A method and system for constructing a cooperative management control loop of unmanned clusters

Abstract

The application discloses a kind of unmanned cluster cooperative management control loop construction method and system, it is related to unmanned cluster, task planning, including: center unmanned ship searches and detects target information, to the target ship group of distribution is perceived;The relevant information of target ship group obtained by sensing is obtained, and initial situation of sea is generated;Establish event adaptive decision response engine, implement two-layer composite decision, wherein the first layer of two-layer composite decision is directed to situation transition update coping strategy, the second layer is according to current strategy and carries out resource allocation calculation and is organized to strategy in strategy set, to complete control loop construction;The target set of strategy calculation and the action plan generated by strategy are sent to each unmanned ship.The application can be used for unmanned surface ship cluster to execute water safety supervision autonomous patrol task organization and decision, improve task organization and decision efficiency.

Description

Technical Field

[0001] This application relates to the fields of unmanned swarms and mission planning, and in particular to a method and system for constructing a collaborative management control loop for unmanned swarms. Background Technology

[0002] Unmanned surface vessels (USVs) have broad application prospects in many fields as a new type of surface vehicle. Typical applications include surveying and mapping, hydrological and meteorological surveys, and environmental monitoring. However, there are relatively few USVs integrated into maritime systems for maritime safety supervision.

[0003] Existing maritime facilities used for safety monitoring systems mainly include Vessel Tracking System (VTS) radar, Automatic Identification System (AIS), and manned patrol vessels. VTS radar can establish a point track based on the target's echo, but it is susceptible to surface light reflection and sea clutter, resulting in false targets. Targets may also be obscured by waves, causing weak or discontinuous echoes, leading to missed detections. AIS systems can provide dynamic and static information such as vessel position, identity, and intent; however, some fishing vessels, small boats, and vessels engaged in illegal activities do not have AIS installed, resulting in incomplete and inaccurate maritime information. Manned patrol vessels have cumbersome preparation processes, require resupply, have limited numbers and sailing time, and are significantly affected by weather, hydrological conditions, and human factors during patrols. They are also relatively large and have high maintenance costs. Therefore, using unmanned surface vessels (USVs) to integrate into the maritime equipment system to form an autonomous patrol force, equipped with radar, photoelectric, infrared, acoustic and other reconnaissance and evidence collection equipment for patrol and law enforcement, and distributed and parallel execution of maritime safety supervision tasks, is an effective means to make up for the insufficient capabilities of the existing maritime equipment system, refine the granularity of maritime supervision, increase the density of control, and improve the effectiveness of safety supervision.

[0004] How to efficiently coordinate and cooperate multiple unmanned surface vessel (USV) platforms to enhance the ability to conduct collaborative reconnaissance and surveillance of multiple targets, and optimize the allocation of reconnaissance and surveillance resources, is a crucial issue in the application of USV swarms for maritime safety surveillance. Traditional swarm collaborative decision-making only allocates resources for a specific type of maritime situation and cannot adaptively adjust strategies for various situations to improve the swarm's ability to respond to emergencies. Summary of the Invention

[0005] This application provides a method and system for constructing a collaborative management control loop for unmanned surface vessels (USVs), which is used in the organization of collaborative maritime safety management tasks for USVs. The decision-making part establishes an event adaptive response decision engine and a situation agent-based policy scheduler, which are applied to collaborative maritime safety management of USVs.

[0006] This application provides a method for constructing a control loop for unmanned cluster collaborative management, including: The central unmanned surface vessel searches for and detects target information in order to perceive the distributed target vessel group; Acquire relevant information about the target vessel group obtained through sensing and generate an initial sea situation; Establish an event adaptive decision response engine and implement a two-layer composite decision-making process. The first layer of the two-layer composite decision-making process updates the response strategy for situational changes, while the second layer calculates resource allocation based on the current strategy and organizes and applies the strategies within the strategy set to complete the construction of the control loop. The target set calculated by the strategy and the action plan generated by the strategy are sent to each unmanned surface vessel.

[0007] This application provides an unmanned cluster collaborative management control loop construction system, including a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the unmanned cluster collaborative management control loop construction method described above.

[0008] The embodiments of this application are used in the organization of unmanned surface vessel (USV) swarm collaborative maritime safety management tasks. The decision-making part establishes an event adaptive response decision engine and a situation agent-based policy scheduler, which are applied to unmanned surface vessel swarm collaborative maritime safety management.

