Air-ground cooperative detection task dispatching method

Abstract

The invention discloses an air-ground cooperative detection task order dispatching method, and the method comprises the steps: constructing a detection position and communication relay position alternative set through combining the detection distance of an air-ground heterogeneous unmanned intelligent agent, the communication relay distance and a target position; taking task timeliness and path cost as objective functions, and constructing an unmanned intelligent agent order sending model oriented to the detection communication maneuvering task; aiming at nonlinear characteristics of the model, combining a set coverage algorithm, a Hungary algorithm and a genetic algorithm, designing a three-stage hybrid algorithm consisting of a detection task order dispatching algorithm, a communication task order dispatching algorithm and a maneuvering task order dispatching algorithm, and solving the model. According to the method, joint decision-making of detection / communication relay position selection and detection / communication relay task allocation is realized, and the cooperative detection task execution efficiency is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent planning and scheduling, and in particular to a method for dispatching air-ground collaborative detection tasks. Background Art

[0002] With the breakthrough innovation and rapid development of unmanned technology, the types of unmanned systems are becoming more and more diverse, the scope of mission areas is constantly expanding, and the task allocation methods of unmanned systems are also widely studied, but few scholars consider the allocation of new tasks and the optimization of existing task sequences at the same time. Due to the real-time dynamic nature of task assignment, decision makers cannot obtain all task information in advance and can only make decisions based on current task information. As the situation changes and task information is continuously issued, the quality of existing task allocation plans may decline. Therefore, in a dynamic environment, the coordinated optimization of new task allocation and existing task reallocation is particularly important.

[0003] For the problem of unmanned system detection task allocation, most studies tend to assume that the detection task location is the target point location, and few scholars consider the selection of detection location and the allocation of detection tasks at the same time. However, the selection of detection task location and the allocation of detection tasks are tightly coupled. If a detection location can perform detection tasks for multiple targets, it will greatly save the task path and time of the unmanned system.

[0004] During the task execution process, the detection unmanned intelligent agent may be out of the communication range and unable to send back detection information and receive task instructions. However, most studies often ignore this problem, or only dispatch a communication relay agent after the communication is disconnected, which affects the efficiency of task execution. Summary of the invention

[0005] In response to the above problems, the present invention takes the air-ground collaborative detection unmanned intelligent agent group as the research object, constructs a list of tasks to be assigned with new tasks and existing tasks that have not yet been executed, and constructs an unmanned intelligent agent dispatching model for detection and communication mobile tasks, which realizes the joint decision-making of detection / communication relay location selection and detection / communication relay task allocation, and can be applied to the air-ground collaborative detection unmanned intelligent agent group composed of detection unmanned vehicles, detection drones, communication relay drones, communication relay unmanned vehicles, command vehicles, etc.

[0006] The present invention provides a method for dispatching air-ground collaborative detection tasks. Based on the status and parameter information of each unmanned intelligent agent in the air-ground collaborative detection unmanned intelligent agent group and the detection task information, a detection position / communication relay position alternative set and an unmanned intelligent agent dispatching model for detection and communication maneuvering tasks are constructed, and a three-stage hybrid algorithm is designed to realize the joint decision-making of detection / communication relay position selection and detection / communication relay task allocation.

[0007] in:

[0008] The unmanned intelligent body status and parameter information includes the unmanned intelligent body's current position, maximum detection distance, communication relay distance, average speed, etc.

[0009] The detection task information includes the detection target location and the latest execution time of the task. If the time to execute the task exceeds the latest execution time, the target is likely to be lost, resulting in task failure.

[0010] S1. Constructing a candidate set of detection locations / communication relay locations

[0011] Based on the state and parameter information of the unmanned agent and the detection task information, a detection position candidate set is constructed. For each detection agent, the optional range of detection positions for each target is determined based on its maximum detection distance for different types of targets. Within the optional range of detection positions, the detection candidate positions for each target are determined based on the safety and concealment of the position, whether there are obstructions between the position and the target, and the distance from the current position of the agent and the distance from the command vehicle, while considering whether the detection position can detect multiple targets to save the driving distance of the agent. Based on the latest execution time of each task, the latest access time of each alternative detection position is determined, thereby constructing a detection position candidate set consisting of the alternative detection position and the latest access time corresponding to the position.

