A distributed dynamic collaborative task allocation method for fixed-wing UAV swarm

Abstract

The present invention discloses a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster. The method comprises the following steps: S1, obtaining the status information of the environment, UAVs, and tasks; S2, constructing a mathematical model for task allocation in combination with the task requirements and the aforementioned status information; S3, each UAV updates the sequence of tasks to be executed according to its own status and task requirements; S4, neighboring UAVs share task information through links and negotiate conflicting tasks; S5, obtaining emergency information; S6, updating the aforementioned status information according to the emergency information and task status; S7, looping through S3-S6 until a consistent task allocation result is obtained, and outputting the final task allocation result. A mathematical model for cluster task allocation is constructed by designing a variety of evaluation indicators and constraints; the mathematical model is solved in combination with interactive information and emergencies; the present invention can reduce cluster communication requirements and improve the robustness of dynamic task allocation results while ensuring the effectiveness of the allocation results.

Description

Technical Field

[0001] The present invention relates to the technical field of fixed-wing unmanned aerial vehicle clusters, and in particular to a method for dynamically coordinating tasks among fixed-wing unmanned aerial vehicle clusters. Background Art

[0002] Unmanned Aerial Vehicle (UAV) has the advantages of being easy to operate, low-cost, and sustainable, and is widely used in various civil and military fields. For example, drones are used for geographic information mapping, agricultural protection, and fire rescue; drones are used for communication relay support and disaster monitoring in natural disaster rescue such as earthquakes and tsunamis.

[0003] Among them, fixed-wing drone clusters have broad application prospects in industry, agriculture, rescue, scientific research and military due to their advantages such as long flight time, long range and ability to perform multiple types of tasks. When the mission scenario is complex and changeable, a single drone cannot effectively guarantee the mission execution effect, and multiple drones are required to perform the mission in coordination. Since the mission requirements may change at any time with the dynamic scenario, it is necessary to adjust the weights of the three evaluation indicators of shortest total flight distance, shortest total time and maximum mission value benefit to obtain the optimal task allocation result; at the same time, the task allocation result also needs to meet the drone performance constraints and related task requirements. The task allocation algorithm needs to quickly plan a reasonable task execution sequence for each drone in a short period of time to meet the rapidity of task allocation in a dynamic environment; in addition, if an emergency occurs, such as drone failure or emergency addition of tasks, the task allocation algorithm needs to respond quickly to minimize the impact of the emergency and meet the robustness requirements of task allocation.

[0004] Therefore, the present invention proposes a distributed dynamic collaborative task allocation method for a fixed-wing UAV cluster. Summary of the invention

[0005] The present invention aims to provide a method for distributed dynamic collaborative task allocation of a fixed-wing UAV cluster, so as to solve the problem of the number of task targets and the task execution order of a cluster of multiple fixed-wing UAVs dynamically collaboratively allocated when there are requirements for economy or rapid and robust task completion.

[0006] The present invention adopts the following technical solution: a method for allocating distributed dynamic collaborative tasks of a fixed-wing UAV cluster, comprising the following steps:

[0007] S1, obtain environmental status information, drone status information, and mission status information;

[0008] S2, combining the above state information to build a mathematical model for task allocation;

[0009] S3, according to the drone status information and task allocation requirements, each drone selectively updates the set of task sequences to be executed;

[0010] S4, neighboring UAVs share mission status information through communication links and negotiate conflicting tasks;

[0011] S5, obtaining emergency information;

[0012] S6, updating the environmental status information, drone status information, and mission status information on each drone according to the current emergency and mission status information;

[0013] S7, execute S3-S6 in a loop until the task allocation results of the cluster drones are consistent, and output the final task allocation result.

[0014] Preferably, the information acquired by the drone in S1 includes status information of the environment, cluster drones and tasks.

[0015] Furthermore, the environmental status information includes geographical information of the mission area and the UAV avoidance area.

[0016] Furthermore, the drone status information includes the current number of drones, the number of each drone, the configuration number, the position coordinates, the cruising speed, the remaining maximum range, the remaining maximum flight time, the health status, and the set of completed mission numbers.

[0017] Furthermore, the task status information includes the number of tasks, the number of each task, the task configuration number, the priority, the task status, the task target, the task location coordinates, the task completion time constraint, and the task value.

[0018] Preferably, constructing the task allocation mathematical model in S2 includes designing evaluation indicators and constraint conditions.

[0019] Furthermore, the evaluation indicators include the total flight range of the swarm UAVs, the mission execution time and the mission value, and each evaluation indicator can be weighted according to the weight parameter.

