CN114860416A - Distributed multi-agent detection task assignment method and device in confrontation scene - Google Patents

Abstract

The application discloses a distributed multi-agent detection task allocation method and device in a confrontation scene, wherein the method comprises the following steps: establishing a multi-agent detection efficiency function according to detection task distribution weight, target detector configuration and detection capacity carried by the multi-agent system, target disguise strategy and multi-agent detection task distribution strategy; establishing a distributed multi-agent detection task distribution model according to preset multi-agent detection capability constraint, target disguising capability constraint and detection task constraint; and alternately solving a task allocation strategy and a corresponding target disguise strategy of the multi-agent system according to the gradient information to generate a distributed multi-agent detection task allocation result. Therefore, the technical problem that the execution result of the detection task is easily influenced by the target disguise strategy due to the fact that the potential disguise countermeasure process of the target cannot be fully considered through a distributed detection task allocation method in the related art is solved, and therefore the acquisition amount of target characteristic information is reduced.

Description

Distributed multi-agent detection task assignment method and device in confrontation scene

technical field

The present application relates to the technical field of intelligent decision-making in multi-agent systems, and in particular, to a method and device for allocating distributed multi-agent detection tasks in confrontation scenarios.

Background technique

Task allocation is a key issue for multi-agents in collaborative target detection. Good task allocation can give full play to the capabilities and advantages of different agents. Especially in confrontation scenarios, the target can disguise its own feature information and cause some detection methods. Therefore, according to the type and capability of the detector carried by the agent, correctly matching the agent set and the detection task set is an important basis for obtaining more target feature information.

Task allocation methods can be divided into two types: centralized and distributed according to the organization of the multi-agent system. In the centralized task assignment method, a central node collects the information of all agents, calculates and sends the task assignment results to all agents. The central node can be an agent or a central base station in a multi-agent system. In the distributed task assignment method, each agent can make autonomous decisions and dynamically adjust the task assignment results according to its own state information and communication results. Compared with the centralized task assignment method, it has higher flexibility and stronger performance. Robustness, less affected by agent size.

However, the distributed detection task assignment method in the related art has not fully considered the potential camouflage confrontation process of the target, and the execution result of the detection task is easily affected by the target camouflage strategy, which needs to be improved.

SUMMARY OF THE INVENTION

The present application provides a distributed multi-agent detection task assignment method and device in a confrontation scenario, so as to solve the problem that the distributed detection task assignment method in the related art fails to fully consider the potential camouflage confrontation process of the target, resulting in the detection task The execution result is easily affected by the target camouflage strategy, thus reducing the technical problem of the acquisition of target feature information.

The embodiment of the first aspect of the present application provides a distributed multi-agent detection task assignment method in a confrontation scenario, including the following steps: assigning weights according to detection tasks, target detector configurations and detection capabilities carried by a multi-agent system, and target detection capabilities. The camouflage strategy and the multi-agent detection task assignment strategy establish the multi-agent detection efficiency function; according to the preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints, the distributed multi-agent detection task assignment model is established; The distributed multi-agent detection task assignment model alternately solves the task assignment strategy of the multi-agent system and the corresponding target camouflage strategy according to the gradient information, and generates a distributed multi-agent detection task assignment result.

Optionally, in an embodiment of the present application, before solving the task allocation strategy and the corresponding target camouflage strategy of the multi-agent system, the method further includes: setting optimization parameters and termination conditions; establishing an augmented pull-out of detection efficiency Grange function, and increase the non-convergence penalty term of the detection efficiency function.

Optionally, in an embodiment of the present application, the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors.

Optionally, in an embodiment of the present application, the generating a distributed multi-agent detection task assignment result includes: using the multi-agent task assignment model to find the maximum value of the detection efficiency function under all camouflage strategies.

Optionally, in an embodiment of the present application, the alternatively solving the task allocation strategy and the corresponding target camouflage strategy of the multi-agent system according to the gradient information includes: solving the optimal target camouflage strategy under the current detection task allocation strategy. ; Solve the optimal task allocation strategy of each agent under the current optimal target camouflage strategy in parallel to solve the dual parameters under the current detection task allocation strategy and the optimal target camouflage strategy in parallel; Based on the dual parameters , solve the residual of the original problem and the residual of the dual problem until the optimization ends, so that the detection task assignment solution of the multi-agent system is the average value of the local detection task assignment solutions of all agents.