[0009] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0010] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of the organizational process for constructing the unmanned cluster collaborative management control loop according to an embodiment of this application; Figure 2 This is a schematic diagram of the event adaptive decision response engine model architecture of the unmanned cluster collaborative management control loop construction method in this application embodiment; Figure 3 This is a schematic diagram of the situation agent policy scheduler architecture of the unmanned cluster collaborative management control loop construction method in this application embodiment. Detailed Implementation

[0011] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0012] The increasing number and frequency of maritime vessels traveling to and from ports, waterways, and coastal areas have put pressure on and challenged maritime safety supervision, creating a demand for maritime management agencies to increase the density, frequency, scope, and duration of regulatory patrols. Using unmanned swarms for collaborative maritime safety supervision is one solution.

[0013] How to rationally and efficiently organize unmanned swarms to perform regulatory tasks is an OOD*A control loop construction problem. Within this problem, how to allocate resources and arrange the patrol target sets of each unmanned vessel to enable them to work in parallel and cope with different situations is a task collaborative decision-making problem and the core of the command and control loop. Traditional resource allocation methods in task collaborative decision-making only address certain typical situations, while the actual maritime situation changes with the removal of false targets, the addition of new targets, the removal of completed targets, and the joining and leaving of unmanned vessels. If the task collaborative decision-making resource allocation strategy is singular, it cannot adaptively adjust to multiple maritime situations. Based on this, this application provides an optimized method for constructing an OOD*A control loop for unmanned swarm collaborative maritime management in the decision-making stage, such as... Figure 1 As shown, it includes the following steps: In step S101, corresponding to the distributed perception of the distributed monitoring vessel group (distributed O phase), the central unmanned surface vessel searches and detects target information in order to perceive the distributed target vessel group. In some examples, the central unmanned surface vessel acts as a flexible mobile perception platform to search and detect target information within a large sea area, while other management unmanned surface vessels in the group perceive and detect surrounding targets.

[0014] In step S102, corresponding to the centralized monitoring situation judgment (centralized O stage), relevant information of the target vessel group obtained by perception is acquired, and an initial sea situation is generated. Specifically, the management unmanned vessel group center will comprehensively process the early warning target information and the internal information of the group, and generate a unified initial sea situation based on the target information.

[0015] In step S103, corresponding to the centralized composite regulatory decision-making (centralized D* stage), an event adaptive decision response engine is established to implement two-layer composite decision-making. The first layer of the two-layer composite decision-making updates the response strategy for situational transitions, and the second layer calculates resource allocation based on the current strategy and organizes and applies the strategies within the strategy set to complete the construction of the control loop.

[0016] In step S104, the target set calculated by the strategy and the action plan generated by the strategy are sent to each unmanned surface vessel.

[0017] The embodiments of this application are used in the organization of unmanned surface vessel (USV) swarm collaborative maritime safety management tasks. The decision-making part establishes an event adaptive response decision engine and a situation agent-based policy scheduler, which are applied to unmanned surface vessel swarm collaborative maritime safety management.

[0018] In practice, situations are constantly changing, and different strategies have their own advantages and disadvantages. A complementary approach is needed to address these shifts. An event-adaptive decision-response engine model is as follows: Figure 2 As shown, the engine is used to achieve adaptive composite decision-making based on received situational information. Its main components are an event-driven policy scheduler and a situational-association policy-based decision calculator. Taking the maritime management of unmanned surface vessel (USV) swarms as an example, this engine runs on the central USV and is used for USV resource allocation. In some embodiments of this application, establishing an event-adaptive decision-response engine includes: After the adaptive decision-response engine runs, the management target data after data processing is extracted. Information on unmanned surface vessel swarms For example, to determine the maritime situation of the unmanned surface vessel cluster at the initial moment.

[0019] The management target data to be processed Information on unmanned surface vessel swarms As input to the engine's first-level decision-making module, it enters n S1, S2, ... It identifies a specific situational node and selects the corresponding response strategy.

[0020] In the event of subsequent changes in the management environment, the situation will be adapted to the current situation based on event-driven scenario migration. Coping strategies , As input to the engine's second-layer decision-making module, it is then processed by the strategy organization and application unit. according to The strategy performs resource allocation calculations and organizes and applies the strategies within the strategy set. They represent n A set of strategies for dealing with each situation.