[0020] Furthermore, the constraints include UAV coordination constraints, UAV remaining range constraints, and UAV maximum mission execution quantity constraints.

[0021] Preferably, updating the set of to-be-executed task sequences in S3 includes constructing an executable task list and task selection.

[0022] Furthermore, the UAV adds the executable task number to the current executable task list according to its own and task status information and the constraints in the mathematical model.

[0023] Furthermore, the UAV adds tasks to the executable task list in order according to the evaluation indicators in the mathematical model and the greedy criterion until no new tasks can be added.

[0024] Preferably, in S4, the UAVs negotiate the conflicting tasks in the set of task sequences to be executed, including shared task information content and negotiation methods.

[0025] Furthermore, the shared task status information includes the winning drone number of each task, the winning bid price, the current status of each task, the arrival time of each task drone, the waiting time of the drone, the task start time, the task execution time, the task end time and the task delay time.

[0026] Furthermore, the content negotiation method is as follows: the neighboring drone receives the task information of other drones, compares it with its own task information, updates its own task information content according to the task allocation consistency rule table, and obtains a task allocation result that meets the consistency requirements.

[0027] Furthermore, the allocation result is the target number and execution order assigned to each UAV.

[0028] Preferably, the emergency event information in S5 includes damage to a certain drone, continuous communication failure, emergency addition or cancellation of a mission.

[0029] Preferably, the elements that need to be updated after an emergency event in S6 include the status information of the environment, the drone, and the task.

[0030] Furthermore, the updated environmental status information includes newly added or cancelled geographic information and avoidance areas;

[0031] Furthermore, the updated drone status information includes the current number of drones, the location coordinates of each drone, the health status, and the set of completed mission numbers;

[0032] Furthermore, the updated task status information includes the current number of tasks, the priority of each task, the task status, the task location coordinates, and the task value.

[0033] Preferably, the task allocation result output in S7 is the target number and execution order assigned to each UAV.

[0034] Compared with the prior art, the present invention has the following beneficial effects: the present invention enriches the evaluation index function and can adjust the optimization target of task allocation in real time according to the actual dynamic task scenario, so that the task allocation result can better adapt to the needs of the dynamic environment; at the same time, the status of each task in the task sequence to be executed is designed, and the impact of key emergencies on the task allocation process is considered, which further improves the robustness and effectiveness of task allocation in the dynamic environment of fixed-wing UAV clusters. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 This is a flow chart of a distributed dynamic collaborative task allocation method for fixed-wing UAV clusters;

[0037] Figure 2 Build a flowchart of executable task lists for each drone;

[0038] Figure 3 Select a flow chart for each drone mission;

[0039] Figure 4 Assigning information to the task receiving drones to negotiate information;

[0040] Figure 5 Schematic diagram of the initial task allocation results of 3 fixed-wing UAV clusters for 5 tasks at time T1;

[0041] Figure 6 This is a schematic diagram of the failure of UAV 02 at time T2 based on time T1;

[0042] Figure 7 Schematic diagram of task allocation results after UAV 02 fails at T3 based on T2;

[0043] Figure 8 Schematic diagram of task allocation results after the urgent addition of task 06 at T4 based on T3. DETAILED DESCRIPTION

[0044] The embodiments of the present invention are described in detail below. It should be understood that these embodiments are used to illustrate the present invention but are not limited to the scope of the present invention.

[0045] Example

[0046] like Figure 1 As shown, a method for allocating distributed dynamic collaborative tasks of a fixed-wing UAV cluster in this embodiment includes the following steps:

[0047] S1, obtain environmental status information, drone status information, and mission status information;

[0048] S2, combining the above state information to build a mathematical model for task allocation;

[0049] S3, according to the drone status information and task allocation requirements, each drone selectively updates the set of task sequences to be executed;

[0050] S4, neighboring UAVs share mission status information through communication links and negotiate conflicting tasks;

[0051] S5, obtaining emergency information;

[0052] S6, updating the environmental status information, drone status information, and mission status information on each drone according to the current emergency and mission status information;

[0053] S7, execute S3-S6 in a loop until the task allocation results of the cluster drones are consistent, and output the current optimal task allocation result.

[0054] Specifically, the information obtained in S1 includes the environment, the task, and the status information of the UAV.

[0055] Furthermore, the environmental status information acquired in S1 includes geographical information of the mission area and the avoidance area. , including the time when the map was last updated , Map location and map categories Drone avoidance areas , including avoidance area locations and avoid time windows .