The embodiment of the second aspect of the present application provides a distributed multi-agent detection task assignment device in a confrontation scenario, including: a first function establishment module for assigning weights according to detection tasks, and the configuration of target detectors carried by the multi-agent system and detection ability, target camouflage strategy, and multi-agent detection task assignment strategy to establish multi-agent detection efficiency function; model building module is used for the preset multi-agent detection ability constraints, target camouflage ability constraints, detection task constraints A distributed multi-agent detection task assignment model is established; the assignment module is used to alternately solve the task assignment strategy and the corresponding target camouflage strategy of the multi-agent system based on the distributed multi-agent detection task assignment model according to the gradient information, and generate Distributed multi-agent detection task assignment results.

Optionally, in an embodiment of the present application, it further includes: a setting module for setting optimization parameters and termination conditions; a second function establishing module for establishing an augmented Lagrangian function of detection efficiency, and adding A non-convergence penalty term for the detection efficiency function.

Optionally, in an embodiment of the present application, the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors.

Optionally, in an embodiment of the present application, the allocation module includes: a first solving unit, configured to use the multi-agent task allocation model to solve the maximum value of the detection efficiency function under all camouflage strategies.

Optionally, in an embodiment of the present application, the allocation module further includes: a second solving unit for solving the optimal target camouflage strategy under the current detection task allocation strategy; a third solving unit for solving in parallel the optimal task allocation strategy of each agent under the current optimal target camouflage strategy to solve the current detection task allocation strategy and the dual parameters under the optimal target camouflage strategy in parallel; The dual parameter is used to solve the residual of the original problem and the residual of the dual problem until the optimization ends, so that the detection task assignment solution of the multi-agent system is the average value of the local detection task assignment solutions of all agents.

An embodiment of a third aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to achieve The distributed multi-agent detection task assignment method in the confrontation scene as described in the above embodiment.

Embodiments of the fourth aspect of the present application provide a computer-readable storage medium, where the computer-readable storage medium stores computer instructions, where the computer instructions are used to cause the computer to execute the distribution in the confrontation scenario described in the foregoing embodiments A multi-agent detection task assignment method.

The embodiment of the present application can use the multi-agent detection efficiency function and distributed multi-agent according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the camouflage strategy of the target, and the multi-agent detection task allocation strategy. The task allocation model of body detection is used to achieve a reasonable allocation of multi-agents with different detection capabilities, thereby minimizing the impact of target camouflage on the detection performance of multi-agents, and then maximizing the collection of target feature information in the confrontation environment. As a result, the distributed detection task assignment method in the related art is solved, because the potential camouflage confrontation process of the target is not fully considered, the execution result of the detection task is easily affected by the target camouflage strategy, thereby reducing the acquisition of target feature information. amount of technical issues.

Additional aspects and advantages of the present application will be set forth, in part, in the following description, and in part will be apparent from the following description, or learned by practice of the present application.

Description of drawings

The above and/or additional aspects and advantages of the present application will become apparent and readily understood from the following description of embodiments taken in conjunction with the accompanying drawings, wherein:

FIG. 1 is a flowchart of a distributed multi-agent detection task assignment method in a confrontation scenario provided according to an embodiment of the present application;

2 is a flowchart of a distributed multi-agent detection task assignment method in a confrontation scenario according to an embodiment of the present application;

3 is a schematic structural diagram of a distributed multi-agent detection task assignment device in a confrontation scenario provided according to an embodiment of the present application;

FIG. 4 is a schematic structural diagram of an electronic device provided according to an embodiment of the present application.

Detailed ways

The following describes in detail the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary, and are intended to be used to explain the present application, but should not be construed as a limitation to the present application.

The method and apparatus for allocating distributed multi-agent detection tasks in a confrontation scenario according to embodiments of the present application will be described below with reference to the accompanying drawings. In view of the distributed detection task allocation method in the related art mentioned in the above background technology center, because the potential camouflage confrontation process of the target is not fully considered, the execution result of the detection task is easily affected by the target camouflage strategy, thereby reducing the target characteristics. The technical problem of the acquisition amount of information, the present application provides a distributed multi-agent detection task assignment method in confrontation scenarios, in which the weight can be assigned according to the detection task, and the target detector configuration carried by the multi-agent system can be configured. and detection ability, target camouflage strategy, multi-agent detection task allocation strategy, and use multi-agent detection efficiency function and distributed multi-agent detection task allocation model to achieve a reasonable allocation of multi-agents with different detection capabilities, so as to maximize the Minimize the impact of target camouflage on the multi-agent detection performance, and then maximize the collection of target feature information in the confrontation environment. As a result, the distributed detection task assignment method in the related art is solved, because the potential camouflage confrontation process of the target is not fully considered, the execution result of the detection task is easily affected by the target camouflage strategy, thereby reducing the acquisition of target feature information. amount of technical issues.

Specifically, FIG. 1 is a schematic flowchart of a distributed multi-agent detection task assignment method in a confrontation scenario provided by an embodiment of the present application.

