A collaborative mission planning and allocation method for multiple fixed-wing long-flight UAVs to multiple targets

Abstract

The present invention discloses a method for planning and allocating collaborative tasks of multiple fixed-wing long-flight UAVs to multiple targets, which includes: obtaining the state information of each of the multiple UAVs and multiple targets; setting the initial parameters and objective function of the genetic algorithm with the state information, and initializing the population with an improved coding method; using the genetic algorithm, based on the pre-set initial parameters, according to the acquired UAV and target information, with the purpose of covering all targets with multiple UAV groups as quickly as possible, performing global decoding and optimization to obtain the optimal allocation result of this round, and re-evolving the population, performing global decoding and optimization to obtain the optimal allocation result of the task to be planned. The present invention is reasonably designed and efficient, can effectively utilize the genetic optimization algorithm, improve the algorithm coding method, and under the constraint of meeting the endurance of the UAV, multiple fixed-wing long-flight UAVs can complete multiple target tasks in the shortest time.

Description

Technical Field

[0001] The invention relates to the technical field of fixed-wing long-flight UAVs, and in particular to a method for planning and allocating collaborative tasks of multiple fixed-wing long-flight UAVs for multiple targets. Background Art

[0002] Drones are widely used in various civil and military fields due to their flexibility, convenience and speed. For example, drones are used to inspect power lines and pipelines; drones are used to replace manpower in geographic surveying and mapping; and drones are used for communication relay support and disaster monitoring in rescue operations after natural disasters such as earthquakes.

[0003] Among them, fixed-wing long-flight UAVs have broad application prospects in industry, agriculture, rescue, scientific research and military due to their advantages of long endurance, long range and ability to perform multiple types of tasks. When the number of available UAVs and targets is large, and their locations are scattered and far away from each other, it is necessary to coordinate the number of targets assigned to each UAV and the execution order to meet certain actual needs.

[0004] In the prior art, various heuristic algorithms are usually used to implement the task allocation planning of drones. However, in the genetic algorithm of the prior art, two-layer encoding is performed separately for multiple drones and multiple targets during encoding, which easily leads to invalid coding combinations and reduces the search efficiency of the algorithm. In addition, the optimization goal is to only consider the minimum sum of the execution time of all drones to achieve the economic efficiency of saving energy consumption, without considering the requirement of rapid coverage of all targets in actual special circumstances. Therefore, the genetic algorithm in the prior art will have the problem of long operation time and fall into local optimality under large-scale problems, and it is difficult to design a more suitable allocation plan under the requirement of rapid coverage of all targets in special circumstances. Summary of the invention

[0005] In view of this, the present invention provides a method for planning and allocating collaborative tasks for multiple fixed-wing long-flight UAVs to multiple targets.

[0006] The present invention discloses a method for planning and allocating collaborative tasks of multiple fixed-wing long-flight UAVs to multiple targets, which comprises:

[0007] Step 1: Obtain the state information of multiple drones and multiple targets; use the state information to set the initial parameters and objective function of the genetic algorithm, and initialize the population in an improved encoding method;

[0008] Step 2: Genetic algorithm is used to perform global decoding and optimization based on the pre-set initial parameters and the acquired drone and target information, with the goal of covering all targets with multiple drone groups as quickly as possible, to obtain the optimal allocation result for this round;

[0009] Step 3: If the optimal result of this round satisfies that the mission time of each drone is less than the endurance time of the drone, then the optimal result update is performed; otherwise, it is determined whether the preset number of iterations has been reached;

[0010] Step 4: If the optimal result of this round is better than the global optimal result, the global optimal result is updated with the optimal result of this round, otherwise it is not updated, and it is determined whether the preset number of iterations is reached;

[0011] Step 5: Re-evolve the population, globally decode, and optimize, repeat steps 2 to 5. If the preset number of iterations is reached, the optimal result at this time is the optimal allocation result of the task to be planned.

