Multi-unmanned aerial vehicle cooperative task allocation method under communication constraint

Abstract

The invention belongs to the technical field of unmanned aerial vehicle command decision, and relates to a multi-unmanned aerial vehicle cooperative task allocation method under communication constraints. The method specifically comprises the steps of 1, determining initial situation information of an unmanned aerial vehicle and a task target; 2, constructing a task beam by adopting a PI algorithm, and introducing an RPI matrix Z for each unmanned aerial vehicle; 3, calculating an unmanned aerial vehicle communication error rate calculation model, and judging whether the two unmanned aerial vehicles can communicate normally or not; and 4, performing consistency processing on the unmanned aerial vehicles capable of performing normal communication: for the same task, if the number of the unmanned aerial vehicles allocated to a certain task in the matrix Z is greater than the number of the unmanned aerial vehicles required by the task, sequentially rejecting the unmanned aerial vehicle with the maximum RPI value until the number of the allocated unmanned aerial vehicles is equal to the number of the unmanned aerial vehicles required by the task, and if the number of the unmanned aerial vehicles allocated to the task is greater than the number of the unmanned aerial vehicles required by the task. Updating a task beam and an RPI matrix Z of each unmanned aerial vehicle; 5, judging whether the unmanned aerial vehicle execution sequence converges or not, and if yes, generating an executable path; and if not, returning to the step 2.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle command decision-making, and in particular relates to a multi-unmanned aerial vehicle collaborative task allocation method under communication constraints. Background Art

[0002] Drone swarms have been widely used in modern warfare due to their low cost, high efficiency, no casualties, easy and simple operation, wide application range, and strong survivability. How to achieve efficient task allocation in a complex battlefield environment has become a hot issue in the current research on drone swarms. Distributed allocation methods have become a hot topic in the field of drone task allocation because they do not rely on a certain node, do not require global task information, and have strong robustness. PI (Performance Impact) algorithm [1] As one of the most representative distributed task allocation methods, it has good solving efficiency and accuracy.

[0003] The PI algorithm has two shortcomings when solving task allocation: (1) The PI algorithm assumes that the communication between drones is stable and reliable and that drones can achieve global communication. However, in a real combat environment, the communication distance of drones is limited, and there are non-ideal communication factors such as packet loss and bit errors; (2) The PI algorithm can effectively solve single-task problems, that is, each task only requires one drone, but when a task requires two or more drones to execute simultaneously, the PI algorithm does not have the ability to solve such problems. Based on the PI algorithm, this patent adds an actual drone communication model and improves the PI algorithm to solve the problem of multi-drone collaborative task allocation under communication-constrained conditions. Summary of the invention

[0004] In view of this, the present invention provides a multi-UAV collaborative task allocation method under communication constraints, which can achieve multi-target multi-UAV collaborative task allocation.

[0005] The technical solution for implementing the present invention is as follows:

[0006] A multi-UAV collaborative task allocation method under communication constraints, the specific process is as follows:

[0007] Step 1: Determine the initial situation information of the UAV and the mission target;

[0008] Step 2: Use the PI algorithm to construct the task bundle. When constructing the task bundle for the first time, introduce the RPI matrix Z for each drone. The matrix Z for the i-th drone is i , the i-th row represents the RPI value of drone i for each task, and the elements of the remaining rows are 0;

[0009] Step 3: Calculate the UAV communication bit error rate calculation model based on the UAV location information and environmental information to determine whether the two UAVs can communicate normally;

[0010] Step 4: For the drones that can communicate normally, the consistency processing is as follows: for the same task, if the number of drones assigned to a task in Z is greater than the number of drones required for the task, the drones with the largest RPI values ​​are eliminated in turn until the number of assigned drones is equal to the number of drones required for the task, and the task bundles and RPI matrix Z of each drone are updated;

[0011] Step 5: Determine whether the drone execution sequence converges. If so, output the execution sequence of each drone and generate an executable path; if not, return to step 2.

[0012] Furthermore, the UAV communication model of the present invention is:

[0013]

[0014] Among them, P b is the bit error rate; M is a constant, γ dB is the signal-to-noise ratio, P T is the signal gain; K dB is a dimensionless constant; d is the distance between UAVs; d0 is the reference distance of the antenna far field; is the path loss at distance d0, δ is the path fading exponent; represents a normally distributed random number, and ω is the bit error rate adjustment factor.