As shown in Figure 1, the distributed multi-agent detection task assignment method in the confrontation scene includes the following steps:

In step S101, a multi-agent detection efficiency function is established according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the target camouflage strategy, and the multi-agent detection task allocation strategy.

In the actual execution process, the embodiment of the present application can establish the multi-agent detection efficiency according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the target camouflage strategy, and the multi-agent detection task allocation strategy function, it is convenient to combine the distributed multi-agent detection task allocation model to achieve a reasonable allocation of multi-agents with different detection capabilities, so as to minimize the impact of target camouflage on the detection performance of multi-agents, and thus maximize the countermeasures in the environment. Collection of target feature information.

Among them, the multi-agent system is a distributed topology, and only the detection task assignment results of the communication neighborhood agents can be obtained.

Optionally, in an embodiment of the present application, the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors.

Specifically, the detection efficiency function of the multi-agent detection task assignment established in the embodiment of the present application may be:

表示智能体与目标之间的分配权重参数；C∈{0,1}^S×N表示智能体系统携带的目标探测器类型，[C]_sn＝1表示智能体n带有传感器s；D∈[0,1]^S×M表示目标伪装策略，[D]_sm表示目标m伪装自身可被传感器s探测的特征信息。Among them, X _i ∈ [0,1] ^M×N represents the estimation of the global task assignment result of the multi-agent system by the agent i, and [X _i ] _mn represents the probability that the target m is assigned to the agent n;

Represents the distribution weight parameter between the agent and the target; C∈{0,1} ^S×N indicates the type of target detector carried by the agent system, [C] _sn = 1 indicates that the agent n has a sensor s; D∈ [0,1] ^S×M represents the target camouflage strategy, and [D] _sm represents the feature information that the target m camouflages itself can be detected by the sensor s.

In step S102, a distributed multi-agent detection task assignment model is established according to preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints.

Further, the embodiment of the present application can establish a distributed multi-agent detection task assignment model in combination with preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints.

Among them, the multi-agent detection capability constraint means that each agent can perform at most one detection task at the same time; the target camouflage capability constraint means that the target can camouflage at most one type of feature information at the same time; the detection task constraint means that all targets are at least Assigned to one agent; the multi-agent consistency constraint means that the global task assignment solution of each agent is the same.

并要求每个智能体至多可执行一个探测任务；Specifically, the embodiment of the present application can establish a multi-agent detection capability constraint condition 0≤X _i ≤1,

And requires each agent to perform at most one detection task;

并要求目标至多伪装自身一类特征信息；Establish the target's camouflage ability constraint 0≤D≤1,

And require the target to disguise at most one type of characteristic information of itself;

并要求所有目标均至少被分配至一个智能体。Establishing Probing Mission Constraints

and requires all targets to be assigned to at least one agent.

并要求所有智能体对多智能体系统的探测任务分配的全局估计需保持一致。Further, the embodiment of the present application can estimate the detection task assignment results of each agent to the agent in the non-communication neighborhood, and combine the detection task assignment results of itself and the communication neighborhood to form a global estimation of the detection task of the multi-agent system. , while increasing the consistency constraints of the multi-agent system

And it is required that the global estimation of the detection task assignment of all agents to the multi-agent system should be consistent.

In step S103, based on the distributed multi-agent detection task assignment model, the task assignment strategy and the corresponding target camouflage strategy of the multi-agent system are alternately solved according to the gradient information, and the distributed multi-agent detection task assignment result is generated.

In the actual execution process, the embodiments of the present application may, based on the distributed multi-agent detection task allocation model, allocate the multi-agent detection task according to the optimal solution, and alternately solve the task allocation strategy and the corresponding multi-agent system according to the gradient information. Target camouflage strategy, so as to generate distributed multi-agent detection task assignment results, to achieve a reasonable allocation of multi-agents with different detection capabilities, so as to minimize the impact of target camouflage on multi-agent detection performance, and maximize the confrontation environment. collection of target feature information.

Optionally, in an embodiment of the present application, before solving the task assignment strategy of the multi-agent system and the corresponding target camouflage strategy, the method further includes: setting optimization parameters and termination conditions; establishing an augmented Lagrang of detection efficiency day function, and increase the non-convergence penalty term of the detection efficiency function.

As a possible way of implementation, the embodiment of the present application can set optimization parameters and termination conditions, establish an augmented Lagrangian function of detection efficiency, thereby increasing the non-convergence penalty term of the detection efficiency function, and alternately solve multiple problems according to gradient information. The task allocation strategy of the agent system and the corresponding target camouflage strategy until the termination conditions are met.