[0021] In some embodiments of this application, a policy scheduler based on a situation agent is also included. The policy scheduler includes a situation agent, events, and migration paths, such as... Figure 2 As shown, where: Each situation agent, represented as a node in the policy scheduler, is configured with a corresponding description, an environmental perception and judgment unit, and a set of response strategies. Any two different situation agent nodes are connected. For example... Figure 2 In the diagram, ellipses represent situation agents, labeled S1, S2, ... The status agent description is represented by P1, P2, ... This represents the units used by the situation agent to perceive and judge the environment, D1, D2, ... This represents the set of response strategies of the situation agent, B1, B2, ... This indicates the organizational unit by which the situation agent applies the strategy.

[0022] An event is used to describe the cause event that causes any situational agent to migrate.

[0023] The migration path, represented as a vector, points from the originating node to the destination node, including migrations between situation agents and self-return migrations between situation agents. Originating situation agents are in a dormant state, while destination situation agents are in an active state. The vector is labeled... The event that causes the migration of the status agent.

[0024] In some embodiments of this application, establishing an event adaptive decision response engine and implementing two-layer composite decision-making further includes: Through the perception and judgment units P1, P2, ... of each situation agent Based on situation information and initial situation assessment conditions , … It determines whether it meets the conditions to decide whether to activate the current situation agent; For the subsequent situation agent i The migration path is determined by its perception and judgment unit. Based on situation information and events Sure.

[0025] In some embodiments of this application, such as Figure 3 As shown, it also includes classifying the situation into four types of nodes—initial, sparse, normal, and saturated—according to threshold rules, and classifying and determining the situation agent through the following steps: Initial situation description This indicates that management decisions will not be made at this time; the conditions for situation assessment are as follows. For: The number of management targets is satisfied ; Sparse situation description words This indicates that the problem space for management decision-making is small and the conditions for situation assessment are limited. For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies ; Conventional situation description words It represents a medium-sized problem space in management decision-making, and the judgment conditions are... For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies ; Saturation state description words It represents a large spatial scale of management decision-making problems and has specific judgment conditions. For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies .

[0026] In some embodiments of this application, it also includes based on events Perform situational agent migration: During the situational shift, the increase in the number of unmanned surface vessel swarms and managed targets was respectively... This means that the following conditions are met: ; After accumulating for a period of time Afterwards, the situation was updated, and the number of unmanned surface vessel clusters and managed targets were respectively represented by... and This means that the following conditions are met: ; Configure unit vector Situation description matrix ; According to the incident Perform migration between situation agents or self-return migration of situation agents.

[0027] In a specific example, due to the event The following is an example of a situational agent: : , indicating the increment of the management target number =0, Still in an empty set, management level unchanged, status quo migrated from S1 to S1, described as follows: ; : The management level has been raised, with the number of management targets increasing. Greater than the number of completed targets The updated unmanned surface vessel swarm Not empty, the situation changes Transfer to , described as ; : The management level is raised, and the sum of the updated cluster size and the target size is increased. The situation has expanded from an initial state to a conventional state, and the situation has changed from... Transfer to , described as ; : This indicates an increase in management level. The situation has expanded from a primary state to a saturated state, and the situation has changed from... Transfer to , described as ; : This indicates a reduction in management level. The situation has been reduced from a sparse state to an elementary state, and the situation has changed from... Transfer to , described as ; : This indicates that the management level has not changed. It remains in a sparse state, and the situation has changed. Transfer to , described as ; : This indicates an increase in management level. The situation has expanded from a sparse state to a normal state, and the situation has changed from... Transfer to , described as ; : This indicates an increase in management level. The situation has expanded from a sparse state to a saturated state, and the situation has changed from... Transfer to , described as ; : This indicates a reduction in management level. The situation has been downgraded from a normal state to an initial state, and the situation has changed from... Transfer to , described as ; : This indicates a reduction in management level. From being in a normal state to being in a sparse state, the state changed from Transfer to , described as ; : This indicates that the management level has not changed. It remains within the normal range, and the situation has changed. Transfer to , described as ; : This indicates an increase in management level. The situation has expanded from a normal state to a saturated state, and the situation has changed from... Transfer to , described as ; : This indicates a reduction in management level. The situation has decreased from a saturated state to a primary state, and the situation has changed from... Transfer to , described as ; : This indicates a reduction in management level. The state decreased from a saturated state to a sparse state, and the state changed from... Transfer to , described as ; : This indicates a reduction in management level. The situation has decreased from a saturated state to a normal state, and the situation has changed from... Transfer to , described as ; : This indicates that the management level has not changed. It is still in the saturation range, and the state has changed. Transfer to , described as .

[0028] In some embodiments of this application, the situation agent is also associated with a policy: During the operation of the policy scheduler, the associated decision function is jointly followed by all situation agents in the event adaptive decision response engine. Determine the situation-policy and output the policy; The associated decision function Specifically, it is used based on the value of the situation status word S and the situation transition identifier. The value of determines the management resource allocation strategy used under the current maritime situation. , for:

[0029] Mapping management resource configuration policy set: D1: The strategy is to not allocate resources. D2: Resource allocation strategy based on Hungarian algorithm for sparse situations; Resource allocation strategies based on genetic computing for routine situations; : Clustering and genetic computing-based classification resource allocation strategies and small-group relay reconnaissance strategies for saturation situations.

[0030] In addition to considering traditional indicators such as distance, payload and task attribute matching degree, and remaining endurance, the decision functions of strategies D2, D3, and D4 also need to consider the backlog of unprocessed tasks of the unmanned surface vessel.

[0031] In some embodiments of this application, it also includes... , , , As respectively The strategy application unit calculates the strategy and organizes and applies the strategies within the strategy set. Some specific examples are as follows: When the target size is small, the requirement is to conduct detailed reconnaissance of each target, that is, the situational attributes are sparse. It is necessary to make accurate resource allocation decisions for the monitored targets—unmanned surface vessels—within a reasonable computation time. Therefore, the Hungarian algorithm is used as the resource allocation strategy D2. Even if the number of targets is greater than the number of unmanned surface vessels, an accurate solution can be obtained after multiple rounds of allocation. As the target size increases and the situational attributes become more conventional, the computation time for obtaining an accurate solution using D2 algorithm after multiple rounds of calculations becomes unacceptable due to the long computation time. In this case, a more computationally efficient evolutionary computation strategy is needed to complete the configuration in one go, allowing for adjustment of solution quality based on population size and iteration time, and providing parameter adjustability for decision quality. Genetic algorithms can achieve a controllable balance between the acceptability of solutions and the acceptability of computation time. Therefore, they can be used to solve resource allocation decision problems under such conventional conditions. Alternatively, other evolutionary computation algorithms or their improvements can be selected. Examples include ant colony optimization and particle swarm optimization, which will not be elaborated on here.

[0032] When faced with large-scale problems and when the situational attributes become saturated, especially in non-equilibrium problems where the number of targets exceeds the number of unmanned surface vessels (USVs), using USVs for detailed reconnaissance of each target would become unacceptably time-consuming. In such cases, a new operational strategy is needed. .

[0033] ① Use a divide-and-conquer strategy: First, use a clustering algorithm to cluster the target groups. The class uses each cluster center as the target coordinate. The resource allocation decision algorithm requires at least one unmanned surface vessel (USV) to be assigned to reconnaissance for each target cluster. ② A relay reconnaissance strategy is employed: each USV is instructed to form a collaborative unit with its onboard drones to reconnoiter the target cluster area. The drones utilize their speed advantage to advance ahead of the USV, and the USV and drones share the cluster area situation via data link, completing air-sea relay collaborative reconnaissance. ③ A collaborative unit action plan is generated: when the drone's electro-optical pod reaches the target cluster area at its maximum effective range, the USV releases a multi-rotor drone, allowing it to advance ahead of the USV using its speed advantage. The drone is set to use an autonomous navigation and reconnaissance mode, employing an electro-optical pod to scan the sea surface from the air and provide information to the USV via data link. Both sides jointly maintain the target list. The drone removes false targets and completed targets from the target list and marks uncompleted targets, requesting the USV to approach for reconnaissance, thus completing the air-sea relay collaborative reconnaissance with the USV.

[0034] Finally, in step S104, the distributed monitoring target handling (distributed A phase) is carried out. The central vessel sends the target set calculated by the strategy and the action plan generated by the strategy to each unmanned surface vessel (USV). The USV and its unmanned surface vessel units perform target reconnaissance. The USV infers and confirms the handling method and implements the handling action based on the reconnaissance information such as target type and intent and internal rules.

[0035] This application also proposes an unmanned cluster collaborative management control loop construction system, including a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the steps of the unmanned cluster collaborative management control loop construction method described above.

[0036] It should be noted that, in the embodiments of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0037] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0038] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0039] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.

