Underwater Target Tracking Method, Device and Computer Equipment

Abstract

The present application relates to an underwater target tracking method, device, and computer device. The method includes: obtaining real-time information of each submersible, real-time information of the target object, and current environmental information; the real-time information includes current position information and current status information; determining the tracking trajectories of the submersibles according to the real-time information of each submersible, the real-time information of the target object, the current environmental information, and a target tracking planning model; and determining that the target tracking is successful when the distance between the real-time information of the target object and the real-time information of the submersible is not greater than a preset distance. Using this method can improve the accuracy of the actual tracking trajectories of the submersibles.

Description

Technical Field

[0001] The present application relates to the technical field of target tracking, and particularly to an underwater target tracking method, device and computer equipment. Background Art

[0002] With the development of marine biological detection technology, marine-related technologies such as underwater acoustic monitoring, seabed mapping, and underwater communication have achieved rapid development. Building a timely, accurate, and effective underwater moving target tracking network can better detect marine organisms and understand more characteristics of marine organisms.

[0003] Currently, traditional underwater moving target tracking methods use an unmanned underwater vehicle (AUV) for tracking. This method regards the AUV as a particle and predicts the tracking trajectory of the AUV to the target in real time. However, for traditional methods, the obtained tracking trajectory of the AUV is relatively complex and has a low error tolerance, resulting in a low accuracy of the optimal tracking trajectory of the underwater moving target. Summary of the Invention

[0004] Based on this, it is necessary to provide an underwater target tracking method, device and computer equipment for the above technical problems.

[0005] In a first aspect, the present application provides an underwater target tracking method. The method includes:

[0006] Obtain the real-time information of each submersible, the real-time information of the target, and the current environmental information; the real-time information includes the current position information and the current state information;

[0007] Determine the tracking trajectory of each submersible according to the real-time information of each submersible, the real-time information of the target, the current environmental information, and the target tracking planning model;

[0008] When the distance between the real-time information of the target and the real-time information of the submersible is not greater than a preset distance, it is determined that the target tracking is successful.

[0009] Optionally, the obtaining the real-time information of the target includes:

[0010] Obtain each detection information of the target through each submersible;

[0011] Obtain the real-time information of the target according to each detection information of the target and the underwater target detection model.

[0012] Optionally, the target tracking and planning model includes a tracking trajectory prediction model, an underwater dynamics model, and a trajectory optimization model. Determining the tracking trajectories of the submarines according to the real-time information of each submarine, the real-time information of the target, the current environmental information, and the target tracking and planning model includes:

[0013] For each submarine, determine the first predicted trajectory of the submarine according to the real-time information of the submarine, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model;

[0014] Determine the second predicted trajectories of the submarines according to the first predicted trajectories of the submarines, the trajectory optimization model, and the tracking trajectory prediction model;

[0015] Determine the tracking trajectories of the submarines according to the real-time information of each submarine, the second predicted trajectories of each submarine, and the underwater dynamics model.

[0016] Optionally, when the trajectory optimization model includes an obstacle avoidance model, determining the second predicted trajectories of the submarines according to the first predicted trajectories of the submarines, the trajectory optimization model, and the tracking trajectory prediction model includes:

[0017] For each submarine, determine the collision probability between the submarine and an adjacent submarine according to the first predicted trajectory of the submarine, the first predicted trajectory of the adjacent submarine, and the obstacle avoidance model;

[0018] When the collision probability is greater than a collision threshold, add the first predicted trajectory of the submarine to the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of the submarine according to the real-time information of the submarine, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model for each submarine;

[0019] When the collision probability is not greater than the collision threshold, mark the first predicted trajectory of the submarine as the second predicted trajectory of the submarine.

[0020] Optionally, when the trajectory optimization model includes a clustering consistency model, determining the second predicted trajectories of the submarines according to the first predicted trajectories of the submarines, the trajectory optimization model, and the tracking trajectory prediction model includes:

[0021] Determine the clustering dispersion value of each submarine and the average dispersion value of each submarine according to the first predicted trajectories of the submarines and the clustering consistency model;

[0022] When the discrete average value of each submersible is greater than the discrete threshold, include the first predicted trajectory of each submersible in the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of each submersible according to the real-time information of the submersible, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model for each submersible;

[0023] When the cluster discrete values of all submersibles are not greater than the discrete threshold, mark the first predicted trajectory of each submersible as the second trajectory of each submersible.

[0024] Optionally, the method further includes:

[0025] When the discrete average value of each submersible is not greater than the discrete threshold and there is a target cluster discrete value greater than the discrete threshold, include the first predicted trajectory of the submersible corresponding to each target cluster discrete value in the taboo list of the tracking trajectory prediction model;

[0026] According to the real-time information of the submersible corresponding to each target cluster discrete value, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectory of the submersible corresponding to each target cluster discrete value, and based on the re-determined first predicted trajectory of the submersible corresponding to each target cluster discrete value, return to execute the step of determining the cluster discrete value of each submersible and the discrete average value of each submersible according to the first predicted trajectory of each submersible and the cluster consistency model.

[0027] In a second aspect, the present application further provides an underwater target tracking device. The device includes:

[0028] An acquisition module, configured to acquire the real-time information of each submersible, the real-time information of the target, and the current environmental information; the real-time information includes the current position information and the current state information;

[0029] A first determination module, configured to determine the tracking trajectory of each submersible according to the real-time information of each submersible, the real-time information of the target, the current environmental information, and the target tracking planning model;

[0030] A second determination module, configured to determine that the target tracking is successful when the distance between the real-time information of the target and the real-time information of the submersible is not greater than a preset distance.

[0031] Optionally, the acquisition module is specifically configured to:

[0032] Obtain each detection information of the target through each submersible detection;

[0033] Based on the detection information of the target object and the underwater target detection model, obtain the real-time information of the target object.

[0034] Optionally, the target tracking and planning model includes a tracking trajectory prediction model, an underwater dynamics model, and a trajectory optimization model. The first determination module is specifically configured to:

[0035] For each submersible, determine the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model;

[0036] Determine the second predicted trajectories of the submersibles according to the first predicted trajectories of the submersibles, the trajectory optimization model, and the tracking trajectory prediction model;

[0037] Determine the tracking trajectories of the submersibles according to the real-time information of the submersibles, the second predicted trajectories of the submersibles, and the underwater dynamics model.

[0038] Optionally, when the trajectory optimization model includes an obstacle avoidance model, the first determination module is specifically configured to:

[0039] For each submersible, determine the collision probability between the submersible and the adjacent submersible according to the first predicted trajectory of the submersible, the first predicted trajectory of the adjacent submersible, and the obstacle avoidance model;

[0040] When the collision probability is greater than the collision threshold, include the first predicted trajectory of the submersible in the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model for each submersible;

[0041] When the collision probability is not greater than the collision threshold, mark the first predicted trajectory of the submersible as the second predicted trajectory of the submersible.

[0042] Optionally, when the trajectory optimization model includes a clustering consistency model, the first determination module is specifically configured to:

[0043] Determine the clustering discrete values of the submersibles and the discrete average values of the submersibles according to the first predicted trajectories of the submersibles and the clustering consistency model;

[0044] When the discrete average value of each submersible vehicle is greater than the discrete threshold, include the first predicted trajectory of each submersible vehicle in the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of each submersible vehicle according to the real-time information of the submersible vehicle, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model for each submersible vehicle;

[0045] When the cluster discrete values of all submersible vehicles are not greater than the discrete threshold, mark the first predicted trajectory of each submersible vehicle as the second trajectory of each submersible vehicle.

[0046] Optionally, in the case where the trajectory optimization model includes a cluster consistency model, the first determination module further includes:

[0047] When the discrete average value of each submersible vehicle is not greater than the discrete threshold and there is a target cluster discrete value greater than the discrete threshold, include the first predicted trajectory of the submersible vehicle corresponding to each target cluster discrete value in the taboo list of the tracking trajectory prediction model;

[0048] According to the real-time information of the submersible vehicle corresponding to each target cluster discrete value, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectory of the submersible vehicle corresponding to each target cluster discrete value, and based on the re-determined first predicted trajectory of the submersible vehicle corresponding to each target cluster discrete value, return to execute the steps of determining the cluster discrete value of each submersible vehicle and the discrete average value of each submersible vehicle according to the first predicted trajectory of each submersible vehicle and the cluster consistency model.

[0049] In a third aspect, the present application provides a computer device. The computer device includes a memory and a processor, and the memory stores a computer program. The computer program, when executed by the processor, is characterized in that the steps of the method according to any one of the first aspects are implemented.

[0050] In a fourth aspect, the present application provides a computer-readable storage medium. The storage medium includes a computer program stored thereon. The computer program, when executed by the processor, is characterized in that the steps of the method according to any one of the first aspects are implemented.

[0051] In a fifth aspect, the present application provides a computer program product. The computer program product includes a computer program. The computer program, when executed by the processor, is characterized in that the steps of the method according to any one of the first aspects are implemented.

[0052] The above-mentioned underwater target tracking method, device, and computer equipment obtain an underwater observation space through an underwater target detection model. The underwater observation space is a space composed of the observation information of each submersible underwater. The observation information of the submersible includes the real-time information of the submersible, the real-time information of the target object, and the current environmental information. The real-time information includes the current position information and the current state information. According to the real-time information of each submersible, the real-time information of the target object, the current environmental information, and a target tracking planning model, the tracking trajectories of each submersible are determined. When the distance between the real-time information of the target object and the real-time information of the submersible is not greater than a preset distance, it is determined that the target tracking is successful. By obtaining the real-time information of each submersible and the target object through the target detection model, considering the current environmental information, and using the target tracking planning model to plan the target tracking trajectory in real time, the target is tracked, thereby improving the accuracy of the actual tracking trajectory of the submersible. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is an application environment diagram of the underwater target tracking method in an embodiment;

[0054] Figure 2 It is a flowchart of the underwater target tracking method in an embodiment;

[0055] Figure 3 It is a flowchart of the steps for determining the tracking trajectories of each submersible in an embodiment;

[0056] Figure 4 It is a flowchart of the steps for determining the second predicted trajectory in an embodiment;

[0057] Figure 5 It is a flowchart of the steps for determining the second predicted trajectory in another embodiment;

[0058] Figure 6 It is a flowchart of the steps for re-determining the second predicted trajectory in an embodiment;

[0059] Figure 7 It is a flowchart of the underwater target tracking method in another embodiment;

[0060] Figure 8 It is a structural block diagram of the underwater target tracking device in an embodiment;

[0061] Figure 9 It is an internal structure diagram of the computer equipment in an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] To make the objectives, technical solutions and advantages of this application more clear and understandable, the following further elaborates on this application in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for explaining this application and are not intended to limit this application.

[0063] The underwater target tracking method provided by the embodiments of this application can be applied to an application environment as Figure 1 shown. This application environment includes an underwater vehicle network system, specifically including a terminal and multiple underwater vehicles arranged underwater. This method can be applied to the terminal, can also be applied to the server, and can also be applied to a system including the terminal and the server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can include, but is not limited to, various personal computers, laptop computers, tablet computers, Internet of Things devices, etc. This terminal is used to obtain the observation values of each underwater vehicle, and through the observation values and the target tracking planning model, plan the tracking trajectories of each underwater vehicle in real time. Finally, when the distance between the real-time information of the target object and the real-time information of the underwater vehicle is not greater than the preset distance, it is determined that the target tracking is successful.

[0064] In one embodiment, as Figure 2 shown, a method for underwater target tracking is provided. Taking the application of this method to the terminal as an example, it includes the following steps:

[0065] Step S201, obtain the real-time information of each underwater vehicle, the real-time information of the target object, and the current environmental information.

[0066] Among them, the real-time information includes the current position information and the current state information.

[0067] In this embodiment, the terminal obtains the real-time information of each underwater vehicle according to the self-real-time information transmitted by each underwater vehicle, and obtains the real-time information of the target object and the current environmental information according to the target detection information of each underwater vehicle.

[0068] Specifically, the underwater vehicle can be an unmanned underwater vehicle (AUV). The underwater vehicle includes an annular acoustic wave generator and an annular hydrophone array. The annular acoustic wave generator and the annular hydrophone array are arranged on the outer surface of the underwater vehicle and are used to detect the target object or underwater vehicle in all directions in real time. Among them, the acoustic wave generator emits acoustic waves in all directions. When the acoustic waves reach the target object or the adjacent underwater vehicle, the target object or the adjacent underwater vehicle will reflect the echo; and after the hydrophone array receives the reflected echoes in all directions, it feeds back the acoustic wave intensity of the received reflected echo and the emitted acoustic wave intensity to the terminal. The terminal detects the current position information and the current state information of the target object or the adjacent underwater vehicle according to the acoustic wave intensity of the reflected echo and the emitted acoustic wave intensity.

[0069] The terminal emits acoustic signals to the sample target through the submersible, and receives the echo reflected by the sample target through the hydrophone array to determine the current environmental information. Among them, the current environmental information includes, but is not limited to, the water flow velocity of the current environment, the underwater depth of the current environment, the water density of the current environment, etc. The sample target is an object that is known to be at a certain distance and direction from the submersible and can reflect the emitted acoustic signal, and this sample target can be used by the submersible to obtain the current environmental information. The position information of the submersible includes the two-dimensional coordinate information of the thruster of the submersible and the two-dimensional coordinate information of the servo of the submersible, and the position information of the target object is the two-dimensional coordinate information of the target object. The two-dimensional coordinate information of the submersible and the two-dimensional coordinate information of the target object are both established in the earth coordinate system. The current state information of the submersible or the target object includes, but is not limited to, the current speed and the current direction, etc.

[0070] The real-time information of each submersible includes the current position information of each submersible and the current state information of each submersible, and the real-time information of the target object includes the current position information of the target object and the current state information of the target object. The real-time information of each submersible, the real-time information of the target object observed by each submersible, and the current environmental information observed by each submersible constitute the observation space O=(o1, o2,..., o I ), O i represents the observation information of each submersible.

[0071] Step S202: Determine the tracking trajectories of each submersible according to the real-time information of each submersible, the real-time information of the target object, the current environmental information, and the target tracking planning model.

[0072] In this embodiment, the terminal inputs the real-time information of each submersible, the real-time information of the target object, and the current environmental information into the target tracking planning model, and outputs the tracking trajectories of each submersible. Among them, the target tracking planning model can be a multi-agent reinforcement learning - MADDPG (Multi-Agent Deep Deterministic Policy Gradient) deep neural network model.

[0073] After the target tracking planning model completes one round of predicting the tracking trajectories of each submersible, the previous input information (the real-time information of the target object, the real-time information of each submersible, and the current environmental information) can be used as the training samples for the target tracking planning model to predict the tracking trajectories of each submersible in the next round, and the target tracking planning model is trained to obtain a new target tracking planning model, so as to ensure that the decisions made by the target tracking planning model in each round are more reasonable than the decisions in the previous round.

[0074] For example, in the first round, the predicted tracking trajectory is completed through the MADDPG deep neural network model. After each submersible completes tracking according to the predicted tracking trajectory, the terminal collects the real-time information of the current target, the real-time information of the current submersible, and the current environmental information, and inputs the real-time information of the current target, the real-time information of the current submersible, the current environmental information, the tracking trajectory of the submersible, the real-time information of the target at the start of the previous round, the real-time information of the submersible at the start of the previous round, and the environmental information at the start of the previous round into the MADDPG deep neural network model to perform real-time training on the MADDPG deep neural network model, obtaining a new MADDPG deep neural network model, and using this MADDPG deep neural network model as the target tracking planning model for predicting the tracking trajectories of each submersible in the next round.

[0075] Step S203: When the distance between the real-time information of the target and the real-time information of the submersible is not greater than the preset distance, it is determined that the target tracking is successful.

[0076] In this embodiment, the terminal uses the distance threshold between the position information of the target and the position information of the submersible as the preset distance. The terminal controls each submersible to execute the command to track the target. During the process of the submersible executing the command to track the target, the terminal continues to obtain the real-time information of the target and determines whether the target has moved. In the case where the target has moved, the terminal controls the submersible to abandon the current tracking trajectory and return to execute step S202; in the case where the target has not moved, the terminal does not issue other instructions, and the submersible continues to execute the current command to track the target.

[0077] When each submersible completes the command to track the target, each submersible sends a command completion signal to the terminal. When the terminal receives this command completion signal, the terminal obtains the real-time information of the current target, the real-time information of the current submersible, and the current environmental state information again, and determines whether the distance between the position information of the target and the position information of the submersible is not greater than the preset distance. When the distance between the position information of the target and the position information of the submersible is greater than the preset distance, it is determined that the target tracking fails, and it returns to execute step S202; when the distance between the position information of the target and the position information of the submersible is not greater than the preset distance, it is determined that the target tracking is successful. Each submersible that has successfully tracked feeds back the target tracking success information to the terminal. The terminal trains the target tracking planning model by inputting the real-time information of each submersible that has successfully tracked, the real-time information of the target, the environmental information, and the tracking trajectories of each submersible that has successfully tracked into the target tracking planning model, obtaining a new target tracking planning model. The terminal predicts the next round of tracking trajectories of each submersible that has failed to track the target through the new target tracking planning model.

[0078] Based on the above solution, real-time information of each submersible and the target object is obtained through the target detection model. Considering the current environmental information, the target tracking and planning model is used to plan the target tracking trajectory in real time, so as to track the target and improve the accuracy of the actual tracking trajectory of the submersible.

[0079] Optionally, obtaining the real-time information of the target object includes: obtaining each detection information of the target object through each submersible; obtaining the real-time information of the target object according to each detection information of the target object and the underwater target detection model.

[0080] In this embodiment, the terminal receives the echo reflected by the target object through the hydrophone array of each submersible and obtains the acoustic waves (i.e., each detection information) emitted by each submersible. The terminal inputs the echo intensity of the target object reflected and the acoustic wave intensity emitted by each submersible into the underwater target detection model. Based on the underwater target detection model, the direction and distance from each submersible to the target object can be obtained first, and then the current position information of the target object can be calculated according to the distance and direction from each submersible to the target object. For example, the underwater target detection model includes an algorithm for the distance d from the submersible to the target object and an algorithm for the propagation loss TL of the acoustic wave in water. The formula for the propagation loss TL of the acoustic wave in water in the underwater target detection model is as follows:

[0081] 2TL = SL - EM + TS - (NL - DI) - DT

[0082] In the above formula, the units of all variables are dB. Among them, SL is the emission acoustic wave intensity of the acoustic wave generator of the submersible. NL is the ocean background noise level. EM is the echo intensity of the target object reflected received by each hydrophone of the submersible. DI is the sonar directivity index, which is related to the number N of hydrophones in the array: DI = 10lg(N). DT is the detection threshold of the active sonar system, usually 0dB. TS is the target strength, which is related to the reflection area of the target object to the acoustic wave and can be obtained through Table 1.

[0083] Table 1: Common underwater target strengths

[0084]

[0085] The algorithm formula for the distance d from the submersible to the target object in the underwater target detection model is as follows:

[0086] TL = 20lg(d) + d × a(f) × l0 -3 ,

[0087]

[0088] In the above formula, f is the center operating frequency of the acoustic wave generator, and α(f) is the empirical formula for the attenuation characteristics of the acoustic wave propagation in water.

[0089] The terminal obtains the position information of the target object at different times by repeating the above steps multiple times. The terminal obtains the state information of the target object at different times based on the position information of the target object at different times, and thus takes the position information of the target object at different times and the state information of the target object at different times as the real-time information of the target object.

[0090] Based on the above solution, the direction and distance from each submersible to the target object are obtained through the underwater target detection model, and the real-time information of the target object is judged by judging the distance and direction from each submersible to the target object, so that the real-time information of the target object can be obtained more accurately.

[0091] Optionally, as Figure 3 shown, the target tracking and planning model includes a tracking trajectory prediction model, an underwater dynamics model, and a trajectory optimization model. According to the real-time information of each submersible, the real-time information of the target object, the current environmental information, and the target tracking and planning model, the tracking trajectories of each submersible are determined, including:

[0092] Step S301, for each submersible, determine the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model.

[0093] In this embodiment, the tracking trajectory prediction model includes a taboo list. The terminal inputs the current position information of the submersible, the current position information of the target object, and the current environmental information into the tracking trajectory prediction model for each submersible, and the tracking trajectory prediction model outputs the first predicted trajectory of the submersible except for the predicted trajectories in the taboo list. The first predicted trajectory of the submersible includes the trajectory route of the submersible and the action information of the submersible. The set of the action information of the submersible is the action space of the submersible cluster, and the submersible cluster includes each submersible. The action information of each submersible includes the power of the thruster of each submersible and the angle of the rudder, as well as other physical constraint conditions of the submersible. Other physical constraint conditions include, but are not limited to, gravity, buoyancy, viscous hydrodynamic force, and inertial hydrodynamic force, etc. For example, A=(a1,a2,...,a I ) is the action space composed of the action information of each submersible in the cluster, and the action information of each submersible is where, represents the thrust of the thruster of the submersible, and δ i represents the angle of the rudder of the submersible. sat(·) represents other physical constraint conditions of the submersible, and its expression is as follows:

[0094]

[0095] where ι is ([[]] δ i)。The tracking trajectory prediction model can be the Actor target policy network in the MADDPG deep neural network model Actor target policy network contains the action space A=(a1,a2,...,a I ), where θ i represents the parameters of the neural network. In each round of training the Actor target policy network in the MADDPG deep neural network model, the θ of this neural network will be optimized i parameters.

[0096] Step S302, determine the second predicted trajectory of each submersible according to the first predicted trajectory of each submersible, the trajectory optimization model and the tracking trajectory prediction model.

[0097] In this embodiment, the terminal preset the evaluation criteria of the trajectory optimization model. According to the first predicted trajectory of each submersible and the trajectory optimization model, the evaluation value of the first predicted trajectory of each submersible is obtained. The terminal judges whether the evaluation value of the first predicted trajectory of each submersible meets the evaluation criteria. On the premise that it does not meet the evaluation criteria, the terminal returns to execute step S301 until the first predicted trajectory of each submersible meets the evaluation criteria, and marks the first predicted trajectory of each submersible as the second predicted trajectory. The trajectory optimization model can be the Critic action value network in the MADDPG deep neural network model Critic action value network contains the action space A=(a1,a2,...,a I ), where v i represents the parameters of this neural network. In each round of training the Critic action value network in the MADDPG deep neural network model, the v of this neural network i parameters are optimized. The trajectory optimization model can be an obstacle avoidance model or a cluster consensus model, or can contain both an obstacle avoidance model and a cluster consensus model. The specific process of optimization through the obstacle avoidance model or the cluster consensus model will be described in detail later.

[0098] In the case where the trajectory optimization model contains both an obstacle avoidance model and a cluster consensus model, the first predicted trajectory of each submersible needs to be judged by the obstacle avoidance model and the cluster consensus model respectively; when the first predicted trajectory meets the evaluation criteria of the first model, then input the first predicted trajectory into the next model for judgment, and the order of the two models is not restricted; and the evaluation value of the first predicted trajectory of each submersible is where i represents the virtual number of each submersible is the target tracking reward, is the obstacle avoidance reward, r cIt is the reward for the consistency of swarm actions. α and β are hyperparameters of this optimization criterion. By setting the magnitudes of the parameters α and β, the reward intensity of different optimization processes can be adjusted to meet the requirements of different tracking conditions.

[0099] Step S303: Determine the tracking trajectories of each submersible according to the real-time information of each submersible, the second predicted trajectory of each submersible, and the underwater dynamics model.

[0100] In this embodiment, for each submersible, the terminal inputs the real-time information of the submersible and the second predicted trajectory of the submersible into the underwater dynamics model according to the real-time information of the submersible and the second predicted trajectory of the submersible. The underwater dynamics model outputs the motion information of the submersible in the second predicted trajectory of the submersible. The terminal determines the tracking trajectory for the submersible according to the motion information in the second predicted trajectory of the submersible and the second predicted trajectory of the submersible. The underwater dynamics model includes kinematic equations and dynamic equations. Among them, the kinematic equation is: Among them is the transformation matrix between the earth coordinate system and the body-fixed coordinate system:

[0101]

[0102] The dynamic equation of the submersible is:

[0103]

[0104] Among them is the inertia matrix of the autonomous submersible, mainly including parameters such as the mass, moment of inertia, and inertia-like hydrodynamic force (the resistance exerted on the submersible to overcome inertia when the water body around the submersible makes a variable-speed motion) of the autonomous submersible. is the Coriolis matrix. φ is the motion speed of the submersible. is the resistance matrix formed by viscous-like hydrodynamic forces. is the arrangement matrix of the thrusters and rudders of the autonomous submersible.

[0105] Based on the above solution, environmental factors, physical factors, etc. are considered in the obtained tracking trajectories of each submersible, and the tracking trajectories are further optimized, thereby improving the reliability of the tracking trajectories of the submersibles.

[0106] Optionally, as Figure 4 shown, when the trajectory optimization model includes an obstacle avoidance model, determining the second predicted trajectory of each submersible according to the first predicted trajectory of each submersible, the trajectory optimization model, and the tracking trajectory prediction model includes:

[0107] Step S401: For each submersible vehicle, determine the collision probability between the submersible vehicle and adjacent submersible vehicles according to the first predicted trajectory of the submersible vehicle, the first predicted trajectories of adjacent submersible vehicles, and the obstacle avoidance model.

[0108] In this embodiment, for each submersible vehicle, the terminal determines the evaluation value of the first predicted trajectory of the submersible vehicle (i.e., determines the collision probability between the submersible vehicle and adjacent submersible vehicles) according to the first predicted trajectory of the submersible vehicle, the first predicted trajectories of adjacent submersible vehicles, and the obstacle avoidance model. The evaluation value of the first predicted trajectory of the submersible vehicle is

[0109] Step S402: When the collision probability is greater than the collision threshold, add the first predicted trajectory of the submersible vehicle to the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of the submersible vehicle for each submersible vehicle according to the real-time information of the submersible vehicle, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model.

[0110] In this embodiment, for each submersible vehicle, the terminal determines the size relationship between the evaluation value of the first predicted trajectory of the submersible vehicle (i.e., the collision probability) and the evaluation criterion (i.e., the collision threshold). When the evaluation value of the first predicted trajectory of the submersible vehicle is greater than the evaluation criterion, the terminal adds the first predicted trajectory of the submersible vehicle to the taboo list of the tracking trajectory prediction model, and returns to execute step S301.

[0111] Step S403: When the collision probability is not greater than the collision threshold, mark the first predicted trajectory of the submersible vehicle as the second predicted trajectory of the submersible vehicle.

[0112] In this embodiment, for each submersible vehicle, when the evaluation value of the first predicted trajectory of the submersible vehicle (i.e., the collision probability) is less than the evaluation criterion (i.e., the collision threshold), the terminal marks the first predicted trajectory of the submersible vehicle as the second predicted trajectory of the submersible vehicle.

[0113] Based on the above solution, the first predicted trajectory is optimized through the obstacle avoidance model, thereby improving the accuracy of the first predicted trajectory.

[0114] Optionally, as Figure 5 shown, when the trajectory optimization model includes a cluster consistency model, determining the second predicted trajectory of each submersible vehicle according to the first predicted trajectories of each submersible vehicle, the trajectory optimization model, and the tracking trajectory prediction model includes:

[0115] Step S501: According to the first predicted trajectories of each submersible vehicle and the cluster consistency model, determine the cluster discrete value of each submersible vehicle and the discrete average value of each submersible vehicle.

[0116] In this embodiment, the terminal inputs the first predicted trajectory of each submersible into the cluster consistency model to determine the evaluation value of the first predicted trajectory of each submersible (i.e., determine the cluster dispersion value of each submersible), and the average evaluation value of each evaluation value (i.e., the average dispersion value of the submersibles). The evaluation value of the first predicted trajectory of the submersible is The cluster consistency model can be established through the underwater communication signal-to-noise ratio of each submersible. For example, the terminal inputs the first predicted trajectory of each submersible into the cluster consistency model to obtain the underwater communication signal-to-noise ratio of each submersible (i.e., the cluster dispersion value), and the average underwater communication signal-to-noise ratio of each submersible (i.e., the average dispersion value). Define two adjacent submersibles as a and b, where the signal-to-noise ratio between i and j is SNR ab , where a, b ∈ N (N is the number of all submersibles), and where SNR ab The expression is as follows:

[0117] SNR ab = SL - TL - NL + DI

[0118]

[0119] In the above formula, SL is the transmitted acoustic intensity of the acoustic wave generator of the submersible, TL is the propagation loss of the transmitted acoustic wave in water, DI is the sonar directivity index, and NL is the ocean background noise level. is the average evaluation value of each evaluation value.

[0120] Step S502: When the average dispersion value of each submersible is greater than the dispersion threshold, include the first predicted trajectory of each submersible in the taboo list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of each submersible according to the real-time information of the submersible, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model for each submersible.

[0121] In this embodiment, the terminal judges the size of the average evaluation value of the first predicted trajectory of each submersible (i.e., the average dispersion value of each submersible) and the evaluation criterion (dispersion threshold). When the average evaluation value of the first predicted trajectory of each submersible is greater than the evaluation criterion, the terminal includes the first predicted trajectory of each submersible in the taboo list of the tracking trajectory prediction model and returns to execute step S301. For example, it can be judged the size of the average underwater communication signal-to-noise ratio (i.e., the average dispersion value of each submersible) and the signal-to-noise ratio threshold DT (dispersion threshold). When the average underwater communication signal-to-noise ratio is greater than the signal-to-noise ratio threshold DT, the terminal includes the first predicted trajectory of each submersible in the taboo list of the tracking trajectory prediction model and returns to execute step S301.

[0122] Step S503: When the cluster discrete values of all the submersibles are not greater than the discrete threshold, mark the first predicted trajectory of each submersible as the second trajectory of each submersible.

[0123] In this embodiment, when the evaluation values of the first predicted trajectories of all the submersibles (i.e., the cluster discrete values of all the submersibles) are not greater than the evaluation criterion (i.e., the discrete threshold), the terminal marks the first predicted trajectory of each submersible as the second predicted trajectory of each submersible.

[0124] Based on the above solution, the first predicted trajectory is optimized through the cluster consistency model, thereby improving the accuracy of the first predicted trajectory.

[0125] Optionally, as Figure 6 shown, the method further includes:

[0126] Step S601: When the discrete average value of each submersible is not greater than the discrete threshold and there is a target cluster discrete value greater than the discrete threshold, include the first predicted trajectory of the submersible corresponding to each target cluster discrete value in the taboo list of the tracking trajectory prediction model.

[0127] In this embodiment, when the average evaluation value (i.e., the discrete average value) of the first predicted trajectory of each submersible is not greater than the evaluation criterion (i.e., the discrete threshold), the terminal selects the target evaluation value (i.e., the target cluster discrete value) of the first predicted trajectory greater than the evaluation criterion among each evaluation value, and includes the first predicted trajectory of the submersible corresponding to each target evaluation value in the taboo list of the tracking trajectory prediction model.

[0128] Step S602: According to the real-time information of the submersible corresponding to each target cluster discrete value, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectory of the submersible corresponding to each target cluster discrete value, and based on the re-determined first predicted trajectory of the submersible corresponding to each target cluster discrete value, return to execute the step of determining the cluster discrete value of each submersible and the discrete average value of each submersible according to the first predicted trajectory of each submersible and the cluster consistency model.

[0129] In this embodiment, the terminal inputs the real-time information of the submersible corresponding to each target evaluation value (i.e., the target cluster discrete value), the real-time information of the target, and the current environmental information into the tracking trajectory prediction model, re-determines the first predicted trajectory of the submersible corresponding to each target evaluation value, and returns to execute step S302 for the first predicted trajectory of the submersible corresponding to each target evaluation value.

[0130] Based on the above solution, by specifically optimizing the first predicted trajectory of each submersible that conforms to the cluster consistency model again, the cluster consistency performance of the first predicted trajectory of each submersible is improved, and the practicality of the tracking trajectory of each submersible obtained is improved indirectly.

[0131] This application also provides an example of underwater target tracking, as Figure 7 shown. The specific processing process includes the following steps:

[0132] Step S701, detect various detection information of the target object through each submersible.

[0133] Step S702, obtain the real-time information of the target object according to the various detection information of the target object and the underwater target detection model.

[0134] Step S703, obtain the real-time information of each submersible and the current environmental information.

[0135] Among them, the real-time information includes the current position information and the current state information.

[0136] Step S704, for each submersible, determine the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model.

[0137] Step S705, for each submersible, determine the collision probability between the submersible and the adjacent submersible according to the first predicted trajectory of the submersible, the first predicted trajectory of the adjacent submersible, and the obstacle avoidance model.

[0138] Step S706, determine whether the collision probability is greater than the collision threshold.

[0139] If so, include the first predicted trajectory of the submersible in the taboo list of the tracking trajectory prediction model, and return to execute step S704; if not, execute step S707.

[0140] Step S707, mark the first predicted trajectory of the submersible as the initial second predicted trajectory of the submersible.

[0141] Step S708, determine the cluster dispersion value of each submersible and the average dispersion value of each submersible according to the initial second predicted trajectory of each submersible and the cluster consistency model.

[0142] Step S709, determine whether the average dispersion value of each submersible is greater than the dispersion threshold.

[0143] If so, include the initial second predicted trajectory of each submersible in the taboo list of the tracking trajectory prediction model, return to execute step S704; if not, execute step S710.

[0144] Step S710, determine whether there is a target cluster dispersion value greater than the dispersion threshold.

[0145] If so, include the initial second predicted trajectories of the submarines corresponding to the discrete values of each target cluster in the tabu list of the tracking trajectory prediction model, and execute step S711; if not, execute step S712.

[0146] Step S711: Based on the real-time information of the submarines corresponding to the discrete values of each target cluster, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectories of the submarines corresponding to the discrete values of each target cluster, and based on the re-determined first predicted trajectories of the submarines corresponding to the discrete values of each target cluster, return to execute step S705.

[0147] Step S712: When the discrete values of all submarines' clusters are not greater than the discrete threshold, mark the initial second predicted trajectories of each submarine as the second trajectories of each submarine.

[0148] Step S713: Determine the tracking trajectories of each submarine according to the real-time information of each submarine, the second predicted trajectory of each submarine, and the underwater dynamics model.

[0149] Step S714: Determine whether the distance between the real-time information of the target object and the real-time information of the submarine is greater than the preset distance.

[0150] If so, return to execute step S704; if not, execute step S715.

[0151] Step S715: Determine that the target tracking is successful.

[0152] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or steps or stages in other steps.

[0153] Based on the same inventive concept, an embodiment of the present application also provides an underwater target tracking device for implementing the above-mentioned underwater target tracking method. The implementation solutions provided by this device to solve problems are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the underwater target tracking device provided below can refer to the limitations on the underwater target tracking method in the above text, and will not be repeated here.

[0154] In one embodiment, as Figure 8 shown, an underwater target tracking device is provided, including: an acquisition module 810, a first determination module 820, and a second determination module 830, where:

[0155] The acquisition module 810 is configured to acquire real-time information of each submersible vehicle, real-time information of the target body, and current environmental information; the real-time information includes current position information and current status information;

[0156] The first determination module 820 is configured to determine the tracking trajectory of each submersible vehicle according to the real-time information of each submersible vehicle, the real-time information of the target body, the current environmental information, and the target tracking planning model;

[0157] The second determination module 830 is configured to determine that the target tracking is successful when the distance between the real-time information of the target body and the real-time information of the submersible vehicle is not greater than a preset distance.

[0158] Optionally, the acquisition module 810 is specifically configured to:

[0159] Obtain each detection information of the target body through each submersible vehicle;

[0160] Obtain the real-time information of the target body according to each detection information of the target body and the underwater target detection model.

[0161] Optionally, the target tracking planning model includes a tracking trajectory prediction model, an underwater dynamics model, and a trajectory optimization model. The first determination module 820 is specifically configured to:

[0162] For each submersible vehicle, determine the first predicted trajectory of the submersible vehicle according to the real-time information of the submersible vehicle, the real-time information of the target body, the current environmental information, and the tracking trajectory prediction model;

[0163] Determine the second predicted trajectory of each submersible vehicle according to the first predicted trajectory of each submersible vehicle, the trajectory optimization model, and the tracking trajectory prediction model;

[0164] Determine the tracking trajectory of each submersible vehicle according to the real-time information of each submersible vehicle, the second predicted trajectory of each submersible vehicle, and the underwater dynamics model.

[0165] Optionally, when the trajectory optimization model includes an obstacle avoidance model, the first determination module 820 is specifically configured to:

[0166] For each submersible vehicle, determine the collision probability between the submersible vehicle and the adjacent submersible vehicle according to the first predicted trajectory of the submersible vehicle, the first predicted trajectory of the adjacent submersible vehicle, and the obstacle avoidance model;

[0167] When the collision probability is greater than the collision threshold, the first predicted trajectory of the submersible is included in the taboo list of the tracking trajectory prediction model, and the step of determining the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model is returned for each submersible;

[0168] When the collision probability is not greater than the collision threshold, the first predicted trajectory of the submersible is marked as the second predicted trajectory of the submersible.

[0169] Optionally, when the trajectory optimization model includes a cluster consistency model, the first determination module 820 is specifically configured to:

[0170] According to the first predicted trajectories of the submersibles and the cluster consistency model, determine the cluster dispersion values of the submersibles and the average dispersion value of the submersibles;

[0171] When the average dispersion value of the submersibles is greater than the dispersion threshold, the first predicted trajectories of the submersibles are included in the taboo list of the tracking trajectory prediction model, and the step of determining the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model is returned for each submersible;

[0172] When the cluster dispersion values of all submersibles are not greater than the dispersion threshold, the first predicted trajectories of the submersibles are marked as the second trajectories of the submersibles.

[0173] Optionally, when the trajectory optimization model includes a cluster consistency model, the first determination module 820 further includes:

[0174] When the average dispersion value of the submersibles is not greater than the dispersion threshold and there are target cluster dispersion values greater than the dispersion threshold, the first predicted trajectories of the submersibles corresponding to the target cluster dispersion values are included in the taboo list of the tracking trajectory prediction model;

[0175] According to the real-time information of the submersibles corresponding to the target cluster dispersion values, the real-time information of the target, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectories of the submersibles corresponding to the target cluster dispersion values, and based on the re-determined first predicted trajectories of the submersibles corresponding to the target cluster dispersion values, return and execute the step of determining the cluster dispersion values of the submersibles and the average dispersion value of the submersibles according to the first predicted trajectories of the submersibles and the cluster consistency model.