[0012] Based on the alternative detection position set, a communication relay position alternative set is constructed. For each alternative detection position in the detection position alternative set, it is determined whether it exceeds the communication range of the command vehicle. If it exceeds the communication range of the command vehicle, when the detection agent performs the detection task at this position, the communication relay agent needs to establish a communication connection between the detection agent and the command vehicle. For each alternative detection position that exceeds the communication range of the command vehicle, the corresponding alternative communication relay position is determined by comprehensively considering the position of the detection agent and the command vehicle, the communication distance of each agent, and the security of the position. At the same time, combined with the latest access time of the alternative detection position, the latest access time of the alternative communication relay position is determined to ensure that before the detection agent arrives at the alternative detection position, the communication relay agent has established a communication connection between the detection agent and the command vehicle, thereby constructing a communication relay position alternative set consisting of the alternative communication relay position and the latest access time corresponding to the position.

[0013] S2. Based on the detection location alternative set and the communication relay location alternative set, an unmanned intelligent agent dispatching model for detection and communication mobile tasks is constructed, a detection / communication relay location is selected for each detection / communication relay agent, and the timing of visiting the task location is determined, so that the time for the agent to perform the task is as close to the task time limit as possible, while minimizing the task execution cost.

[0014] The decision variables of the unmanned intelligent agent dispatching model for detection and communication mobile tasks are the selection of alternative detection locations in the detection location alternative set by each detection intelligent agent, the selection of alternative communication relay locations in the communication relay location alternative set by each communication relay intelligent agent, and the order and time of each detection / communication relay intelligent agent's visit to the detection / communication relay location.

[0015] The unmanned intelligent agent dispatching model for detection and communication maneuvering tasks sets the objective function by comprehensively considering the path cost of executing the task and the penalty for exceeding the latest access time of the task location, so as to ensure that the execution time of all tasks is as close as possible to the task time limit, while minimizing the total mileage of executing the task.

[0016] The constraints of the unmanned agent dispatching model for the detection and communication maneuvering task include: detection location selection constraints, coupling constraints between the detection location and the communication relay location, coupling constraints between the task location selection and the task path, path continuity constraints, coupling constraints between time and path, and agent range constraints. Among them:

[0017] The detection location selection constraint ensures that all detection tasks must be assigned to detection agents;

[0018] The coupling constraint between the detection position and the communication relay position ensures that when the detection agent selects a detection position beyond the communication distance of the command vehicle, a communication relay agent will go to the corresponding communication relay position to establish a communication connection;

[0019] The coupling constraint between task location selection and task path ensures the consistency of two types of decision variables: task location selection and task location visit order.

[0020] The path continuity constraint ensures that the agent visits the task location in a continuous path;

[0021] The intelligent agent range constraint ensures that the path planned for each intelligent agent is within its remaining range. S3, in view of the nonlinear characteristics of the unmanned intelligent agent dispatch model for detection, communication and mobility tasks, adopts the idea of ​​segmented solution to decompose it into detection task dispatch model, communication task dispatch model, and mobility task dispatch model, and designs algorithms to solve them respectively. Among them:

[0022] The detection task dispatching model linearizes the penalty function for exceeding the latest access time of the task location, takes minimizing the matching cost between the detection agent and the detection location as the goal, and ensures that all detection tasks are assigned to the detection agent as the constraint. The minimum weighted set cover model is constructed, and the branch and bound method or greedy algorithm is designed to solve the model and decide the detection agent's selection plan for the detection location.

