) and kinematics into consideration. The combat target is a stationary ground target with time-first constraints:
1) Let N be T The task targets, target sets are as follows:
Figure RE-GSB0000197473780000011
2) Each target needs to execute 3 tasks respectively, and the task sets are as follows:
M T ＝{C，A，V} (2)
wherein C represents a investigation and identification task, A represents an attack task, and V represents a damage evaluation task. And the a task needs to be performed after the C task is completed, and the V task needs to be performed after the a task is completed.
4) Let there be N V The individual fixed wing drones together perform the sea task, the aggregate of which is:
Figure RE-GSB0000197473780000021
wherein t represents the type of unmanned aerial vehicle, and there are 3 heterogeneous unmanned aerial vehicles in unmanned aerial vehicle cluster, and unmanned aerial vehicle type has reconnaissance unmanned aerial vehicle (Surveillance), attack unmanned aerial vehicle (Combat), ammunition unmanned aerial vehicle (station), and different unmanned aerial vehicle functions and tasks that can be carried out see table one.
Figure RE-GSB0000197473780000022
Watch unmanned aerial vehicle type
4) To more closely match the real flight path, the method introduces a Dubins curve to generate the flight path of the drone. The configuration of the drone can be represented by 3 state variables, namely the planar cartesian inertial coordinates x and y, and the heading of the UAV
Figure RE-GSB0000197473780000023
The kinematic equation is as follows:
Figure RE-GSB0000197473780000024
wherein in the formula V u For the speed of the unmanned aerial vehicle u,
Figure RE-GSB0000197473780000025
and u is the minimum turning radius, and c is the steering engine operation coefficient. The flight path derived from this model, the Dubins curve, is used for path generation for the drone.
5) The course angle discretization method is adopted in the chapter to discretize the course angle of the unmanned aerial vehicle, and the CMTAP problem graph theory description method is adopted to describe the problem. The method discretizes the course angle of the unmanned aerial vehicle into fixed integer angle values, correspondingly expresses the possible configuration of the UAV and the path thereof in a directed graph mode, and certain basic expressions are as follows:
Figure RE-GSB0000197473780000031
wherein ,
Figure RE-GSB0000197473780000032
to the extent that the heading angle is discrete, i.e. how many angles the 2 pi angle is evenly dispersed into; h is a course discrete angle set; v (V) T Representing the coordinates of a combat target and the spatial configuration of the unmanned aerial vehicle when approaching the target for a node set; n (N) 1 Is V (V) T Is a potential of (2); v (V) U Is the initial position of the unmanned aerial vehicle; v is a node set of a CMTAP directed graph, and represents all possible configurations of the unmanned aerial vehicle; n (N) 2 The potential of V, the number of all nodes of the directed graph; e is the edge set of the CMTAP directed graph, pointing from the node in V to V T Representing all possible unmanned aerial vehicle paths; n (N) E Is the potential of E.
6) The optimization objective of this party is the total time of task execution, the objective function is as follows:
Figure RE-GSB0000197473780000033
Figure RE-GSB0000197473780000034
(2) In order to quickly search a solution space and find a solution with a target function value as low as possible, a multi-type gene chromosome genetic algorithm is designed, and a deadlock detection and release algorithm based on a task execution priority directed graph is designed in consideration of a chromosome which may cause deadlock in the iterative process of the genetic algorithm.
1) Encoded version of the solution:
the chromosomal coding of multiple types of genes is a key part of genetic algorithms. We encode the form of the task allocation scheme, i.e. the chromosome, into a 5 row N column matrix. N is a base factor. Each column of chromosomes is a gene, a gene has three parts: the first part is a matrix with two rows and two lines, and represents a combined sequence of a target and a task; the second part is a third row and a fourth row of the matrix, the third row represents which unmanned aerial vehicle is used for executing the task determined by two rows, and the fourth row targets approach the angle
Figure RE-GSB0000197473780000041