[0176] Each module in the above-mentioned underwater target tracking device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in the processor of the computer device in hardware form or be independent of it, or can be stored in the memory of the computer device in software form, so as to facilitate the processor to call and execute the operations corresponding to each of the above modules.

[0177] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structural diagram can be as Figure 9 shown. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be achieved through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements an underwater target tracking method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0178] Those skilled in the art can understand that Figure 9 the structure shown in

[0179] is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0180] In an embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, it implements the steps in the above method embodiments.

[0181] In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, it implements the steps in the above method embodiments.

[0182] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0183] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., and are not limited thereto. The processors involved in the embodiments provided in this application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., and are not limited thereto.

[0184] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0185] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. An underwater target tracking method, characterized in that, The method includes: Obtaining the real-time information of each submersible, the real-time information of the target object, and the current environmental information; the real-time information includes the current position information and the current state information; For each submersible, determining the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model; When the trajectory optimization model includes a cluster consistency model, determining the cluster dispersion value of each submersible and the average dispersion value of each submersible according to the first predicted trajectories of each submersible and the cluster consistency model; When the average dispersion value of each submersible is greater than the dispersion threshold, adding the first predicted trajectories of each submersible to the taboo list of the tracking trajectory prediction model, and returning to execute the step of determining the first predicted trajectory of each submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model; When the cluster dispersion values of all submersibles are not greater than the dispersion threshold, marking the first predicted trajectories of each submersible as the second trajectories of each submersible; Determining the tracking trajectories of each submersible according to the real-time information of each submersible, the second predicted trajectories of each submersible, and the underwater dynamics model; When the distance between the real-time information of the target object and the real-time information of the submersible is not greater than the preset distance, it is determined that the target tracking is successful.

2. The method according to claim 1, wherein The obtaining of the real-time information of the target object includes: Obtaining each detection information of the target object through each submersible; Obtaining the real-time information of the target object according to each detection information of the target object and the underwater target detection model.

3. The method according to claim 1, wherein When the trajectory optimization model includes an obstacle avoidance model, after determining the first predicted trajectory of the submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model, it further includes: For each submersible, determining the collision probability between the submersible and the adjacent submersible according to the first predicted trajectory of the submersible, the first predicted trajectory of the adjacent submersible, and the obstacle avoidance model; When the collision probability is greater than the collision threshold, adding the first predicted trajectory of the submersible to the taboo list of the tracking trajectory prediction model, and returning to execute the step of determining the first predicted trajectory of each submersible according to the real-time information of the submersible, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model; When the collision probability is not greater than the collision threshold, marking the first predicted trajectory of the submersible as the second predicted trajectory of the submersible.

4. The method according to claim 1, wherein The method further includes: When the average dispersion value of each submersible is not greater than the dispersion threshold and there are target cluster dispersion values greater than the dispersion threshold, adding the first predicted trajectories of the submersibles corresponding to each target cluster dispersion value to the taboo list of the tracking trajectory prediction model; Based on the real-time information of the submarines corresponding to each of the target cluster discrete values, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model, re-determine the first predicted trajectory of the submarines corresponding to each of the target cluster discrete values, and based on the first predicted trajectories of the submarines corresponding to each re-determined target cluster discrete value, return to execute the step of determining the cluster discrete value of each submarine and the discrete average value of each submarine according to the first predicted trajectory of each submarine and the cluster consistency model.

5. An underwater target tracking device, characterized in that, The device includes: An acquisition module, configured to acquire the real-time information of each submarine, the real-time information of the target object, and the current environmental information; the real-time information includes the current position information and the current status information; A first determination module, for each submarine, according to the real-time information of the submarine, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model, determine the first predicted trajectory of the submarine; in the case where the trajectory optimization model includes a cluster consistency model, according to the first predicted trajectories of each submarine and the cluster consistency model, determine the cluster discrete value of each submarine and the discrete average value of each submarine; when the discrete average value of each submarine is greater than the discrete threshold, include the first predicted trajectory of each submarine in the tabu list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of each submarine according to the real-time information of the submarine, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model for each submarine; when the cluster discrete values of all submarines are not greater than the discrete threshold, mark the first predicted trajectory of each submarine as the second trajectory of each submarine; according to the real-time information of each submarine, the second predicted trajectory of each submarine, and the underwater dynamics model, determine the tracking trajectory of each submarine; A second determination module, configured to determine that the target tracking is successful when the distance between the real-time information of the target object and the real-time information of the submarine is not greater than a preset distance.

6. The device according to claim 5, wherein The acquisition module is specifically configured to: Obtain each detection information of the target object through each submarine; According to each detection information of the target object and the underwater target detection model, obtain the real-time information of the target object.

7. The device according to claim 5, characterized in that, In the case where the trajectory optimization model includes an obstacle avoidance model, the first determination module is specifically configured to: For each submarine, according to the first predicted trajectory of the submarine, the first predicted trajectory of the adjacent submarine, and the obstacle avoidance model, determine the collision probability between the submarine and the adjacent submarine; When the collision probability is greater than the collision threshold, include the first predicted trajectory of the submarine in the tabu list of the tracking trajectory prediction model, and return to execute the step of determining the first predicted trajectory of each submarine according to the real-time information of the submarine, the real-time information of the target object, the current environmental information, and the tracking trajectory prediction model for each submarine; When the collision probability is not greater than the collision threshold, mark the first predicted trajectory of the submersible as the second predicted trajectory of the submersible.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method described in any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When this computer program is executed by a processor, it implements the steps of the method described in any one of claims 1 to 4.

Images