[0056] Furthermore, the task status information obtained in S1 includes the number of tasks , the number of each task , Task type number , Priority , Task Status 、Mission Objectives , Task location coordinates , Task completion time constraints 、Task value Among them, the task status Elements 0-7 represent the task status: unassigned, assigned but not executed, need to be reallocated, executing, suspended, completed, failed, and other.

[0057] Furthermore, the drone status information in S1 includes the current number of drones , the serial number of each drone , Configuration No. , location coordinates , Cruising speed , Remaining maximum range , Remaining maximum flight time , health status , Completed task number collection .

[0058] Specifically, the mathematical model of task allocation constructed in S2 includes the design of evaluation indicators and constraints.

[0059] Furthermore, the evaluation index includes three types of evaluation index function designs and normalization processing.

[0060] The three types of evaluation index functions include the total flight range of the UAV, the mission execution time and the total mission value.

[0061] Evaluation index of total range of UAV As shown in the following formula, For the moment Drones Execute the task Required flight distance, is the decision variable, and Drones Execute the task Its value is 1, otherwise its value is 0.

[0062]

[0063] Task execution time evaluation index As shown in the following formula, For drones The time required to complete the task sequence, It is a parameter much smaller than 1, which is used to balance the execution time of a single UAV task and the execution time of a cluster of UAV tasks.

[0064]

[0065]

[0066] Evaluation index of total mission value As shown in the following formula, For drones Successfully executed the mission probability.

[0067]

[0068] The normalization of the above three types of evaluation indicators is shown in the following formula:

[0069]

[0070] Task allocation evaluation indicators As shown in the following formula, are the weight values ​​of the three types of evaluation indicators, satisfying {\omega}_{1}+{\omega}_{2}+{\omega}_{3}=1,{\omega}_{k}\in \left [ {0,1} \right ],k=1,2,3 The goal of task allocation optimization is to maximize the total value of the task while minimizing the total flight range and mission time of the UAV.

[0071]

[0072] Furthermore, the constraints include UAV coordination constraints, UAV remaining range constraints, and UAV maximum mission execution quantity constraints.

[0073] The UAV coordination constraints are as follows: ,time At most one drone Perform tasks on it:

[0074]

[0075] The remaining range of the drone is constrained. ,time The required range of the task allocation result cannot exceed the remaining maximum range of the drone:

[0076]

[0077] The maximum number of missions that can be executed by drones is limited. is the maximum number of missions to be executed for the drone ,time The number of task assignments cannot exceed its maximum number of task executions.

[0078]

[0079] Furthermore, the mathematical model of task allocation in S2 is constructed as follows:

[0080]

[0081] Specifically, update the set of pending task sequences in S3 Includes building a list of executable tasks and task selection.

[0082] Furthermore, if Figure 2 As shown in the figure, each drone traverses the list of tasks to be assigned and builds an executable task list based on the constraints in the above mathematical model and the current task requirements. .

[0083] Furthermore, if Figure 3 As shown in the figure, each drone traverses the executable task list, calculates the optimal bid for each task based on the evaluation index, and compares it with the current bids of other drones for the task recorded by itself. If it is better, it will be added to the task sequence set to be executed. , until all executable task lists are traversed.

[0084] Specifically, in S4, the UAVs negotiate the conflicting tasks in the set of task sequences to be executed, including task information content and task negotiation.

[0085] Furthermore, the negotiated information includes the winning drone number of each task in the list of tasks to be performed {\mathbb{Z}}_{i}=\left [ {{\mathbb{Z}}_{i,1,}...,{\mathbb{Z}}_{i,{N}_{c}}} \right ] , Winning bid price{\mathbb{Y}}_{i}=\left [ {{\mathbb{Y}}_{i,1},...,{\mathbb{Y}}_{i,{N}_{c}}} \right ] , task status {\mathbb{S}}_{i}=\left [ {{\mathbb{S}}_{i,1},...,{\mathbb{S}}_{i,{N}_{c}}} \right ] and negotiate information receiving time

[0086] Mission information update modes include "update, reset, no operation" and the update contents of receiving drones in each mode are: , and For example, drones Received drone About the task Distribution information of drones Your own information With drones of Compare and update its own task allocation information through the above three methods.

[0087] Negotiation guidelines such as Figure 4 As shown in the figure, neighboring drones continuously coordinate and update their own task allocation results by sharing task information with each other. Think Task Executed by yourself ( ), and the message receiving drone Think Task You should execute it yourself ( ), resulting in task allocation conflicts. Bid for the task Higher than drones of , then the message receiving drone Need to update its own mission information to drone of: , By analogy, the cluster UAVs share information with each other and continuously coordinate and update the task allocation results, so that the cluster task allocation results tend to be consistent.