[0023] The communication task dispatching model linearizes the objective function of the unmanned intelligent agent dispatching model for detection and communication maneuvering tasks, retains the coupling constraints between the detection position and the communication relay position, and builds a bipartite graph matching model based on the solution results of the detection task dispatching model. The Hungarian algorithm is designed to solve it and obtain the communication relay agent's selection plan for the communication relay position.

[0024] The mobile task dispatching model, based on the solution results of the detection task dispatching model and the communication task dispatching model, decomposes the model into a task path planning model for each intelligent agent, and further transforms it into a traveling salesman problem model with a time window. It uses a genetic algorithm to solve it and obtain the order and time of each intelligent agent's visit to each task location.

[0025] Compared with the prior art, the beneficial effects of the present invention are: (1) a list of tasks to be assigned is constructed by combining new tasks with existing tasks that have not yet been executed, and an unmanned intelligent agent dispatching model for detection, communication and mobile tasks is constructed, thereby realizing joint decision-making on detection / communication relay position selection and detection / communication relay task allocation; (2) one detection position can perform detection tasks for multiple targets, which can greatly save the task path and time of the unmanned system; (3) the communication relay intelligent agent is taken into account when allocating tasks, thus avoiding the impact of communication disconnection on task execution efficiency; (4) it has a wide range of applications and can be applied to air-to-ground collaborative detection unmanned intelligent agent groups composed of detection unmanned vehicles, detection drones, communication relay drones, communication relay unmanned vehicles, command vehicles, etc.; (5) model construction and solution methods for dispatching detection tasks, communication tasks and mobile tasks are respectively given. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other objects, features and advantages of the present disclosure will become more apparent through a more detailed description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, wherein like reference numerals generally represent like components throughout the exemplary embodiments of the present disclosure.

[0027] Figure 1 A schematic diagram for constructing a candidate set of detection positions according to an exemplary embodiment of the present disclosure;

[0028] Figure 2 A schematic diagram is constructed for a candidate set of communication relay locations according to an exemplary embodiment of the present disclosure;

[0029] Figure 3 It is an overall flow chart of the present disclosure. DETAILED DESCRIPTION

[0030] The preferred embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the preferred embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0031] The present invention provides a method for dispatching air-ground collaborative detection tasks, which combines the detection distance, communication relay distance and target position of air-ground heterogeneous unmanned intelligent bodies to construct a detection position and communication relay position candidate set; takes meeting the task timeliness and path cost as the objective function, takes the detection position selection constraints, the coupling constraints of the detection position and the communication relay position, the coupling constraints of the task position selection and the task path, the path continuity constraints, and the coupling constraints of the task execution time and the path as the constraints, and constructs an unmanned intelligent body dispatching model for detection, communication and maneuvering tasks; and designs a three-stage hybrid algorithm consisting of a detection task dispatching algorithm, a communication task dispatching algorithm and a maneuvering task dispatching algorithm in combination with a set covering algorithm, a Hungarian algorithm and a genetic algorithm in view of the nonlinear characteristics of the model, so as to achieve efficient solution of the model.

[0032] Flowchart as attached Figure 3 shown.

[0033] Exemplary embodiments:

[0034] Taking the air-ground collaborative detection unmanned intelligent agent group consisting of a command vehicle, a detection unmanned vehicle / machine, and a communication relay unmanned vehicle / machine performing collaborative detection tasks as the application scenario, the application of the proposed air-ground collaborative detection task dispatching method is further explained.

[0035] In this embodiment, the command vehicle receives the detection tasks and intelligence support from the superior, as well as the status information of the unmanned intelligent body in real time, and accordingly dispatches the detection task order to the detection unmanned system within its communication range in a timely manner, and creates a communication relay task order to dispatch it to the communication relay unmanned system. new ={m1,m2,...,m M The position and timeliness of each target are known. The air-ground collaborative detection unmanned intelligent group includes the command vehicle k0, the detection unmanned vehicle / machine K TL 、Communication relay unmanned vehicle / machine K TX .

[0036] The command vehicle k0 remains stationary during the mission execution, and its position is The current positions of the remaining agents k∈K are known, which are o k After completing all tasks, the location to return to is d kThe performance of each agent, such as maximum detection distance, communication relay distance, speed, etc., is known.

[0037] At the current moment, the set of detection tasks that all agents have accepted but not yet executed is M old At this time, not only the newly arrived detection task M new In addition, the detection tasks that have been assigned but not yet executed need to be reallocated according to the real-time situation. The set of detection tasks to be assigned is M ZC =M new ∪M old .

[0038] For each detection task m∈M ZC , according to the support intelligence, the target location and the latest execution time t can be obtained m If the time to execute the task exceeds The target is likely to be lost, causing the mission to fail.

[0039] Step 1: For each detection agent k∈K TL , create a candidate set of detection locations for it

[0040] As attached Figure 1 As shown, two detection agents k1 and k2 need to detect three targets m1, m2, and m3.

[0041] According to the detection distance of each intelligent agent for different types of targets, the optional range of detection positions for each target can be determined, as shown by the hollow circle of the target point in the figure.

[0042] For each target m∈M ZC , in the agent k∈K TL The optional detection position of the target is determined based on the safety and concealment of the position, whether there are any obstructions between the target and the target, the distance from the current position of the agent k, and the distance from the command vehicle. At the same time, the detection position can detect multiple targets to save the driving distance of the agent.

[0043] For example, for target m1, the optional detection positions of agent k1 are: The optional detection positions of agent k2 are: In addition, a detection position may be able to detect multiple targets. Build a detectable target set for it For example,

[0044] Since each task has a latest execution time, a latest access time needs to be set for each candidate detection location. TL Select detection position i and execute detection task Then the agent needs to Arrive at the detection position before the latest execution time of each task in, that is, the latest access time of detection position i is

[0045] For example, Figure 1 In the above example, if the detection agent k1 chooses to detect position 2, the latest visit time of this position is

[0046] Step 2: Create a communication relay location candidate set

[0047] If the alternative detection position exceeds the communication distance of the command vehicle When the detection agent performs the detection task at this location, the communication relay agent is required to establish a communication connection between the detection agent and the command vehicle. The alternative communication relay location is determined by comprehensively considering the location of the detection agent and the command vehicle, the communication distance of each agent, and the safety of the location. For the convenience of description, this type of communication relay task is called the first type of communication relay task, and this type of candidate communication relay position is called the first type of candidate communication relay position. Figure 2 In the figure, position 3 is the alternative detection position of the detection agent k1, which is outside the communication distance of the command vehicle, so an alternative communication relay position needs to be set.

[0048] For all alternative detection positions beyond the command vehicle's communication range Constructing alternative communication relay locations For example, in the attached Figure 2 The set of detection locations beyond the communication distance is The first type of candidate communication relay location is

[0049] Since the access time to the detection location is constrained by the latest access time, the access time to its associated communication relay location is also constrained, i.e. For example, Figure 2 In the example, the candidate detection position 1 and its associated candidate communication relay position 3 are consistent, that is, tl1=tl3.

[0050] Step 3: Build an unmanned agent dispatch model for detection, communication and mobility

[0051] Based on the above-mentioned alternative task locations and access time constraints, the task dispatch problem is transformed into: selecting a detection / communication relay location for each detection / communication relay agent and determining the sequence of visiting the task locations so that the agent can perform tasks as much as possible within the task time limit while minimizing the task execution cost. Thus, an unmanned agent dispatch model for detection and communication maneuvers is constructed.

[0052] 1) Decision variables

[0053] The decision variables of the unmanned agent dispatching model for the detection and communication mobile task are: the selection of the candidate detection position in the detection position candidate set by each detection agent, the selection of the candidate communication relay position in the communication relay position candidate set by each communication relay agent, and the order and time of each detection / communication relay agent visiting the detection / communication relay position, as shown in equations (1)-(3). Indicates whether agent k chooses to visit position i. If so, otherwise, For detecting unmanned agent k∈K TL , At the same time, it indicates whether to select the detection task For the communication relay unmanned agent k∈K TX , Also indicates whether to select the communication relay task represents the time when agent k visits location i. Indicates whether agent k chooses path (i, j). If so, otherwise,

[0054]

[0055] 2) Objective Function

[0056] The unmanned agent dispatching model for detection and communication maneuvering tasks comprehensively considers the path cost of executing the task and the penalty for exceeding the latest access time of the task location to set the objective function as shown in formula (5). The objective function consists of two parts. The first part is the penalty for violating the time constraint to ensure that the execution time of all tasks is as close to the task time limit as possible, where is the penalty value for agent k visiting location i beyond the latest visit time, as shown in formula (4). The second part is the total path cost of executing the task. In formula (5), μ time , μ length are the weight coefficients of the two parts of the objective function respectively.

[0057]

[0058] 3) Constraints

[0059] The constraints of the unmanned intelligent agent dispatching model for detection and communication maneuvering tasks include: detection location selection constraints, coupling constraints between the detection location and the communication relay location, coupling constraints between task location selection and task path, path continuity constraints, coupling constraints between time and path, and intelligent agent range constraints.

[0060] The detection location selection constraint is shown in Equation (6), which means that all detection tasks must be assigned to detection agents, that is, for each detection task m∈M ZC , an alternative detection position that can detect the target Selected by at least one detection agent.

[0061]

[0062] The coupling constraint between the detection position and the communication relay position is shown in (7), which means that if the detection unmanned system k∈K TL Select a detection position beyond the communication distance of the command vehicle, and a communication relay is required The drones head to their corresponding communication relay locations To ensure that the detection information can be returned in time.

[0063]

[0064] The coupling constraints between task location selection and task path are shown in equations (8)-(9). If the candidate detection / communication relay location is not selected by the agent, The path passing through this location should not be selected.

[0065]

[0066] The path continuity constraints are shown in equations (10)-(12), which constrain each agent to start from the initial position (i.e., equation (10)), pass through the task position (i.e., equation (11)), and finally return to the final position (i.e., equation (12)).

[0067]

[0068] The coupling constraints of time and path are shown in equations (13)-(15). Equation (13) indicates that each agent k∈K TL ∪K TX From their free time Start; Formula (14) indicates that if agent k chooses path (i, j), the time to visit location i, j must satisfy (meaning departure time + path time ≤ latest arrival time, M in the formula represents a maximum value); Formula (15) means that the agent completes all tasks and finally returns to dk The time is within the preset time T.

[0069]

[0070]

[0071] The agent range constraint is shown in formula (16). For each agent k∈K TL ∪K TX , the total mileage of the planned path is the remaining range of the agent Within.

[0072]

[0073] Step 4: Solve the model in sections

[0074] Since the unmanned intelligent agent dispatching model for detection, communication and mobility tasks is a nonlinear programming model and difficult to solve in polynomial time, the idea of ​​piecewise solution is adopted to decompose it into detection task dispatching model, communication task dispatching model and mobility task dispatching model, and algorithms are designed to solve them respectively.

[0075] 1) Detection task dispatch model

[0076] The penalty function for exceeding the latest access time of the task location is linearized, and the penalty function in formula (5) is Approximately a constant As shown in formula (17). The model is further transformed into a minimum weighted set coverage model, where the set of selected detection positions must cover all target points and minimize the total cost of the selected set, as shown in formulas (18)-(19).

[0077] When the number of detection tasks is small, the branch and bound method is used to obtain the exact solution; when the number of detection tasks is large, the greedy algorithm is used to solve it.

[0078]

[0079]

[0080] 2) Communication task dispatch model

[0081] The detection task dispatching model will Approximately a constant To linearize the objective function, the constraint condition (16) is approximated as equation (23), so that each communication relay agent will not have insufficient range due to too many tasks. Thus, the communication task dispatching model is constructed as shown in equations (20)-(23).

[0082] According to formula (21), Corresponding Remove from the model, that is, if the detection agent does not choose the communication position i, there is no need to consider the communication relay task corresponding to this position According to formula (23), we can construct At the same time, a virtual communication relay task is constructed to make the number of communication agents consistent with the communication relay task. Then, the model can be further transformed into a bipartite graph matching model, which can be solved by the Hungarian algorithm.

[0083]

[0084] 3) Mobile task dispatch model

[0085] In the detection task dispatching model and algorithm, communication task dispatching model and algorithm, Based on this, the tasks of each agent can be decoupled, that is, the model can be decomposed into a task path planning model for each agent, which can support parallel computing of multiple models. As a result, unselected detection locations and communication relay locations can be eliminated to satisfy constraints (8)-(9).

[0086] In addition, for each agent k∈K TL ∪K TX , since there is no need to consider other agents, the variable Rewrite as y i,j ,t i , then its mobile task dispatching model is as shown in equations (24)-(31). The path planning model of each agent can be regarded as a traveling salesman problem model with a time window, which can be solved by genetic algorithm.

[0087]

[0088] The above technical scheme is only an exemplary embodiment of the present invention. For those skilled in the art, it is easy to make various types of improvements or modifications based on the application methods and principles disclosed in the present invention, and it is not limited to the method described in the above specific embodiment of the present invention. Therefore, the method described above is only preferred and does not have a restrictive meaning.

Claims

1. A method for dispatching an air-ground collaborative detection task, comprising the following steps: S1, based on the status and parameter information of each unmanned intelligent agent in the air-ground collaborative detection unmanned intelligent agent group and the detection task information, build a candidate set of detection positions and communication relay positions; S2, based on the detection position candidate set and the communication relay position candidate set, construct an unmanned agent dispatching model for the detection and communication mobile task, which is used to select a detection / communication relay position for each detection / communication relay agent and determine the timing of visiting the task position, so that the time for the agent to perform the task is as much as possible within the task time limit, while minimizing the task execution cost; S3, decomposing the unmanned intelligent agent dispatch model into: a detection task dispatch model, a communication task dispatch model, and a mobile task dispatch model, and using corresponding methods to solve them in sections.

2. The method according to claim 1, characterized in that In step S1: The unmanned intelligent body state and parameter information include: the unmanned intelligent body's current position, maximum detection distance, communication relay distance, and average speed; The detection task information includes the detection target location and the latest execution time of the task; The specific method of step S1 includes: S11, based on the state and parameter information of the unmanned intelligent agent and the detection task information, construct a detection location candidate set consisting of a detection candidate location of each target and a latest access time corresponding to the location; S12, based on the detection position candidate set, construct a communication relay position candidate set consisting of a candidate communication relay position for each candidate detection position and a latest access time corresponding to the position.

3. The method according to claim 2, characterized in that In step S1, the method for establishing the detection position candidate set includes: For each detection agent, the optional range of detection positions for each target is determined based on its maximum detection distance for different types of targets; Within the optional range of detection positions, determine the detection candidate position for each target based on the safety and concealment of the position, whether there are any obstructions between the position and the target, and the distance from the current position of the agent and the distance from the command vehicle, while considering whether the detection position can detect multiple targets to save the driving distance of the agent; Based on the latest execution time of each task, the latest access time of each candidate detection position is determined, so as to construct a detection position candidate set consisting of the candidate detection positions and the latest access time corresponding to the positions.

4. The method according to claim 2, characterized in that: In step S1, the method for establishing the communication relay location candidate set includes: For each candidate detection position in the detection position candidate set, determine whether it is beyond the communication range of the command vehicle. If it is beyond the communication range of the command vehicle, when the detection agent performs the detection task at this position, the communication relay agent needs to establish a communication connection between the detection agent and the command vehicle; For each candidate detection position that is beyond the communication range of the command vehicle, determine the corresponding candidate communication relay position based on the position of the detection agent and the command vehicle, the communication distance of each agent, and the safety of the position; At the same time, according to the latest access time of the candidate detection position, the latest access time of the candidate communication relay position is determined to ensure that before the detection agent arrives at the candidate detection position, the communication relay agent has established a communication connection between the detection agent and the command vehicle; In this way, a communication relay location candidate set consisting of the candidate communication relay location and the latest access time corresponding to the location is constructed.

5. The method according to claim 1, characterized in that: The unmanned agent dispatching model for the detection communication maneuvering task constructed in step S2 includes: Decision variables include: the selection of candidate detection positions in the detection position candidate set by each detection agent, the selection of candidate communication relay positions in the communication relay position candidate set by each communication relay agent, and the order and time of each detection / communication relay agent visiting the detection / communication relay position; Objective function: The objective function is set based on the path cost of executing the task and the penalty for exceeding the latest access time of the task location to ensure that the execution time of all tasks is within the task time limit as much as possible, while minimizing the total mileage of executing the task; Constraints include: detection location selection constraints, coupling constraints between detection location and communication relay location, coupling constraints between task location selection and task path, path continuity constraints, coupling constraints between task execution time and path, and intelligent agent range constraints, among which: The detection location selection constraint ensures that all detection tasks must be assigned to detection agents; The coupling constraint between the detection position and the communication relay position ensures that when the detection agent selects a detection position beyond the communication distance of the command vehicle, a communication relay agent will go to the corresponding communication relay position to establish a communication connection; The coupling constraint between task location selection and task path ensures the consistency of two types of decision variables: task location selection and task location visit order. The path continuity constraint ensures that the agent visits the task location in a continuous path; The coupling constraint between task execution time and path ensures the consistency between task path and task time; The agent range constraint ensures that the path planned for each agent is within its remaining range.

6. The method according to claim 5, characterized in that The constraints of the unmanned agent dispatch model in step S2 specifically include: Detection location selection constraint: For each detection task, the candidate detection location that can detect the target must be selected by at least one detection agent; Coupling constraints between the detection position and the communication relay position: If the detection unmanned system selects a detection position that is beyond the communication distance of the command vehicle, a communication relay UAV will go to its corresponding communication relay position; Coupling constraints between task location selection and task path: If the alternative detection / communication relay location is not selected by the agent, the path passing through this location should not be selected either; Path continuity constraint: Each agent starts from the initial position, passes through the task position, and finally returns to the final position; The coupling constraints of task execution time and path are: departure time + path time ≤ latest arrival time, and the time when the agent completes all tasks and finally returns to the final destination is within the preset time; Agent range constraint: For each agent, the total mileage of the path planned for it is within the remaining range of the agent.

7. The method according to claim 5 or 6, characterized in that: The construction and solution method of the detection task dispatching model in step S3 specifically includes: The penalty function for exceeding the latest access time of the task location is linearized, with the goal of minimizing the matching cost between the detection agent and the detection location, and with the constraint of ensuring that all detection tasks are assigned to detection agents, a minimum weighted set cover model is constructed to decide the selection plan of the detection agent for the detection location; The model is solved using the branch and bound method or the greedy algorithm.

8. The method according to claim 7, characterized in that The construction and solution method of the communication task dispatching model in step S3 specifically includes: The objective function of the unmanned agent dispatch model for detection and communication maneuvering tasks is linearized, and the coupling constraints between the detection position and the communication relay position are retained. Based on the solution of the detection task dispatch model, a bipartite graph matching model is constructed to obtain the selection plan of the communication relay agent for the communication relay position. The Hungarian algorithm is used to solve the problem.

9. The method according to claim 8, characterized in that: The construction and solution method of the mobile task dispatching model in step S3 specifically includes: Based on the results of the detection task dispatch model and the communication task dispatch model, the unmanned agent dispatch model is decomposed into a task path planning model for each agent, and further transformed into a traveling salesman problem model with a time window to obtain the order and time of each agent's visit to each task location; Genetic algorithm is used to solve the problem.

Images