[0012] Furthermore, in step 1:

[0013] The status information of multiple drones obtained includes the number of drones , the serial number of each drone , take-off and landing position coordinates ( , Cruising speed and battery life , and the drones can be considered to take off at the same time; ;

[0014] The status information of multiple targets is obtained, including the number of targets. , the number of each target , location coordinates( , the time required for the drone to complete each target mission , where the number of targets Greater than the number of drones , the target position coordinates remain unchanged during the task execution; .

[0015] Furthermore, in step 1, the initial parameters include the number of populations, the number of iterations, the gene length and the objective function, wherein a group of genes of individuals in the population represents an allocation scheme, and a one-dimensional integer array is used to represent an allocation scheme.

[0016] Furthermore, in the improved encoding method, a length of A one-dimensional integer array is used as a gene code to represent a distribution scheme; The information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone.

[0017] Furthermore, in the improved encoding method, the rule for initializing the genes in the population is: the position of each gene in a gene encoding group is recorded as , ; For the front genes, in [1, ] to round up for initialization; the subsequent The genes at positions [1, ] are rounded for initialization.

[0018] Furthermore, in step 2, the purpose is to cover all targets as quickly as possible with a group of multiple drones, and the specific purpose is to minimize the time from the drone taking off to the completion of the last target; The time from the take-off of a UAV to the completion of the last target assigned to the UAV is ;make Minimum, the global optimal result of output , that is, the task execution route sequence of each UAV is obtained, the target number assigned to the first UAV and the execution order are , the target number assigned to the second UAV and the execution order are ;in, is the number of columns of the i-th row of data in the Cell, that is, the number of all elements in the i-th row of the Cell; The numbers assigned to the i-th drone are and The distance between two targets is based on the position coordinates of the two targets ( Calculated; For superposition operation The starting value of Start, to End the accumulation operation; is the flight speed of the i-th UAV; k is the cumulative operation The starting value of Start to End the accumulation operation; The number assigned to the i-th drone is The goal requires time to complete; is the element in the i-th row and j-th column of Cell, is the element in the i-th row and j+1-th column in Cell.

[0019] Furthermore, the global decoding and optimization are performed to obtain the optimal allocation result of this round, including:

[0020] Step 21: For a possible genetic code [u,v,w,x,y,…], the first The information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone; u, v, w, x, and y are all genes, and their values ​​are all integers;

[0021] Step 22: From The subsequent genes start decoding and obtain the selection order of the UAV mission targets;

[0022] Step 23: Front The genes are decoded to obtain the mission objectives corresponding to each drone;

[0023] Step 24: Put the numbers corresponding to the drones in order in front of each row of the drone mission target decoding array, and you can get the allocation result of this decoding gene, decode all the genes in the population, and calculate the objective function. The maximum value of the objective function is the optimal allocation result of this round.

[0024] Further, the step 22 comprises:

[0025] set up Gene [v,w,x,y] is decoded; the integer array in the gene contains the information of the target selection order, and the number array of the UAV mission target is established in order first [u1,v1,w1,x1,y1], decode the array of the UAV mission target obtained after decoding select Initialize []; then start decoding from gene v step by step, in the first step v means to get the target number array rest The second target is labeled v1 is the target decoding array select The first goal is [v1], then rest The target number selected in the first step is deleted, that is, after decoding, , [u1,w1,x1,y1]; in the second step w represents the target number array rest The first target is labeled u1 is the target decoding array select The second goal is [u1], then rest The target number selected in the second step is deleted, that is, after decoding, , [w1,x1,y1]; in the third step y represents the target number array rest The third target is labeled w1 as the target decoding array select The third goal is [w1], then rest The target number selected in the third step is deleted, that is, after decoding, , [x1,y1]; in the fourth step y represents the target number array rest The second target is labeled y1 is the target decoding array select The fourth goal is [y1], then rest The target number selected in the third step is deleted, that is, after decoding, , [x1]; at this point, the information contained in the subsequent genes has been completely decoded and automatically rest The number of the last target in the select In the end , indicating the selection order of UAV mission targets.