该惩罚项对非{0,1}的分配结果进行惩罚，从而保证分布式任务分配结果收敛至整数集合{0,1}。Specifically, in this embodiment of the present application, a non-convergence penalty term may be added

The penalty term penalizes the assignment results other than {0,1}, so as to ensure that the distributed task assignment results converge to the integer set {0,1}.

Further, the augmented Lagrangian function of the multi-agent detection efficiency function can be:

Among them, g(X _i ) is the indicator function of the target allocation strategy Xi _, and h(D) is the indicator function of the camouflage strategy:

Optionally, in an embodiment of the present application, generating a distributed multi-agent detection task assignment result includes: using a multi-agent task assignment model to find the maximum value of the detection efficiency function under all camouflage strategies.

Further, using the multi-agent task assignment model to solve the maximum value of the detection efficiency function under all camouflage strategies can be:

0≤D≤1,

Optionally, in an embodiment of the present application, alternately solving the task allocation strategy and the corresponding target camouflage strategy of the multi-agent system according to the gradient information, including: solving the optimal target camouflage strategy under the current detection task allocation strategy; parallel Solve the optimal task allocation strategy of each agent under the current optimal target camouflage strategy, and solve the dual parameters of the current detection task allocation strategy and the optimal target camouflage strategy in parallel; based on the dual parameters, solve the residual of the original problem and the residual of the dual problem difference until the end of the optimization, so that the detection task assignment solution of the multi-agent system is the average value of the local detection task assignment solution of all agents.

其中V＝{1,…,N}为多智能体集合，

为多智能体通信边，初始化探测任务分配策略{X_i}⁰，初始化目标伪装策略D⁰，设置优化终止条件∈_pri、∈_dual。Specifically, the embodiment of the present application can initialize the optimization step size ρ, initialize the multi-agent detection task parameters B, C, Λ, and set the multi-agent communication topology

where V={1,...,N} is a multi-agent set,

For the multi-agent communication edge, initialize the detection task assignment strategy {X _i } ⁰ , initialize the target camouflage strategy D ⁰ , and set the optimization termination conditions ∈ _pri and ∈ _dual .

Further, the embodiment of the present application can solve the optimal target camouflage strategy under the current detection task allocation strategy:

Further, the embodiment of the present application can solve the optimal task allocation strategy of each agent under the current optimal target camouflage strategy in parallel:

in,

Further, the embodiment of the present application can solve the dual parameters under the current detection task allocation strategy and the optimal target camouflage strategy in parallel:

及对偶问题残差

若同时满足

及

则终止优化，否则重复上述过程直至优化时间消耗完毕。Further, the embodiment of the present application can solve the residual of the original problem

and dual problem residuals

If both meet

and

Then terminate the optimization, otherwise repeat the above process until the optimization time is exhausted.

Further, in the embodiment of the present application, the detection task assignment solution of the multi-agent system can be the average value of the local detection task assignment solutions of all agents:

To sum up, compared with the related art, the embodiment of the present application solves the problem of the influence of the camouflage ability of the target on the detection efficiency in the multi-target detection task, has strong robustness, and is suitable for multi-intelligence with connected communication topology. In the volume distributed detection task, it has a high convergence rate, and can search for a reasonable detection task assignment result under the target camouflage in a short time.

Specifically, with reference to FIG. 2 , the working principle of the distributed multi-agent detection task assignment method in the confrontation scenario according to the embodiment of the present application will be described in detail with a specific embodiment.

As shown in FIG. 2, this embodiment of the present application may include the following steps:

初始化探测任务分配权重B、多智能体系统携带的目标探测器配置参数C、不同探测器的目标探测能力参数Λ、目标伪装策略D⁰、多智能体探测任务分配策略

Step S201: Initialize parameters and set termination conditions. This embodiment of the present application can set a multi-agent communication topology

Initialized detection task assignment weight B, target detector configuration parameter C carried by the multi-agent system, target detection capability parameter Λ of different detectors, target camouflage strategy D ⁰ , multi-agent detection task assignment strategy

表示多智能体系统的通信拓扑结构，其中，V＝{1,…,N}为多智能体集合，

为多智能体通信边集合。Multi-agent communication topology

represents the communication topology of the multi-agent system, where V={1,...,N} is the multi-agent set,

Sets of edges for multi-agent communication.

为连通图。In the embodiment of the present application, according to the communication range of the current multi-agent system, if the distance between the agent i and the agent j is less than the communication distance threshold, it is considered that the agent i can communicate continuously and stably with the agent j. In the present invention think communication topology

is a connected graph.

It should be noted that the communication distance threshold can be set by those skilled in the art according to actual conditions, and no specific limitation is made here.