To facilitate the writing of crossover and mutation operators, we add a third part, the fifth row of the matrix: gene order. There are two types of genes in the chromosome. The first type is an attack type, with a second number of rows greater than 0. While the number of offending genes in the second row means how many drones are needed to attack the target together. The other is the investigation task genes whose second row is less than 0. -1 represents a investigation identification task, -2 represents a damage assessment task.
2) Initializing an operator:
the initialisation operator is used to obtain an initial population for the evolution process.
The details of the initialization operator are as follows:
step one: based on the first two lines of chromosomes are generated. Then randomly selecting a corresponding function of the unmanned aerial vehicle and a target approach angle for each gene from the set
Figure RE-GSB0000197473780000042
The third and fourth rows used to generate genes. Note that if one task requires two or moreMore unmanned aerial vehicles are executed together, the genes generated by the task should select different unmanned aerial vehicles as the third row of the unmanned aerial vehicles, and the number of generated genes is equal to the number of unmanned aerial vehicles required to jointly complete the hitting task. After all tasks are assigned, a chromosome based on the target sequence is generated.
Step two: and sequencing the chromosomes according to the serial numbers of the unmanned aerial vehicles to generate the chromosomes based on the unmanned aerial vehicle sequence.
Step three: and (3) running a deadlock-free algorithm to ensure the feasibility of the chromosome.
Step four: repeating the steps one to three N p Next, we then get an N p An initial population of viable individuals.
3) Crossover operator:
during crossover, two parent chromosomes are randomly selected by roulette methods to generate a pair of child chromosomes. The probability of selecting a parent chromosome is determined by the fitness function value of the parent chromosome. We used a single point hybridization method to generate the daughter chromosome. The initialization method is as follows:
step one: and converting the chromosome based on the unmanned aerial vehicle sequence into a chromosome based on the target sequence.
Step two: the start and end points of the chromosomal crossover were randomly selected.
Step three: genes are exchanged between the start and end points, producing two daughter chromosomes.
Step four: if the exchanged chromosomes violate the resource constraint of the unmanned aerial vehicle or the gene of multiple attacks is executed by the same unmanned aerial vehicle, a corresponding unmanned aerial vehicle with another corresponding function is randomly selected for the gene. The resource table is updated.
Step five: and running a deadlock detection and release algorithm on the two chromosomes.
Step six: and converting the chromosome based on the target sequence into a chromosome based on the unmanned plane sequence according to the fifth row of the chromosome and the unmanned plane sequence.
4) Mutation operator:
mutation operators can avoid the algorithm converging to local optimum and help the genetic algorithmThe algorithm converges to a globally optimal solution. Unmanned aerial vehicle with gene changed by our mutation method and target approach angle
Figure RE-GSB0000197473780000051
To create new chromosomes when the drone allocation changes, the drone resource constraints should be ensured. For multiple attack tasks, the unmanned aerial vehicle assigned after mutation should be different from the unmanned aerial vehicle assigned before mutation to perform this task.
5) Deadlock detection and release algorithm:
step one: based on the chromosome, a time-series priority directed graph is generated.
Step two: the directed graph is traversed in depth first, whether a loop exists in the directed graph is detected, if yes, the chromosome is a deadlock solution, and if not, the original chromosome is directly returned.
Step three: from the chromosome, it is determined whether each edge in the loop is generated by an execution sequence of the drone or a timing priority constraint.
Step four: the edges of the unmanned aerial vehicle execution sequence generation are reversed.
Step five: updating the chromosome according to the modified time-first directed graph.
(3) Aiming at the found optimal chromosome, designing a path extension algorithm according to the Dubins curve and the time sequence constraint of the task to generate a real flight path for the unmanned aerial vehicle cluster:
if the UAV performs non-simultaneous tasks, adding a plurality of waiting circles (namely minimum turning circles) taking the minimum turning radius of the UAV as the radius after the UAV reaches a specified target until the UAV meets the time requirement; if a concurrent task is performed, it is relatively complex: if the UAV needs to lengthen the path length not smaller than the circumference of the minimum turning circle, a plurality of radii are added in [ ρ ] min ，2ρ min ) The radius and the number of waiting circles of the interval are determined by the length of the path which needs to be prolonged; if the unmanned aerial vehicle needs to have an extended path length less than its minimum turning circumference, an intermediate point is created through which the UAV flies around while flying to the target, to meet the time requirement.
Drawings
FIG. 1 is an algorithm flow chart of a multi-type gene chromosome genetic algorithm.
FIG. 2 is a coding strategy for a genetic algorithm for a multi-type gene chromosome.
FIG. 3 is an initialization operator for a multi-type genetic chromosomal genetic algorithm.
FIG. 4 is an example of a deadlocked chromosome.
FIG. 5 is a task timing priority directed graph derived based on the deadlock chromosome of FIG. 3.
FIG. 6 is a task timing priority directed graph of a chromosome after execution of an unlock.
FIG. 7 is a graph of unmanned aerial vehicle cluster trajectories after distribution of a multi-type genetic chromosome genetic algorithm.
Detailed Description
The invention will be further described with reference to the accompanying drawings.
Fig. 1 is a flowchart of an algorithm, and first, we need to set some necessary parameters of the algorithm, such as iteration times, population size, and probability of cross variation. An initial population is then generated using an initialization operator. The next step runs a deadlock detection and release algorithm to ensure that there are no deadlocked chromosomes in the initial population. The next step is to calculate the objective function value for each chromosome, i.e. the feasible solution, and select the crossing parent chromosomes based on the objective function values. The deadlock resolution algorithm is run again. Repeating the steps until the iteration times are reached. And finally, generating a flight path of the unmanned aerial vehicle cluster according to the optimal solution.
FIG. 2 shows an example of chromosome coding:
as shown in the figure, the first row of the chromosome is a target sequence number in the SEAD task, the second row is a task type of the target, 3 tasks need to be executed under each target, the task code number is-1 for investigation and identification, the attack task is greater than 0, and the number of the attack task is the number of unmanned aerial vehicles which need to jointly execute the task and the damage evaluation task code number is-2. Orange genes are aggressive genes in the figure. The third row is the serial number of the unmanned aerial vehicle, and the function of the unmanned aerial vehicle corresponds to the corresponding task type, for example, the unmanned aerial vehicle 1 is a reconnaissance unmanned aerial vehicle only capable of executing the reconnaissance task, the unmanned aerial vehicle 2 is an attack unmanned aerial vehicle only capable of executing the attack task, and the unmanned aerial vehicle 3 is a hybrid unmanned aerial vehicle capable of executing the reconnaissance task and the attack task. The fourth row x represents the target approach angle of the unmanned aerial vehicle, and the target approach angle of the unmanned aerial vehicle is (2 pi/18) x. The fifth row of chromosomes is the order of genes.
FIG. 3 is a flow chart of an initialization operator for a multi-type gene chromosome genetic algorithm.
The purpose of the initializer is to generate an initial population of genetic algorithms, as shown in fig. 3, first we generate all combinations of target-task types, i.e. the first two lines of chromosomes, from the target information. Then randomly selecting a corresponding unmanned aerial vehicle and a target approach angle according to the task type
Figure RE-GSB0000197473780000061
The third and fourth rows used to generate genes. Note that if a task requires two or more drones to perform together, then the genes generated by the task should select different drones as their third row, the number of genes generated being equal to the number of drones that need to together complete the hit task. As shown in fig. 2, the attack task of the target 2 needs 2 unmanned aerial vehicles to be completed together, so it is necessary to generate two genes, respectively, gene 4 and gene 5, and the third row of the genes 4 and 5, that is, the unmanned aerial vehicle performing the task, cannot be the same, respectively, unmanned aerial vehicle 2 and unmanned aerial vehicle 3. After all tasks are assigned, a chromosome based on the target sequence is generated. And sequencing the chromosomes according to the serial numbers of the unmanned aerial vehicles to generate the chromosomes based on the unmanned aerial vehicle sequence. And (3) running a deadlock-free algorithm to ensure the feasibility of the chromosome. Finally repeating the steps one to three N p Next, we then get an N p An initial population of viable individuals.