[0015] Furthermore, the constant M of the present invention has a value range of 3.0-4.2, and the bit error rate adjustment factor ω has a value range of 2.0-4.5.

[0016] Furthermore, a random number is randomly generated in [0,1], and the random number is compared with the bit error rate P b The size of P is used to determine whether the two drones can communicate normally. If the generated value is greater than P b , indicating that the two drones can communicate normally, otherwise, the drones cannot communicate normally.

[0017] Beneficial Effects

[0018] First, the present invention introduces an RPI matrix Z for each UAV, and designs the matrix Z as a two-dimensional matrix, which can retain several RPI values ​​with small size during consistency processing, thereby achieving multi-machine collaboration.

[0019] Second, the present invention introduces a bit error rate communication model and adjusts the bit error rate adjustment factor according to different communication environments, so that the task allocation result is close to the actual application requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 This is a flow chart of a multi-UAV collaborative task allocation method under communication constraints.

[0022] Figure 2 Initial distribution map of drones and missions.

[0023] Figure 3 Assign results and executable paths to tasks. DETAILED DESCRIPTION

[0024] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] It should be noted that the following embodiments and features in the embodiments may be combined with each other in the absence of conflict; and, based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making any creative work are within the scope of protection of the present disclosure.

[0026] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0027] like Figure 1 As shown, the embodiment of the present application is a method for allocating multi-UAV collaborative tasks under communication constraints, and the specific process is as follows:

[0028] Step 1: Given the known initial situation information, including the initial position P of the drone uav , Maximum range D uav , Speed ​​V uav , communication capability, quantity N a , the number of targets N T , initial position P task, the number of drones required is NUM.

[0029] Step 2: Based on the information in step 1, the PI algorithm is used to construct the task bundle. When constructing the task bundle, each drone introduces the RPI matrix Z, and the size of Z is N a ×N T , record the RPI value of each drone for each task, such as the i-th row and j-th column element Z in Z ij Represents the RPI value of drone i for task j. Except for the i-th row, the elements in the other rows are all 0.

[0030] Step 3: Based on the drone location information, environmental information, etc., introduce the drone communication bit error rate calculation model to determine whether the two drones can communicate normally. The drone communication model is:

[0031]

[0032] Among them, P b is the bit error rate; M is a constant, ranging from 3.0 to 4.2; γ dB is the signal-to-noise ratio; P T is the signal gain; K dB is a dimensionless constant; d is the distance between UAVs; d0 is the reference distance of the antenna far field; is the path loss at distance d0, δ is the path fading exponent, and its value range is 2-6; represents a normally distributed random number, μ dB =0,σ dB =[3,9]. ω is the bit error rate adjustment factor, and its value range is generally 2.0-4.5. The value of the bit error rate is closely related to the communication environment of the drone. The worse the communication environment, the larger the ω value.

[0033] A random number is randomly generated in [0,1], and the random number is compared with the bit error rate P b The size of P is used to determine whether the two drones can communicate normally. b , indicating that the two drones can communicate normally, otherwise, the drones cannot communicate normally.

[0034] Step 4: According to the drone communication results calculated in step 3, if the two drones can communicate normally, the consistency processing method between drones is as follows: the drone RPI matrix Z records the RPI values ​​of all drones for each task. For the same task, the smaller the RPI value, the better. Therefore, if the number of drones assigned to a task in Z is greater than the number of drones required for the task, the drones with the largest RPI values ​​are eliminated in turn until the number of assigned drones is equal to the number of drones required for the task, that is, the requirements for the number of drones for the multi-machine collaborative task are met. Finally, the execution sequence of each drone is updated.

[0035] Taking task j as an example, if two drones are required to perform the task collaboratively, the two drones with smaller RPI will perform the task, and the other drones will voluntarily give up the task.

[0036] The consistency processing method in this step is different from the PI algorithm. The PI algorithm only considers tasks that need to be performed by a single drone and can only retain the drone with the smallest RPI. Since the RPI matrix size in the PI algorithm is 1×N T , can only record the RPI value of a single UAV for each task, and cannot solve the problem of multi-machine coordination. This step uses the RPI matrix proposed in step 2 to retain several RPI values ​​with small values, thereby achieving multi-machine coordination.