[B]_mn表示智能体n执行探测任务m的权重大小，本申请实施例通过智能体执行探测任务的权重，可表示多智能体与目标的距离等因素产生的探测任务的执行成本，权重值越小表示执行成本越高。The detection task assignment weight B represents the weight of the agent to perform the detection task, where [B] _mn is a scalar value, and [B] _mn ≥ 0 and

[B] _mn represents the weight of the detection task m performed by the agent n. In this embodiment of the application, the weight of the detection task performed by the agent can represent the execution cost of the detection task caused by factors such as the distance between the multi-agent and the target, and the weight value Smaller means higher execution cost.

The target detector configuration parameter C represents the types of detectors carried by different agents, such as infrared detectors, lidars, cameras, etc. If the multi-agent system has S-type detectors, take [C] _sn ∈ {0,1}, [C] _sn = 1 means that the agent n has s-type detectors, otherwise it is 0.

The target detection ability parameter Λ of different detectors represents the ability of the detector to acquire different characteristic information of the target, and takes Λ=diag(λ ₁ ,...,λ _S ), where λ _s represents the target detection ability of the s-type detector. The detection capability parameter Λ can represent the influence of various complex meteorological conditions, such as low visibility and night time, on the capability of the detector. The larger the parameter value, the stronger the detection capability.

The target camouflage strategy D represents the type of self-feature information corresponding to the target detector type of target camouflage, where [D] _ms ∈ [0,1], indicating that the feature information type of target m camouflage is s. Under this camouflage strategy, Target detector s cannot detect target m. In the initialization of the target camouflage strategy D, the embodiment of the present application may sample the S×M dimension D ⁰ with a uniform distribution in the interval [0,1].

The multi-agent detection task assignment strategy {X _i } represents the detection task assignment result of the agent i, where [X _i ] _mn ∈[0,1] represents the probability that the detection task m is assigned to the agent n by the agent i. This embodiment of the present application may take the initialization of the multi-agent detection task assignment strategy {X _i }, and sample N M×N dimensions with a uniform distribution in the interval [0,1]

更新目标伪装策略D^k+1。目标伪装策略的目标是最小化探测效能函数，同时满足目标伪装能力约束条件0≤D≤1,

结合增广拉格朗日函数，可以将当前探测任务分配策略

代入目标伪装优化函数，得到k+1时刻的目标伪装策略。Step S202: Optimizing the target camouflage strategy. In this embodiment of the present application, the current detection task allocation strategy can be input

Update the target camouflage policy D ^k+1 . The goal of the target camouflage strategy is to minimize the detection efficiency function while satisfying the target camouflage capability constraint 0≤D≤1,

Combined with the augmented Lagrangian function, the current detection task assignment strategy can be

Substitute into the target camouflage optimization function to obtain the target camouflage strategy at time k+1.

并通过以下公式进行更新：The optimization method of target camouflage strategy is based on the alternating direction multiplier method, and iteratively solves the target camouflage strategy at time k+1. Specifically, the embodiment of the present application may firstly initialize the optimal target camouflage strategy Dk at time ^k as the initial value, that is, Dk ^+1,0 = ^Dk ; secondly, alternately optimize the objective function of the camouflage strategy and the local part of the constraint function The optimal value, where the initial value of the local optimal value of the objective function of the camouflage strategy is

and updated by the following formula:

U_d＝0_M×N，约束函数的局部最优值初值为

并通过以下公式进行更新：in,

U _d =0 _M×N , the initial value of the local optimal value of the constraint function is

and updated by the following formula:

Prox is a near-end operator, which can be defined as follows:

Again, this embodiment of the present application can update step t+1

以及

足够小时(通常取10^-3)，令k+1时刻的目标伪装策略

Finally, when the convergence error

as well as

Small enough (usually 10 ^-3 ) to make the target camouflage strategy at time k+1

多智能体探测任务分配策略的目标是最大化探测效能函数，同时满足多智能体能力约束条件0≤X_i≤1,

满足探测任务约束条件|X^m|_∞＝1，结合增广拉格朗日函数，将k+1步的目标伪装策略D^k+1代入探测任务分配策略优化函数，得到k+1时刻的探测任务分配策略。Step S203: Optimizing the agent detection task assignment strategy. In this embodiment of the present application, the target camouflage strategy D ^k+1 of k+1 steps can be input, and the k+1 multi-agent detection task assignment strategy can be updated in parallel

The goal of the multi-agent detection task assignment strategy is to maximize the detection efficiency function while satisfying the multi-agent capability constraint 0≤X _i ≤1,

Satisfy the detection task constraint |X ^m | _∞ =1, combined with the augmented Lagrangian function, substitute the target camouflage strategy D ^k+1 of k+1 steps into the detection task assignment strategy optimization function, and obtain the detection at time k+1 task assignment strategy.