FIG. 4 is an example of a deadlocked chromosome.
Unmanned aerial vehicles wait each other for their tasks to complete in order to begin their respective next tasks, but they cannot complete the tasks assigned to them due to the time conflicts between the tasks assigned to them, thus becoming involved in endless waiting.
FIG. 5 is a task timing priority directed graph derived based on the deadlock chromosome of FIG. 4.
First, a task execution priority directed graph is generated from the chromosome in fig. 3. The task execution timing constraint is composed of the task type and the sequence of the unmanned aerial vehicle execution tasks, as shown in fig. 4, the unmanned aerial vehicle 3 is required to complete the tasks in the gene 5 first, the task in the gene 5 is completed, namely the damage evaluation task of the target 2 is required to complete the attack tasks of the target 2, namely the tasks in the gene 4 and the task in the gene 7, however, the unmanned aerial vehicle 3 is required to complete the tasks in the gene 6 first when the tasks in the gene 7 are executed, so that the genes 5,6 and 7 sink into the waiting dead office.
In order to remove the deadlock, the deadlock is discovered first, the directed graph is traversed in depth first, whether the directed graph has a loop or not is detected, and if the loop exists, the chromosome is the deadlock solution. As shown in fig. 5, we find the ring in the directed graph. The execution sequence of the unmanned aerial vehicle can be changed, but the time sequence of the target task cannot be changed, so that each edge in the cycle needs to be determined according to the chromosome in the next step, and whether the edge is generated by the execution sequence of the unmanned aerial vehicle or the time sequence priority constraint is determined. The edges of the directed graph brought by the unmanned aerial vehicle execution sequence are 5-6, 6-7. The edges of the unmanned aerial vehicle execution sequence generation are reversed. Updating the chromosome according to the modified time-first directed graph.
FIG. 6 is a task timing priority directed graph of a chromosome after execution of an unlock.
FIG. 7 is a graph of unmanned aerial vehicle cluster trajectories after distribution of a multi-type genetic chromosome genetic algorithm.
As shown in fig. 7, there are 5 targets and 7 unmanned aerial vehicles in total in this scenario, the target 1 is executed by the unmanned aerial vehicle 3 alone, and since the fixed wing unmanned aerial vehicle cannot hover, each time the unmanned aerial vehicle 3 completes execution of one task of the task set (C, a, V) on the target 1, it is required to fly around the target 1 one turn with a minimum turning radius, and then the next task is executed. The C task on the target 2 is executed by the unmanned aerial vehicle 1, and the A task on the target 2 is executed simultaneously by two unmanned aerial vehicles, so that the unmanned aerial vehicle 1 and the unmanned aerial vehicle 6 need to arrive at the target 2 simultaneously, and therefore, the flight path of the unmanned aerial vehicle 6 needs to be prolonged to meet the time requirement, and the path of the unmanned aerial vehicle 6 flying from the target 5 to the target 2 is prolonged, so that the unmanned aerial vehicle 1 and the unmanned aerial vehicle 6 arrive at the target 2 simultaneously to execute the A task. The task set on the same management target 3 is completed by the unmanned aerial vehicle 4 and the unmanned aerial vehicle 5 together; the task set of the target 4 is completed by the unmanned aerial vehicle 3, the unmanned aerial vehicle 2 and the unmanned aerial vehicle 7 together. The task set of the target 5 is jointly completed by the unmanned aerial vehicle 6 and the unmanned aerial vehicle 7 and the unmanned aerial vehicle 2.
The foregoing is only a preferred embodiment of the invention, it being noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the present invention, and such modifications and adaptations are intended to be comprehended within the scope of the invention.

