Unmanned submersible cluster uncontrolled self-running task execution system and method

Abstract

The invention discloses an unmanned submersible cluster uncontrolled self-running task execution system and method, and the system comprises a plurality of AUVs carrying resources, and each AUV carrying resources is provided with an edge calculation module, a multi-mode sensing module, a communication module and a resource configuration module. Each AUV carrying resources corresponds to a group of target identification information, the target identification information comprises a target frequency and a target confidence coefficient, and the multi-modal sensing module is configured with a passive sonar array, a multi-beam sonar MBS, an inertial measurement unit IMU and a depth sensor and is used for obtaining sensing data and sending the sensing data to the AUV carrying resources. The perception data comprises a target voiceprint feature vector, a target relative distance, a speed vector and an environment parameter. According to the uncontrolled self-running task execution system and method for the unmanned submersible cluster, the adaptability of the cluster to a complex underwater environment is improved, and optimal resource configuration and efficient task execution are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater unmanned equipment cluster control, and in particular to a system and method for an unmanned submersible cluster to self-execute tasks without control. Background Art

[0002] Autonomous underwater vehicle (AUV) swarms have significant application value in military underwater attack missions, but the complexity of the underwater environment poses a significant challenge to their coordinated capabilities. Existing technologies have the following limitations: centralized control systems rely on real-time commands from ground stations or motherships, making them poorly adaptable to the high-latency, low-bandwidth (<5 Kbps) underwater environment; communication interruptions can lead to mission failure. Existing AUV systems (such as the Sailfish-324) focus on detection and monitoring, lacking attack capabilities and dynamic mission management capabilities, making them incapable of handling high-level confrontation scenarios. Furthermore, swarms struggle to adjust their strategies in real time when faced with changes in target maneuverability or the incapacitation of individual AUVs, resulting in insufficient mission robustness. Therefore, an AUV swarm system that achieves full autonomy after launch and adapts to low-bandwidth communications and dynamic confrontation environments is urgently needed. Summary of the Invention

[0003] In order to solve the technical problems existing in the background technology, the present invention proposes a system and method for the unmanned underwater vehicle cluster to autonomously execute tasks without control.

[0004] The present invention proposes an unmanned underwater vehicle cluster uncontrolled self-operation execution task system, comprising: multiple autonomous underwater vehicles (AUVs) carrying resources, each of which is equipped with an edge computing module, a multimodal perception module, a communication module, and a resource configuration module. Each AUV carrying resources corresponds to a set of target identification information, and the target identification information includes target frequency and target confidence. The multimodal perception module is equipped with a passive sonar array, a multi-beam sonar MBS, an inertial measurement unit (IMU), and a depth sensor to obtain perception data, including target voiceprint feature vectors, target relative distance, velocity vector, and environmental parameters. The edge computing module is used to receive and process the sensory data, and then output the task allocation vector, attack instruction sequence and trajectory planning matrix based on the processed sensory data and target identification information to drive the resource configuration module to execute the corresponding task; The communication module is used to periodically obtain the state vector of the current autonomous underwater vehicle (AUV) carrying resources and the state matrix of the autonomous underwater vehicle (AUV) carrying resources in the neighboring area. After signal demodulation, CRC check and data alignment, it generates a global state diagram of the cluster and synchronizes it to the edge computing module to support decision-making and planning; The resource configuration module is used to record the equipment type and quantity of the current autonomous underwater vehicle (AUV) carrying resources, and use the corresponding equipment according to the driving instructions of the edge computing module.

[0005] Preferably, the edge computing module includes: The target recognition submodule is used to map the voiceprint feature vector to the target type probability distribution T based on the pre-trained MobileNet model to output the confidence matrix C; The trajectory prediction submodule is used to process the relative distance and velocity vectors using the Kalman filter algorithm to generate a three-dimensional position prediction matrix P at the next time t; The decision-making planning submodule is used to solve the optimal task allocation vector A based on the confidence matrix C, relative distance, velocity vector and environmental parameters through the particle swarm optimization algorithm.