[0037] Step 5: Determine whether the drone execution sequence in step 4 converges. If so, output the execution sequence of each drone to generate an executable path; if not, jump to step 2.

[0038] In this embodiment, N a Drones perform N T Taking the task objective as an example, the process of the present invention is described in detail:

[0039] Step 1: Given the initial situation information, including the initial position of the drone Maximum range Speed Communication parameters (constant M, signal gain P T , dimensionless constant K dB , reference distance d0 of the antenna far field, path loss at distance d0 Path fading index δ), number N a , the number of targets N T , initial position Number of drones required

[0040] Step 2: Based on the information in step 1, the PI algorithm is used to construct the task bundle and generate the RPI matrix Z. The result of the task bundle construction is the execution task bundle of each drone. RPI matrix for each drone

[0041] Step 3: Bit error rate adjustment factor ω, calculate the UAV communication bit error rate P of the two UAVs based on the information in step 1 b , randomly generate a value in [0,1]. If the generated value is greater than P b , indicating that the drones can communicate normally, otherwise, the drones cannot communicate normally;

[0042] Step 4: If the drones can communicate normally, then for each task, the drone with a smaller RPI value will execute it, and the drone with a larger RPI value will actively give up the task until the number of drones executing the task meets the multi-machine coordination constraint. That is, if the number of drones assigned to a task in Z is greater than the number of drones required for the task, the drone with the largest RPI value will be eliminated in turn until the number of assigned drones is equal to the number of drones required for the task, and p will be updated. i ,i∈[1,N a ],Z i ,i∈[1,N a ].

[0043] Step 5: Determine p in step 4 i ,i∈[1,N a ],Z i ,i∈[1,N a ] Whether it converges. If so, output the execution sequence of each drone to generate an executable path; if not, jump to step 2.

[0044] The PI algorithm includes two processes: task bundle construction and consistency processing. Through the continuous iteration of the above two processes, a certain objective function no longer changes, that is, the algorithm converges. When step 2 is executed for the first time, the task bundle, RPI matrix Z, etc. are all empty sets. After consistency processing (step 4), the task bundle and RPI matrix Z are not empty. On this basis, the task bundle construction and consistency processing continue.

[0045] Take drone 1 as an example:

[0046] When the task execution sequence of UAV 1 is empty, that is, p1 = [], the RPI value of the UAV for task 2 is a1; when the task execution sequence of UAV 1 is not empty, that is, p1 = [task 1], the RPI value of the UAV for task 2 is a2; because a1 is not equal to a2, when returning to step 2, the task bundle, RPI matrix Z, etc. may all be non-empty sets, and the results of executing step 2 twice are different.

[0047] Example:

[0048] This embodiment adopts a multi-UAV collaborative task allocation method under communication constraints provided by the present invention to solve the multi-UAV collaborative task allocation problem under communication constraints. The specific execution steps are as follows:

[0049] Step 1: Given the initial situation information, within a range of 10km×10km, 5 drones collaborate to perform 5 tasks, each task requires 1, 2, 3, 4, and 5 drones to perform. The maximum range of the drones is 20km, the speed of the drones is 100m / s, and the communication parameters are as follows: μ dB =0,σ dB =7M=3,P T =30dBm,K dB =30dB, d = 100m. Randomly generate N a = 5 drones and N T = 5 mission coordinate points, the initial distribution map of drones and missions is as follows Figure 2 shown.

[0050] Step 2: Based on the information in step 1, the PI algorithm is used to construct the mission bundle and generate the RPI matrix Z. The mission bundle and RPI matrix of each drone are as follows:

[0051] UAV 1 mission bundle, RPI matrix:

[0052] p1 = [1-4-2-3-5]

[0053]

[0054] Drone 2 mission bundle, RPI matrix:

[0055] p2 = [1-4-2-3-5]

[0056]

[0057] Drone 3 mission bundle, RPI matrix:

[0058] p3 = [1-4-2-3-5]

[0059]

[0060] Drone 4 mission bundle, RPI matrix:

[0061] p4 = [1-4-2-3-5]

[0062]

[0063] Drone 5 mission bundle, RPI matrix:

[0064] p5 = [2-3-5-1-4]

[0065]