[0026] Further, the step 23 comprises:

[0027] Decode the first gene. , indicating the decoding array of the task target select In the Separate at each position, select Center front The target is assigned to the drone numbered 1. select Middle The targets after that are assigned to the drone numbered 2. When .

[0028] Further, the step 24 comprises:

[0029] Put the corresponding numbers of the drones into the select In front of each line, you can get the final distribution result , for an element in Cell ,when hour, For the number The first The number of the target, the distribution result The order of arrangement in is the execution order of the targets. The corresponding allocation result is ;in, = , The total number of non-empty elements in is equal to , , , represents the number of UAVs participating in task allocation, represents the target number of all participating tasks, is the element in the i-th row and the first column of the Cell, is the element in the i-th row and j-th column of Cell.

[0030] Due to the adoption of the above technical solution, the present invention has the following advantages:

[0031] 1. The genetic algorithm adopted by the present invention improves the encoding and decoding method of genes, and uses a one-dimensional integer array as a group of gene sequences to replace the separate double-layer encoding for multiple drones and multiple targets in the prior art solution, thereby avoiding invalid coding combinations, effectively solving the problems of long computing time and falling into local optimality, and having the advantages of fast and high efficiency.

[0032] 2. The present invention aims to quickly complete all predetermined mission objectives while satisfying the UAV endurance time constraint, and improves and supplements the prior art which only considers minimizing the sum of mission execution time of all UAVs to achieve the economy of saving energy consumption, and effectively solves the task planning and scheduling problem of large-scale multi-UAVs for multi-target situations requiring rapid coverage of all targets.

[0033] 3. The present invention can not only provide a better task planning allocation plan in a shorter time, but also search for a better allocation plan more accurately, which is more desirable in practical applications.

[0034] 4. The present invention is reasonably designed and efficient, and can effectively utilize genetic optimization algorithms and improve algorithm coding methods. Under the constraint of the UAV's endurance, it solves the problem of the fastest time coverage faced by multiple fixed-wing long-flight UAVs for multiple targets, ensures that the task is completed quickly, and can be widely used in the field of UAVs. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0036] Figure 1 This is a flow chart of a method for collaborative task planning and allocation of multiple fixed-wing long-flight UAVs to multiple targets according to an embodiment of the present invention;

[0037] Figure 2The limitations and examples of gene initialization when two fixed-wing long-flight UAVs according to an embodiment of the present invention use the genetic algorithm of the present invention to perform collaborative task planning and allocation for five task targets;

[0038] Figure 3 The process of decoding one gene when two fixed-wing long-flight UAVs use the genetic algorithm of the present invention to perform collaborative task planning and allocation for five task targets in an embodiment of the present invention;

[0039] Figure 4 This is a schematic diagram of the collaborative task planning and allocation results of two fixed-wing long-flight UAVs with shown position distribution for five task targets to complete the task coverage as quickly as possible according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The present invention is further described in conjunction with the accompanying drawings and embodiments, and the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field should fall within the scope of protection of the embodiments of the present invention.

[0041] like Figure 1 As shown, a method for planning and allocating collaborative tasks for multiple fixed-wing long-flight UAVs to multiple targets in this embodiment includes the following steps:

[0042] 1) Obtain the status information of multiple drones and multiple targets; and use this to set initial parameters, objective functions, and initialize the population;

[0043] 2) Using genetic algorithms, based on pre-set initial parameters, according to the acquired drone and target information, with the goal of multiple drone swarms covering all targets as quickly as possible, global decoding and optimization are performed, and the optimal allocation result of this round is selected from the obtained results;

[0044] 3) If the optimal result of this round satisfies that the mission time of each drone is less than the endurance time of the drone, then the optimal result update is performed, otherwise the number of iterations is compared;

[0045] 4) If the optimal result of this round is better than the global optimal result, the global optimal result is updated with the optimal result of this round, otherwise it is not updated and the number of iterations is compared;

[0046] 5) Re-evolve the population, globally decode, and optimize. Repeat 2), 3), 4), and 5) until the preset number of iterations is reached. The optimal result at this time is the optimal allocation result of the task to be planned.