[0006] Preferably, the state vector of the current resource-carrying autonomous underwater vehicle AUV includes: an AUV identifier ID, a position vector, an energy state, an attack resource vector, and a task allocation identifier, wherein each resource-carrying autonomous underwater vehicle AUV corresponds to a unique AUV identifier ID.

[0007] Preferably, the communication module adopts a broadcast time division multiple access protocol, broadcasts status information in a 32-byte standard PDU or a 64-byte extended PDU format, and the bandwidth is controlled at <5Kbps.

[0008] The present invention proposes a method for an unmanned submersible swarm to autonomously execute a task without control, which is applied to any of the above-mentioned unmanned submersible swarm autonomously execute a task system, and the method comprises the following steps: S1: Obtain target detection data and status flags corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and control multiple AUVs to broadcast their own target detection data to other AUVs in the cluster for the first time. The target detection data includes target position, target speed, target shape, and target frequency. S2: synchronously correcting target detection data corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and broadcasting the corrected target detection data to multiple autonomous underwater vehicles (AUVs) carrying resources for a second time; S3: Calculating utility scores for the multiple autonomous underwater vehicles (AUVs) carrying resources one by one, and grouping the multiple autonomous underwater vehicles (AUVs) carrying resources according to the utility scores and a preset grouping strategy to obtain multiple preset groups, each group corresponding to one execution task; S4: Control the autonomous underwater vehicles (AUVs) carrying resources in each group to complete the mission according to the preset multiple groups.

[0009] Preferably, the calculation formula for the utility score is: ; in, For distance weight, a fixed parameter can be preset to emphasize the importance of being close to the target; For the action energy weight, fixed parameters can be preset to ensure continuous combat capability; For attack resource weight, a fixed parameter can be preset to represent the sufficiency of ammunition; is the attack resource type weight; is the normalized ammunition vector; is the ammunition item index, a fixed parameter that controls the degree of coupling between ammunition types; is the ratio of the target distance to the maximum distance of the surviving AUV to the same target, It is the ratio of the action energy percentage to the maximum energy; Score the utility of the i-th autonomous underwater vehicle (AUV) carrying resources.

[0010] Preferably, the preset grouping strategy specifically includes: Rank multiple autonomous underwater vehicles (AUVs) carrying resources in descending order of utility scores; Preset scoring thresholds corresponding to the plurality of groups are obtained, and the sorted plurality of autonomous underwater vehicles (AUVs) carrying resources are divided into groups according to the preset scoring thresholds to obtain a plurality of groups, wherein the groups include but are not limited to an attack group and a feint group.

[0011] Preferably, it also includes: calculating the cluster resource value corresponding to the unmanned submersible cluster in real time , when the cluster resource value When the value is less than the preset termination threshold, the task execution is terminated.

[0012] Preferably, the cluster resource value The specific calculation process is: ; in, Autonomous underwater vehicles (AUVs) that carry resources are available; The total amount of resources carried by the autonomous underwater vehicle AUV.

[0013] Preferably, the AUV status flags include but are not limited to idle, attack, feint, and terminate.

[0014] The present invention proposes a system and method for autonomously executing tasks in a swarm of unmanned underwater vehicles (AUVs). Efficient underwater task execution is achieved through distributed sensing, dynamic grouping, and collaborative control. The method includes: acquiring target detection data from multiple AUVs and broadcasting it initially; performing spatiotemporal consistency corrections on the detection data and broadcasting it a second time; calculating utility scores based on distance, energy, and resource status, and grouping them according to a preset strategy; and controlling each group of AUVs to perform the corresponding task. Through multiple rounds of data synchronization and a dynamic grouping strategy, the present invention significantly improves the swarm's adaptability to complex underwater environments, achieving optimized resource allocation and efficient task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the system architecture of a system for unmanned underwater vehicle clusters to autonomously execute tasks without control, as proposed by the present invention; Figure 2 This is a schematic diagram of the workflow of a method for unmanned underwater vehicle clusters to autonomously execute tasks without control, as proposed by the present invention. DETAILED DESCRIPTION

[0016] Reference Figure 1 and Figure 2 The present invention proposes an unmanned underwater vehicle cluster uncontrolled self-operation execution task system, comprising: a plurality of resource-carrying autonomous underwater vehicles (AUVs), each of which is equipped with an edge computing module, a multimodal perception module, a communication module, and a resource configuration module. Each resource-carrying autonomous underwater vehicle (AUV) corresponds to a set of target identification information, and the target identification information includes target frequency and target confidence. The multimodal perception module is equipped with a passive sonar array, a multi-beam sonar MBS, an inertial measurement unit (IMU), and a depth sensor to obtain perception data, including target voiceprint feature vectors, target relative distance, velocity vector, and environmental parameters. The edge computing module is used to receive and process the sensory data, and then output the task allocation vector, attack instruction sequence and trajectory planning matrix based on the processed sensory data and target identification information to drive the resource configuration module to execute the corresponding task; The communication module is used to periodically obtain the state vector of the current autonomous underwater vehicle (AUV) carrying resources and the state matrix of the autonomous underwater vehicle (AUV) carrying resources in the neighboring area. After signal demodulation, CRC check and data alignment, it generates a global state diagram of the cluster and synchronizes it to the edge computing module to support decision-making and planning; The resource configuration module is used to record the equipment type and quantity of the current autonomous underwater vehicle (AUV) carrying resources, and use the corresponding equipment according to the driving instructions of the edge computing module.

[0017] In this embodiment, the edge computing module includes: The target recognition submodule is used to map the voiceprint feature vector to the target type probability distribution T based on the pre-trained MobileNet model to output the confidence matrix C; The trajectory prediction submodule is used to process the relative distance and velocity vectors using the Kalman filter algorithm to generate a three-dimensional position prediction matrix P at the next time t; The decision-making planning submodule is used to solve the optimal task allocation vector A based on the confidence matrix C, relative distance, velocity vector and environmental parameters through the particle swarm optimization algorithm.

[0018] Specifically, the target state is evaluated using the existing MBS image and acoustic signals to generate a comprehensive confidence C.

[0019] In this embodiment, the state vector of the current resource-carrying autonomous underwater vehicle AUV includes: an AUV identifier ID, a position vector, an energy state, an attack resource vector, and a task allocation identifier, wherein each resource-carrying autonomous underwater vehicle AUV corresponds to a unique AUV identifier ID.

[0020] In this embodiment, the communication module adopts a broadcast time division multiple access protocol to broadcast status information in a 32-byte standard PDU or a 64-byte extended PDU format, and the bandwidth is controlled at <5Kbps.

[0021] Specifically, the contents of the PDU format are shown in the following table:

[0022] Field Storage (B) describe Data Source AUV ID 1 Unique identifier (0-255) System initialization allocation, underwater acoustic communication standard uses broadcast mode, no destination address is required Target distance D 4 Sonar ranging distance (meters, single-precision floating point) Existing sonar systems (active sonar ranging) Action Energy P 1 Action energy percentage (0-100) AUV Energy Management System Attack Resource Vectors 6 For example, torpedoes, decoys, and explosive bombs (2 bytes each) AUV Attack Resource Management System <![CDATA[Attack resource stock a i,k > 2 Attack resource storage value, k is the resource type Initialize full load, calculate after attack <![CDATA[Utility score U i > 4 Utility score (floating point number, 0-1) Local calculation, single-precision floating point Target Information 8 Compressed sonar characteristics (such as main frequency) Existing sonar system (feature extraction) Status Flags 1 Status (0 Idle, 1 Attack, 2 Feint, 3 Terminate) AUV state machine CRC check 2 Data integrity check CRC32 algorithm Target ID 2 Target identifier (0-65535) Sonar feature clustering Confidence 4 Target distance confidence (0-1) Sonar signal strength Group Confirmation 2 Group confirmation (1-byte target ID, 0-255; 1-byte role, 1 attack, 2 feints) Group calculation

[0023] PDU size: D = 32 bytes.

[0024] Specifically, before launch, the AUV cluster completes the following initialization through the ground station: 1) Loading Target Information: For mobile targets (such as enemy ships), load sonar signatures (such as propeller noise spectrum, 20-100 Hz). Specify the approximate target area (5 km radius), without precise coordinates. Target information is stored in each AUV's edge computing module, occupying less than 1 KB of memory.

[0025] 2) Each AUV carries a variety of attack resources and their initial quantity, for example: Torpedo (T): High power, used for attacking ships, each occupies 5 storage units, initially n (storage value 5n). Decoy (D): Interference defense, each occupies 2 storage units, initially m (storage value 2m). Precision Bomb (P): High-precision strike, each occupies 4 storage units, initially k (storage value 4k). i,k: The current storage value of attack resource type k (k∈{T, D, P, …}) of AUV i (unit: storage unit). i,kmax : The maximum storage value of attack resource type k. a i : The total attack resource storage value of AUV i, defined as a i =∑ k∈{T，D，P，……} a i,k , subject to the total capacity a max =∑ k∈{T,D,P，……} a i,kmax constraint.

[0026] 3) Configure the communication protocol: Time Division Multiple Access (TDMA) acoustic communication, frame length T frame =10 seconds. The number of time slots N is equal to the number of AUVs, such as N=20, the time slot length T slot =T frame / N = 10 / 20 = 0.5 seconds. PDU size: 24 bytes, bandwidth: 192 bps.

[0027] 4) Synchronous clock: Each AUV is equipped with a high-precision quartz clock with a drift of <1ms / h, and the frame start time is synchronized through the ground station.

[0028] 5) Initialize edge computing: Deploy the MobileNet model, which has approximately 4.2MB of parameters and takes up less than 10MB of storage, for target recognition. Configure the A* path planning algorithm and a terrain map resolution of 50 meters per pixel. Load the Kalman filter algorithm for target trajectory prediction.

[0029] 6) Hardware Check: Verify that the passive sonar (10-100 Hz), MBS (20 kHz, 0.1 meter resolution), IMU (0.01° accuracy), and depth sensor (0.1 meter accuracy) are functioning properly. Initial operational resources are 100%.

[0030] Specifically, the AUV uses passive sonar to monitor the target's main frequency and active sonar short pulse ranging to calculate the target distance d i =t×1500 / 2 (t is the round trip time), speed v i =f d ×1500 / (2×20000)(f d is the Doppler frequency shift); MBS generates a 3D acoustic image, which is then input into the MobileNet model after Gaussian filtering and noise reduction, and the target type and confidence C are output. target , if C target >0.8, confirm the target ID.

[0031] It should be noted that when there are differences in the ID recognition of the same target by multiple AUVs, the following steps are performed: Extract the target main frequency f detected by each AUV jand confidence C d,j ; K-means clustering (threshold 10Hz) was used. When the main frequency difference was <10Hz, C was selected. d,j The highest ID is used as the unified target ID consensus ; Broadcast the corrected target ID to ensure cluster data consistency.

[0032] Specifically, ; For target ID and main frequency f j (PDU target information field) performs K-means clustering based on C d,j Weighted. If the main frequency difference is less than 10 Hz, it is unified into a single target ID (take C d, j highest ID), get the consistent target ID consensus Otherwise, the subgroup is divided and a new ID is assigned (0-255, PDU Group Identification field). The PDU Target ID field (16-bit integer) is updated and the correction result is broadcast.

[0033] Reference Figure 1 and Figure 2 The present invention proposes an unmanned submersible swarm uncontrolled self-operation task execution system, which is applied to any of the unmanned submersible swarm uncontrolled self-operation task execution systems described above. The method includes the following steps: S1: Obtain target detection data and status flags corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and control multiple AUVs to broadcast their own target detection data to other AUVs in the cluster for the first time. The target detection data includes target position, target speed, target shape, and target frequency. S2: synchronously correcting target detection data corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and broadcasting the corrected target detection data to multiple autonomous underwater vehicles (AUVs) carrying resources for a second time; S3: Calculating utility scores for the multiple autonomous underwater vehicles (AUVs) carrying resources one by one, and grouping the multiple autonomous underwater vehicles (AUVs) carrying resources according to the utility scores and a preset grouping strategy to obtain multiple preset groups, each group corresponding to one execution task; S4: Control the autonomous underwater vehicles (AUVs) carrying resources in each group to complete the mission according to the preset multiple groups.

[0034] In this embodiment, the calculation formula for the utility score is: ; in, For distance weight, a fixed parameter can be preset to emphasize the importance of being close to the target; For the action energy weight, fixed parameters can be preset to ensure continuous combat capability; For attack resource weight, a fixed parameter can be preset to represent the sufficiency of ammunition; is the attack resource type weight; is the normalized ammunition vector; is the ammunition item index, a fixed parameter that controls the degree of coupling between ammunition types; is the ratio of the target distance to the maximum distance of the surviving AUV to the same target, It is the ratio of the action energy percentage to the maximum energy; is the utility score of the i-th autonomous underwater vehicle AUV carrying resources, ,like, =0.4, =0.3, =0.3.

[0035] Specifically, ; in, Normalized ammunition vector, based on a i,k (PDU attack resource type vector) a i,kmax (Initialization configuration). Normalize the ammunition amount to facilitate vector operations. i,k The stock of attack resource type k (e.g., T: torpedo, D: decoy, P: explosive bomb), a i,kmax The maximum storage value of the attack resource type k, sourced from the PDU attack resource vector field.

[0036] In this embodiment, the preset grouping strategy specifically includes: Rank multiple autonomous underwater vehicles (AUVs) carrying resources in descending order of utility scores; Preset scoring thresholds corresponding to the plurality of groups are obtained, and the sorted plurality of autonomous underwater vehicles (AUVs) carrying resources are divided into groups according to the preset scoring thresholds to obtain a plurality of groups, including but not limited to an attack group and a feint group.

[0037] In this embodiment, it also includes: calculating the cluster resource value corresponding to the unmanned underwater vehicle cluster in real time , when the cluster resource value When the value is less than the preset termination threshold, the task execution is terminated.

[0038] In this embodiment, after completing an attack mission, the surviving AUVs calculate their remaining ammunition, compare it to the attack requirement, and regroup based on the remaining ammunition and utility score, continuing to attack undestroyed targets. This grouping process repeats. Based on the system's preset AUV state parameters, the system analyzes its own operational resources and total resources, deriving new utility scores and regrouping. This continues until the sonar determines the mission is complete or the surviving AUVs' UI values ​​are insufficient to complete the mission, at which point they evacuate according to the pre-set plan.

[0039] In this embodiment, the cluster resource value The specific calculation process is: ; in, Autonomous underwater vehicles (AUVs) that carry resources are available; The total amount of resources carried by the autonomous underwater vehicle AUV.

[0040] In this embodiment, the AUV status flags include but are not limited to idle, attack, feint, and terminate.

[0041] Example 1: Single Target: 12 AUVs, target (ID=1), AUV1 measured d1=2000m, v1=0.375m / s, C d,1 =0.9. Number of time slots N = 12, time slot duration Tslot = 12 / 12 = 1 second. Bandwidth: R AUV =32 / 12=2.667B / s=21.33b / s, R total =12×21.33=255.96bps<5Kb / s, meeting underwater communication requirements.

[0042] Single target: 12 AUVs, target ID = 1. AUV1: U1 = 0.78, AUV2: U2 = 0.73, AUV3 not broadcast (excluded), AUV4: U4 = 0.75, AUV5: U5 = 0.70, and other AUVs are similar. AUV1 collects PDUs, extracts and sorts them, and obtains [(AUV1, 0.78), (AUV2, 0.73), (AUV4, 0.75), (AUV5, 0.70), ...].

[0043] Total resourcesa i =a max , resource vector, maximum storage value a i,kmax System initialization configuration and normalization of attack resources.

[0044] Example 2: Scenario 2 (multi-target): 8 AUVs, target A (ID=1), target B (ID=2), AUV1 measures target A’s d1=2500m, C d,1 = 0.9. Number of time slots N = 8, time slot duration Tslot = 12 / 8 = 1.5 seconds. Total bandwidth Rtotal = 8 × 21.33 = 170.64 b / s < 5 Kb / s, meeting underwater communication requirements.

[0045] Multiple targets: 8 AUVs, target A (ID=1) and target B (ID=2).

[0046] AUV1: U1, A=0.755, U 1,B =0.72; AUV2: U2,A=0.746,U 2,B =0.71; AUV3:U 3,A =0.74, target B was not detected; AUV4:U 4,A =0.73, U 4,B =0.70; AUV5:a i,D =0 excluded.

[0047] AUV6: Target A, U not detected 6,B =0.73; AUV7: Target A, U not detected 7,B =0.70; AUV8:U 8,A =0.72, U 8,B =0.69; AUV1 generates two tables, and other AUVs generate their own priority tables in a similar manner: Objective A: PriorityTable 1,A =[(AUV1,0.755,ID=1),(AUV2,0.746,ID=1),(AUV3,0.74,ID=1),(AUV4,0.73,ID=1),(AUV8,0.72,ID=1),(AUV6,0,ID=1),(AUV7,0,ID=1),…].

[0048] Objective B: PriorityTable 1,B=[(AUV6,0.73,ID=2),(AUV1,0.72,ID=2),(AUV2,0.71,ID=2),(AUV4,0.70,ID=2),(AUV7,0.70,ID=2),(AUV8,0.69,ID=2),(AUV3,0,ID=2),…].

[0049] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A system for unmanned underwater vehicle swarms to autonomously execute tasks without control, characterized in that: include: Multiple autonomous underwater vehicles (AUVs) carrying resources, each of which is equipped with an edge computing module, a multimodal perception module, a communication module, and a resource configuration module. Each of which corresponds to a set of target identification information, including target frequency and target confidence. The multimodal perception module is equipped with a passive sonar array, a multi-beam sonar MBS, an inertial measurement unit (IMU), and a depth sensor to obtain perception data, including target voiceprint feature vectors, target relative distance, velocity vector, and environmental parameters. The edge computing module is used to receive and process the sensory data, and then output the task allocation vector, attack instruction sequence and trajectory planning matrix based on the processed sensory data and target identification information to drive the resource configuration module to execute the corresponding task; The communication module is used to periodically obtain the state vector of the current autonomous underwater vehicle (AUV) carrying resources and the state matrix of the autonomous underwater vehicle (AUV) carrying resources in the neighboring area. After signal demodulation, CRC check and data alignment, it generates a global state diagram of the cluster and synchronizes it to the edge computing module to support decision-making and planning; The resource configuration module is used to record the equipment type and quantity of the current autonomous underwater vehicle (AUV) carrying resources, and use the corresponding equipment according to the driving instructions of the edge computing module.

2. The unmanned underwater vehicle swarm uncontrolled self-operation task execution system according to claim 1 is characterized in that: The edge computing module includes: The target recognition submodule is used to map the voiceprint feature vector to the target type probability distribution T based on the pre-trained MobileNet model to output the confidence matrix C; The trajectory prediction submodule is used to process the relative distance and velocity vectors using the Kalman filter algorithm to generate a three-dimensional position prediction matrix P at the next time t; The decision-making planning submodule is used to solve the optimal task allocation vector A through the particle swarm optimization algorithm based on the confidence matrix C, relative distance, velocity vector and environmental parameters.