为初始值，即

其次交替优化探测任务分配策略的目标函数和约束函数的局部最优值，其中探测任务分配的目标函数的局部最优值初值为

并通过以下公式进行更新：For the detection task allocation strategy optimization method, the embodiment of the present application may be based on the alternate direction multiplier method to iteratively solve the detection task allocation strategy at time k+1. Specifically, for agent i, this embodiment of the present application may first initialize the optimal detection task assignment strategy at time k

is the initial value, that is

Secondly, the objective function of the detection task assignment strategy and the local optimal value of the constraint function are alternately optimized, wherein the initial value of the local optimal value of the objective function of the detection task assignment is

and updated by the following formula:

in,

并可以通过以下公式进行更新：The initial value of the local optimal value of the constraint function of the detection task assignment strategy is

and can be updated by the following formula:

Again, update t+1 step

以及

足够小时(通常取10^-3)，令k+1时刻的探测任务分配策略

Finally, when the convergence error

as well as

is small enough (usually 10 ^-3 ), so that the detection task allocation strategy at time k+1

及对偶问题残差

In this embodiment of the present application, the solution residual can be solved according to the following formula

and dual problem residuals

及

则终止优化，否则跳至上述步骤直至优化时间消耗完毕。If both meet

and

Then terminate the optimization, otherwise skip to the above steps until the optimization time is exhausted.

以及对应的最差目标伪装策略D^*＝^k。This embodiment of the present application can output an optimal detection task allocation strategy

and the corresponding worst target camouflage strategy D ^* = ^k .

Step S204: Determine whether the maximum number of optimization steps is reached. If the maximum number of steps is reached, go to step S206, otherwise go to step S205.

Step S205: Determine whether the threshold conditions of the consistency error and the optimality error are satisfied. If satisfied, go to step S206, otherwise go to step S203.

Step S206: Return to the optimal solution of multi-agent detection task assignment.

According to the distributed multi-agent detection task assignment method in the confrontation scenario proposed by the embodiment of the present application, the weight of the detection task can be assigned, the target detector configuration and detection capability carried by the multi-agent system, the camouflage strategy of the target, the multi-agent The detection task allocation strategy uses the multi-agent detection efficiency function and the distributed multi-agent detection task allocation model to achieve a reasonable allocation of multi-agents with different detection capabilities, thereby minimizing the impact of target camouflage on multi-agent detection performance. , and then maximize the collection of target feature information in the adversarial environment. As a result, the distributed detection task assignment method in the related art is solved, because the potential camouflage confrontation process of the target is not fully considered, the execution result of the detection task is easily affected by the target camouflage strategy, thereby reducing the acquisition of target feature information. amount of technical issues.

Next, the distributed multi-agent detection task assignment device in the confrontation scenario proposed according to the embodiment of the present application will be described with reference to the accompanying drawings.

FIG. 3 is a schematic block diagram of a distributed multi-agent detection task assignment device in a confrontation scenario according to an embodiment of the present application.

As shown in FIG. 3 , the distributed multi-agent detection task assignment device 10 in the confrontation scenario includes: a first function establishment module 100 , a model establishment module 200 and an assignment module 300 .

Specifically, the first function establishment module 100 is used to establish the multi-agent detection efficiency according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the target camouflage strategy, and the multi-agent detection task allocation strategy function.

The model establishment module 200 is configured to establish a distributed multi-agent detection task assignment model according to preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints.

The allocation module 300 is configured to alternately solve the task allocation strategy and the corresponding target camouflage strategy of the multi-agent system according to the gradient information based on the distributed multi-agent detection task allocation model, and generate a distributed multi-agent detection task allocation result.

Optionally, in an embodiment of the present application, the distributed multi-agent detection task assignment device 10 in the confrontation scene further includes: a setting module and a second function establishment module.

Among them, the setting module is used to set optimization parameters and termination conditions.

The second function establishment module is used for establishing the augmented Lagrangian function of the detection efficiency, and adding a non-convergence penalty term of the detection efficiency function.

Optionally, in an embodiment of the present application, the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors.

Optionally, in an embodiment of the present application, the allocation module 300 includes: a first solving unit.

Wherein, the first solving unit is used to solve the maximum value of the detection efficiency function under all camouflage strategies by using the multi-agent task assignment model.

Optionally, in an embodiment of the present application, the allocation module 300 further includes: a second solving unit, a third solving unit, and an allocating unit.

Among them, the second solving unit is used to solve the optimal target camouflage strategy under the current detection task allocation strategy.

