Dynamic real-time planning method for cooperative search scene

Abstract

A dynamic real-time planning method for collaborative search scenarios is proposed. Based on the initial position of the swarm and the boundary of the unfamiliar area to be explored, the method estimates the total search path length of each unmanned surface vessel (USV) unit through an optimization algorithm and autonomously arranges the search areas of each unit in the USV swarm. Compared with existing technologies, this invention enables the USV swarm to grasp the global situational information of the area to be explored as quickly as possible, while autonomously deciding the task status and path planning of each unit in the swarm based on the current situation. This mitigates the impact of uncertainties in the actual exploration process and ultimately minimizes the total time for multi-target search.

Description

Technical Field

[0001] This invention relates to a technology in the field of unmanned swarm control, specifically a dynamic real-time planning method for collaborative search scenarios. Background Technology

[0002] Unmanned surface vessel (USV) swarm search missions typically begin with an unfamiliar area search. This involves the USV swarm conducting a comprehensive, blind-spot-free search of a designated area to pinpoint the exact locations of targets. Once a target is detected, the USV swarm dynamically plans the tasks and trajectories of each USV unit based on known target numbers, attributes, the detection range of each USV's payload, and the remaining verification base. This ensures effective detection of all targets within the unfamiliar area before returning to the initial position. In collaborative search, the USV swarm autonomously arranges the trajectories of each USV unit to complete the search mission in the unfamiliar area and determines the verification strategies for each USV unit, ensuring optimal or near-optimal verification of all targets within the area.

[0003] The current technical challenges of unmanned swarms performing collaborative exploration in unfamiliar areas include: in actual tasks, the location and number of targets to be explored are often unknown in advance, requiring the design of a reasonable initial area search allocation strategy; for the exploration of multiple targets, the feasible path action space is often an NP-hard problem, making it difficult to solve the optimal path planning result in a short time. Summary of the Invention

[0004] To address the aforementioned shortcomings of existing technologies, this invention proposes a dynamic real-time planning method for collaborative exploration scenarios. This method enables unmanned swarms to quickly grasp the global situational information of the area to be explored, while autonomously deciding the task status and path planning of each unit in the swarm based on the current situation. This mitigates the impact of uncertainties in the actual exploration process and ultimately minimizes the total time for multi-target exploration.

[0005] This invention is achieved through the following technical solution:

[0006] This invention relates to a dynamic real-time planning method for cooperative search scenarios, comprising:

[0007] Step 1) Scene initialization: Generate the initial task area and scene elements, and set the initial parameters for obstacles, targets to be investigated, unmanned clusters and each unmanned surface vessel unit.

[0008] Step 2) Further divide the exploration area for the unmanned surface vessel (USV) units assigned to the sub-region: Based on the initial position of the USV cluster and the boundary size of the area to be explored, generate the initial search task area for each USV unit through a task allocation planning method.

[0009] Step 3) Environment map update: Add the locations of targets and obstacles discovered during the search to the map and update the environment map information.

[0010] Step 4) Target status update: After the unmanned surface vessel unit has successfully verified the target, it updates the target verification information.

[0011] Step 5) Dynamic real-time path planning for collaborative verification of unmanned swarms: Based on the information of available unmanned surface vessel (USV) units that have not yet been assigned a target, as well as the environmental map and target status information, a collaborative verification path planning objective function is established, and the target assignment result for each USV unit is calculated. Based on the position of the USV unit to be assigned and the position of the target to be verified at any given time, available USV units are assigned to the next target point through a planning algorithm.

[0012] Step 6) Unmanned surface vessel (USV) unit mission state switching: During the planning process, the USV unit may be in idle, mission execution, or disconnected states. Define the mission state and state transition relationship.

[0013] Step 7) Determine whether the collaborative search task is complete.

[0014] This invention relates to a dynamic real-time planning system for implementing the above-mentioned method, comprising: an unmanned surface vessel (USV) unit simulation module, a system configuration module, a cluster task allocation module, and a cluster planning module, wherein: the USV unit simulation module models the USV unit based on its physical dimensions, speed, acceleration, and turning angular velocity, and sets the static and dynamic model parameters of the USV unit; the system configuration module configures environmental parameters and USV unit parameters; the cluster task allocation module obtains the initial exploration area allocation results for each USV unit based on the initial task allocation algorithm, according to the environmental parameters and USV unit parameters; and the cluster planning module performs real-time task replanning processing based on the target information obtained from exploration, and obtains the next exploration path planning results for each USV unit.

[0015] Technical effect

[0016] Compared to existing initial task allocation techniques, which often involve manually determining the exploration areas of each unmanned surface vessel (USV) unit and are highly subjective, this invention autonomously allocates the exploration areas of each USV unit based on the initial position of the cluster and the boundaries of the unfamiliar area to be explored, using an optimization algorithm to estimate the total search path length of each USV unit. In contrast to existing global path planning methods, which require significant computation time and cannot guarantee real-time task allocation, this invention dynamically allocates idle USV units to the target for the next verification based on the current task status and target information (location, remaining verification attempts) of each unit in the USV cluster. This minimizes the total time for the cluster to complete the exploration of all unknown areas. The USV cluster can autonomously decide the task status and path planning of each unit based on the current situation, with a task replanning time of less than 1 second, demonstrating high real-time performance. This dynamic planning method can absorb the uncertainties in the actual exploration process while possessing high decision-making efficiency, ultimately minimizing the total time for the USV cluster to complete multi-target exploration of unfamiliar areas. Attached Figure Description

[0017] Figure 1 A schematic diagram illustrating the planning process for unmanned swarm collaborative search and exploration missions;

[0018] Figure 2 Diagram illustrating the initial task allocation for collaborative search;

[0019] Figure 3 The diagram illustrates the finite state machine for an unmanned surface vessel (USV) unit task.

[0020] Figure 4 A schematic diagram of the unmanned surface vessel (USV) unit trajectory generation and process;

[0021] Figure 5 This is a simulation diagram illustrating the initial planning of the task.

[0022] Figure 6 A simulation diagram for searching unfamiliar areas;

[0023] Figure 7 A schematic diagram illustrating the state transitions for area search and multi-target collaborative search missions;

[0024] Figure 8 A schematic diagram illustrating real-time planning for multi-target collaborative search;

[0025] Figure 9 This is a diagram illustrating the completion of the task. Detailed Implementation

[0026] like Figure 1 As shown, this embodiment relates to a dynamic real-time planning method for cooperative search scenarios, including:

[0027] Step 1) Initial area allocation considers the area size and the detection range of the unmanned boat units: Determine the number of unmanned boat units to be allocated to each sub-area. When the number of unmanned boat units allocated to the left area is N1 and the number of unmanned boat units allocated to the right area is N2, the calculation method is as follows: Where: x i ,

[0033] , , ,

[0032] is the position of the center point of the unmanned cluster on the x-axis in the Cartesian coordinate system, and x0 and x1 are the x-axis coordinates of the lower left corner and the lower right corner of the area to be explored, respectively; is rounding up; N is the total number of unmanned boat units in the cluster, N1 is the number of unmanned boats allocated to the left side of the starting area, N2 is the number of unmanned boats allocated to the right side of the starting area, and N = N1 + N2.

[0028] Step 2) Further divide the exploration area of the unmanned boat units allocated to the sub-areas, specifically including: Check the relationship between the total width in the X direction of the sub-area and the detection diameter of the unmanned boat unit. When the sum of the detection diameters of the unmanned boat units is greater than the total width in the X direction of the sub-area, it means that only some unmanned boat units are needed to complete the area search task. Then arrange the first m (m < nk) unmanned boat units to participate in the mapping, and the remaining nk - m unmanned boat units are directly assigned to the verification work; when the range of sub-area k is large and the planned unmanned boat units need to complete the area search work by turning back, the division of the exploration area of the unmanned boat units needs to consider the total navigation distance of the unmanned boat units. The goal is to minimize the search time for all unknown areas as short as possible, that is, to satisfy the min-max mixed integer linear programming problem: Solve to obtain the value of x im That is, obtain the number of channels allocated to each unmanned boat unit, and its constraints include:

[0029] i) The total navigation distance of unmanned boat unit i, including the total length of the channels in the Y direction covered by unmanned boat unit i and the navigation distance in the X direction required to reach the target channel, specifically:

[0030] ii) Each channel must be planned with an unmanned boat unit to cover it, specifically:

[0031] iii) The total width of the channels covered by the unmanned boat unit should be greater than the width in the X direction of the sub-area, specifically:

[0032] iv) When the unmanned boat unit participates in the search of unknown areas, the boolean variable z i is equal to 1, that is, unmanned boat unit i is assigned a search task, specifically:

[0033] v) The total number of unmanned surface vessel (USV) units participating in the search mission within a sub-region shall not exceed the total number of USV units allocated to that region, specifically:

[0034] vi) Only one unmanned surface vessel (USV) unit is planned to complete the search for each waterway, specifically: Where: the set of cluster units within the subregion is N = {1, 2, ... n} k}, the set of waterway numbers D is the detection diameter of unmanned surface vessel (USV) unit i, i.e., the channel within the sub-region. The maximum number of channels that USV unit i can navigate without turning back in a single trip along the Y-axis of the map does not exceed the total width divided by the minimum detection diameter of the USV unit. K is the set of sub-regions, x im Let x be a Boolean variable. When unmanned surface vessel unit i is assigned to the m-th channel, then x... im =1. The objective function means that we want the maximum travel distance of each unmanned surface vessel (USV) unit in the sub-region to be minimized. Since we need to ensure that the target to be explored is covered to the maximum extent, the search process often requires that the speed of the USV units cannot be too fast. When the maximum allowable speed of all USV units in the cluster is the same, it is equivalent to finding the shortest maximum total time and finding the shortest maximum total distance.

[0035] To ensure comprehensive coverage of the search area boundaries, the route furthest from the initial position is planned to make an additional turnaround. Once any unmanned surface vessel (USV) unit completes its assigned area search, it will directly enter the target verification phase.

[0036] Step 3) Environment map update: Based on the search results of each unmanned surface vessel (USV) unit in Step 2), the environmental information is updated. Specifically, when any USV unit detects a previously undiscovered target or ordinary obstacle, the target or obstacle is updated into the map environment information.

[0037] The environmental information includes: the state of the target to be investigated and the location information of obstacles.

[0038] Step 4) Dynamic update of target status: Check if any unmanned surface vessel (USV) unit has completed a valid verification. If an USV unit has completed a valid verification, update the verification status of the corresponding target.

[0039] Step 5) Dynamic real-time path planning for collaborative verification of unmanned swarms: Based on the information of available unmanned surface vessel (USV) units that have not yet been assigned a target, as well as the environmental map and target status information, a collaborative verification path planning objective function is established, and the target assignment result for each USV unit is calculated. Based on the position of the USV unit to be assigned and the position of the target to be verified at any given time, available USV units are assigned to the next target point through a planning algorithm.

[0040] The constraints of the collaborative verification dynamic real-time path planning include:

[0041] a) Revisit time constraint during second verification by unmanned surface vessel (USV) unit: When any USV unit verifies any target for the second time, the time interval between the two verifications must be greater than a certain threshold in order to be considered a valid verification.

[0042] b) Revisit distance constraint during secondary verification by unmanned surface vessel (USV) units: When any USV unit verifies any target for the second time, it must first leave the target beyond a certain distance threshold before proceeding to the target for verification.

[0043] The objective function for the collaborative verification path planning is: Where: optimization variable x ij The value is an integer; a value of 1 indicates that the i-th unmanned surface vessel (USV) will proceed to the target j to be verified, while a value of 0 indicates that the USV will not proceed to the target j. The objective function coefficient is a. ij Let x be the verification cost for the i-th unmanned surface vessel (USV) unit to travel from its current location to the target j to be verified. ij When it involves secondary verification and the revisit distance constraint is satisfied, the cost is: a ij =max(t) now +dis ij / v i ,t last +Δt j )-t now Otherwise, the cost is a. ij =max(t) now +(2Δd j -dis ij ) / v i ,t last +Δt j )-t now When x ij If it does not involve secondary verification, then the cost is: a ij =dis ij / v i , where: dis ij v is the distance between unmanned surface vessel unit i and target j at the current moment. i Let t be the average velocity of unmanned surface vessel unit i. now Let t be the current time. last Δt is the time when unmanned surface vessel unit i last verified j. j Let Δd be the revisit time threshold for target j. j Let be the revisit distance threshold for target j.

[0044] The constraints of the objective function for the collaborative verification path planning include: Where: V notLet V be the set of targets that have not yet been fully verified, and b be the set of targets that have been fully verified. i For the single-transaction verification capability of unmanned surface vessel unit i, m j The heuristic threshold for target j is equal to the remaining verification base of target j + h, where h is a heuristic function determined based on the current cluster and the status of the remaining targets. Finally, the target allocation result for each unmanned surface vessel unit is obtained through planning, and the command is issued to the trajectory planning and tracking layer for execution.

[0045] Step 6) Switching the unmanned surface vessel (USV) unit's mission status: (e.g., ...) Figure 3 As shown, each unmanned surface vessel (USV) executes the next task according to the current task instructions. A finite state machine is used to represent the system based on the different tasks performed by each unit. Task = 8 indicates the USV is performing the area search phase; Task = 0 indicates the USV is currently in a state where no target has been assigned and target assignment is needed; Task = 1 indicates the USV has been assigned a target and is in the process of moving to verify that target; Task = -1 indicates the USV is out of contact. The switching conditions between each state are as follows: Start: When the scene is initialized, all USVs trigger this condition and enter the Task = 8 phase. A: When the USV has completed tracking its assigned search path, this switching condition is triggered, and the USV directly enters the Task = 0 state, waiting for the target assignment module to assign it a target. B: When the USV loses contact due to a malfunction, this condition is triggered, and the USV enters the out-of-contact state and stops moving. C: When the USV is assigned a target to be verified, this condition is triggered, and the USV enters the Task = 1 verification execution phase. D: When the unmanned surface vessel unit has completed the effective verification of the assigned target, this condition is triggered, and the unmanned surface vessel unit enters the task=0 stage, waiting to be assigned the next verification target.

[0046] Step 7) Determine the end of the collaborative search task: Determine whether the scenario task has ended. If the entire unfamiliar area has been searched and all targets have been verified, the task ends; otherwise, the task continues.

[0047] Through specific practical experiments, the unfamiliar area was set to a size of 6500*3300 meters, with an initial number of 6 unmanned surface vessel (USV) units and a speed of 4 m / s. Nine targets were set to be searched. In the initial phase, the USV swarm began its mission in the lower left corner of the search area. The simulation results are as follows: Figure 5-9 As shown.

[0048] like Figure 5The image shows a simulation diagram of the initial mission planning, where the dashed lines represent the channel allocation information generated by the initial mission planning. It can be seen that areas closer to the initial position are allocated more channels, which conforms to the principle that the maximum coverage distance of a single unmanned surface vessel (USV) unit should be as short as possible.

[0049] like Figure 6 The image shows a schematic diagram of the search process in an unfamiliar area.

[0050] like Figure 7 The diagram illustrates the state transition for area search and multi-target collaborative exploration tasks. When an unmanned surface vessel (USV) unit finishes its area search task, it immediately transitions to the multi-target collaborative exploration task, while USV units that have not completed their search continue to complete the search task in unfamiliar areas.

[0051] like Figure 8 The diagram illustrates real-time planning for multi-target collaborative search, planning the movement of each unmanned surface vessel (USV) to the next target point based on its current mission status. Figure 8 It can be seen that after completing the current exploration task, the unmanned surface vessel (USV) unit is basically assigned to the nearest node to perform the next exploration task. A small number of longer planned paths traverse certain targets to be explored. This is because the targets traversed have reached their exploration threshold and no longer require exploration.

[0052] like Figure 9 As shown, this illustrates the end of the mission. Once all targets within the target area have been searched, the mission is considered complete, and all unmanned surface vessel (USV) units autonomously return to their initial areas. In summary, the dynamic real-time planning method for collaborative unmanned surface vessel (USV) swarm search proposed in this invention can efficiently plan USV swarms and complete multi-target search tasks in unfamiliar areas with high efficiency through autonomous collaboration.

[0053] Compared with existing technologies, this method proposes an initial search area allocation planning method for collaborative exploration tasks in unfamiliar environments. The proposed method theoretically minimizes the total time for the unmanned swarm to complete the search of a designated unfamiliar area. Between the search and collaborative verification phases, a task transfer strategy is designed for the unmanned system swarm from the search phase to the verification phase, enabling seamless connection between the two processes and improving the exploration efficiency of the unmanned swarm. In the multi-target collaborative exploration phase, the collaborative exploration task is decomposed into several path planning sub-tasks. Based on the task status and location of each unmanned surface vessel (USV), a local swarm path planning method is adopted, which significantly improves the solution speed compared to the global path planning method, with a task replanning time of <1 second, meeting the real-time requirements of swarm path planning. This method can autonomously arrange the USVs and target points participating in the next planning step in real time based on the current task status of each USV. Therefore, it can mitigate the impact of environmental uncertainties on the task allocation of USVs; that is, if any USV has not completed its current task, it will not participate in the next planning step.

[0054] The above-described specific implementations can be partially adjusted by those skilled in the art in different ways without departing from the principles and purpose of the present invention. The scope of protection of the present invention is defined by the claims and is not limited to the above-described specific implementations. All implementation schemes within the scope of the claims are bound by the present invention.

Claims

1. A dynamic real-time planning method for collaborative search scenarios, characterized in that, include: Step 1) Scene initialization: Generate the initial task area and scene elements, and set the initial parameters for obstacles, targets to be investigated, unmanned swarms, and each unmanned surface vessel unit; Step 2) Allocate and generate search paths: According to the initial positions of the unmanned cluster and the boundary size of the to-be-explored area, generate the initial search task areas for each unmanned boat unit through the task allocation planning method, specifically including: checking the relationship between the total width in the X direction of the sub-area and the detection diameter of the unmanned boat unit. When the sum of the detection diameters of the unmanned boat units is greater than the total width in the X direction of the sub-area, it means that only some unmanned boat units are needed to complete the area search task. Then arrange the first m (m < nk) unmanned boat units to participate in the mapping, and directly allocate the verification work to the remaining nk - m unmanned boat units. When the range of sub-area k is large and the planned unmanned boat units need to complete the area search work by turning back, the division of the exploration area of the unmanned boat unit needs to consider the total sailing distance of the unmanned boat unit. The goal is to minimize the search time for all unfamiliar areas, that is, to satisfy the min-max mixed integer linear programming problem: , solve to obtain the value of x im , that is, obtain the number of channels allocated to each unmanned boat unit, and its constraints include: i) The total travel distance of unmanned surface vessel (USV) unit i includes the total length of the Y-direction channel covered by USV unit i and the travel distance in the X-direction required to reach the target channel, specifically: ; ii) Each waterway must be covered by one unmanned surface vessel (USV) unit, specifically: ; iii) The total width of the channel covered by the unmanned surface vessel unit should be greater than the width of the sub-region in the X direction, specifically: ; iv) When an unmanned surface vessel unit participates in the search and exploration of unfamiliar areas, the Boolean variable z i The value equals 1, meaning that unmanned surface vessel unit i has been assigned a search task, specifically: ; v) The total number of unmanned surface vessel (USV) units participating in the search mission within a sub-region shall not exceed the total number of USV units allocated to that region, specifically: ; vi) Only one unmanned surface vessel (USV) unit is planned to complete the search for each waterway, specifically: Where: the set of cluster units within a sub-region The set of waterway numbers D is the detection diameter of unmanned surface vessel (USV) unit i, i.e., the channel within the sub-region. The maximum number of channels that USV unit i can navigate without turning back in a single trip along the Y direction of the map does not exceed the total width divided by the minimum detection diameter of the USV unit. K is the set of sub-regions, x im Let x be a Boolean variable. When unmanned surface vessel unit i is assigned to the m-th channel, then x... im =1, the objective function means that we want the maximum travel distance of each unmanned surface vessel (USV) unit in the sub-region to be minimized. Since we need to ensure that the target to be explored is covered to the maximum extent, the search process often requires that the speed of the USV units cannot be too fast. When the maximum allowable speed of all USV units in the cluster is the same, it is equivalent to finding the shortest maximum total time and finding the shortest maximum total distance. To ensure that the search area boundary is covered without any blind spots, the route furthest from the initial position is planned to make an extra turnaround; once any unmanned surface vessel unit completes the search of its assigned area, it will directly enter the verification phase of the target to be searched; Step 3) Environment map update: Add the locations of targets and obstacles detected during the search to the map and update the environment map information; Step 4) Target Status Update: After the unmanned surface vessel unit has successfully verified the target, it updates the target verification information. Step 5) Plan and assign the next exploration target: Based on the information of available unmanned surface vessel (USV) units that have not yet been assigned a target, as well as the environmental map and target status information, establish a collaborative verification path planning objective function and calculate the target assignment result for each USV unit; based on the position of the USV cluster units to be assigned and the position of the target to be verified at any given time, use a planning algorithm to assign available USV units to the next target point; Step 6) Unmanned surface vessel (USV) unit mission state switching: During the planning process, the USV unit may be in an idle, mission execution, or disconnected state. Define the mission state and state transition relationship. Step 7) Determine whether the collaborative search task is complete.

2. The dynamic real-time planning method for cooperative search scenarios according to claim 1, characterized in that, Step 1 specifically includes: determining the number of unmanned surface vessel (USV) units to be allocated to each sub-region; when the number of USV units allocated to the left region is N1 and the number of USV units allocated to the right region is N2, the calculation method is as follows: ,in: Let X be the position of the center point of the unmanned cluster on the x-axis in the Cartesian coordinate system. , These are the x-coordinates of the lower left and lower right corners of the area to be explored, respectively. To round up; N is the total number of unmanned surface vessel (USV) units in the cluster, N1 is the number of USVs assigned to the left side of the starting area, and N2 is the number of USVs assigned to the right side of the starting area, and satisfies N=N1+N2.