[0047] Specifically, obtain the status information of multiple drones, including the number of drones , the serial number of each drone ( ) and take-off and landing position coordinates ( ), cruising speed , Battery life , and the drones can be considered to take off at approximately the same time.

[0048] Specifically, obtain the status information of multiple targets, including the number of targets , the number of each target ( ) and position coordinates ( ), the time required for each target task to be completed by the drone , where the number of targets should be greater than the number of drones (i.e. ), the target position coordinates remain unchanged during the task execution.

[0049] In the genetic algorithm, the initial parameters include the population size, the number of iterations, the gene length and the objective function, where a group of genes of individuals in the population represents an allocation scheme, and a one-dimensional integer array represents an allocation scheme.

[0050] Specifically, , As an example, the initialization, encoding and decoding methods of the genes of the genetic algorithm are shown; the length is , that is, a one-dimensional integer array of length 5 is used as a gene code to represent a distribution scheme; That is, the information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone.

[0051] like Figure 2 As shown, the initialization rule is, record the position of each gene in a gene coding group as , in this example ; For the front genes, in [1, ] for initialization, that is, the first gene is rounded between [1,6]; the subsequent The genes at positions [1, ] for initialization, that is, the second gene is initialized by rounding the integer between [1,5], the third gene is initialized by rounding the integer between [1,4], the fourth gene is initialized by rounding the integer between [1,3], and the fifth gene is initialized by rounding the integer between [1,2].

[0052] like Figure 3 As shown, the decoding rule is: , For example, for a possible gene encoding [3,2,1,3,2] is used to explain the decoding method; That is, the information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone. For ease of understanding, we start decoding from the subsequent genes, that is, first decode [2,1,3,2] is decoded. The integer array in the gene contains the information of the target selection order. First, the number array of the drone mission target is established in order. [3,4,5,6,7] and decoded array of the decoded drone mission target select Initialize []. Then start decoding from the second gene step by step. 2 means to get the target number array rest The second target is labeled 4 as the target decoding array select The first goal is [4], then rest The target number selected in the first step is deleted, that is, after decoding, , [3,5,6,7]; in the second step 1 means to get the target number array rest The first target is labeled 3 as the target decoding array select The second goal is [3], then rest The target number selected in the second step is deleted, that is, after decoding, , [5,6,7]; in the third step 3 means to get the target number array rest The third target is labeled 7 as the target decoding array select The third goal is [7], then rest The target number selected in the third step is deleted, that is, after decoding, , [5,6]; in the fourth step 2 means to get the target number array rest The second target is labeled 6 as the target decoding array select The fourth goal is [6], then rest The target number selected in the third step is deleted, that is, after decoding, , [5]; at this point, the information contained in the subsequent genes has been completely decoded and automatically rest The number of the last target in the select In the end , indicating the order of selecting the UAV mission targets. Then the first gene is decoded. , indicating the decoding array of the task target select Separate at the third position, select The first two targets are assigned to the drone numbered 1. select The third and subsequent targets (including the third) are assigned to the drone numbered 2, i.e. . Then put the corresponding numbers of the drones into the select In front of each line, you can get the final distribution result , for an element in Cell , represents the number of UAVs participating in task allocation, represents the target number of all participating tasks, where hour, For the number The first The number of the target, the distribution result The order of arrangement in is the execution order of the targets. = , The total number of non-empty elements in is equal to , and the corresponding distribution result is .