The third solving unit is used to solve the optimal task allocation strategy of each agent under the current optimal target camouflage strategy in parallel, so as to solve the dual parameters of the current detection task allocation strategy and the optimal target camouflage strategy in parallel.

The allocation unit is used to solve the residual of the original problem and the residual of the dual problem based on the dual parameters until the end of the optimization, so that the detection task allocation solution of the multi-agent system is the average value of the local detection task allocation solutions of all agents.

It should be noted that the foregoing explanation of the embodiment of the distributed multi-agent detection task assignment method in the confrontation scene is also applicable to the distributed multi-agent detection task assignment device in the confrontation scene of this embodiment, which is not repeated here. Repeat.

According to the distributed multi-agent detection task assignment device in the confrontation scenario proposed by the embodiment of the present application, the weight of detection tasks can be assigned, the target detector configuration and detection capability carried by the multi-agent system, the camouflage strategy of the target, and the multi-agent system. The detection task allocation strategy uses the multi-agent detection efficiency function and the distributed multi-agent detection task allocation model to achieve a reasonable allocation of multi-agents with different detection capabilities, thereby minimizing the impact of target camouflage on multi-agent detection performance. , and then maximize the collection of target feature information in the adversarial environment. As a result, the distributed detection task assignment method in the related art is solved, because the potential camouflage confrontation process of the target is not fully considered, the execution result of the detection task is easily affected by the target camouflage strategy, thereby reducing the acquisition of target feature information. amount of technical issues.

FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The electronic device may include:

Memory 401 , processor 402 , and computer programs stored on memory 401 and executable on processor 402 .

When the processor 402 executes the program, the distributed multi-agent detection task assignment method in the confrontation scenario provided in the above embodiment is implemented.

Further, the electronic device also includes:

The communication interface 403 is used for communication between the memory 401 and the processor 402 .

The memory 401 is used to store computer programs that can be executed on the processor 402 .

The memory 401 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

If the memory 401, the processor 402 and the communication interface 403 are independently implemented, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and complete communication with each other. The bus may be an Industry Standard Architecture (referred to as ISA) bus, a Peripheral Component (referred to as PCI) bus, or an Extended Industry Standard Architecture (referred to as EISA) bus or the like. The bus can be divided into address bus, data bus, control bus and so on. For ease of presentation, only one thick line is used in FIG. 4, but it does not mean that there is only one bus or one type of bus.

Optionally, in terms of specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on one chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.

The processor 402 may be a central processing unit (Central Processing Unit, CPU for short), or a specific integrated circuit (Application Specific Integrated Circuit, ASIC for short), or is configured to implement one or more of the embodiments of the present application integrated circuit.

This embodiment also provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, implements the above distributed multi-agent detection task assignment method in a confrontation scenario.

In the description of this specification, description with reference to the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples", etc., mean specific features described in connection with the embodiment or example , structure, material or feature is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials or characteristics described may be combined in any suitable manner in any one or N of the embodiments or examples. Furthermore, those skilled in the art may combine and combine the different embodiments or examples described in this specification, as well as the features of the different embodiments or examples, without conflicting each other.

In addition, the terms "first" and "second" are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with "first", "second" may expressly or implicitly include at least one of that feature. In the description of the present application, "N" means at least two, such as two, three, etc., unless otherwise expressly and specifically defined.

Any process or method description in the flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or N more executable instructions for implementing custom logical functions or steps of the process , and the scope of the preferred embodiments of the present application includes alternative implementations in which the functions may be performed out of the order shown or discussed, including performing the functions substantially concurrently or in the reverse order depending upon the functions involved, which should It is understood by those skilled in the art to which the embodiments of the present application belong.

The logic and/or steps represented in flowcharts or otherwise described herein, for example, may be considered an ordered listing of executable instructions for implementing the logical functions, may be embodied in any computer-readable medium, For use with, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other system that can fetch instructions from and execute instructions from an instruction execution system, apparatus, or apparatus) or equipment. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport the program for use by or in conjunction with an instruction execution system, apparatus, or apparatus. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections (electronic devices) with one or N wires, portable computer disk cartridges (magnetic devices), random access memory (RAM), Read Only Memory (ROM), Erasable Editable Read Only Memory (EPROM or Flash Memory), Fiber Optic Devices, and Portable Compact Disc Read Only Memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, followed by editing, interpretation, or other suitable medium as necessary process to obtain the program electronically and then store it in computer memory.

It should be understood that various parts of this application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented by any one of the following techniques known in the art, or a combination thereof: discrete with logic gates for implementing logic functions on data signals Logic circuits, application specific integrated circuits with suitable combinational logic gates, Programmable Gate Arrays (PGA), Field Programmable Gate Arrays (FPGA), etc.