Claims (2)

  1. A heterogeneous multi-unmanned aerial vehicle task planning method based on a multi-type gene chromosome genetic algorithm in an SEAD task scene is characterized by comprising the following steps:
    (1) And constructing a system combination optimization model, wherein the model mainly aims at the collaborative combat problem of the heterogeneous fixed wing unmanned aerial vehicle cluster, and takes the limitation of limited resources (such as ammunition resources) and kinematics into consideration. The combat target is a stationary ground target with time-first constraints:
    1) Let N be T The task targets, target sets are as follows:
    Figure FSA0000255282220000015
    2) Each target needs to execute 3 tasks respectively, and the task sets are as follows:
    M T ＝{C，A，V} (2)
    wherein C represents a investigation and identification task, A represents an attack task, and V represents a damage evaluation task. And the a task needs to be performed after the C task is completed, and the V task needs to be performed after the a task is completed.
    3) Let there be N V The fixed wings are not provided withThe man-machine performs the SEAD task together, their set is:
    Figure FSA0000255282220000011
    wherein t represents the type of unmanned aerial vehicle, and there are 3 heterogeneous unmanned aerial vehicles in unmanned aerial vehicle cluster, and unmanned aerial vehicle type has reconnaissance unmanned aerial vehicle (Surveillance), attack unmanned aerial vehicle (Combat), ammunition unmanned aerial vehicle (station), and different unmanned aerial vehicle functions and tasks that can be carried out see table one.
    Figure FSA0000255282220000012
    Form-unmanned aerial vehicle type
    4) To more closely match the real flight path, the method introduces a Dubins curve to generate the flight path of the drone. The configuration of the drone can be represented by 3 state variables, namely the planar cartesian inertial coordinates x and y, and the heading of the UAV
    Figure FSA0000255282220000016
    The kinematic equation is as follows:
    Figure FSA0000255282220000013
    wherein in the formula V u For the speed of the unmanned aerial vehicle u,
    Figure FSA0000255282220000014
    and u is the minimum turning radius, and c is the steering engine operation coefficient. The flight path derived from this model, the Dubins curve, is used for path generation for the drone.
    5) The course angle discretization method is adopted in the chapter to discretize the course angle of the unmanned aerial vehicle, and the CMTAP problem graph theory description method is adopted to describe the problem. The method discretizes the course angle of the unmanned aerial vehicle into fixed integer angle values, correspondingly expresses the possible configuration of the UAV and the path thereof in a directed graph mode, and certain basic expressions are as follows:
    Figure FSA0000255282220000021
    wherein ,
    Figure FSA0000255282220000022
    to the extent that the heading angle is discrete, i.e. how many angles the 2 pi angle is evenly dispersed into; h is a course discrete angle set; v (V) T Representing the coordinates of a combat target and the spatial configuration of the unmanned aerial vehicle when approaching the target for a node set; n (N) 1 Is V (V) T Is a potential of (2); v (V) U Is the initial position of the unmanned aerial vehicle; v is a node set of a CMTAP directed graph, and represents all possible configurations of the unmanned aerial vehicle; n (N) 2 The potential of V, the number of all nodes of the directed graph; e is the edge set of the CMTAP directed graph, pointing from the node in V to V T Representing all possible unmanned aerial vehicle paths; n (N) E Is the potential of E.
    6) The optimization objective of this party is the total time of task execution, the objective function is as follows:
    Figure FSA0000255282220000023
    Figure FSA0000255282220000024
    (2) In order to quickly search a solution space and find a solution with a target function value as low as possible, a multi-type gene chromosome genetic algorithm is designed, and a deadlock detection and release algorithm based on a task execution priority directed graph is designed in consideration of a chromosome which may cause deadlock in the iterative process of the genetic algorithm.
    1) Encoded version of the solution:
    the chromosomal coding of multiple types of genes is a key part of genetic algorithms. We encode the form of the task allocation scheme, i.e. the chromosome, into a 5 row N column matrix. N is a base factor. Each column of chromosomes is a gene, a gene has three parts: the first part is a matrix with two rows and two lines, and represents a combined sequence of a target and a task; the second part is a third row and a fourth row of the matrix, the third row represents which unmanned aerial vehicle is used for executing the task determined by two rows, and the fourth row targets approach the angle
    Figure FSA0000255282220000033
    To facilitate the writing of crossover and mutation operators, we add a third part, the fifth row of the matrix: gene order. There are two types of genes in the chromosome. The first type is an attack type, with a second number of rows greater than 0. While the number of offending genes in the second row means how many drones are needed to attack the target together. The other is the investigation task genes whose second row is less than 0. -1 represents a investigation identification task, -2 represents a damage assessment task. />
    2) Initializing an operator:
    the initialisation operator is used to obtain an initial population for the evolution process.