[0053] Specifically, the traditional collaborative mission planning aims to minimize the total time taken by multiple drones to complete all goals. The specific goal is to minimize the sum of the flight time of all drones from the time the drones take off to the time they complete all tasks and return to the landing point, so as to achieve the economic benefit of saving energy. UAV takeoff and landing or target location and number UAV takeoff and landing or target location Coordinates calculate the distance between two locations , then when hour, The purpose of the traditional method is to sum up the flight time of all drones The minimum, among which, is the number of columns of the i-th row of data in the Cell, that is, the number of all elements in the i-th row of the Cell; The numbers assigned to the i-th drone are and The distance between two targets is based on the position coordinates of the two targets ( Calculated; For accumulation operation The starting value of Start to End the accumulation operation; is the flight speed of the i-th UAV; k is the cumulative operation The starting value of Start to End the accumulation operation; The number assigned to the i-th drone is The goal requires time to complete; is the element in the i-th row and j-th column of Cell, is the element in the i-th row and j+1-th column in Cell.

[0054] In this embodiment, the solution proposed in this article is different from the traditional method. It aims to cover all targets as quickly as possible with a group of multiple drones. The specific purpose is to minimize the time from the drone taking off to the completion of the last target. The time from the take-off of a UAV to the completion of the last target assigned to the UAV is The purpose of this method is to make ,in is a function that finds the largest element in an array. , the UAV and target positions are distributed as follows Figure 4 In the example described, the global optimal result output by the method in this paper is: ,Right now Figure 4 The order of the mission execution routes of drones No. 1 and No. 2 is shown in the figure.

[0055] In this embodiment, when using the heuristic optimization algorithm for collaborative task planning and allocation, a constraint should be added that the time required for each drone to complete all tasks according to the assigned target set from takeoff to return and land should be less than the drone's endurance time. The time from the takeoff of a drone to the completion of all assigned targets and return to land is , should be less than the flight time of the drone ; That is, to require hour, .

[0056] The present invention is reasonably designed and efficient. When dealing with the problem of allocating multiple targets to multiple fixed-wing long-flight UAVs, it fully considers the requirement for rapid task completion while meeting the UAV endurance performance requirement.