3. The dynamic real-time planning method for cooperative search scenarios according to claim 1, characterized in that, Step 3 specifically includes: updating the environmental information based on the search results of each unmanned surface vessel (USV) unit in step 2), specifically: when any USV unit detects a previously undiscovered target or ordinary obstacle, the target or obstacle is updated to the map environmental information. The environmental information includes: the state of the target to be investigated and the location information of obstacles.

4. The dynamic real-time planning method for cooperative search scenarios according to claim 1, characterized in that, The constraints of the collaborative verification dynamic real-time path planning include: a) Revisit time constraint during second verification by unmanned surface vessel (USV) unit: When any USV unit verifies any target for the second time, the time interval between the two verifications must be greater than a certain threshold in order to be considered a valid verification. b) Revisit distance constraint during secondary verification by unmanned surface vessel (USV) units: When any USV unit verifies any target for the second time, it must first leave the target beyond a certain distance threshold before proceeding to the target for verification.

5. The dynamic real-time planning method for cooperative search scenarios according to claim 1, characterized in that, The objective function for the collaborative verification path planning is: , where: optimization variables The integer represents the value of 1, indicating that the i-th unmanned surface vessel (USV) will proceed to the target j to be verified, while a value of 0 indicates that the USV will not proceed to the target j. The objective function coefficients are... Let be the verification cost for the i-th unmanned surface vessel (USV) unit to travel from its current location to the target j to be verified. When it involves secondary verification and the revisit distance constraint is satisfied, the cost is: Otherwise the cost is ,when If it does not involve secondary verification, then the cost is: ,in: Let i be the distance between unmanned surface vessel unit i and target j at the current moment. Let i be the average velocity of the unmanned surface vessel unit. The current time. The last time unmanned surface vessel unit i verified j, Let be the revisit time threshold for target j. Let be the revisit distance threshold for target j; The constraints of the objective function for the collaborative verification path planning include: , , , ,in: This is the set of targets that have not yet been fully verified. The set of targets that have been verified. For the single-transaction verification capability of unmanned surface vessel unit i, The heuristic threshold for target j is equal to the remaining verification base of target j + h, where h is a heuristic function determined based on the current cluster and the status of the remaining targets. Finally, the target allocation result for each unmanned surface vessel unit is obtained through planning, and the command is issued to the trajectory planning and tracking layer for execution.

6. The dynamic real-time planning method for cooperative search scenarios according to claim 1, characterized in that, Step 6 specifically includes: each unmanned surface vessel (USV) unit executes the next task according to the current task instruction. Depending on the task executed by each unit, a finite state machine is used to implement the system state, where task=8 indicates the USV unit is performing area search; task=0 indicates the USV unit is currently in a state where no target has been assigned and target assignment is needed; task=1 indicates the USV unit has been assigned a target and is in the process of moving to that target for verification; task=-1 indicates the USV unit is out of contact. The switching conditions between each state are: Start: When scene initialization is performed, all USV units trigger this condition and enter task= Phase 8; A: When the unmanned surface vessel (USV) has completed tracking its assigned search path, the switching condition is triggered, and the USV directly enters task=0 state, waiting for the target allocation module to assign it a target; B: When the USV loses contact due to a malfunction, the condition is triggered, and the USV enters a disconnected state and stops moving; C: When the USV is assigned a target to be verified, the condition is triggered, and the USV enters the verification execution phase of task=1; D: When the USV has completed effective verification of the assigned target, the condition is triggered, and the USV enters the task=0 phase, waiting to be assigned the next verification target.

7. A dynamic real-time planning system for implementing the dynamic real-time planning method for cooperative search scenarios as described in any one of claims 1-6, characterized in that, include: The system comprises an unmanned surface vessel (USV) unit simulation module, a system configuration module, a cluster task allocation module, and a cluster planning module. Specifically: the USV unit simulation module models the USV unit based on its physical dimensions, speed, acceleration, and turning angular velocity, and sets the static and dynamic model parameters of the USV unit; the system configuration module configures environmental parameters and USV unit parameters; and the cluster task allocation module, based on an initial task allocation algorithm, obtains the initial exploration area allocation results for each USV unit according to the environmental parameters and USV unit parameters. The cluster planning module performs real-time task replanning based on the target information obtained from the exploration, and obtains the next exploration path planning results for each unmanned surface vessel unit.

Images