    The details of the initialization operator are as follows:
    step one: based on the first two lines of chromosomes are generated. Then randomly selecting a corresponding function of the unmanned aerial vehicle and a target approach angle for each gene from the set
    Figure FSA0000255282220000032
    The third and fourth rows used to generate genes. Note that if a task requires two or more drones to perform together, then the genes generated by the task should select different drones as their third row, the number of genes generated being equal to the number of drones that need to together complete the hit task. After all tasks are assigned, a chromosome based on the target sequence is generated.
    Step two: and sequencing the chromosomes according to the serial numbers of the unmanned aerial vehicles to generate the chromosomes based on the unmanned aerial vehicle sequence.
    Step three: and (3) running a deadlock-free algorithm to ensure the feasibility of the chromosome.
    Step four: repeating the steps one to three N p Next, we then get an N p An initial population of viable individuals.
    3) Crossover operator:
    during crossover, two parent chromosomes are randomly selected by roulette methods to generate a pair of child chromosomes. The probability of selecting a parent chromosome is determined by the fitness function value of the parent chromosome. We used a single point hybridization method to generate the daughter chromosome. The initialization method is as follows:
    step one: and converting the chromosome based on the unmanned aerial vehicle sequence into a chromosome based on the target sequence.
    Step two: the start and end points of the chromosomal crossover were randomly selected.
    Step three: genes are exchanged between the start and end points, producing two daughter chromosomes.
    Step four: if the exchanged chromosomes violate the resource constraint of the unmanned aerial vehicle or the gene of multiple attacks is executed by the same unmanned aerial vehicle, a corresponding unmanned aerial vehicle with another corresponding function is randomly selected for the gene. The resource table is updated.
    Step five: and running a deadlock detection and release algorithm on the two chromosomes.
    Step six: and converting the chromosome based on the target sequence into a chromosome based on the unmanned plane sequence according to the fifth row of the chromosome and the unmanned plane sequence.
    4) Mutation operator:
    the mutation operator can avoid the algorithm to converge to the local optimum and help the genetic algorithm to converge to the global optimum solution. Unmanned aerial vehicle with gene changed by our mutation method and target approach angle
    Figure FSA0000255282220000041
    To create new chromosomes when the drone allocation changes, the drone resource constraints should be ensured. For multiple attack tasks, after mutationThe unmanned aerial vehicle assigned should be different from the unmanned aerial vehicle assigned before mutation to perform this task.
    5) Deadlock detection and release algorithm:
    step one: based on the chromosome, a time-series priority directed graph is generated.
    Step two: the directed graph is traversed in depth first, whether a loop exists in the directed graph is detected, if yes, the chromosome is a deadlock solution, and if not, the original chromosome is directly returned.
    Step three: from the chromosome, it is determined whether each edge in the loop is generated by an execution sequence of the drone or a timing priority constraint.
    Step four: the edges of the unmanned aerial vehicle execution sequence generation are reversed.
    Step five: updating the chromosome according to the modified time-first directed graph.
    (3) Aiming at the found optimal chromosome, designing a path extension algorithm according to the Dubins curve and the time sequence constraint of the task to generate a real flight path for the unmanned aerial vehicle cluster:
    if the UAV performs non-simultaneous tasks, adding a plurality of waiting circles (namely minimum turning circles) taking the minimum turning radius of the UAV as the radius after the UAV reaches a specified target until the UAV meets the time requirement; if a concurrent task is performed, it is relatively complex: if the UAV needs to lengthen the path length not smaller than the circumference of the minimum turning circle, a plurality of radii are added in [ ρ ] min ，2ρ min ) The radius and the number of waiting circles of the interval are determined by the length of the path which needs to be prolonged; if the unmanned aerial vehicle needs to have an extended path length less than its minimum turning circumference, an intermediate point is created through which the UAV flies around while flying to the target, to meet the time requirement.
  2. 2. The heterogeneous multi-unmanned aerial vehicle task planning method based on the multi-type gene chromosome genetic algorithm in the SEAD task scene of claim 1 is characterized in that reasonable task sequences are distributed and real flight paths of unmanned aerial vehicle clusters are generated on the premise that the task completion time is low for the unmanned aerial vehicle clusters in a limited time.
CN202111218159.7A 2021-10-19 2021-10-19 Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene Pending CN116009569A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111218159.7A CN116009569A (en) 2021-10-19 2021-10-19 Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111218159.7A CN116009569A (en) 2021-10-19 2021-10-19 Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene

Publications (1)

Publication Number Publication Date
CN116009569A true CN116009569A (en) 2023-04-25

Family

ID=86027203

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111218159.7A Pending CN116009569A (en) 2021-10-19 2021-10-19 Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene

Country Status (1)

Country Link
CN (1) CN116009569A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116542468A (en) * 2023-05-06 2023-08-04 中国人民解放军32370部队 Unmanned aerial vehicle cluster task planning method
CN118012079A (en) * 2024-04-10 2024-05-10 西安现代控制技术研究所 Multi-angle attack lateral nominal track generation method based on overload capacity

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116542468A (en) * 2023-05-06 2023-08-04 中国人民解放军32370部队 Unmanned aerial vehicle cluster task planning method
CN116542468B (en) * 2023-05-06 2023-10-20 中国人民解放军32370部队 Unmanned aerial vehicle cluster task planning method
CN118012079A (en) * 2024-04-10 2024-05-10 西安现代控制技术研究所 Multi-angle attack lateral nominal track generation method based on overload capacity

Similar Documents

Publication Publication Date Title
Jia et al. Cooperative multiple task assignment problem with stochastic velocities and time windows for heterogeneous unmanned aerial vehicles using a genetic algorithm
CN106990792B (en) Multi-unmanned aerial vehicle collaborative time sequence coupling task allocation method based on hybrid gravity search algorithm
CN110865653B (en) Distributed cluster unmanned aerial vehicle formation transformation method
CN102506863B (en) Universal gravitation search-based unmanned plane air route planning method
CN116009569A (en) Heterogeneous multi-unmanned aerial vehicle task planning method based on multi-type gene chromosome genetic algorithm in SEAD task scene
Huang et al. A new dynamic path planning approach for unmanned aerial vehicles
CN115494873A (en) Heterogeneous multi-unmanned aerial vehicle cooperative task allocation method based on Monte Carlo tree search architecture under time sequence constraint
CN111240211A (en) Dynamic recovery method for unmanned aerial vehicle group
CN114417735A (en) Multi-unmanned aerial vehicle cooperative task planning method in cross-regional combined combat
CN112926825A (en) Multi-unmanned aerial vehicle task allocation method based on multi-target quantum shrimp swarm mechanism
CN115471110A (en) Conditional probability-based multi-heterogeneous unmanned aerial vehicle system cooperative task allocation method considering multi-objective evolutionary algorithm
CN113985915B (en) Multi-leader clustering formation tracking control method for high-order group system
CN114897215A (en) Method for optimizing multi-unmanned aerial vehicle reconnaissance task allocation based on unsupervised learning discrete pigeon flock
Shen Bionic communication network and binary pigeon-inspired optimization for multiagent cooperative task allocation
Wei et al. Multi-UAVs cooperative reconnaissance task allocation under heterogeneous target values
Jia et al. Automatic path planning of unmanned combat aerial vehicle based on double-layer coding method with enhanced grey wolf optimizer
CN117113274A (en) Heterogeneous network data-free fusion method and system based on federal distillation
CN116088586A (en) Method for planning on-line tasks in unmanned aerial vehicle combat process
CN114333429B (en) Rule extraction method for multi-unmanned aerial vehicle target coverage task
CN115829230A (en) Heterogeneous unmanned aerial vehicle group cooperative attack task allocation method
CN115509256A (en) Unmanned aerial vehicle cluster collaborative planning and autonomous scheduling method
Senthilnath et al. Cooperative communication of UAV to perform multi-task using nature inspired techniques
Huang et al. Autonomous Reconnaissance Action of Swarm Unmanned System Driven by Behavior Tree
Fan et al. Research on mission planning for MAV/UAV cooperative combat
He et al. Function chain-based mission planning method for hybrid combat SoS

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination