Unmanned ship cluster maneuvering target tracking control method in complex dynamic environment

Abstract

The invention discloses an unmanned surface vehicle cluster maneuvering target tracking control method in a complex dynamic environment, and belongs to the field of unmanned surface vehicle cooperative control, and the method comprises the steps: configuring an extended state observer for each unmanned surface vehicle, and enabling the extended state observers to be used for estimating the speed and control input of an unknown maneuvering target online, and compensating external bounded disturbance; a distributed cooperative controller is designed, the controller integrates target attraction, inter-boat collision avoidance, environment obstacle avoidance and speed cooperation functions, an unmanned boat cluster can be driven to autonomously form and dynamically adjust a stable tracking configuration surrounding a target under the condition that formation parameters do not need to be preset, and meanwhile, the stability of the unmanned boat cluster is improved. A distributed event triggering mechanism deeply fused with a controller is designed, the communication opportunity is dynamically determined according to local measurement errors and neighbor state deviation, and the inter-boat communication frequency is remarkably reduced. According to the invention, the cooperative tracking problem of the unmanned ship cluster under multiple constraints of unknown state, interference, intermittent communication, obstacles and the like in a complex dynamic environment is effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned surface vehicle (USV) cooperative control, and in particular, to a USV swarm maneuvering target tracking control method in complex dynamic environments. BACKGROUND

[0002] Swarm cooperative control has become an important research direction in the field of intelligent unmanned systems due to its high efficiency and robustness in executing complex tasks. This technology has shown broad application prospects in scenarios such as environmental monitoring, disaster rescue, logistics distribution, and security patrol. Compared with manned platforms, USVs have significant advantages such as small size, low cost, flexible deployment, and controllable risk, making them ideal platforms for executing tasks such as wide-area reconnaissance, target search and positioning, mobile target tracking, and cooperative interception suppression. Among them, the cooperative tracking control of USV swarm for maneuvering targets is particularly critical.

[0003] Existing technologies usually face the following challenges: (1) Unknown target state: In actual tasks, the target is often non-cooperative, and its speed, acceleration, and other state information are difficult to obtain; (2) Environmental disturbances and obstacles: External disturbances such as wind, waves, and currents in the ocean environment, as well as static / dynamic obstacles, seriously affect control accuracy and safety; (3) Limited communication resources: Continuous high-speed communication between swarm members can bring huge bandwidth pressure and energy consumption, especially in harsh communication conditions such as the open sea; (4) Strong dependence on formation: Traditional methods usually require a fixed geometric formation (such as a circle or polygon), lacking the ability to adapt to target maneuvers and environmental changes.

[0004] Although existing research has attempted to address some of these problems, such as using disturbance observers to handle disturbances, using artificial potential field methods (APF) to achieve obstacle avoidance, and introducing event-triggered control (ETC) to reduce communication load, these methods are often isolated and fail to organically integrate state estimation, disturbance rejection, self-organizing tracking, collision and obstacle avoidance, and intermittent communication optimization. In particular, there is a lack of a comprehensive control scheme that can break free from the dependence on pre-set formations, spontaneously form tracking configurations through internal potential field mechanisms, and deeply cooperate with event-triggered communication. SUMMARY

[0005] The purpose of the present application is to provide a USV swarm maneuvering target tracking control method in complex dynamic environments to solve the problem of cooperative tracking under multiple complex constraints such as unknown target state, external disturbances, obstacles, and limited communication.

[0006] To achieve the above purpose, the present application provides the following solutions: The present application provides a USV swarm maneuvering target tracking control method in complex dynamic environments, comprising: S1, configuring an extended state observer and a sliding mode disturbance observer for each unmanned surface vehicle in the swarm respectively; S2, obtaining position information of the target; S3, estimating unknown velocity and control input of the target in real time based on the position information of the target by using the extended state observer, to obtain an estimated value of the target velocity and an estimated value of the target control input; S4, estimating and compensating external bounded disturbance acting on the unmanned surface vehicle in real time by using the sliding mode disturbance observer, to obtain an estimated value of the disturbance to the unmanned surface vehicle; S5, designing a distributed cooperative controller based on the estimated value of the target velocity, the estimated value of the target control input, and the estimated value of the disturbance to the unmanned surface vehicle; S6, designing a distributed event-triggering mechanism, defining the state error of the unmanned surface vehicle as the difference between its current velocity and the velocity broadcast to the neighbors at the last triggering time; when the norm of the state error of the unmanned surface vehicle meets the preset triggering condition, activating communication and broadcasting the velocity information of the current unmanned surface vehicle to the neighbor unmanned surface vehicles; otherwise, keeping communication silent; S7, taking the velocity information broadcast at the last triggering time in step S6 as the input of the distributed cooperative controller, executing the distributed cooperative controller to drive the unmanned surface vehicle swarm to autonomously form and maintain a stable tracking configuration around the target without presetting formation parameters, while avoiding inter-boat collision and environmental obstacles.

[0007] Optionally, the position information of the unknown target is obtained based on the position information of the target obtained from the sensor of the unmanned surface vehicle.

[0008] Optionally, the distributed cooperative controller includes a target attractive term, an inter-boat collision avoidance term, an environmental obstacle avoidance term, and a velocity cooperation term, wherein the inter-boat collision avoidance term and the environmental obstacle avoidance term are respectively generated by the negative gradient of the collision avoidance potential field function and the obstacle avoidance potential field function, and the velocity cooperation term is used to drive the velocities of all unmanned surface vehicles to be consistent with the target velocity.

[0009] Optionally, the expression of the extended state observer is as follows: ; wherein, denotes the derivative of the unmanned surface vehicle with respect to the estimated value of the target position; denotes the derivative of the unmanned surface vehicle with respect to the estimated value of the target velocity; denotes the measurable position of the target; denotes the derivative of the unmanned surface vehicle with respect to the estimated value of the target position; denotes the derivative of the unmanned surface vehicle with respect to the estimated value of the target velocity; for an unmanned surface vehicle an estimate of the target control input; denotes an unmanned surface vehicle a derivative of the estimate of the target control input, is an extended state observer gain.

[0010] Optionally, the expression of the sliding mode disturbance observer is as follows: ; wherein, for an unmanned surface vehicle an estimate of the disturbance; , denotes a disturbance observer gain, denotes a sliding surface.

[0011] Optionally, the expression of the distributed cooperative controller is as follows: ; wherein, is a distributed controller; is a control gain; denotes an unmanned surface vehicle a position of; denotes a velocity of the unmanned surface vehicle at the latest triggering time ; denotes the unmanned surface vehicle; denotes a dynamic neighbor set of the unmanned surface vehicle; denotes a communication weight between the unmanned surface vehicle and the unmanned surface vehicle; denotes a velocity of the unmanned surface vehicle at the latest triggering time ; is a gradient operator; is an obstacle avoidance potential field function; denotes a relative distance between the unmanned surface vehicle and the unmanned surface vehicle; is an obstacle avoidance potential field function; denotes an obstacle; denotes a set of obstacles; denotes a relative distance between the unmanned surface vehicle and the obstacle; denotes a triggering time; denotes a cooperative tracking control term; denotes a velocity cooperation term; denotes an obstacle avoidance term; for unmanned surface vehicle an estimated value of the disturbance.

[0012] Optionally, the expression of the collision avoidance potential function is as follows: ; wherein, is a gradient operator; is a collision avoidance potential function; , is a collision avoidance constant; denotes a relative distance between the unmanned surface vehicle and another unmanned surface vehicle, is a safe collision avoidance distance; is a body radius; denotes a position of the unmanned surface vehicle ; denotes a position of the unmanned surface vehicle .

[0013] Optionally, the expression of the obstacle avoidance potential function is as follows: ; wherein, is a gradient operator, is an obstacle avoidance constant; denotes a relative distance between the unmanned surface vehicle and an obstacle; is a safe obstacle avoidance distance, is a radius of the obstacle; is a body radius; denotes a position of the unmanned surface vehicle ; denotes a position of the obstacle.

[0014] Optionally, the expression of the preset triggering condition is as follows: ; ; wherein, denotes a triggering time; denotes a current time; denotes a triggering time; denotes a triggering function, denotes a state error, denotes a to-be-designed parameter, denotes a cardinality of a neighbor set of the unmanned surface vehicle; denotes The speed coordination term at the triggering moment.

[0015] According to the specific embodiments provided in the application, the application has the following technical effects: The application provides a unmanned ship cluster mobile target tracking control method in a complex dynamic environment, and has the following advantages compared with the prior art: Break away from the dependence on formation and realize self-organizing tracking: through the potential field gradient driving mechanism, the cluster can spontaneously form a stable configuration around the target without any preset formation geometric parameters, greatly improving the self-adaptation and robustness of the system to target mobility and environmental changes; Strong anti-interference and high precision estimation: through the joint design of extended state observer ESO and sliding mode disturbance observer SMDO, the two core challenges of unknown target state and external environmental disturbance can be effectively handled, laying a solid foundation for high-precision tracking control; Efficient use of communication resources: the event triggering mechanism proposed can dynamically adjust the communication frequency according to the actual needs of the system, significantly reducing the network communication load while ensuring control performance, and is particularly suitable for bandwidth-limited marine application scenarios; High safety: through the designed collision avoidance potential field and obstacle avoidance potential field, the anti-collision between unmanned ships and the effective avoidance of obstacles in the environment can be ensured at the same time, ensuring the safety of task execution; Theoretical completeness and engineering feasibility: the stability and non-singular behavior of the system are proved by strict Lyapunov theory, which provides a solid theoretical guarantee for the actual deployment of the algorithm. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description only constitute some embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0017] Figure 1 It is a cooperative target tracking task schematic diagram in an embodiment of the application; Figure 2 It is a multi-unmanned ship system target tracking control framework schematic diagram in an embodiment of the application; Figure 3 It is a flowchart of a kind of unmanned ship cluster mobile target tracking control method in complex dynamic environment in an embodiment of the application; Figure 4 It is a collision avoidance potential field schematic diagram in an embodiment of the application; Figure 5 It is a target speed estimation error schematic diagram in an embodiment of the application; Figure 6 Target acceleration estimation error diagram for an embodiment of the present application; Figure 7 Interference estimation error diagram for an embodiment of the present application; Figure 8 Unmanned vehicle cluster tracking trajectory diagram for an embodiment of the present application; Figure 9 Unmanned vehicle spatial distribution diagram at different times for an embodiment of the present application; Figure 10 Tracking error diagram for an embodiment of the present application; Figure 11 Minimum inter-vehicle distance diagram for an embodiment of the present application; Figure 12 Unmanned vehicle velocity diagram for an embodiment of the present application; Figure 13 Trigger distribution diagram for an embodiment of the present application; Figure 14 Unmanned vehicle cluster tracking trajectory diagram in an obstacle environment for an embodiment of the present application; Figure 15 Unmanned vehicle spatial distribution diagram at different times in an obstacle environment for an embodiment of the present application; Figure 16 Tracking error diagram in an obstacle environment for an embodiment of the present application; Figure 17 Minimum inter-vehicle distance diagram in an obstacle environment for an embodiment of the present application; Figure 18 Velocity diagram in an obstacle environment for an embodiment of the present application; Figure 19 Trigger distribution diagram in an obstacle environment for an embodiment of the present application; Figure 20 Trigger distribution diagram in an obstacle environment for an embodiment of the present application using a prior art solution; Figure 21 Trigger distribution diagram in an obstacle environment for an embodiment of the present application using a prior art solution. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0019] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which:

[0020] Suppose the USV system moves in two-dimensional space. Consider a system consisting of one USV and one target, the first The model of the first USV is: (1) where, denotes the derivative of the position of the USV axis, denotes the derivative of the position of the USV axis, denotes the derivative of the velocity of the USV, denotes the derivative of the heading angle, denotes the position vector and velocity vector of the USV, denotes the control input of the USV, denotes the heading angle, denotes the position of the USV axis, denotes the position of the USV axis, denotes the velocity of the USV axis, denotes the velocity of the USV axis, denotes the control input of the USV axis, denotes the control input of the USV axis.

[0021] For subsequent analysis, the dynamic feedback linearization method is used to convert equation (1) into a double integral model, which is the kinematic model of the USV, and its expression is: (2) where, denotes the derivative of the position of the USV; denotes the derivative of the velocity of the USV; denotes the position of the USV.

[0022] Then the USV model in equation (1) can be rewritten as: (3) where, denotes the derivative of the velocity of the USV; denotes the velocity of the USV.

[0023] Assumption 1: External unknown disturbance is continuous and differentiable, and satisfies , , , is the upper bound of the disturbance, is the upper bound of the rate of change of the disturbance, denotes the rate of change of the disturbance.

[0024] In addition, since the motion pattern of the non-cooperative target is unknown, it can be modeled as: (4) where, denotes the derivative of the target position; denotes the derivative of the target velocity; denotes the measurable position of the target, and denotes the unknown velocity and control input of the target, denotes the position of the target axis, denotes the position of the target axis, denotes the velocity of the target axis, denotes the velocity of the target axis.

[0025] Assumption 2: Assume that the rate of change of the target control input is bounded, i.e. , is the upper bound of the rate of change of the control input, denotes the rate of change of the target control input.

[0026] In the tracking control problem, it is usually assumed that the motion state of the target is completely known. This application considers the potential non-cooperative nature of the target, assuming that both its velocity and acceleration information are unknown, and only the target position information is obtained from the sensors carried by the boat.

[0027] Use the graph to describe the information interaction between the unmanned boats, where is the point set of the graph, used to represent the set of unmanned boats in the cluster, is the edge set of the graph, used to represent the interaction relationship between the unmanned boats, is the adjacency matrix. If the unmanned boat can receive information from the unmanned boat , then ; otherwise, . In particular, it is assumed that . The degree matrix of the graph is , where , denotes the in-degree of the unmanned boat , then the graph The Laplacian matrix of the graph G is defined as The dynamic neighbor set of the USV is defined as: (5) wherein, is the dynamic neighbor set of the USV; denotes the i-th USV; denotes the i-th USV; denotes the i-th USV; denotes the i-th USV; denotes the relative distance between the USVs, denotes the sensing range of the USVs.

[0028] Compared with the fixed topology, in the dynamic network topology, the USVs can detect the existence of the neighbor USVs according to the sensors, form a completely distributed self-organizing communication architecture, and more meet the actual dynamic task environment.

[0029] The core target of the present application is to design a distributed control protocol, so that the multi-USV cluster can cooperatively track the maneuvering target with unknown state information in a complex dynamic environment. As shown in Figure 1 , the tracking task requires the USVs not only to dynamically converge around the target, but also to autonomously form and maintain a stable configuration in the obstacle and interference environment. To overcome the challenges of unknown target motion state (velocity, acceleration) and external disturbance, an online estimation mechanism needs to be established. At the same time, in the face of intermittent communication constraints and the urgent need for anti-collision / obstacle avoidance, the protocol must significantly reduce the communication burden through a triggering mechanism, and realize the organic integration of the potential field gradient-driven self-organizing configuration generation and the anti-collision / obstacle avoidance capability.

[0030] The present application designs a disturbance observer and a state observer to cope with environmental disturbance and unknown target motion, respectively; the key innovation is to use a distributed cooperative controller of anti-collision and obstacle avoidance potential field gradient, which can autonomously form and maintain a stable tracking configuration around the target without presetting geometric parameters; further, an event-triggered mechanism deeply integrated with the controller is proposed, which dynamically adjusts the communication between USVs based on local measurement error and neighbor state deviation, significantly reducing the network load; the stability proof based on Lyapunov guarantees the performance of the closed-loop system and excludes Zeno behavior. The overall cooperative framework is shown in Figure 2 In an exemplary embodiment, as shown in Figure 3 , a flowchart of a USV cluster maneuvering target tracking control method in a complex dynamic environment is provided, which is executed by a computer device, specifically, can be executed by a terminal or a server, etc. computer device alone, or can be executed by a terminal and a server together, in the embodiment of the present application, the following steps S1 to S7 are included. Wherein: ​S1. Configure an extended state observer and a sliding mode disturbance observer for each unmanned surface vessel in the cluster.

[0031] S2. Obtain the target's location information.

[0032] S3. Using the extended state observer and based on the target's position information, estimate the unknown velocity and control input of the target in real time to obtain the estimated value of the target velocity and the estimated value of the target control input.

[0033] S4. Using the sliding mode disturbance observer, estimate and compensate for the external bounded disturbances acting on the unmanned surface vessel in real time, and obtain the estimated value of the disturbance by the unmanned surface vessel.

[0034] The extended state observer and the sliding mode perturbation observer are introduced below: Extended State Observer Considering that the target's position can be measured by sensors, and assuming that the target's velocity and control input are unknown, this application, based on the extended state observer theory, designs a separate state observer for each unmanned surface vessel (USV) to estimate the target's velocity and control input. (6) in, For observer gain; For unmanned surface vessels An estimate of the target's location; For unmanned surface vessels An estimate of the target velocity; For unmanned surface vessels Estimates of the target control input; unmanned surface vessel The derivative of the target location estimate, unmanned surface vessel The derivative of the target velocity estimate, unmanned surface vessel The derivative of the target control input estimate.

[0035] Before performing the tracking mission, the observer can be integrated into the onboard computer of the unmanned surface vessel. When the onboard sensors detect the target position, the speed and control input can be quickly calculated using equation (6).

[0036] Define position estimation error Speed ​​estimation error and control input estimation error for: (7) The dynamic equation for the estimation error is then: (8) where, denotes the unmanned surface vehicle error in the target position estimate, denotes the unmanned surface vehicle error in the target velocity estimate, denotes the unmanned surface vehicle error in the target control input estimate.

[0037] Define the estimation error vector The error dynamics can be represented as (9) where, error system matrix, is the error input matrix, is the derivative of the target state estimation error, is the rate of change of the target control input, is the dimensional identity matrix, , , are all gains.

[0038] Theorem 1: Consider the Lyapunov equation: (10) where, is a positive definite symmetric matrix, , denotes a symmetric matrix. Then, based on the observer (6), the velocity and control input of the target can be estimated.

[0039] Proof: Define a candidate Lyapunov function as follows: (11) where, denotes the estimation error energy function, denotes the target state estimation error.

[0040] To make Hurwitz matrix (all eigenvalues have negative real parts), select the gain such that the characteristic equation of the matrix satisfies: all eigenvalues are located in the left half complex plane, denotes the variable of the characteristic equation, denotes the identity matrix.

[0041] Taking the derivative of the Lyapunov function gives: (12) where, denotes the derivative of the estimation error function, denotes the derivative of the target state estimation error.

[0042] According to Lyapunov equation and , we have (13) where , denotes the upper bound of the control input rate.

[0043] Let , using Young inequality: (14) where denotes the upper bound of the norm, denotes the minimum eigenvalue.

[0044] Thus we have (15) This shows that the target state estimation error is ultimately uniformly bounded, and when (i.e., the target acceleration is constant), the estimation error converges exponentially to 0.

[0045] Sliding mode disturbance observer design (SMDO) Define an auxiliary state variable with the dynamic equation: (16) where is the sliding surface, is the observer gain, denotes the distributed controller, denotes the speed of the unmanned surface vehicle , denotes the auxiliary state variable.

[0046] Based on (16), the disturbance observer is designed as: (17) Taking the derivative of the sliding surface , we have: (18) where denotes , denotes the derivative of the sliding surface.

[0047] Define the disturbance estimation error : (19) In the formula, For unmanned surface vessels An estimate of the disturbance. Further, we can obtain... .

[0048] Theorem 2: If Assumption 1 holds and the observer gain satisfies Then the system disturbance can be accurately estimated for the observer (17).

[0049] prove: Candidate Lyapunov functions are defined as follows: (20) in, This represents the energy function related to the perturbation.

[0050] Differentiating (20), we get (twenty one) Considering, , Substituting into (19), we get... (twenty two) Furthermore, we can obtain This indicates that, The exponent converges to 0. According to the equivalent control principle on the sliding surface, once the system reaches and maintains its position on the sliding surface, it means... ,Right now .

[0051] S5. Based on the estimated target speed, the estimated target control input, and the estimated disturbance response of the unmanned surface vessel, design a distributed cooperative controller.

[0052] The self-organizing tracking controller designed in this application mainly includes four main modules: driving the unmanned surface vessel to approach the target. Achieving speed coordination between unmanned surface vessels Maintain distance between machines And autonomous obstacle avoidance Therefore, this application designs a controller with the following form: (twenty three) in, It is a distributed controller; To control the gain; unmanned surface vessel Location; express Unmanned surface vessel at the most recent trigger moment speed; Indicates the first an unmanned surface vehicle; denotes a dynamic neighborhood of the unmanned surface vehicle; denotes a set of unmanned surface vehicles and a set of communication weights between the unmanned surface vehicles; denotes a velocity of the unmanned surface vehicle at a nearest trigger time; is a gradient operator; is an obstacle avoidance potential function; denotes a relative distance between the unmanned surface vehicle and a set of unmanned surface vehicles; is an obstacle avoidance potential function; denotes a set of obstacles; denotes a set of obstacles; denotes a relative distance between the unmanned surface vehicle and a set of obstacles; denotes a trigger time; denotes a cooperative pursuit control term; denotes a velocity cooperation term; denotes an obstacle avoidance term; is an estimate of the disturbance by the unmanned surface vehicle .

[0053] The design has the form (24) where , is the body radius, is the safe collision avoidance distance, , is the collision avoidance constant. Further, we have (25) for the obstacle avoidance potential : (26) where , is the obstacle avoidance potential activation distance, is the safe obstacle avoidance distance, is the radius of an obstacle, is the obstacle avoidance constant. Taking the gradient, we have (27) ​S6, a distributed event-triggered mechanism is designed, and a state error of the unmanned ship is defined as a difference between a current speed of the unmanned ship and a speed broadcast to a neighbor at a last triggering time; when a norm of the state error of the unmanned ship meets a preset triggering condition, communication is activated, and speed information of the current unmanned ship is broadcast to the neighbor unmanned ship; otherwise, the communication is kept silent.

[0054] S7, the speed information broadcast at the last triggering time in step S6 is taken as an input of the distributed cooperative controller, the distributed cooperative controller is executed, and the unmanned ship cluster is driven to autonomously form and maintain a stable tracking configuration around the target without presetting formation parameters, while avoiding inter-ship collision and environmental obstacles.

[0055] In the traditional control strategy, the information interaction between unmanned ships is continuous, that is, data transmission is performed through a fixed update step. However, when the consistency error of the unmanned ship is in a relatively stable state, the controller does not need to be calculated through the update of the state. Therefore, the application proposes a tracking controller (23) based on an event-triggered mechanism. The design of the event-triggering rule is as follows.

[0056] For unmanned ships, a state error (28) When the following triggering condition is met, the unmanned ship transmits its speed information to the neighbor through communication.

[0057] (29) (30) In the formula, is a triggering function, is a parameter to be designed, represents a cardinality of a neighbor set of the unmanned ship.

[0058] Theorem 3 considers that the unmanned ship model with external disturbance satisfies (3), if the parameters satisfy , , , then under the action of the observer (6), (17) and the triggering controller (23), the unmanned ship formation can realize tracking of the target, and no Zeno behavior occurs.

[0059] Next, first, the formation tracking performance in the obstacle-free environment is proved.

[0060] Define the relative position error and the relative speed error : , .

[0061] Define Lyapunov candidate functions as follows (31) in, Denotes Lyapunov candidate functions. Indicates the number of unmanned surface vessels. Indicates the first A single unmanned surface vessel Indicates control gain. express The relative positional error between the unmanned surface vessel and the target. express The relative speed error between the unmanned surface vessel and the target. Represents the dynamic neighbor set of the unmanned surface vessel. Indicates the first A single unmanned surface vessel express Unmanned surface vessels and Communication weights between unmanned surface vessels express Unmanned surface vessels and Distance difference between unmanned surface vessels; Taking its derivative, we get, (32) in, Indicates the trigger time The relative speed error between the unmanned surface vessel and the target. Indicates the trigger time The relative speed error between the unmanned surface vessel and the target. Indicates control gain. Indicates control gain; Combining (28), (32) can be further simplified to (33) in, , , express The state error of the unmanned surface vessel express The state error of the unmanned surface vessel express The relative speed error between the unmanned surface vessel and the target. express The relative speed error between the unmanned surface vessel and the target. Represents the state error vector. Indicates the constant value of the inequality. denotes the Laplacian matrix, denotes the three-dimensional identity matrix, denotes the velocity error vector. According to (29), we have (34) According to (30), we have (35) where, denotes the triggering parameter.

[0062] It is known that, (36) where, denotes the maximum eigenvalue.

[0063] Then (34) can be transformed into (37) Therefore, . Combining (30) and Theorem 1, we can know that the velocity of the swarm system can converge to the target velocity uniformly, and no collision occurs. Further, we can obtain that the USV formation can autonomously form a stable configuration and track the target.

[0064] Next, we prove that the USV formation can achieve autonomous obstacle avoidance.

[0065] Define the Lyapunov function (38) where, denotes the Lyapunov function, denotes the position difference between the USV and the obstacle; Taking the derivative of (38), we have (39) where, denotes the derivative of the Lyapunov function.

[0066] Considering that the time for the USV to enter the obstacle activation region is finite, and except for , the changes in the remaining state quantities are bounded. According to the definition of the obstacle avoidance potential function, , if . Then, when the USV approaches the obstacle, the following inequality can be obtained: (40) Further, we have . The following inequality holds ​ (41) Then by adjusting , can be obtained Therefore, based on the above theoretical analysis, it can be concluded that during the process of unmanned surface vessels forming a configuration and tracking, they can autonomously avoid obstacles.

[0067] We will now demonstrate that Zeno behavior does not exist in the system under the triggering strategy.

[0068] When state error Exceed When the next trigger time is activated, that is... (42) exist Inside, to Finding the derivative of Dini, we get (43) Furthermore, it can be seen that, (44) make , , This represents the upper bound of the rate of change of state error. Let represent the upper bound of the state error, then (45) At the triggering time, there is According to the principle of comparison, we can obtain (46) Furthermore, when At that time, it can be obtained (47) in, express The velocity coordination term at the trigger moment.

[0069] when At that time, obviously .like Then there is However, when the collaborative task is not completed, there are... This indicates Under triggering rules (29)-(30), the system does not have Zeno behavior.

[0070] By implementing steps S1 to S7 above, the collision avoidance and obstacle avoidance potential field gradients spontaneously drive the cluster to form a stable tracking configuration around the target. This intrinsic mechanism eliminates the need for any externally designed formation function, thereby enabling the multi-unmanned surface vessel system to obtain strong self-adjustment and adaptability. The emerging configuration can be dynamically adjusted according to the target's maneuverability and environmental obstacles, and can still maintain the integrity of the tracking structure when some members are missing.

[0071] Simulation and Analysis This application verifies the algorithm through simulation experiments. First, the effectiveness of the Extended State Observer (ESO) and Sliding Mode Disturbance Observer (SMDO) is verified; second, the effectiveness of the controller is verified in both scenarios with and without obstacles; finally, the superiority of the proposed triggering mechanism is verified through comparative simulations.

[0072] The simulation parameters are set as follows. State observer gain: , , Perturbation observer gain: , Control gain: , , , Number of unmanned surface vessels: Potential field parameters: , , , Triggering parameters: , Upper bound of perturbation: Upper bound of acceleration: The control input for selecting the target is: .

[0073] Target state estimation and disturbance estimation Simulation results of the two observers are as follows Figures 5-7 As shown. Figure 5 The results show that the target velocity estimation errors of each unmanned surface vessel converge to the zero neighborhood, indicating that the designed extended state observer can quickly track the unknown velocity of the target and satisfies the exponential convergence characteristic described in Theorem 1. Figure 6 The robust estimation capability of the extended state observer for time-varying control inputs was verified. Figure 7 This indicates that the disturbance estimation error remains stable within a small range, proving that the sliding mode disturbance observer can effectively compensate for time-varying disturbances.

[0074] Collaborative target tracking in accessible environments Results of collaborative target tracking in an accessible environment, such as Figures 8-13 As shown. Figure 8The motion trajectory of the target (dashed line) and the motion trajectory of the swarm (solid line) in two-dimensional space are shown. Figure 9 The spatial distribution of the six time instants shows that the unmanned surface vehicles can self-organize into a stable cluster configuration without pre-defined geometric formation functions under the controller, and the self-regulating ability of the anti-collision potential field is verified. Figure 10 The cluster tracking error asymptotically converges to a neighborhood of zero, proving that the proposed distributed controller can effectively achieve the expected goal. Figure 11 The minimum distance curve between the agents is given, and it can be seen that the unmanned surface vehicles always meet the collision avoidance constraints, proving the effectiveness of the potential field design. Figure 12 It shows that the speed of all slave vehicles can be synchronized to the target speed. Based on the above analysis, the unmanned surface vehicle swarm realizes effective and stable tracking of the target. Figure 13 The event-triggered communication distribution shown in Table 1 shows that compared with continuous communication, the communication frequency between unmanned surface vehicles is significantly reduced, verifying that the event-triggered mechanism effectively reduces the communication burden while maintaining performance.

[0075] Table 1 Comparison of trigger times in an obstacle-free environment Cooperative target tracking in an obstacle environment Through Figures 14-19 The effectiveness of the proposed control strategy in the obstacle avoidance scenario is verified. Figure 14 The target trajectory (dashed line) and the unmanned surface vehicle trajectory (solid line) in a two-dimensional space containing a spherical obstacle (red area) are shown. It can be observed that when the swarm approaches the obstacle, the unmanned surface vehicles autonomously generate a smooth collision-free detour path through the obstacle avoidance potential field. Figure 15 The spatial distribution of the six time instants shows that the unmanned surface vehicle formation can successfully avoid obstacles and restore a stable configuration in a complex environment, verifying the self-organizing ability without pre-defined geometric constraints. Figure 16 The tracking error shows transient fluctuations during obstacle avoidance, but ultimately converges to an acceptable range, proving the robustness of the controller to obstacle disturbances. Figure 17 The minimum distance curve between the unmanned surface vehicles is given, confirming the effectiveness of the anti-collision potential field in the obstacle environment. Figure 18 It shows that the slave vehicles produce transient speed fluctuations due to avoidance maneuvers, but ultimately remain synchronized with the target speed, meeting the cooperative speed requirement. Figure 19 The event-triggered communication distribution shown in Table 2 shows that the communication frequency of all unmanned surface vehicles remains at a low level, and Zeno behavior is excluded.

[0076] Table 2 Comparison of trigger times in an obstacle environment Comparison of different trigger function simulations To further demonstrate the superiority of the event-triggered control method designed in this paper, simulations are performed under the conditions of no obstacles and the presence of obstacles, respectively, and the triggering function of the prior art [Yang, D.; Ren, W.; Liu, X. Decentralized event-triggered consensus for linear multi-agent systems under general directed graphs. Automatica 2016, 69, 242-249.] is used for comparative simulation. The simulation results are as follows Figures 20-21 and Tables 3-4.

[0077] Table 3 Comparison of triggering times of different schemes under obstacle-free environment Table 4 Comparison of triggering times of different schemes under obstacle environment Compared with the event-triggered control method designed in this application, the event-triggered times of the prior art [Yang, D.; Ren, W.; Liu, X. Decentralized event-triggered consensus for linear multi-agent systems under general directed graphs. Automatica 2016, 69, 242-249.] are significantly higher. The simulation results show that the target tracking control algorithm design based on event triggering can effectively reduce the communication times, realize intermittent communication, and achieve target tracking.

[0078] In summary, this application proposes a distributed event-triggered control framework for solving the cooperative tracking problem of multiple unmanned surface vehicles in dynamic uncertain environments. Through observer design, dynamic estimation of the unknown motion state of the leader and external disturbance is achieved; a self-organizing cooperative control architecture is proposed, which synchronously solves the problems of formation configuration generation, collision avoidance between unmanned surface vehicles and environmental obstacle avoidance through potential field gradient driving mechanism; an event-triggered strategy based on neighbor state error is constructed to regulate communication behavior, which significantly optimizes network load while ensuring system performance. Based on Lyapunov method, the global asymptotic stability of the closed-loop system is strictly proved, and it is proved that the triggering interval has a strict positive lower bound, which fundamentally excludes Zeno behavior. Future work will study complex formation tracking control under network attacks.

[0079] Any technical features in the above embodiments can be combined, and for the sake of brevity, not all possible combinations are described above, however, it should be understood that the application encompasses all possible combinations of the technical features described above.

[0080] The principles and implementation manners of the present application are described herein by using specific examples, and the above embodiments are only used to help understand the method of the present application and its core idea; meanwhile, according to the idea of the present application, the specific implementation manners and application scopes will be changed by those skilled in the art. In conclusion, the content of the present specification should not be understood as a limitation of the present application.

Claims

1. A method for controlling a maneuvering target tracking of an unmanned surface vehicle (USV) cluster in a complex dynamic environment, characterized in that, The unmanned vehicle cluster mobile target tracking control method in the complex dynamic environment comprises the following steps: S1, respectively configuring an extended state observer and a sliding mode disturbance observer for each unmanned vehicle in the cluster; S2, obtaining position information of the target; S3, using the extended state observer and based on the position information of the target, estimating the unknown speed and control input of the target in real time to obtain an estimated value of the target speed and an estimated value of the target control input; S4, using the sliding mode disturbance observer, estimating and compensating the external bounded disturbance acting on the unmanned vehicle in real time to obtain an estimated value of the disturbance to the unmanned vehicle; S5, based on the estimated value of the target speed, the estimated value of the target control input and the estimated value of the disturbance to the unmanned vehicle, designing a distributed cooperative controller; S6, designing a distributed event triggering mechanism, defining the state error of the unmanned vehicle as the difference between its current speed and the speed broadcast to the neighbors at the last triggering time; when the norm of the state error of the unmanned vehicle meets the preset triggering condition, activating communication and broadcasting the speed information of the current unmanned vehicle to the neighbor unmanned vehicles; otherwise, keeping communication silent; S7, taking the speed information broadcast at the last triggering time in step S6 as the input of the distributed cooperative controller, executing the distributed cooperative controller to drive the unmanned vehicle cluster to autonomously form and maintain a stable tracking configuration around the target without presetting formation parameters, while avoiding inter-boat collision and environmental obstacles.

2. The method of claim 1, wherein, The position information of the target is obtained based on the position information of the target obtained from the sensor of the unmanned vehicle itself.

3. The method of claim 1, wherein, The distributed cooperative controller comprises a target attraction term, an inter-boat collision avoidance term, an environmental obstacle avoidance term and a speed cooperation term, wherein the inter-boat collision avoidance term and the environmental obstacle avoidance term are respectively generated by the negative gradient of the collision avoidance potential field function and the obstacle avoidance potential field function, and the speed cooperation term is used to drive the speed of all unmanned vehicles to be consistent with the target speed.

4. The method of claim 1, wherein, The expression of the extended state observer is as follows: ； wherein, denotes an unmanned surface vehicle derivative of the estimate of the target position; is an unmanned surface vehicle estimate of the target velocity; denotes a measurable position of the target; is an unmanned surface vehicle estimate of the target position; denotes an unmanned surface vehicle derivative of the estimate of the target velocity; is an unmanned surface vehicle estimate of the target control input; denotes an unmanned surface vehicle derivative of the estimate of the target control input, is an extended state observer gain.

5. The method of claim 1, wherein, The expression of the sliding mode disturbance observer is as follows: ； wherein, unmanned surface vehicle an estimate of the disturbance; , denotes a disturbance observer gain, denotes a sliding surface.

6. The method of claim 4, wherein, The expression of the distributed cooperative controller is as follows: ； wherein, is a distributed controller; is a control gain; denotes an unmanned surface vehicle position; denotes the velocity of the unmanned surface vehicle at the last triggered time instant ; denotes the unmanned surface vehicle; denotes the dynamic neighbor set of the unmanned surface vehicle; denotes the communication weight between the unmanned surface vehicle and the unmanned surface vehicle; denotes the velocity of the unmanned surface vehicle at the last triggered time instant ; is a gradient operator; is an obstacle avoidance potential field function; denotes the relative distance between the unmanned surface vehicle and the unmanned surface vehicle; is an obstacle avoidance potential field function; denotes the obstacle; denotes the set of obstacles; denotes the relative distance between the unmanned surface vehicle and the obstacle; denotes the triggered time instant; denotes the cooperative pursuit control term; denotes the velocity cooperation term; denotes the obstacle avoidance term; is an estimate of the disturbance by the unmanned surface vehicle .

7. The method of claim 3, wherein, The expression of the collision avoidance potential field function is as follows: ； wherein, is a gradient operator; is a collision avoidance potential function; , is a collision avoidance constant; denotes the relative distance between the USV and another USV, is a safe collision avoidance distance; is a body radius; denotes the position of the USV ; denotes the position of the USV .

8. The method of claim 3, wherein, The expression of the obstacle avoidance potential field function is as follows: ； wherein, is a gradient operator, is an obstacle avoidance constant; denotes an unmanned surface vehicle and a relative distance to an obstacle; is a safe obstacle avoidance distance, is a radius of an obstacle; is a radius of the vehicle; denotes a position of the unmanned surface vehicle ; denotes a position of an obstacle.

9. The method of claim 1, wherein, The expression of the preset triggering condition is as follows: ； ； wherein, denotes triggering time instant; denotes the current time instant; denotes triggering time instant; denotes a triggering function, denotes a state error, denotes a parameter to be designed, denotes cardinality of the unmanned surface vehicle neighborhood set; denotes velocity coordination term at the triggering time instant.

Images