[0088] Specifically, the emergency information in S5 includes the start time of the emergency, the event subject, , event impact time window. and environment Related event types are defined as the addition or cancellation of an avoidance zone; Related event types are defined as new drones, damage, communication failure, and navigation failure; The relevant time types are defined as task addition, task cancellation, and task suspension. Each emergency event has an estimated impact time window, including the impact start time. The impact ends .

[0089] Specifically, after the emergency and the preceding tasks are assigned or executed in S6, each UAV needs to update the environment, tasks, and UAV status information.

[0090] Furthermore, the updated content of the environmental status information includes newly added or cancelled avoidance areas. , .

[0091] Furthermore, the task status information update content includes the task quantity , Task Priority , , Task location coordinates 、Task value .

[0092] Furthermore, the drone onboard status update information includes the number of drones , UAV location coordinates , health status , Completed task number collection , the sequence of tasks to be executed is {TaskPath}_{i}=\left [ {{task}_{1},...,{task}_{M}} \right ],{task}_{k}\in \left \{{\varnothing ,1,2,...,{N}_{c}} \right \} ,k=1,2...,M , the arrival time of the drone to be executed sequence task {\mathbb{T}}^{arrive}_{i}=\left [ {{\mathbb{T}}^{arrive}_{i,1},...,{\mathbb{T}}^{arrive}_{i,{N}_{c}}} \right ] , drone waiting time {\mathbb{T}}^{wait}_{i}=\left [ {{\mathbb{T}}^{wait}_{i,1},...,{\mathbb{T}}^{wait}_{i,{N}_{c}}} \right ] , task start time {\mathbb{T}}^{start}_{i}=\left [ {{\mathbb{T}}^{start}_{i,1},...,{\mathbb{T}}^{start}_{i,{N}_{c}}} \right ] , task execution time {\mathbb{T}}^{do}_{i}=\left [ {{\mathbb{T}}^{do}_{i,1},...,{\mathbb{T}}^{do}_{i,{N}_{c}}} \right ] , task end time {\mathbb{T}}^{end}_{i}=\left [ {{\mathbb{T}}^{end}_{i,1},...,{\mathbb{T}}^{end}_{i,{N}_{c}}} \right ] , task delay time {\mathbb{T}}^{delay}_{i}=\left [ {{\mathbb{T}}^{delay}_{i,1},...,{\mathbb{T}}^{delay}_{i,{N}_{c}}} \right ] .

[0093] Specifically, Figure 5-Figure 8As shown, during the task allocation, task execution and emergency process in S7, the above S3-S6 are executed cyclically, and the task allocation results of each UAV are dynamically output according to the current status of each fixed-wing UAV and the task requirements. The task allocation results include the task number and the execution order of each task.

[0094] Further, Figure 5 The initial task allocation results for 3 fixed-wing UAVs to 5 mission targets at time T1 are as follows: UAV 01 performs tasks 02 and 01 successively; UAV 02 performs tasks 03 and 04 successively; UAV 03 performs task 05.

[0095] Further, Figure 6 The following is a schematic diagram of the failure of drone 02 at time T2. While completing mission 03 and heading to mission 04, drone 02 encountered an emergency and failed. At this time, drone 01 had completed mission 02 and was flying to mission 01, and drone 03 was on its way to mission 05. If drones 01 and 03 did not receive the communication heartbeat packet from drone 02 for a period of time, drone 02 was considered to be faulty. The pending mission 04 of drone 02 was reassigned as a pending mission and executed by the remaining drones 01 and 03 in the cluster.

[0096] Further, Figure 7 The diagram is a diagram of the task allocation result after the failure of UAV 02 at time T3. UAVs 01 and 03 calculate the cost of executing task 04 respectively, obtain the local task allocation results, and negotiate conflicts through the data link, and finally determine that UAV 03 will execute task 04. After UAV 03 completes task 05, it bypasses the avoidance zone according to the environmental information to execute task 04.

[0097] Further, Figure 8 This is a diagram of the task allocation result after the urgent addition of task 06 at time T4. Through the negotiation of the cluster drones, the location of the newly added task 06 is closest to drone 01, and the task execution efficiency is optimal. Therefore, drone 01 executes the urgently added task 06.

[0098] The present invention is reasonably designed and efficient. In dealing with the problem of distributed dynamic collaborative task allocation of multiple fixed-wing UAV clusters, it fully takes into account the requirements of rapid task completion and robustness, and at the same time conforms to the dynamic adjustment of energy conservation, task execution efficiency, and task value requirements in a dynamic environment, so that the task allocation process reduces the communication requirements between UAVs as much as possible, and can coordinate according to sudden dynamic situations, making the task allocation results more flexible and effective.