[0057] Finally, 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for collaborative task planning and allocation of multiple fixed-wing long-flight UAVs to multiple targets, characterized in that: include: Step 1: Obtain the status information of multiple drones and multiple targets; The initial parameters and objective function of the genetic algorithm are set with the state information, and the population is initialized with the improved encoding method; Step 2: Genetic algorithm is used to perform global decoding and optimization based on the pre-set initial parameters and the acquired drone and target information, with the goal of covering all targets with multiple drone groups as quickly as possible, to obtain the optimal allocation result for this round; Step 3: If the optimal result of this round satisfies that the mission time of each drone is less than the endurance time of the drone, then the optimal result update is performed; otherwise, it is determined whether the preset number of iterations has been reached; Step 4: If the optimal result of this round is better than the global optimal result, the global optimal result is updated with the optimal result of this round, otherwise it is not updated, and it is determined whether the preset number of iterations is reached; Step 5: Re-evolve the population, globally decode, and optimize, and repeat steps 2 to 5. If the preset number of iterations is reached, the optimal result at this time is the optimal allocation result of the task to be planned; In step 1, the initial parameters include the number of populations, the number of iterations, the gene length and the objective function, wherein a group of genes of individuals in the population represents an allocation scheme, and a one-dimensional integer array represents an allocation scheme; In the improved encoding method, the length is A one-dimensional integer array is used as a gene code to represent a distribution scheme; The information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone; In the improved encoding method, the rule for initializing the genes in the population is: the position of each gene in a gene encoding group is recorded as , ; For the front genes, in [1, ] to round up for initialization; the subsequent The genes at positions [1, ] to round up for initialization; The global decoding and optimization are performed to obtain the optimal allocation result of this round, including: Step 21: For a possible genetic code [u,v,w,x,y,…], the first The information contained in the first gene is the number of targets assigned to each drone, and the information contained in the subsequent genes is the specific number and execution order of the targets assigned to each drone; u, v, w, x, and y are all genes, and their values ​​are all integers; is the number of drones; Step 22: Start decoding from the subsequent genes to obtain the selection order of the UAV mission targets; Step 23: Front The genes are decoded to obtain the mission objectives corresponding to each drone; Step 24: Put the numbers corresponding to the drones in order in front of each row of the drone mission target decoding array, and you can get the allocation result of this decoding gene. Decode all the genes in the population and calculate the objective function. The maximum value of the objective function is the optimal allocation result of this round. The step 22 comprises: set up , for genes [v,w,x,y] is decoded; the integer array in the gene contains the information of the target selection order, first establish the number array of the drone mission target in order [u1,v1,w1,x1,y1], decode the array of the UAV mission target obtained after decoding select Initialize []; then start decoding from gene v step by step, in the first step v means to get the target number array rest The second target is labeled v1 is the target decoding array select The first goal is [v1], then rest The target number selected in the first step is deleted, that is, after decoding, , [u1,w1,x1,y1]; in the second step w represents the target number array rest The first target is labeled u1 is the target decoding array select The second goal is [u1], then rest The target number selected in the second step is deleted, that is, after decoding, , [w1,x1,y1]; in the third step y represents the target number array rest The third target is labeled w1 as the target decoding array select The third goal is [w1], then rest The target number selected in the third step is deleted, that is, after decoding, , [x1,y1]; in the fourth step y represents the target number array rest The second target is labeled y1 is the target decoding array select The fourth goal is [y1], then rest The target number selected in the third step is deleted, that is, after decoding, , [x1]; at this point, the information contained in the subsequent genes has been completely decoded and automatically rest The number of the last target in the select In the end , represents the selection order of UAV mission targets; The step 23 comprises: Decode the first gene. , indicating the decoding array of the task target select In the Separate at each position, select Center front The target is assigned to the drone numbered 1. select Middle The targets after that are assigned to the drone numbered 2. When ; The step 24 comprises: Put the corresponding numbers of the drones into the select In front of each line, you can get the final distribution result , for an element in Cell when hour, For the number The first The number of the target, the distribution result The order of arrangement in is the execution order of the target, and the corresponding allocation result is ;in, = , The total number of non-empty elements in is equal to , , , represents the number of UAVs participating in task allocation, represents the target number of all participating tasks, is the element in the i-th row and the first column of the Cell, is the element in the i-th row and j-th column of Cell.

2. The method according to claim 1, characterized in that In step 1: The status information of multiple drones obtained includes the number of drones , the serial number of each drone , take-off and landing position coordinates ( , Cruising speed and battery life , and the drones can be considered to take off at the same time; ; The status information of multiple targets is obtained, including the number of targets. , the number of each target , location coordinates( , the time required for the drone to complete each target mission , where the target number Greater than the number of drones , the target position coordinates remain unchanged during the task execution; .

3. The method according to claim 1, characterized in that In step 2, the purpose is to cover all targets as quickly as possible with a swarm of drones, specifically, to minimize the time from the drone taking off to the completion of the last target. The time from the take-off of a UAV to the completion of the last target assigned to the UAV is ; make Minimum, the global optimal result of output , that is, the task execution route sequence of each UAV is obtained, the target number assigned to the first UAV and the execution order are , the target number assigned to the second UAV and the execution order are ;in, is the number of columns of the i-th row of data in the Cell, that is, the number of all elements in the i-th row of the Cell; The numbers assigned to the i-th drone are and The distance between two targets is based on the position coordinates of the two targets ( Calculated; For accumulation operation The starting value of Start, to End the accumulation operation; is the flight speed of the i-th UAV; k is the cumulative operation The starting value of Start, to End the accumulation operation; The number assigned to the i-th drone is The goal requires time to complete; is the element in the i-th row and j-th column of Cell, is the element in the i-th row and j+1-th column in Cell.

Images