Those skilled in the art can understand that all or part of the steps carried by the methods of the above embodiments can be completed by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium, and the program can be stored in a computer-readable storage medium. When executed, one or a combination of the steps of the method embodiment is included.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing module, or each unit may exist physically alone, or two or more units may be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware, and can also be implemented in the form of software function modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.

The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, and the like. Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limitations to the present application. Embodiments are subject to variations, modifications, substitutions and variations.

Claims (12)

1. a distributed multi-agent detection task assignment method in a confrontation scene, is characterized in that, comprises the following steps: The multi-agent detection efficiency function is established according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the target camouflage strategy, and the multi-agent detection task allocation strategy; Establish a distributed multi-agent detection task assignment model according to the preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints; Based on the distributed multi-agent detection task assignment model, the task assignment strategy of the multi-agent system and the corresponding target camouflage strategy are alternately solved according to the gradient information, and the distributed multi-agent detection task assignment result is generated. 2. The method according to claim 1, characterized in that, before solving the task allocation strategy of the multi-agent system and the corresponding target camouflage strategy, further comprising: Set optimization parameters and termination conditions; The augmented Lagrangian function of the detection efficiency is established, and the non-convergence penalty term of the detection efficiency function is added. 3. The method according to claim 1, wherein the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors. 4. The method according to claim 1, wherein the generating a distributed multi-agent detection task assignment result comprises: The multi-agent task assignment model is used to solve the maximum value of the detection efficiency function under all camouflage strategies. 5. The method according to any one of claims 1-4, wherein the alternately solving the task allocation strategy and the corresponding target camouflage strategy of the multi-agent system according to the gradient information, comprising: Solve the optimal target camouflage strategy under the current detection task assignment strategy; Solve the optimal task assignment strategy of each agent under the current optimal target camouflage strategy in parallel, so as to solve the current detection task assignment strategy and dual parameters under the optimal target camouflage strategy in parallel; Based on the dual parameters, the original problem residual and the dual problem residual are solved until the optimization ends, so that the detection task assignment solution of the multi-agent system is the average value of the local detection task assignment solutions of all agents. 6. A distributed multi-agent detection task assignment device in a confrontation scene, characterized in that, comprising: The first function establishment module is used to establish the multi-agent detection efficiency function according to the detection task allocation weight, the target detector configuration and detection capability carried by the multi-agent system, the target camouflage strategy, and the multi-agent detection task allocation strategy; The model establishment module is used to establish a distributed multi-agent detection task assignment model according to preset multi-agent detection capability constraints, target camouflage capability constraints, and detection task constraints; The assignment module is configured to alternately solve the task assignment strategy and the corresponding target camouflage strategy of the multi-agent system according to the gradient information based on the distributed multi-agent detection task assignment model, and generate a distributed multi-agent detection task assignment result. 7. The apparatus of claim 6, further comprising: Setting module, used to set optimization parameters and termination conditions; The second function establishment module is used for establishing the augmented Lagrangian function of the detection efficiency, and adding a non-convergence penalty term of the detection efficiency function. 8. The apparatus according to claim 6, wherein the multi-agent detection efficiency function is:

Among them, X _i represents the estimation of the global task assignment result of the multi-agent system by the agent i, B represents the distribution weight parameter between the agent and the target, C represents the target detector type carried by the agent system, and D represents the target camouflage strategy , E _i represents a matrix with M rows and N columns in which the i-th column is all 1 and the rest are 0, and Λ is the target detection capability parameter of different detectors. 9. The apparatus according to claim 6, wherein the distribution module comprises: The first solving unit is configured to use the multi-agent task assignment model to solve the maximum value of the detection efficiency function under all camouflage strategies. 10. The device according to any one of claims 6-9, wherein the distribution module further comprises: The second solving unit is used to solve the optimal target camouflage strategy under the current detection task allocation strategy; The third solving unit is used to solve the optimal task allocation strategy of each agent under the current optimal target camouflage strategy in parallel, so as to solve the current detection task allocation strategy and the dual parameters of the optimal target camouflage strategy in parallel ; an allocation unit, configured to solve the residual of the original problem and the residual of the dual problem based on the dual parameters, until the optimization ends, so that the detection task allocation solution of the multi-agent system is the average value of the local detection task allocation solutions of all agents . 11. An electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the program as claimed in the claim Distributed multi-agent detection task assignment method in confrontation scene according to any one of requirements 1-5. 12. A computer-readable storage medium on which a computer program is stored, characterized in that the program is executed by a processor to implement the distributed system in the confrontation scenario according to any one of claims 1-5. A multi-agent detection task assignment method.

Images