[0099] The above embodiments are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any changes made by adopting the design principles of the present invention and performing non-creative work on this basis should fall within the protection scope of the present invention.

Claims

1. A method for allocating distributed dynamic collaborative tasks of a fixed-wing UAV cluster, characterized in that the steps include: S1, obtain environmental status information, drone status information, and mission status information; S2, combining the above state information to build a mathematical model for task allocation; The mathematical model of task allocation is constructed, including evaluation indicators and constraints. The evaluation indicators include the design of three types of evaluation indicator functions and normalization processing. The three types of evaluation indicator functions include the total range of the UAV, the task execution time and the total value of the task. The evaluation index of the total range of the UAV d i，j (t) is the distance required for UAV i to perform mission j at time t, x i，j (t) is the decision variable. When UAV i executes task j at time t, its value is 1, otherwise its value is 0. Task execution time evaluation index is the time required for UAV i to complete the task sequence, β is a parameter much smaller than 1, which is used to balance the task execution time of a single UAV and the task execution time of a cluster of UAVs. Evaluation index of total mission value p i，j is the probability that UAV i successfully performs task j. The normalization of the three types of evaluation indicators is shown in the following formula: Task allocation evaluation indicators ω1, ω2, and ω3 are the weight values ​​of the three types of evaluation indicators, respectively, satisfying ω1+ω2+ω3=1, ω k ∈[0,1],k=1,2,3; S3, according to the drone status information and task allocation requirements, each drone selectively updates the set of task sequences to be executed; S4, neighboring UAVs share mission status information through communication links and negotiate conflicting tasks; S5, obtaining emergency information; S6, updating the environmental status information, drone status information, and mission status information on each drone according to the current emergency and mission status information; S7, execute S3-S6 in a loop until the task allocation results of the cluster drones are consistent, and output the current optimal task allocation result.

2. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The S1 includes: The environmental status information obtained includes geographical information of the mission area and the UAV avoidance area; The drone status information includes the current number of drones, the number of each drone, the configuration number, the position coordinates, the cruising speed, the remaining maximum range, the remaining maximum flight time, the health status, and the set of completed mission numbers; The task status information includes the number of tasks, the number of each task, the task configuration number, the priority, the task status, the task target, the task location coordinates, the task completion time constraint, and the task value.

3. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The S2 includes: According to the complexity of the dynamic environment, a variety of evaluation indicators are designed, including the shortest total flight range of swarm drones, the shortest mission execution time and the maximum mission value. Each evaluation indicator is normalized and the weight parameters are dynamically adjusted according to the mission characteristics and actual dynamic emergencies. According to the mission execution capabilities and onboard energy consumption of heterogeneous UAVs, combined with the actual mission types and requirements, multiple types of constraints are designed, including cluster UAV coordination constraints, each UAV remaining range constraints, and maximum mission execution quantity constraints; The evaluation indicators and constraints are combined to construct a mathematical model for task allocation.

4. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The S3 includes: Each drone adds the executable task number to the current executable task list based on its own and task status information. The tasks in the list to be executed must meet the constraints in the above mathematical model; According to the greedy criterion, each drone selects tasks in turn according to the optimal principle of evaluation indicators and adds them to the list of tasks to be executed until no new tasks can be added again.

5. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The S4 includes: Each drone shares mission information through a communication link. The shared mission information includes: the winning drone number of each mission, the winning bid price, the current status of each mission, the arrival time of each mission drone, the waiting time of the drone, the mission start time, the mission execution time, the mission end time and the mission delay time; The neighboring drone receives the mission information of other drones and compares it with its own mission information. According to the task allocation consistency rule table, it updates its own task information content to obtain a task allocation result that meets the consistency requirements.

6. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The emergency information includes damage to a drone, persistent communication failure, and emergency addition or cancellation of missions.

7. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The S6 includes: The drones that monitor or detect emergencies broadcast the current emergency information. The drones that receive the message update the onboard environmental status information, drone status information, and mission status information based on the previous mission status information. Updated environmental status information includes newly added or cancelled geographic information and avoidance areas; The updated drone status information includes the current number of drones, the location coordinates of each drone, the health status, and the set of completed mission numbers; The updated task status information includes the current number of tasks, the priority of each task, the task status, the task location coordinates, and the task value.

8. According to claim 1, a method for allocating distributed dynamic collaborative tasks for a fixed-wing UAV cluster is characterized in that: The task allocation result is the task number and task execution order of each UAV.

Images