Claims

1. A method for constructing a collaborative management and control loop for unmanned clusters, characterized in that, include: The central unmanned surface vessel searches for and detects target information in order to perceive the distributed target vessel group; Acquire relevant information about the target vessel group obtained through sensing and generate an initial sea situation; Establish an event adaptive decision response engine and implement a two-layer composite decision-making process. The first layer of the two-layer composite decision-making process updates the response strategy for situational changes, while the second layer calculates resource allocation based on the current strategy and organizes and applies the strategies within the strategy set to complete the construction of the control loop. The target set calculated by the strategy and the action plan generated by the strategy are sent to each unmanned surface vessel; Building an event-adaptive decision-response engine includes: After the adaptive decision-response engine runs, the management target data after data processing is extracted. Information on unmanned surface vessel swarms ; The management target data to be processed Information on unmanned surface vessel swarms As input to the engine's first-level decision-making module, it enters n S1, S2, ... A specific situational node is identified, and the corresponding response strategy is selected. In the event of subsequent changes in the management environment, a situational shift will be initiated based on the event-driven mechanism, and an option will be selected to adapt to the current situation. Coping strategies , As input to the engine's second-layer decision-making module, it is then processed by the strategy organization and application unit. according to The strategy performs resource allocation calculations and organizes and applies the strategies within the strategy set. They represent n A set of strategies for dealing with each situation; It also includes classifying the situation into four types of nodes—initial, sparse, normal, and saturated—according to threshold rules, and classifying and determining the situation agent through the following steps: Initial situation description This indicates that management decisions will not be made at this time; the conditions for situation assessment are as follows. For: The number of management targets is satisfied ; Sparse situation description words This indicates that the problem space for management decision-making is small and the conditions for situation assessment are limited. For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies ; Conventional situation description words It represents a medium-sized problem space in management decision-making, and the judgment conditions are... For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies ; Saturation state description words It represents a large spatial scale of management decision-making problems and has specific judgment conditions. For: The sum of the number of unmanned surface vessel swarms and the number of managed targets satisfies .

2. The method for constructing a collaborative management and control loop for unmanned clusters as described in claim 1, characterized in that, This also includes establishing a policy scheduler based on the situation agent. The policy scheduler includes the situation agent, events, and migration paths, where: Any situation agent is represented as a node in the policy scheduling machine, and is configured with a corresponding description, an environmental perception and judgment unit, and a set of response strategies. Any two different situation agent nodes are connected. An event is used to describe the cause event that causes any situational agent to migrate; Migration paths, represented as vectors, point from the outgoing node to the incoming node, including migrations between situation agents and self-return migrations of situation agents.

3. The method for constructing a collaborative management and control loop for unmanned clusters as described in claim 2, characterized in that, Establishing an event-adaptive decision-making and response engine and implementing two-layer composite decision-making also includes: Through the perception and judgment units P1, P2, ... of each situation agent Based on situation information and initial situation assessment conditions , … It determines whether it meets the conditions to decide whether to activate the current situation agent; For the subsequent situation agent i The migration path is determined by its perception and judgment unit. Based on situation information and events Sure.

4. The method for constructing a collaborative management and control loop for unmanned clusters as described in claim 1, characterized in that, Also includes based on the event Perform situational agent migration: During the situational shift, the increase in the number of unmanned surface vessel swarms and managed targets was respectively... This means that the following conditions are met: ; After accumulating for a period of time Afterwards, the situation was updated, and the number of unmanned surface vessel clusters and managed targets were respectively represented by... and This means that the following conditions are met: ; Configure unit vector Situation description matrix ; According to the incident Perform migration between situation agents or self-return migration of situation agents.

5. The method for constructing a collaborative management and control loop for unmanned clusters as described in claim 3, characterized in that, This also includes associating the situation agent with the strategy: During the operation of the policy scheduler, the associated decision function is jointly followed by all situation agents in the event adaptive decision response engine. Determine the situation-policy and output the policy; The associated decision function Specifically, it is used based on the value of the situation status word S and the situation transition identifier. The value of determines the management resource allocation strategy used under the current maritime situation. , for: Mapping management resource configuration policy set: D1: The strategy is to not allocate resources. D2: Resource allocation strategy based on Hungarian algorithm for sparse situations; Resource allocation strategies based on genetic computing for routine situations; : Clustering and genetic computing-based classification resource allocation strategies and small-group relay reconnaissance strategies for saturation situations.

6. The method for constructing a collaborative management and control loop for unmanned clusters as described in claim 5, characterized in that, It also includes , , , As respectively The strategy application unit calculates the strategy and organizes and applies the strategies within the strategy set.

7. A system for constructing a collaborative management and control loop for unmanned clusters, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, which, when executed by the processor, implements the steps of the unmanned cluster collaborative management control loop construction method as described in any one of claims 1 to 6.

Images