3. The unmanned underwater vehicle swarm uncontrolled self-operation task execution system according to claim 1 is characterized in that: The state vector of the current resource-carrying autonomous underwater vehicle AUV includes: an AUV identifier ID, a position vector, an energy state, an attack resource vector, and a task allocation identifier, wherein each resource-carrying autonomous underwater vehicle AUV corresponds to a unique AUV identifier ID.

4. The unmanned underwater vehicle swarm uncontrolled self-operation task execution system according to claim 1 is characterized in that: The communication module adopts broadcast time division multiple access protocol, broadcasts status information in 32-byte standard PDU or 64-byte extended PDU format, and the bandwidth is controlled at <5Kbps.

5. A method for unmanned underwater vehicle clusters to autonomously execute tasks without control, characterized in that: Applied to the unmanned underwater vehicle swarm uncontrolled self-operation task execution system according to any one of claims 1 to 4, the method comprises the following steps: S1: Obtain target detection data and status flags corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and control multiple AUVs to broadcast their own target detection data to other AUVs in the cluster for the first time. The target detection data includes target position, target speed, target shape, and target frequency. S2: synchronously correcting target detection data corresponding to multiple autonomous underwater vehicles (AUVs) carrying resources, and broadcasting the corrected target detection data to multiple autonomous underwater vehicles (AUVs) carrying resources for a second time; S3: Calculating utility scores for the multiple autonomous underwater vehicles (AUVs) carrying resources one by one, and grouping the multiple autonomous underwater vehicles (AUVs) carrying resources according to the utility scores and a preset grouping strategy to obtain multiple preset groups, each group corresponding to one execution task; S4: Control the autonomous underwater vehicles (AUVs) carrying resources in each group to complete the mission according to the preset multiple groups.

6. The method for executing tasks by unmanned underwater vehicle swarms without control and autonomously according to claim 5, characterized in that: The utility score is calculated as follows: ； in, For distance weight, a fixed parameter can be preset to emphasize the importance of being close to the target; For the action energy weight, fixed parameters can be preset to ensure continuous combat capability; For attack resource weight, a fixed parameter can be preset to represent the sufficiency of ammunition; is the attack resource type weight; is the normalized ammunition vector; is the ammunition item index, a fixed parameter that controls the degree of coupling between ammunition types; is the ratio of the target distance to the maximum distance of the surviving AUV to the same target, It is the ratio of the action energy percentage to the maximum energy; Score the utility of the i-th autonomous underwater vehicle (AUV) carrying resources.

7. The method for executing tasks by unmanned underwater vehicle swarms without control and autonomously according to claim 5, characterized in that: The preset grouping strategy specifically includes: Rank multiple autonomous underwater vehicles (AUVs) carrying resources in descending order of utility scores; Preset scoring thresholds corresponding to the plurality of groups are obtained, and the sorted plurality of autonomous underwater vehicles (AUVs) carrying resources are divided into groups according to the preset scoring thresholds to obtain a plurality of groups, wherein the groups include but are not limited to an attack group and a feint group.

8. The method for executing tasks by unmanned underwater vehicle swarms without control and autonomously according to claim 5, characterized in that: Also includes: Calculate the cluster resource value corresponding to the unmanned underwater vehicle cluster in real time , when the cluster resource value When the value is less than the preset termination threshold, the task execution is terminated.

9. The method for executing tasks by unmanned underwater vehicle swarms without control and autonomously according to claim 8, characterized in that: The cluster resource value The specific calculation process is: ； in, Autonomous underwater vehicles (AUVs) that carry resources are available; The total amount of resources carried by the autonomous underwater vehicle AUV.

10. The method for executing tasks by unmanned underwater vehicle swarms without control and autonomously according to claim 5, characterized in that: The AUV status flags include but are not limited to idle, attack, feint, and terminate.

Images