[0066] Step 3: Bit error rate adjustment factor ω = 0.2, calculate the UAV communication bit error rate P of the two UAVs based on the information in step 1 b , the calculation formula is as follows:

[0067]

[0068] Generate a random value in [0,1], if the generated value is greater than P b , indicating that the drones can communicate normally, otherwise, the drones cannot communicate normally. After calculation, drone 3 and drone 5 cannot communicate with each other, and the remaining drones can communicate directly;

[0069] Step 4: According to the communication results calculated in step 3, for each task, the drone with a higher RPI value will actively give up the task, and the redundant drones will be eliminated until the drone multi-machine constraint is met. Taking the consistency processing of drones for task 3 as an example, according to step 1, it can be known that task 3 requires three drones to cooperate in execution. According to the calculation results of step 2, the RPI values ​​of drones 1-5 for task 3 are 203.4262, 211.5654, 220.7260, 203.0717 and 17.8494 respectively. Therefore, drones 2 and 3 give up task 3, and drones 1, 4 and 5 will cooperate in execution. That is, the element values ​​of the 3rd column of row 2 and the 3rd column of row 3 of the RPI matrix are 0, and task 3 is eliminated from p2 and p3. At the same time, the execution task bundle p is updated. i ,i∈[1,N a ] and the RPI matrix Z i ,i∈[1,N a ]. The mission bundle and RPI matrix of UAV 1 are as follows:

[0070] UAV 1 mission bundle, RPI matrix:

[0071] p1 = [4-2-3-5]

[0072]

[0073] Drone 2 mission bundle, RPI matrix:

[0074] p2=[4-5]

[0075]

[0076] Drone 3 mission bundle, RPI matrix:

[0077] p3=[5]

[0078]

[0079] Drone 4 mission bundle, RPI matrix:

[0080] p4 = [1-4-2-3-5]

[0081]

[0082] Drone 5 mission bundle, RPI matrix:

[0083] p5 = [2-3-5-4]

[0084]

[0085] Step 5: Determine p in step 4 i ,i∈[1,N a ],Z i ,i∈[1,N a ] has not converged, jump to step 2. Repeat the above process until the algorithm converges, and output the task allocation results and executable path as follows Figure 3 shown.

[0086] In summary, the above are only embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-UAV collaborative task allocation method under communication constraints, characterized by: The specific process is: Step 1: Determine the initial situation information of the UAV and the mission target; Step 2: Use the PI algorithm to construct the task bundle. When constructing the task bundle for the first time, introduce the RPI matrix Z for each UAV. The matrix Z for the i-th UAV is i , where the i-th row represents the RPI value of drone i for each task, and the elements of the remaining rows are 0; Step 3: Calculate the UAV communication bit error rate calculation model based on the UAV location information and environmental information to determine whether the two UAVs can communicate normally; Step 4: For the drones that can communicate normally, consistency processing is performed as follows: for the same task, if the number of drones assigned to a task in the matrix Z is greater than the number of drones required for the task, the drones with the largest RPI values ​​are eliminated in sequence until the number of assigned drones equals the number of drones required for the task, and the task bundles and RPI matrix Z of each drone are updated; Step 5: Determine whether the drone execution sequence converges. If so, output the execution sequence of each drone and generate an executable path. If not, return to step 2.

2. The multi-UAV collaborative task allocation method under communication constraints according to claim 1 is characterized in that: The specific process is: The drone communication model is: Among them, P b is the bit error rate; M is a constant, γ dB is the signal-to-noise ratio, P T is the signal gain; K dB is a dimensionless constant; d is the distance between UAVs; d0 is the reference distance of the antenna far field; is the path loss at distance d0, δ is the path fading exponent; represents a normally distributed random number, and ω is the bit error rate adjustment factor.

3. The multi-UAV collaborative task allocation method under communication constraints according to claim 2 is characterized in that: The constant M has a value range of 3.0-4.2, and the bit error rate adjustment factor ω has a value range of 2.0-4.

5.

4. The multi-UAV collaborative task allocation method under communication constraints according to claim 2 is characterized in that: Generate a random number in [0,1] and compare the random number with the bit error rate P b The size of P is used to determine whether the two drones can communicate normally. If the generated value is greater than P b , indicating that the two drones can communicate normally, otherwise, the drones cannot communicate normally.

Images